Kinetic resolution of (: RS)-1-chloro-3-(4-(2-methoxyethyl)phenoxy)propan-2-ol: a metoprolol intermediate and its validation through homology model of Pseudomonas fluorescens lipase

Article abstract of DOI:10.1039/c7ra06499c

In the present study Pseudomonas fluorescens lipase (PFL) was screened as a time efficient biocatalyst for the kinetic resolution of a racemic intermediate [(RS)-1-chloro-3-(4-(2-methoxyethyl)phenoxy)propan-2-ol] of metoprolol, an important selective β₁-blocker drug. PFL selectively acylated the R-form of this racemic intermediate in a short duration of 3 h. Different reaction parameters were optimized to achieve maximum enantioselectivity. It was found that at 30 °C, enzyme activity of 400 units and substrate concentration of 10 mM gave a high enantioselectivity and conversion in an optimum time of 3 hours (C = 50.5%, ee_p = 97.2%, ee_s = 95.4%, E = 182). To validate these experimental results, the 3D structure of PFL was built using homology modelling. Validation of the model through Ramachandran plot (92.7% in favored region), Errat plot (overall quality factor, 79.27%), Verify-3D score (86.19) and ProSA-Z score (-6.24) depicted the overall good quality of the model. Molecular docking of the R- and S-enantiomers of the intermediate, which was performed on this model, demonstrated a strong H-bond interaction (1.6 ?) between the hydroxyl group of the R-enantiomer and Arg54, a key binding residue of the catalytic site of PFL, while no significant interaction with the S-enantiomer was observed. Thus, the outcome of this docking study was in agreement with the experimental data, clarifying that PFL preferentially catalysed the transesterification of the R-enantiomer into the corresponding ester, leaving the S-enantiomer intact.

Full text of DOI:10.1039/c7ra06499c

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Kinetic resolution of (RS)-1-chloro-3-(4-(2-
methoxyethyl)phenoxy)propan-2-ol: a metoprolol
intermediate and its validation through homology
model of Pseudomonas fluorescens lipase†
Cite this: RSC Adv., 2017, 7, 36566
Surbhi Soni, Bharat P. Dwivedee, Vishnu K. Sharma and Uttam C. Banerjee *b
a
b
c
In the present study Pseudomonas fluorescens lipase (PFL) was screened as a time efficient biocatalyst for
the kinetic resolution of a racemic intermediate [(RS)-1-chloro-3-(4-(2-methoxyethyl)phenoxy)propan-2-
1
ol] of metoprolol, an important selective b -blocker drug. PFL selectively acylated the R-form of this
racemic intermediate in a short duration of 3 h. Different reaction parameters were optimized to achieve
ꢀ
maximum enantioselectivity. It was found that at 30 C, enzyme activity of 400 units and substrate
concentration of 10 mM gave a high enantioselectivity and conversion in an optimum time of 3 hours (C
¼
50.5%, ee
p
¼ 97.2%, ee ¼ 95.4%, E ¼ 182). To validate these experimental results, the 3D structure of
s
PFL was built using homology modelling. Validation of the model through Ramachandran plot (92.7% in
favored region), Errat plot (overall quality factor, 79.27%), Verify-3D score (86.19) and ProSA_Z score
(ꢁ6.24) depicted the overall good quality of the model. Molecular docking of the R- and S-enantiomers
of the intermediate, which was performed on this model, demonstrated a strong H-bond interaction (1.6
A˚ ) between the hydroxyl group of the R-enantiomer and Arg54, a key binding residue of the catalytic site
of PFL, while no significant interaction with the S-enantiomer was observed. Thus, the outcome of this
docking study was in agreement with the experimental data, clarifying that PFL preferentially catalysed
the transesterification of the R-enantiomer into the corresponding ester, leaving the S-enantiomer intact.
Received 10th June 2017
Accepted 5th July 2017
DOI: 10.1039/c7ra06499c
rsc.li/rsc-advances
6
their excellent chemo-, regio- and enantioselectivity. Lipases
obtained from different sources belong to the a/b-hydrolase
Introduction
The growing demand of optically pure compounds for appli- folding group. Specic sequences of the a-helices and b-strands
7
cations in pharmaceutical, chemical, agricultural and cosmetic are responsible for the hydrolyzing activity of lipases. Lipases
industries is well understood. The increased interest in these are widely applied in kinetic resolution, where the lipase
compounds can be clearly attributed to the better under- discriminates between two enantiomers of the racemic mixture
1,2
standing of chirality and its importance in synthetic reactions.
and selectively catalyzes a faster conversion of one enantiomer
As the predominant outlook of the industry is always on to the product. This is a convenient method for separation of
8
economically expedient, dependable and scalable processes both enantiomers from a racemic mixture. The rationality of
with minimal waste generation, biocatalysis has become lipase catalyzed biocatalysis as a compatible method for enan-
3
a cornerstone for the synthesis of chiral intermediates. Bio- tiopure synthesis is further augmented by the incorporation of
9
catalysis involves the application of enzymes in a suitable form computational strategies. Computational chemistry provides
(
whole-cell, immobilized or commercial preparation) to catalyze an effective set of tools to simulate the biocatalytic reactions in
4,5
chemical reactions.
Lipases as biocatalysts have earned silico and help in the assessment of the effectiveness of the
10
a distinguished position in the eld of chiral synthesis due to process, thereby validating the experimentation. One key
computational methodology is the docking of small molecules
11
into the active site of lipase. The utilization of docking studies
enables the fast evaluation of enantioselectivity and is of prime
interest to suggest a plausible biotransformation reaction. This
enables expensive experimentation to be directed and focused,
a
Department of Biotechnology, National Institute of Pharmaceutical Education and
Research, S.A.S. Nagar, 160062, Punjab, India
b
Department of Pharmaceutical Technology (Biotechnology), National Institute of
Pharmaceutical Education and Research, S.A.S. Nagar 160062, Punjab, India.
12
thereby speeding up bioprocess development.
E-mail: ucbanerjee@niper.ac.in
c
Department of Pharmacoinformatics, National Institute of Pharmaceutical Education
b-Aryloxyalcohols are useful building blocks in the synthesis
of several biologically active compounds. Metoprolol, 1-[4-(2-
methoxyethyl)phenoxy]-3-(propan-2-ylamino)propan-2-ol, a selective
and Research, S.A.S. Nagar 160062, Punjab, India
†
Electronic supplementary information (ESI) available. See DOI:
0.1039/c7ra06499c
1
36566 | RSC Adv., 2017, 7, 36566–36574
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
1
b -adrenergic blocking agent, is used in the treatment of angina
and hypertension. It is well established that the desirable
therapeutic activity of metoprolol resides mainly in the (S)-
13
enantiomer. A number of chemical methods have been
described in the literature for the synthesis of enantiopure
14–22
metoprolol.
However, most of the chemical synthesis
Scheme 1 Kinetic resolution of (RS)-4 using various lipases.
methods suffer from some disadvantages, such as poor enan-
tioselectivity, long reaction time, hazardous reaction condi-
tions, formation of by products, and higher cost. Biocatalytic
approach has also been used for the synthesis of chiral inter-
good enantioselectivity and conversion. Moreover, PFL cata-
lyzed the kinetic resolution of the racemate in a short time. This
behaviour of PFL towards the rapid enantioselective conversion
may be due to the favourable interactions between the amino
acid residues in the active site and the ligands.
23
mediates of metoprolol.
To the best of our knowledge, herein, we report for the rst
time the homology model of PFL and its application to validate
the R-selectivity of PFL in the kinetic resolution of a racemic
intermediate of metoprolol. In the present study, the racemic
intermediate [(RS)-1-chloro-3-(4-(2-methoxyethyl)phenoxy)propan-
Optimization of reaction parameters for kinetic resolution
2-ol] of metoprolol was chemically synthesized. Various To achieve a successful and efficient lipase catalyzed reaction,
commercial lipase preparations were screened for the kinetic conditions favouring rapid progress and irreversibility must be
resolution of this racemic intermediate. Among them, PFL was taken into account. Various reaction parameters such as type of
selected as the most suitable lipase. The reaction parameters were solvents, reaction time and temperature, type of acyl donors,
optimized using ‘one-factor at a time’ approach. Furthermore, to substrate and enzyme concentrations etc. were optimized. The
validate the experimental results, a homology model of PFL was substrate concentration (10 mM), enzyme concentration (400
built since it was not available. The developed homology model IU), solvent (toluene), acyl donor (vinyl acetate), temperature (30
ꢀ
was utilized for the molecular docking analysis wherein experi-
C), and reaction time (3 h) were optimized on the basis of high
mental observations were validated by docking the R and S form enantiomeric excess (ee & ee ), high conversion rate (>49%)
s
p
of intermediate into the active site of PFL.
and enantiomeric ratio (E # 200) [detailed discussion on each
parameter is given in the next section]. Under these optimized
conditions, PFL performed the acylation of (R)-enantiomer of
Results and discussion
Chemical synthesis of racemic intermediate, (RS)-4
(
RS)-4 with a nal result of C ¼ 50.5%, ee
p
¼ 97.2%, ee ¼
s
95.4%, E ¼ 182.
b-Aryloxyalcohols are key intermediates in the preparation of
metoprolol. The fundamental structure of metoprolol Effect of solvents
comprises aryloxypropanolamine. The aromatic ring substitu-
Several examples of improved regioselectivity and enantiose-
lectivity of lipases in organic reactions have been observed.
Lipases show good stability in organic solvents. One of the
ents govern the hydrophobicity of the molecule. Branched alkyl
amine substituents present in metoprolol are essential for the
31
24
b-antagonist activity. A convenient protocol for the prepara-
tion of b-aryloxyalcohols requires the synthesis of epoxide
important reasons for such a behaviour is that hydrophilic
solvents have a higher tendency to strip the tightly bound water,
which is necessary for the catalytic activity of lipase. Organic
solvents have a lesser tendency for this, making them a better
2
5–27
intermediates with phenols
the epoxides.
and subsequent ring-opening of
28,29
The racemic intermediate (RS)-4 of metoprolol
was prepared via a chemical route. Detailed discussion
regarding the synthetic schemes has been furnished in the
Experimental section.
32
reaction medium for lipase catalyzed reactions. Considering
these facts, different organic solvents with a diverse range of
log P values were examined for the kinetic resolution of (RS)-4
with vinyl acetate using PFL. The best result was obtained using
toluene with the highest percentage conversion (50.5%), enan-
Screening of suitable biocatalysts using experimental
approach
s p
tiomeric excess (97.4% ee & 95.5% ee ) and E-value 188
(
Fig. 1A).
Lipases are interfacial enzymes and display prodigious stereo-,
chemo-, and regioselectivity. In the present study, (RS)-4 was
subjected to lipase catalyzed kinetic resolution (Scheme 1). Five
30
Effect of acyl donors
commercially available lipases [Thermomyces lanuginosus (TLL), Enantioselectivity in the lipase mediated reactions originates
Candida rugosa (CRL), Pseudomonas uorescens (PFL), Candida during the deacylation step with the nucleophilic attack of the
antarctica lipase A (CALA) and Candida antarctica lipase B alcohol leading to the formation of the rst acyl–enzyme
33
(
CALB)] were screened for the kinetic resolution of (RS)-4 using complex. Hence, the type of acyl donor also has a pivotal role
vinyl acetate as an acyl donor and toluene as a solvent. Inter- in the progress of the reaction. Among the different acyl donors
estingly, Pseudomonas uorescens lipase (PFL) was found to be [BA ¼ benzyl acetate, EA ¼ ethyl acetate, IPA ¼ isopropenyl
the most suitable biocatalyst (Table 1). PFL resolved (RS)-4 to acetate, VA ¼ vinyl acetate], vinyl acetate showed the best results
the corresponding (R)-ester and (S)-alcohol products, displaying in the kinetic resolution of (RS)-4. Vinyl acetate is an enol ester
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 36566–36574 | 36567

View Article Online
RSC Advances
Paper
a
Table 1 Screening of lipases for the transesterification of (RS)-4 with vinyl acetate
b
c
s
d
p
e
S. no.
Lipase
% C
ee
ee
E
1
2
3
4
5
Thermomyces lanuginosus (TLL)
Candida antarctica immobilized (CAL B)
Candida antarctica immobilized (CAL-A)
Pseudomonas uorescens (PFL)
Candida rugosa (CRL)
50.0 ꢂ 1.21
50.3 ꢂ 2.98
51.2 ꢂ 1.27
50.0 ꢂ 1.58
53.5 ꢂ 2.25
74.4 ꢂ 2.95
70.8 ꢂ 3.23
93.7 ꢂ 3.01
96.2 ꢂ 2.69
95.2 ꢂ 2.84
74.3 ꢂ 2.31
69.9 ꢂ 1.99
89.5 ꢂ 2.08
96.1 ꢂ 3.26
82.8 ꢂ 2.76
14.9 ꢂ 3.51
11.7 ꢂ 3.78
62.9 ꢂ 1.76
200 ꢂ 2.65
39.3 ꢂ 1.82
a
ꢀ
ꢁ1
Conditions: (RS)-4 (20 mM) in toluene (1 mL) was treated with vinyl acetate (5.4 mmol) at 30 C in the presence of different lipases (300 IU mL ).
conversion were calculated from the enantiomeric excess (ee) of (S)-4 and (R)-5 as follows: conversion (C) ¼ ee
b
c
%
s s p
/(ee + ee ). Enantiomeric excess
ꢁ1
of (S)-4 was determined by HPLC analysis (Daicel Chiralcel OD-H column) 90 : 10; hexane : IPA, ow rate of 1.0 mL min , detected at 254 nm.
d
Enantiomeric excess of (R)-5 was determined by HPLC analysis (Daicel Chiralcel OD-H column) 90 : 10; hexane : IPA, ow rate of 1.0 mL
ꢁ1
e
35
min , detected at 254 nm. E value were calculated using the formula: E ¼ ln[ee
p
s p s p s p s
(1 ꢁ ee )/(ee + ee )]/ln[ee (1 + ee )/(ee + ee )].
Fig. 1 Effect of (A) solvents [Tol ¼ toluene, Hex ¼ n-hexane, Hep ¼ n-heptane, DEE ¼ diethyl ether, But ¼ n-butanol, Iso ¼ isooctane]; (B)
reaction time on PFL catalyzed kinetic resolution of [(RS)-4].
that forms unstable enol as a by-product and rapidly tauto- conversion along with enantiomeric excess were low at the early
merizes to form the corresponding aldehydes or ketones onset of the reaction, PFL exhibited the best results promptly.
34
making the reaction completely irreversible. The highest Such high reactivity of PFL towards the intermediate is justied
percentage conversion (50.1%), enantiomeric excess (96.4% ees with the interaction of the (R)-enantiomer at the active site of
&
96.1% ee
p
) and E-value (205) of PFL catalysed kinetic resolu- the enzyme, as later validated by the docking study. During the
tion of (RS)-4 in toluene were obtained with vinyl acetate as the course of the reaction, the enantioselectivity gradually
acyl donor (Fig. 2D).
decreased and PFL exhibited non-selective behaviour. In
a recent study on kinetic resolution of racemic atenolol, PFL has
36
been reported to give (S)-enantiomer in 24 h. In another study,
PFL catalyzed the kinetic resolution of some novel antifungal N-
Effect of reaction time
37
substituted benzimidazole derivatives in more than 72 h.
The most interesting part of the study is the short duration in
which PFL catalyzes the enantioselective reaction of (RS)-4. In
the reaction of (RS)-4, PFL took 3 h for 50.0% conversion with
Thus, the rapid action of PFL on (RS)-4 with high degree of
enantioselectivity and rate of conversion reported here is
suggestive of strong interaction between (RS)-4 and PFL, thereby
revealing a better chemoenzymatic process that could be used
for the synthesis of enantiopure metoprolol in a shorter time.
9
5.5% ee , 96.0% ee , and E-value of 183.5 (Fig. 1B). PFL started
s p
to display good enantioselectivity and conversion rate very early
in the reaction, in only over 30 min to deliver the best result.
Although the values of enantioselective ratio and percentage
36568 | RSC Adv., 2017, 7, 36566–36574
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
Fig. 2 Effect of (A) enzyme concentration; (B) substrate concentration; (C) temperature; and (D) acyl donors [BA ¼ benzyl acetate, EA ¼ ethyl
acetate, IPA ¼ isopropenyl acetate, VA ¼ vinyl acetate] on PFL catalyzed kinetic resolution of [(RS)-4].
ꢀ
Effect of substrate concentration
(RS)-4, the conversion at 30 C was 49.3%, with 93.8% ee
s
, 96.5%
p
ee and 201 E-value aer 3 h of reaction (Fig. 2C).
The effect of substrate concentration on the enantioselectivity
and conversion of an enzymatic reaction has been extensively
38
studied in the literature. Thus, we performed the reaction
using a range of substrate concentration from 1 to 30 mM. The
optimal substrate concentration was 10 mM, at which PFL
displayed good conversion rate of 49.2% along with the ee , ee
s p
and E-value of 93.4%, 96.3% and 187, respectively (Fig. 2B).
Validation of suitable biocatalyst using computational
approach
The structural information of several lipases is well exploited to
give an insight into the catalytic mechanism and enantiopre-
40
ference of the biocatalytic reaction. During the experimental
study, PFL exhibited biocatalytic reaction in only three hours.
This outcome encouraged us to explore the interaction between
PFL and the substrate. Homology modeling and docking
studies of two enantiomers into the active site of PFL were
carried out to substantiate the experimental outcomes of this
study (Table 2).
Effect of enzyme concentration
A biocatalytic transformation occurs at the active sites of the
enzyme. Owing to the kinetic consideration of the enzymes, an
optimal enzyme concentration, where all the active sites are
acquired by substrate molecules, should be identied. This has
a direct effect on the cost effectiveness of the process. Thus, we
optimized the enzyme concentration to obtain the highest
enantioselectivity. The reaction of (RS)-4 was carried out using
different concentrations (100, 200, 300, 400, 600 IU) of PFL in
toluene. At the concentration of 400 IU, PFL showed the highest
39
Development of homology model for Pseudomonas uo-
rescens lipase. Since the 3D crystal structure information of PFL
was not available in the Protein Data Bank (PDB) [http://
www.rcsb.org], we focused on the development of its
homology model. The protein sequence of (PFL) was retrieved
s p
conversion 50.6% and ee , ee and E-value were found to be
41
from UniProt Knowledge Base webserver [primary accession
no. P26504 (LIPA_PSEFL)] with a total length of 449 amino
acids. This sequence was utilized as a target or a query to nd
out the suitable template sequence and crystal structure within
the Protein Data Bank using PSI-BLAST online program. The
PSI-BLAST search program revealed Pseudomonas sp. MIS38
97.4%, 95.1% and 173, respectively (Fig. 2A). Further increase in
enzyme concentration reduced the enantioselectivity of the
reaction, probably due to the non-selective interaction of the
enzymes with the substrates at higher enzyme concentration.
Effect of temperature
(PDB code 2Z8X) as the best template for PFL. The template PDB
The effect of temperature on the degree of enantioselectivity (2Z8X) was selected on the basis of query cover (93%), percent
and percentage conversion was investigated by performing the identity (63%) and resolution of the template PDB crystal
˚
reactions at a range of temperatures (4, 20, 25, 30, 40, 50 and 60 structure (1.48 A). Alignment of the target and template
ꢀ
C). In the reactions of (RS)-4, PFL catalyzed the conversion with sequence was carried out using Modeller 9.5v. A total of 50
ꢀ
the highest enantioselectivity at 30 C. The increase in models were generated using model building script of Modeller
temperature showed a gradual demise in the enantioselective 9.5v. The best model was selected on the basis of DOPE score
resolution of the (R) and (S) forms of the racemic alcohol. For and the alignment of the built model with the template PDB.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 36566–36574 | 36569

View Article Online
RSC Advances
Paper
a
Table 2 Optimum reaction conditions for PFL catalysed kinetic resolution of (RS)-4
a
b
c
d
Physicochemical parameter
Optimum parameter
C (%)
ee
s
(%)
ee
p
(%)
E
Solvent
Acyl donor
Reaction time
Temperature
Substrate conc.
Enzyme conc.
Toluene
Vinyl acetate
3 h
30 C
10 mM
50.5 ꢂ 2.80
50.1 ꢂ 2.21
49.9 ꢂ 2.00
49.3 ꢂ 2.34
49.2 ꢂ 3.00
50.6 ꢂ 2.54
97.4 ꢂ 2.50
96.4 ꢂ 2.17
95.5 ꢂ 2.17
93.8 ꢂ 2.16
93.4 ꢂ 2.06
97.4 ꢂ 2.40
95.5 ꢂ 2.33
96.1 ꢂ 2.35
96.0 ꢂ 3.07
96.5 ꢂ 2.03
96.3 ꢂ 1.91
95.1 ꢂ 2.27
188 ꢂ 3.79
205 ꢂ 2.65
184 ꢂ 3.51
201 ꢂ 4.04
187 ꢂ 3.61
173 ꢂ 3.06
ꢀ
ꢁ1
400 IU (20 mg mL
)
a
a, b, c, d are the reaction conditions and equations described in Table 1.
Ramachandran plot was generated for the selected model, and function over 0.2 are considered to be reliable. Furthermore,
the residues constituting the part of outlier region were rened 86.2% of the residues had a score over 0.2. This validated the
45
using loop-renement script of Modeller 9.5v.
good quality of the predicted model as shown in Fig. 3D.
To validate the quality of the developed PFL homology Furthermore the structure was analyzed using the ProSA Z-score
model, various online validation tools like ERRAT plot, Ram- to assess the quality of the model. The ProSA Z-score value for
achandran plot, ProSA, ProQ and Verify-3D etc. were utilized. In PFL was ꢁ6.24, which falls within the range for native set of
42
the Ramachandran contour plot of PFL (Fig. 3A) 92.4% resi- proteins having the same size (Fig. 3B). If the Z-score of a model
dues were in favored regions, 5.6% in additionally allowed structure is located outside the range characteristic for native
regions and 2.0% residues in outlier regions. The modeled proteins from different sources (X-ray and NMR), it indicates an
43
46
structures were also validated by ERRAT plot, which showed erroneous structure. Thus ProSA Z-score indicated a good
an overall quality factor of 79.27 as illustrated in Fig. 3C. This correlation between the modeled PFL and the native structure
was close to the overall quality factor of the template protein (PDB-2Z8X) in almost of the parts of the sequence. The ProQ LG
(
81.85). Verify-3D program was used to validate the stereo- score, which also predicts the quality of a model protein, was
44
chemical quality of the model. The residues with a scoring 3.73. It indicated that the developed model is of good quality
Fig. 3 Validation of homology model of PFL: (A) Ramachandran plot (B) ProSA Z-score (C) Errat plot (D) Verify-3D plot.
36570 | RSC Adv., 2017, 7, 36566–36574
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
Table 3 Comparison of the model validation parameters for target interactions between the carbonyl O atom of the acyl group and
and template proteins
the –NH groups present in the oxyanion hole.
It is followed by the nucleophilic attack by an alcohol,
leading to the formation of an ester. The acyl–enzyme inter-
Validation parameters
2Z8X
PFL
50
Ramachandran plot
Favoured
Allowed
mediate was prepared by the acylation of the hydroxyl group of
92.4% Ser206. Molecular docking of R and S enantiomers of (RS)-4 was
98.4%
1.5%
0.2%
81.85
98.54
ꢁ10.66
5.85
5.6%
2.0%
79.27
86.19
performed into the acylated enzyme active site, using Schro-
Outliers
51
dinger's Glide module. It has been well established that the R
and S enantiomers of the substrate interact differently with the
Errat plot (overall quality factor)
Verify-3D
ProSA_Z score
ProQ server_LG score
26
ꢁ6.24 residues at the active site of the lipase. Stereochemical rules
3.73
have been proposed for the determination of enantioselectivity.
A large number of studies have established an empirical rule
that states the R-preference of the stereochemical trend
observed for lipase catalyzed kinetic resolution of secondary
(LG score > 3). Comparative analysis of template protein (PDB
52
alcohols. The R-enantiomer (R)-4 showed strong interaction
code 2Z8X) and target protein (PFL) is given in Table 3. Finally,
energy minimization of the homology model of PFL was per-
formed using Prime module of Schrodinger. The overall results
from these model validation tools revealed good quality of the
model for PFL.
with Arg54 residue, which forms a part of the oxyanion hole of
˚
PFL, with a H-bond distance of 1.6 A (Fig. 4A), while the S
enantiomer (S)-4 showed no signicant interactions needed for
the lipase catalyzed reaction (Fig. 4B). The G score for the R-
enantiomer was ꢁ5.33, while the S-enantiomers had a lower
score of ꢁ4.50. This outcome of the docking study is in agree-
ment with experimental data, clarifying that PFL preferentially
catalyzed the transesterication of the R-alcohol. However, this
selectivity was time dependent. Initially, PFL catalyzed the
reaction very fast in favor of transesterication of the R-enan-
tiomer. The maximum conversion of (RS)-4 into the corre-
sponding ester (R)-5 was achieved in 3 h. The result obtained in
Molecular docking of R and S enantiomers of (RS)-4 in the
active site of Pseudomonas uorescens lipase. The information
regarding the active site of PFL was obtained from literature.
The X-ray crystal structure of lipases shows the catalytic triad
composed of Ser–His–Asp/Glu residues and an oxyanion hole
47
containing the –NH groups of the peptide backbone. Ser206,
Asp254 and His312 constituted the catalytic triad of PFL, which
was identical to the catalytic triad (Ser207, Asp255, and His313)
3
h was C ¼ 50.0%, ee ¼ 95.5%, ee ¼ 96.0%, and E ¼ 183.5.
s
p
48
of Pseudomonas sp. MIS38 lipase. Tyr29, Gly53, and Arg54 are
the parts of the oxyanion hole in PFL. According to the well-
known ‘ping-pong bi-bi’ mechanism of enantioselective trans- Conclusions
49
esterication by lipases, the hydroxyl group of Ser residue acts
as a nucleophile, and the imidazole moiety of His residue acts
as a base. The acyl–enzyme intermediate (a tetrahedral inter-
mediate) is stabilized through hydrogen bonds and electrostatic
The methodology of complementing a biocatalytic reaction with
in silico studies, which is corroborated by this work, is a time-
efficient and reliable process. The present study involves
screening of various lipases and selecting Pseudomonas
Fig. 4 Interaction of ligands (A) [(R)-4] and (B) [(S)-4] with Pseudomonas fluorescens lipase (PFL).
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 36566–36574 | 36571

View Article Online
RSC Advances
Paper

uorescens lipase as an expeditious biocatalyst, performing the dd), 3.37 (3H, s), 3.68–3.57 (2H, d), 3.79–3.70 (2H, d), 4.08 (2H,
1
3
kinetic resolution of the racemic intermediate of metoprolol in d), 4.2 (1H, tt), 6.80 (2H, ddd), 7.11 (2H, dd). C NMR (100 MHz,
a short duration of three hours. The biocatalytic reaction yields CDCl ): d (ppm): 35.21, 46.02, 58.65, 69.67, 69.85, 73.81, 114.67,
3
the (S)-alcohol, which can be directly utilized in the synthesis of 114.86, 129.62, 130.15, 156.80.
(S)-metoprolol. A suitable homology model for PFL was
successfully developed, validated and used for docking to
analyse the difference in interaction of two enantiomers with
PFL. The R-enantiomer presented a higher docking score and
stronger H-bond interaction than the S-enantiomer. Both
approaches proved to be advantageous to select a suitable lipase
for the kinetic resolution. Under the optimized conditions, PFL
exhibited selective acylation of the (R)-enantiomer of (RS)-4 [C ¼
Analysis
All biocatalytic reactions were incubated in an incubator shaker
(Kuhner, Switzerland) at 200 rpm. The products from these
1
13
chemical reactions were analyzed using H and C NMR with
Bruker DPX 400 ( H 400 MHz and C 100 MHz) in MeOD and
1
13
CDCl , respectively. Chemical shi values were expressed in
3
d (ppm) units relative to the tetra methyl silane (TMS) as an
internal standard. The chiral intermediates were analyzed by
HPLC (Shimadzu, Japan) using a Chiralcel OD-H column (4.6
mm ꢃ 250 mm, Daicel Chemical, Japan). Mobile phase con-
sisted of hexane : 2-propanol at 90 : 10 (v/v) with a ow rate of
50.5%, ee
p
¼ 97.2%, ee ¼ 95.4%, E ¼ 182]. The study presents
s
a versatile biocatalyst giving speedy and good result for the
enantiopure synthesis of the metoprolol intermediate. This
simple, cost effective and green approach may likewise be
explored for the synthesis of various enantiopure drugs using
Pseudomonas uorescens lipase as a biocatalyst.
ꢁ1
1.0 mL min . At 254 nm, the substrate and product were
detected. At the abovementioned condition, the retention times
of (R)-4 and (S)-4 were 12.2 and 13.3 min, respectively. All
samples were run under the same conditions as stated
previously.
Enantioselectivity was expressed as E value and calculated by
s
eqn (1), substrate enantiomeric excess (ee ) was calculated by
Experimental
Materials
4-(2-Methoxyethyl)phenol, (RS)-epichlorohydrin, and (R)-
epichlorohydrin were purchased from Sigma-Aldrich Corpora-
tion (St. Louis, MO, USA). Anhydrous Na SO and K CO were
eqn (2), and substrate conversion (C) by eqn (3).
2
4
2
3
obtained from Merck (Germany). HPLC grade solvents were
obtained from Merck (Germany). Candida antarctica immobi-
lized (CALA and CALB), Candida rugosa lipase, Thermomyces
lanuginosus lipase and Pseudomonas uorescens lipase were
purchased from SIGMA (St. Louis, Missouri, USA). TLC plates
were purchased from Merck (Germany). Silica gel (60–120
mesh) for column chromatography was obtained from SRL
E ¼ [ln(1 ꢁ C(1 + ee ))]/[ln(1 ꢁ C(1 ꢁ ee ))]
(1)
(2)
(3)
p
p
ee
s
¼ (S ꢁ R)/(S + R)
C ¼ ee
s s p
/(ee + ee )
where ee and eep represent the enantiomeric excess of the
s
substrate and the product, respectively, and (S) and (R) are the
(India).
35
enantiomers of racemic substrate (RS)-4.
Synthesis of 1-chloro-3-(4-(2-methoxyethyl)phenoxy)propan-2-
ol (RS)-4
General procedure for the enantioselective transesterication
of (RS)-4
4
2
3
-(2-Methoxyethyl)phenol (1) (10 mM, 1 eq.) and K CO (10 mM,
eq.) were mixed and reuxed at 85 C in acetonitrile. Aer
0 min, (RS)-epichlorohydrin [(RS)-2] (10 mM, 1.5 eq.) was
2 3
(
RS)-4 (20 mmol, 1 eq.) was subjected to transesterication with
ꢀ
vinyl acetate (acyl donor) (100 mL) in toluene (900 mL) using
commercial lipase preparations. All lipase preparations were
individually added; asks were tightly capped and placed in the
added dropwise into the reaction mixture. Reaction progression
was constantly monitored through TLC. The reaction was
completed in 24 h. Recovered product (RS)-3 was acetylated
using acetyl chloride (6.48 mM, 1.5 eq.) in dichloromethane : -
water (1 : 1) mixture. The reaction was completed in 12 h.
Recovered product (RS)-4 was analyzed by NMR spectroscopy
ꢀ
shaker at 30 C and 200 rpm. Aer completion, the reaction
(Scheme 2). Using the same method, the compound 1 was
treated with (R)-2 to afford the epoxide intermediate (S)-3. Upon
ring opening, using acetyl chloride, (S)-3 afforded (S)-4.
(
RS)-2-((4-(2-Methoxyethyl)phenoxy)methyl)oxirane
(RS-3),
1
H NMR (400 MHz; MeOD; Me
4
Si): d (ppm): 2.87 (2H, dd), 2.95
(
(
1H, dd), 3.37 (3H, s), 3.6 (2H, t) 4.38 (2H, d), 6.91 (2H, ddd), 7.28
1
3
2H, ddd). C NMR (100 MHz, CDCl ): d (ppm): 35.30, 43.40,
3
5
1
0.10, 58.68, 68.78, 68.86, 114.18, 114.80, 129.58, 130.10,
31.60, 156.99.
(
RS)-1-Chloro-3-(4-(2-methoxyethyl)phenoxy)propan-2-ol (RS-
1
4
), H NMR (400 MHz; MeOD; Me
4
Si) d (ppm): 2.87–2.83 (2H, Scheme 2 Synthesis of (RS)-4.
36572 | RSC Adv., 2017, 7, 36566–36574
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
mixture was extracted with toluene and centrifuged, and the
supernatant was evaporated under vacuum. Samples for HPLC
Conflict of interest
analyses were ltered and added to 2-propanol, and the enan- There are no conicts of interest to declare.
tiomeric excess and ratio were determined.
Acknowledgements
Optimization of process parameters for kinetic resolution
SS would like to thank the Department of Science and Tech-
nology, Govt. of India for providing DST-INSPIRE fellowship.
BPD and VKS acknowledge NIPER, Mohali for providing
fellowship. Authors acknowledge the Department of Pharma-
coinformatics, National Institute of Pharmaceutical Education
and Research, S.A.S. Nagar 160062, Punjab, India for providing
facilities for in silico study.
The inuence of various physicochemical parameters on the
percentage conversion and enantioselectivity of the lipase
catalyzed reaction was investigated. The effect of various
organic solvents on the conversion and enantiomeric excess of
lipase catalyzed resolution of (RS)-4 was investigated. Kinetic
resolution investigation was carried out in 1 mL of the organic
solvent at the optimum values of all other reaction parameters.
The effect of reaction time on enantioselectivity and percentage
conversion was studied by collecting samples at different time
intervals till maximum conversion was achieved. The optimi-
zation of the acyl donor was performed using various acyl
donors (benzyl acetate, ethyl acetate, isopropenyl acetate, and
vinyl acetate) for the kinetic resolution of (RS)-4. While studying
the effect of acyl donors, all other reaction parameters were kept
at their optimum values. The percentage conversion and
enantiomeric excess of the product and residual substrate were
determined by chiral HPLC. The optimum substrate concen-
trations for the resolution of (RS)-4 through lipase catalyzed
transesterication, using vinyl acetate/hexane and vinyl acetate/
toluene, respectively, were determined at different concentra-
tions of substrate (1 to 30 mM). The effect of enzyme concen-
tration on the resolution of (RS)-4 was investigated using
References
1 B. M. Nestl, B. A. Nebel and B. Hauer, Curr. Opin. Chem. Biol.,
2011, 15, 187–193.
2 D. Mu n˜ oz Solano, P. Hoyos, M. J. Hern ´a iz, A. R. Alc ´a ntara
and J. M. S ´a nchez-Montero, Bioresour. Technol., 2012, 115,
196–207.
3 A. Schmid, J. S. Dordick, B. Hauer, A. Kiener, M. Wubbolts
and B. Witholt, Nature, 2001, 409, 258–268.
4 S. Sayin, E. Akoz and M. Yilmaz, Org. Biomol. Chem., 2014, 12,
6634–6642.
5 C. M. Kisukuri and L. H. Andrade, Org. Biomol. Chem., 2015,
13, 10086–10107.
6 B. P. Dwivedee, J. Bhaumik, S. K. Rai, J. K. Laha and
C. Banerjee, Bioresour. Technol., 2017, 239, 464–471.
7 B. Joseph, P. W. Ramteke and G. Thomas, Biotechnol. Adv.,
2008, 26, 457–470.
8 A. Kamal, M. A. Azhar, T. Krishnaji, M. S. Malik and
S. Azeeza, Coord. Chem. Rev., 2008, 252, 569–592.
9 T. Ema, Y. Nakano, D. Yoshida, S. Kamata and T. Sakai, Org.
Biomol. Chem., 2012, 10, 6299.
ꢁ
1
different concentrations of lipases (100 to 600 IU mL ) for the
reaction. Conversion and enantiomeric excess of the product
and the residual substrate were determined by chiral HPLC. The
determination of the optimum temperature for the trans-
esterication of (RS)-4 was carried out at different temperatures
ꢀ
ꢀ
from 4 C to 60 C. Samples were collected from the reaction
mixture at regular time intervals and conversion and enantio-
meric excess was determined using chiral HPLC. All other 10 A. Ta, A. Van Almsick, F. Corelli, M. Crusco, K. E. Laumen,
reaction parameters were kept at their optimum values as ob-
tained in earlier experiments.
M. P. Schneider and M. Botta, J. Org. Chem., 2000, 65, 3659–
3665.
11 D. B. Kitchen, H. Decornez, J. R. Furr and J. Bajorath, Nat.
Rev. Drug Discovery, 2004, 3, 935–949.
Methodology for molecular docking
12 G. Sin, J. M. Woodley and K. V. Gernaey, Biotechnol. Prog.,
2009, 25, 1529–1538.
Glide module of Schr ¨o dinger soware incorporated with the
maestro graphical user interface (GUI) was utilized for the 13 J. Agustian, A. H. Kamaruddin and S. Bhatia, Process
molecular docking study. PFL structure was prepared using Biochem., 2010, 45, 1587–1604.
protein preparation wizard incorporating OPLS_2005 force 14 A. V. R. Rao, M. K. Gurjar and S. V. Joshi, Tetrahedron:
eld. The grid for PFL was generated using the key binding Asymmetry, 1990, 1, 697–698.
residues mentioned in the literature. Moreover, (R)-4 and (S)-4 15 M. K. Gurjar, K. Sadalapure, S. Adhikari, B. V. N. B. S. Sarma,

ligand structures were drawn using ChemDraw tools and saved
in MOL format. These MOL les were opened and prepared
A. Talukdar and M. S. Chorghade, Heterocycles, 1998, 48,
1471–1476.
using LigPrep module of Schr ¨o dinger soware and the possible 16 S. H. Jung, T. L. Pham, H. K. Lim, H. J. Kim, K. H. Kim and
ionization states were generated at neutral pH (7.0 ꢂ 0.5). J. S. Kang, Arch. Pharmacal Res., 2000, 23, 226–229.
Finally, prepared ligands were docked in the active site of the 17 M. Muthukrishnan, D. R. Garud, R. R. Joshi and R. A. Joshi,
lipase using the standard precision (SP) protocol from the glide Tetrahedron, 2007, 63, 1872–1876.
module of Schrodinger soware and 20 poses per ligand were 18 S. Cheng, X. Liu, P. Wang, X. Li, W. He and S. Zhang, Lett.
generated. Finally, the best docked conformations (poses) were
Org. Chem., 2012, 9, 516–519.
selected on the basis of G score.
19 H. U. Shetty and W. L. Nelson, J. Med. Chem., 1988, 31, 55–59.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 36566–36574 | 36573

View Article Online
RSC Advances
Paper
20 H. Sasai, T. Suzuki, N. Itoh and M. Shibasaki, Appl. 37 E. Ł. Chojnacka, M. Staniszewska, M. Bondaryk, J. K. Maurin
Organomet. Chem., 1995, 9, 421–426. and M. Bretner, Chirality, 2016, 28, 347–354.
21 S. Cheng, X. Liu, P. Wang, X. Li, W. He and S. Zhang, Lett. 38 M. D. Romero, L. Calvo, C. Alba and A. Daneshfar, J.
Org. Chem., 2012, 9, 516–519. Biotechnol., 2007, 127, 269–277.
22 T. Roy, S. Barik, M. Kumar, R. I. Kureshy, B. Ganguly, 39 L. E. Janes, R. J. Kazlauskas and S. S. West, J. Org. Chem.,
N. H. Khan, S. H. R. Abdi and H. C. Bajaj, Catal. Sci.
Technol., 2014, 4, 3899–3908.
3 A. Kaler, V. S. Meena, M. Singh, B. Pujala, A. K. Chakraborti
1997, 3, 4560–4561.
40 V. Ferrario, C. Ebert, L. Knapic, D. Fattor, A. Basso, P. Spizzo
and L. Gardossi, Adv. Synth. Catal., 2011, 353, 2466–2480.
41 http://www.uniprot.org/uniprot/P26504.
2
and U. C. Banerjee, Tetrahedron Lett., 2011, 52, 5355–5358.
24 D. Zelaszczyk and K. Kiec, Curr. Med. Chem., 2007, 14, 53–65. 42 J. M. T. R. A. Laskowski, M. W. MacArthur and D. S. Moss, J.
25 S. P. Panchgalle, R. G. Gore, S. P. Chavan and U. R. Kalkote,
Appl. Crystallogr., 1993, 283–291.
Tetrahedron: Asymmetry, 2009, 20, 1767–1770.
43 C. Colovos and T. O. Yeates, Protein Sci., 1993, 2, 1511–1519.
2
2
2
2
3
3
6 J. Y. Zhang, H. M. Liu, H. W. Xu and L. H. Shan, Tetrahedron: 44 J. U. B. D. Eisenberg and R. Lothy, Methods Enzymol., 1997,
Asymmetry, 2008, 19, 512–517. 277, 396–404.
7 N. S. Thakur, J. Bhaumik, B. Sooram, L. Banoth and 45 J. Bowie, R. Luthy and D. Eisenberg, Science, 1991, 253, 164–
U. C. Banerjee, ChemistrySelect, 2016, 1, 871–876. 170.
8 B. P. Dwivedee, S. Ghosh, J. Bhaumik, L. Banoth and 46 M. Wiederstein and M. J. Sippl, Nucleic Acids Res., 2007, 35,
U. C. Banerjee, RSC Adv., 2015, 5, 15850–15860. 407–410.
9 L. Banoth, N. S. Thakur, J. Bhaumik and U. C. Banerjee, 47 K. K. Kim, H. K. Song, D. H. Shin, K. Y. Hwang and S. W. Suh,
Chirality, 2015, 27, 382–391. Structure, 1997, 5, 173–185.
0 S. Hari Krishna and N. G. Karanth, Catal. Rev., 2002, 44, 161– 48 C. Angkawidjaja, D. Ju You, H. Matsumura, K. Kuwahara,
4
940.
1 A. Zaks and A. M. Klibanov, J. Biol. Chem., 1988, 263, 3194–
201.
Y. Koga, K. Takano and S. Kanaya, Fed. Eur. Biochem. Soc.,
Lett., 2007, 581, 5060–5064.
49 A. Ghanem, Tetrahedron, 2007, 63, 1721–1754.
50 D. Guieysse, J. Cort ´e s, S. Puech-Guenot, S. Barbe,
V. Lafaqui `e re, P. Monsan, T. Sim ´e on, I. Andr ´e and
M. Remaud-Sim ´e on, ChemBioChem, 2008, 9, 1308–1317.
3
32 A. M. Klibanov, Nature, 2001, 409, 9–11.
33 M. Paravidino and U. Hanefeld, Green Chem., 2011, 13, 2651–
2657.
3
3
3
4 J. E. Puskas, K. S. Seo and M. Y. Sen, Eur. Polym. J., 2011, 47, 51 C. Llc, S. S. Francisco and A. Guttman, Chem. Genomics,
24–534. 2004, 59–60.
5 A. J. J. Straathof and J. A. Jongejan, Enzyme Microb. Technol., 52 M. H ¨o hne, S. Sch ¨a tzle, H. Jochens, K. Robins and
5
1997, 21, 559–571.
U. T. Bornscheuer, Nat. Chem. Biol., 2010, 6, 807–813.
6 J. Agustian, A. H. Kamaruddin and H. Y. Aboul-Enein,
Chirality, 2017, 1–10.
36574 | RSC Adv., 2017, 7, 36566–36574
This journal is © The Royal Society of Chemistry 2017