Chemoenzymatic Synthesis of Proxyphylline Enantiomers

Article abstract of DOI:10.1021/acs.joc.5b01840

A novel synthetic route for preparation of proxyphylline enantiomers using a kinetic resolution (KR) procedure as the key step is presented. The reactions were catalyzed by immobilized Candida antarctica lipase B in acetonitrile. Three types of reactions were examined: (i) enantioselective transesterification of racemic proxyphylline with vinyl acetate as well as (ii) hydrolysis and (iii) methanolysis of its esters. The influence of reaction conditions on the substrate conversion and enantiomeric purity of the products were investigated. Studies on analytical scale reactions revealed that the titled API enantiomers could be successfully obtained with excellent enantiomeric excess (up to >99% ee). The process was easily conducted on a 5 g scale at 100 g/L. In a preparative-scale reaction, unreacted (S)-(+)-butanoate (97% ee) and (R)-(-)-alcohol (96% ee) were obtained after 2 days in yields of 45% and 46%, respectively. When the reaction time was extended to 6 days, (S)-(+)-butanoate was isolated in >99% ee and acceptable high enantioselectivity (E = 90). Importantly, the KR's products could be conveniently isolated by exploiting varying solubility of the ester/alcohol in acetonitrile at room temperature. In addition, a chiral preference of the CAL-B active site for the R-enantiomer was rationalized by in sillico docking studies.

Full text of DOI:10.1021/acs.joc.5b01840

Subscriber access provided by UNIV LAVAL
Article
Chemoenzymatic synthesis of proxyphylline enantiomers
Pawe# Borowiecki, Daniel Paprocki, Agnieszka Dudzik, and Jan Plenkiewicz
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.5b01840 • Publication Date (Web): 30 Oct 2015
Downloaded from http://pubs.acs.org on November 2, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Chemoenzymatic synthesis of proxyphylline enantiomers
†
,
†
‡
†
Paweł Borowiecki, * Daniel Paprocki, Agnieszka Dudzik, and Jan Plenkiewicz
†
Warsaw University of Technology, Faculty of Chemistry, Institute of Biotechnology, Koszykowa St. 3, 00ꢀ664 Warsaw, Poland.
Jerzy Haber Institute of Catalysis and Surface Chemistry, Polish Academy of Sciences, Niezapominajek St. 8, 30ꢀ239 Cracow, Poland.
Corresponding author: pawel_borowiecki@onet.eu or pborowiecki@ch.pw.edu.pl
‡
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
*
ABSTRACT: A novel synthetic route for preparation of proxyphylline enantiomers using a
kinetic resolution (KR) procedure as the key step is presented. The reactions were catalyzed
by immobilized Candida antarctica lipase B in acetonitrile. Three types of reactions were
examined: (i) enantioselective transesterification of racemic proxyphylline with vinyl acetate
as well as (ii) hydrolysis and (iii) methanolysis of its esters. The influence of reaction
conditions on the substrate conversion and enantiomeric purity of the products were
investigated. Studies on analytical scale reactions revealed that the titled API enantiomers
could be successfully obtained with excellent enantiomeric excess (up to >99% ee). The
process was easily conducted on a 5 g scale at 100 g/L. In a preparativeꢀscale reaction,
unreacted (S)ꢀ(+)ꢀbutanoate (97% ee) and (R)ꢀ(ꢀ)ꢀalcohol (96% ee) were obtained after 2 days
in yields of 45% and 46%, respectively. When the reaction time was extended to 6 days (S)ꢀ
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(+)ꢀbutanoate was isolated in >99% ee and acceptable high enantioselectivity (E=90).
Importantly, the KR’s products could be conveniently isolated by exploiting different
solubility of the ester/alcohol in acetonitrile at room temperature. In addition, a chiral
preference of the CALꢀB active site for the Rꢀenantiomer was rationalized by in sillico
docking studies.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Key words: Lipase Catalysis; Kinetic Resolution; Proxyphylline; lipase B from Candida
antarctica, 1,3ꢀDimethylxanthines, Docking.
1. INTRODUCTION
Chirality is one of the immanent characteristics of nature and plays a key role in metabolic
1
processes as well as in many areas of science and technology. Therefore, synthesis of
enantiomerically pure drugs is of prime concern for medicinal chemistry, especially as single
stereoisomers of active pharmaceutical ingredients (APIs) can act differently in living systems
2
like human beings or pathogens. This stems mainly from particular stereoselective
interactions at biological receptors (enzymes, hormones etc.), which are build up of chiral
molecules, such as amino acids, sugars, and steroids. In order to avert adverse side effects that
may pose danger to a patient’s life application of chiral pharmaceuticals should has been
preceded by thorough biological and clinical evaluations conducted on both racemate and its
3
respective enantiomers. Moreover, administering chiral substances as single enantiomers
ACS Paragon Plus Environment

Page 3 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
diminishes the amount of drug to achieve the expected therapeutic effect. For this reason, vast
4
number of racemate separation techniques including direct preferential crystallization of
enantiomeric mixtures (homoꢀ and heterochiral aggregates), fractional crystallization of
diastereomeric salts (classical resolution), chromatography on chiral phases, and kinetic
resolution mediated by chiral selectors/auxiliaries are essential prerequisites to prepare
compounds in enantiomerically pure forms or at least in a very high enantioenrichment.
Among aforementioned procedures, especially the protocols employing enzymes as catalysts
for kinetic resolution of enantiomers has passed to the canon of synthetic organic chemistry as
one of the most elegant and sustainable strategies.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Hydrolases are most commonly employed enzymes, which has received great attention
5
these days. Among the hydrolytic enzymes, lipases (triacylglycerol ester hydrolases, EC
3.1.1.3) are beyond any doubt the most versatile and practical biocatalysts used in organic
6
chemistry due to their selectivity in action, wide substrate specificity, commercial availability
in both free and immobilized forms, low price, easy handling, high reaction rates, and ability
7
to retain almost full catalytic activity even in nearly anhydrous organic solvents, in nonꢀ
conventional media such as neoteric solvents [i.e. ionic liquids (ILs), supercritical carbon
8
dioxide (scCO ), fluorous solvents (FSs) and liquid polymers (LPs)] as well as in solventꢀ
2
9
free systems. Moreover, reactions catalyzed by lipases can be carried out: at high substrate
concentration (this prevents misusing of organic solvents and thus improve wasteꢀsolvent
management), without usage of expensive cofactors, and in ordinary stirred tank reactors
(what eliminate restructuring requirements of the conventional chemical apparatus). Lipases
can also catalyze transformations of sensitive substrates or complex reactions, for which
standard chemical methods are not accessible. Last but not least, it is well documented that
lipases are also capable to catalyze asymmetric and nonꢀasymmetric carbon–carbon bond
formation what become attractive as very promising alternatives to more traditional synthetic
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
0
routes. Apart from improving important bondꢀforming reactions, soꢀcalled ‘lipase
promiscuity’ can also allow to develop new types of reactivity, which can be exploited to
provide shorter and more efficient reaction pathways.
Beside the above mentioned characteristics, a number of important developments have
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
helped to increase the utility of lipases, including protein engineering, protein purification
1
2
13
methods and immobilization techniques, which improve enzyme stability, activity, and
enantioselectivity as well as allow simple catalyst recovery, thereby influencing on the overall
efficiency of the reactions. In this context, lipases serve as highly valuable and efficient
1
4
catalysts for numerous applications including modification of fats and oils, synthesis of
1
5
16
17
18
19
pharmaceuticals, agrochemicals, natural products, vitamins, cosmetics, fragrances
2
0
21
22
23
and flavors, preparation of modified foods additives, nutraceuticals, detergents,
24
25
26
biodegradable polymers, advanced materials and biodiesel. In addition, the usage of
27
lipases in the processing of renewable raw materials, in disposing of waste oils and
28
29
lubricants, and in biodegradation of toxic xenobiotics is also steadily increasing.
Although lipases are broadly applicable biocatalysts for enantioselective
30
transesterification of structurally different secondary alcohols, direct esterification of
31
32
33
34
carboxylic acids, hydrolysis of esters, amides, and lactams, desymmetrisation of
3
5
36
37
prochiral diols/diesters, mesoꢀ1,2ꢀdiamines, and pentaneꢀ1,5ꢀdiamines as well as
3
8
stereoselective synthesis of secꢀamines and amino acid derivatives, to the best of our
knowledge, hydrolytic enzymes of this type have not been used toward 1,3ꢀdimethylxanthines
until now. Moreover, according to the literature, optical resolution of proxyphylline was
achieved only for its diastereoisomeric camphanates, and carried out by thinꢀlayer
39
chromatography (TLC) and fractional crystallization from methanol or reversedꢀphase liquid
4
0
chromatography mainly for the analytical purposes. In this paper, we present the simple and
efficient chemoenzymatic synthesis of both proxyphylline enantiomers in a preparative scale.
ACS Paragon Plus Environment

Page 5 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
. RESULTS AND DISCUSSION
2.1. Synthesis of racemic proxyphylline (±)ꢀ3 and its esters (±)ꢀ4aꢀc
The principal objective of this work was to develop simple and efficient chemoenzymatic
procedure for the sustainable synthesis of wellꢀknown pharmaceutically valuable compound
with cardiac stimulant, vasodilator, and bronchodilator activities – proxyphylline [(±)ꢀ3, 7ꢀ(2ꢀ
hydroxypropyl)ꢀ1,3ꢀdimethylꢀ3,7ꢀdihydroꢀ1Hꢀpurineꢀ2,6ꢀdione]. In the first step, racemic
starting material (±)ꢀ3 was synthesized by triethylamineꢀmediated regioselective ringꢀopening
of propylene oxide 2 with theophylline (1, 1,3ꢀdimethylꢀ7Hꢀpurineꢀ2,6ꢀdione) in boiling
methanol (Scheme 1). The reaction proceeded smoothly, and after 4 h, and recrystallization
from methanol, the product (±)ꢀ3 was prepared in good 74% yield. Interestingly, despite the
41
fact that proxyphylline has been widely used in clinical practice for over 60 years, the
nuclear magnetic resonance spectra of (±)ꢀ3 has not been reported until now. Next, to obtain
racemic esters (±)ꢀ4aꢀc requested for the preparative enzymatic hydrolysis/methanolysis
attempts and robust analytical studies of high performance liquid chromatography (HPLC)
separation, the aforeꢀprepared alcohol (±)ꢀ3 was treated with acetic anhydride or with the
appropriate acyl chloride in dry dichloromethane in the presence of triethylamine and a
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
catalytic amount of 4ꢀ(N,N)ꢀdimethylaminopyridine (DMAP). Under ‘classical’ esterification
conditions, esters (±)ꢀ4aꢀc could be synthesized in the range of high 83ꢀ95% yields.
Notably, during the course of this study we have observed quite interesting
phenomenon regarding spontaneous polymorphic transformation of one of the racemic esters.
Chromatographic purification and subsequent recrystallization of the acetate (±)ꢀ4a from
methanol yielded three independent crops of crystalline solids of different, but very narrow
melting points range, respectively. At first, we thought that fractions of those products are
structurally different, although after analysis it turned out as all these solids are different
crystalline structures of proxyphilline acetate (±)ꢀ4a. This feature might be considered very
interesting from a pharmaceutical point of view due to its potential influence on drug
processing as well as drug quality, safety, and overall performance since the
pharmacokinetics, delivery, and bioavailability of biologically active substances is
significantly dependent on crystal polymorphism.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. Chemoenzymatic synthesis of both enantiomers of proxyphylline (S)ꢀ(+)ꢀ3 and (R)ꢀ(ꢀ)ꢀ3
Reagents and conditions: (i) propylene oxide 2 (2 equiv), Et N (cat.), MeOH, 4 h at reflux; (iia) Ac O (6.5
3
2
equiv), DMAP (0.2 equiv), dry CH
2
Cl
2
, 12 h at rt; (iib) C
3
H
7
COCl (1.5 equiv) or C
9
H19COCl (1.5 equiv), NEt
3
(1.5 equiv), DMAP (0.1 equiv), dry CH Cl , 12 h at rt; (iic) vinyl butanoate (4.5 equiv), Novozym 435 (5%
2
2
ACS Paragon Plus Environment

Page 7 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
w/w), MTBE, 12 h at rt, 1000 rpm; (iii) lipaseꢀcatalyzed transesterification: vinyl ester, lipase (20% w/w),
solvent, 25 °C, 500 rpm; (iv) lipaseꢀcatalyzed hydrolysis/methanolysis: H
solvent, 25 °C, 500 rpm.
2
O or TrisꢀHCl Buffer/MeOH, lipase,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Moreover, surprisingly if the alcohol (±)ꢀ3 was treated with vinyl butanoate in the
presence of Novozym 435 suspended in methyl tertꢀbutyl ether (MTBE), and vigorously
stirred (1000 rpm) at room temperature overnight, Oꢀbutyryl ester (±)ꢀ4b was afforded in
excellent yield (94%) under mild reaction conditions. Although substrate (±)ꢀ3 could not be
fully dissolved in neat MTBE, we have found that in the presence of vinyl butyrate solubility
of (±)ꢀ3 was considerably enhanced, and thus the esterification process proceeded
successfully, providing even better results when compared to ‘classical’ approach. It is
important to mention that changing the synthetic strategy toward preparation of (±)ꢀ4b not
only improved the reaction yield, but also helped to ‘greener’ synthesis pathway by
eliminating both usage of methylene chloride as the solvent and extraction procedure during
the workꢀup.
2.2. KR of (±)ꢀ3 using lipaseꢀcatalyzed transesterification – lipase selection
The development and optimization of a versatile enzymatic methodology is not a simple task,
and often lead through many laborious stages including the appropriate lipase selection, effect
of enzyme loading, pretreatment, and its conditioning as well as the adequate medium
adjustment, proper choice of the acyl group type of donor/acceptor reagent, mutual substrateꢀ
toꢀbiocatalystꢀtoꢀacyl group donor/acceptor molar/weight ratio, reaction time and temperature
etc. Rational planning of enzymatic reaction conditions is critical to meet sustainability
criteria.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The water removal process is tedious and expensive at industrial scale. Hence, it is
recommended to use synthetic approaches that allow the reactions to be carried out in pure
easyꢀtoꢀevaporate organic solvent or at least in an aqueousꢀorganic twoꢀphase system with the
lowest possible water content. The realization of lipaseꢀcatalyzed (trans)esterification
processes by means of an activated esters (e.g. enol esters, trifluoroethyl butyrate, Sꢀethyl
thiooctanoate) seems to be the most suitable solution to overcome this problem. Therefore, at
the outset of our study, the ability of lipases to catalyze the transesterification of
proxyphylline (±)ꢀ3 with 11.5ꢀfold molar excess of vinyl acetate as the acetyl transfer reagent
in an anhydrous environment was investigated (Table 1). The enzymatic KR of the racemate
(±)ꢀ3 using excess of vinyl acetate as the acetyl transfer reagent and standard protocol was
examined in chloroform since surprisingly only this solvent form homogenous solution of the
appropriate substrate concentration at 25 ºC. None of the other tested solvents including less
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
polar than CHCl [hexane, pentane, toluene, cyclohexane, 2ꢀmethylꢀ2ꢀbutanol (tertꢀamyl
alcohol)] or more polar [diisopropyl ether, dichloromethane, MTBE, diethyl ether, vinyl
acetate, tetrahydrofurane, acetone, acetonitrile, 1,4ꢀdioxane] were suitable for enzymatic
acetylation of (±)ꢀ3. Most of the reactions carried out in these solvents resulted in no substrate
conversion mainly because of low solubility of (±)ꢀ3 in those media. Enzyme screening was
attempted using a representative set of 8 lipases isolated from various microorganisms
including: Candida antarctica B (Novozym 435, Chirazyme Lꢀ2, c.ꢀf., C2, Lyo., Chirazyme
Lꢀ2, c.ꢀf., C3, Lyo.), Burkholderia cepacia (Amano PS, Amano PSꢀIM), Pseudomonas
fluorescens (Amano AK), Thermomyces lanuginosus (Lipozyme TL IM), and Rhizomucor
miehei (Lipozyme RM IM).
Since the substrate (±)ꢀ3 and product (±)ꢀ4a enantiomers do not separate on available
Chiralcel ODꢀH column, determination of the enantiomeric excess values directly from the
crude reaction mixture over the course of the process was impossible. Therefore, the progress
ACS Paragon Plus Environment

Page 9 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
of the enzymatic reactions was at first traced by means of nonꢀchiral gas chromatography
(GC) in respect to the appropriate calibration curve for quantitative analyses. Next, the
reactions were stopped as close to 50% substrate conversion as possible, subsequently
isolation by column chromatography was performed, and only then the eeꢀvalues were
determined by chiral HPLC. In addition, for determination of enantiopurity of title API an
extra derivatization procedure of the alcohol into corresponding acetate was required. For that
reason, chemical protection of the remaining free hydroxyl functional group was performed
by addition of acetic anhydride over catalytic amount of DMAP in dry dichloromethane at
room temperature, thus obtaining optically active acetate (S)ꢀ(+)ꢀ4a or (R)ꢀ(ꢀ)ꢀ4a in
quantitative yields from (S)ꢀ(+)ꢀ3 or (R)ꢀ(ꢀ)ꢀ3, respectively.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Lipase screening for the enantioselective Oꢀacetylation of (±)ꢀ3 under kinetically controlled
conditions in CHCl
3
a
b
c
d
Entry
Lipase preparation
t [d]
2
2
1
7
7
1
1
1
Conv. [%]
ee
s
[%]/(config.) ee
59/(S)
35/(S)
p
[%]/(config.)
52/(R)
59/(R)
E
e
1
2
3
4
5
6
7
8
Novozym 435
53
37
6
5
4
N.D.
9
N.D.
2
e
Chirazyme Lꢀ2, Cꢀ2
Chirazyme Lꢀ2, Cꢀ3
Amano PS
Amano PSꢀIM
Amano AK
Lipozyme TL IM
Lipozyme RM IM
e
26
18/(S)
52/(R)
f
g
g
g
g
g
5
N.D.
N.D.
e
19
18/(S)
76/(R)
g
g
0
30
N.D.
N.D.
e
13/(S)
30/(R)
g
g
0
N.D.
N.D.
N.D.
a
Conditions: (±)ꢀ3 100 mg, lipase 20 mg, CHCl
3
1 mL, vinyl acetate 415 mg, 0.5 mL (11.5 equiv), 25 ºC, 500 rpm (magnetic stirrer).
b
Determined by chiral HPLC analysis of corresponding alcohol obtained after derivatization of alcohol (S)ꢀ(+)ꢀ3 into the corresponding
2
acetate (S)ꢀ(+)ꢀ4a, which was performed by addition of DMAP and Ac O (5 equiv) since direct analysis of (S)ꢀ(+)ꢀ3 with Chiralcel ODꢀH
column was unsatisfactory.
c
Determined by chiral HPLC analysis by using a Chiralcel ODꢀH column.
d
e
42
Calculated according to Chen et al., using the equation: E = {ln[(1 – conv.)(1 – ee
Based on GC, for confirmation the % conversion was calculated from the enantiomeric excess of the unreacted alcohol (ee
ee ) according to the formula conv. = ee /(ee + ee ).
s s
)]}/{ln[(1 – conv.)(1 + ee )]}.
s
) and the product
(
p
s
s
p
f
Based on gas chromatography (GC).
Not determined.
g
Among the investigated panel of commerciallyꢀavailable preparations, immobilized
lipases of a fungal origin such as Novozym 435, Chirazyme Lꢀ2, Cꢀ2, and Chirazyme Lꢀ2, Cꢀ
were established as the most potent biocatalysts for enantioselective transesterification.
However, the lipaseꢀcatalyzed acetylation of (±)ꢀ3 in CHCl was not successful as the reaction
3
3
rates and enantioselectivities (E=2ꢀ9) were far from ideal. Moreover, the enantiomeric
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
excesses of the desired products [(S)ꢀ(+)ꢀ3 and (R)ꢀ(ꢀ)ꢀ4a] were low (18ꢀ76% ee), and in
consequence evaluation of an enzymatic transesterification of proxyphylline (±)ꢀ3 approach
was discontinued. The rest of the lipases (Amano PS, Amano AK, Lipozyme RM IM) proved
to be catalytically inactive even after 7 days of proceeding the reactions, while Amano PSꢀIM
and Lipozyme TL IM preparations showed poor selectivity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2.3. KR of (±)ꢀ4a using lipaseꢀcatalyzed alcoholysis – lipase selection
Since acceptable reaction Eꢀfactors were impossible to comply with Oꢀacetylation procedure
(Table 1), we have opted to use the reverse process, that is lipaseꢀcatalyzed alcoholysis of the
acetate (±)ꢀ4a (Table 2). The enzymatic alcoholysis was conducted with methanol as an
acetylꢀacceptor reagent used in a relatively high ratio to the substrate (±)ꢀ4a (10 equiv) in
spite of wellꢀknown fact, that most enzymes are inactivated by lowꢀmolecularꢀweight
alcohols, particularly by methanol and ethanol. This stems from propensity of
polar/hydrophilic solvents to strip off the essential layer of water molecules from the protein
(the structural water), thereby disrupting its native structure responsible for catalysis, and
eventually denaturing the enzyme. However, this decision was made intentionally since we
have observed positive impact of methanol on the substrate (±)ꢀ4a solubility. Initially, the
same panel of lipases (Table 1) was screened as catalysts of asymmetric methanolysis of
racemic proxyphylline acetate (±)ꢀ4a at 25 ºC. A detailed survey on selection of the optimal
lipase catalyst was carried out using acetonitrile as the solvent. Aliquots of the samples were
withdrawn from the reaction mixture at 1 day intervals and the samples were analyzed by GC.
Next, the chromatographic purification, and appropriate alcohol derivatization were
performed, and the eeꢀvalues were measured by means of chiral HPLC. However, preliminary
studies gave disappointing results. In most of the studied enzyme systems no reaction was
ACS Paragon Plus Environment

Page 11 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
observed (Table 2, entries 4ꢀ8) or the lipaseꢀmediated KR proceeded sluggishly achieving
barely 2% of the conversion after 24 h (Table 2, entry 3). Interestingly, Novozym 435 and
Chirazyme Lꢀ2, Cꢀ2 catalyzed methanolysis of the acetate (±)ꢀ4a, showing promising
selective enrichment of the products (Table 2, entry 1 and 2). However, the enantiomeric
excess values of the liberated alcohol (R)ꢀ(ꢀ)ꢀ3 (72ꢀ73% ee) as well as the reaction
enantioselectivities (E=16ꢀ22) in both cases were poorꢀtoꢀmoderate, and the requested 50%
conversions were obtained after as much as 8 days. Although slightly better results in terms of
enantioselectivity were obtained in alcoholysis protocol compared to transesterification, still
the process proceeded within an unreasonable timeꢀscale. In turn, as the results presented in
Table 2 clearly indicate that (Novozym 435)ꢀcatalyzed methanolysis of (±)ꢀ4a yielded
unreacted acetate (S)ꢀ(+)ꢀ4a with higher optical purity (95% ee) than Chirazyme Lꢀ2, Cꢀ2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(81% ee), thereby we decided to assess immobilized CALꢀB as a potential catalyst for further
studies.
Table 2. Lipase screening for the enantioselective methanolysis of (±)ꢀ4a under kinetically controlled
conditions in CH
Lipase preparation
Novozym 435
3
CN a
b
c
d
Entry
t [d]
8
8
1
1
1
1
1
1
Conv. [%] ee
s
[%]/(config.) ee
95/(S)
p
[%]/(config.)
72/(R)
E
e
1
2
3
4
5
6
7
8
57
22
16
e
Chirazyme Lꢀ2, Cꢀ2
Chirazyme Lꢀ2, Cꢀ3
Amano PS
Amano PSꢀIM
Amano AK
Lipozyme TL IM
Lipozyme RM IM
53
81/(S)
73/(R)
f
g
g
g
g
g
g
g
g
2
0
0
0
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
f
g
g
N.D.
f
g
g
N.D.
f
g
g
N.D.
f
g
g
0
N.D.
f
g
g
0
N.D.
a
b
c
Conditions: (±)ꢀ4a 100 mg, lipase 20 mg, CH
3
CN 1 mL, methanol 114 mg, 0.15 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
Determined by chiral HPLC analysis by using a Chiralcel ODꢀH column.
Determined by chiral HPLC analysis of corresponding alcohol obtained after derivatization of alcohol (R)ꢀ(ꢀ)ꢀ3 into the corresponding
2
acetate (R)ꢀ(ꢀ)ꢀ4a, which was performed by addition of DMAP and Ac O (5 equiv) since direct analysis of (R)ꢀ(ꢀ)ꢀ3 with Chiralcel ODꢀH
column was unsatisfactory.
d
e
42
Calculated according to Chen et al., using the equation: E = {ln[(1 – conv.)(1 – ee
Based on GC, for confirmation the % conversion was calculated from the enantiomeric excess of the unreacted ester (ee
ee ) according to the formula conv. = ee /(ee + ee ).
Based on gas chromatography (GC).
Not determined.
s s
)]}/{ln[(1 – conv.)(1 + ee )]}.
s
) and the product
(
p
s
s
p
f
g
2.4. KR of (±)ꢀ4a using lipaseꢀcatalyzed alcoholysis – the solvent effect
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A further important criteria when setting up an enzymatic catalysis with lipases for a given
process is the choice of the reaction medium. Selection of an appropriate solvent system (soꢀ
called ‘medium engineering’) for a particular enzymeꢀsubstrate pair is often very laborious
task, but if properly performed it can significantly improve activity, stability, and selectivity
of the biocatalyst, and in addition, it may even influence on stereochemical preference of the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
43
biocatalyst.
This stage of our study has been aimed to optimize the (Novozym 435)ꢀcatalyzed KR
in alcoholysis of (±)ꢀ4a with methanol mainly by assessment of the influence of solvent on
the conversion rate and stereochemical outcome. For this purpose, an alternative ‘green
solvents’ (CH
3
CN, 1,4ꢀdioxane, and acetone), a traditional organic solvents with low ecoꢀ
toxicity (PhCH , THF), and chlorinated solvents of a bad reputation (CHCl , CH Cl ) were
3
3
2
2
examined as the reaction media. In other commonlyꢀused solvents, wellꢀknown for their
compatibility with lipases, such as MTBE, diisopropyl ether (DIPE), nꢀhexane and 2ꢀ
methylbutanꢀ2ꢀol (tertꢀamyl alcohol, TAA), most of the acetate (±)ꢀ4a remained in suspension
due to poor solubility, and therefore for obvious reasons the reactions were not performed. A
screen of media showed that this method also failed to give high enantioselectivities as the
best results were obtained still with Novozym 435 suspended in CH CN. However, a few
3
observations are noteworthy to comment. Unlikely to common rule that in lipaseꢀcatalyzed
reactions hydrophobic solvents provide higher reaction rate than hydrophilic one, in our
investigations we observed the opposite tendency. This is a surprising result because
Novozym 435 was found to retain its catalytic potency in the polar solvents (Table 3, entries
1ꢀ3), showing very similar enantioselectivities (E=20ꢀ23), while in most of the employed less
polar solvents it was inactive toward (±)ꢀ4a (Table 3, entries 4ꢀ6). However, the last result
(Table 3, entry 4) indicates that activity of the lipases might be affected by the solvent
without correlation to the logP, and therefore other parameters such as solvent–enzyme–
ACS Paragon Plus Environment

Page 13 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
substrate interactions cannot be ignored, i.e. the accessibility of the substrate to the enzyme
active site might be induced by diffusional limitations, changes in protein flexibility or some
allosteric regulation controlled by the solvent. Although in the reaction conducted in toluene,
unreacted acetate (S)ꢀ(+)ꢀ4a was obtained with high 92% ee, however both the degree of
conversion and the enantioselectivity were very low indeed. Notably, when comparing both
types of the studied lipaseꢀcatalyzed reactions, one can see some characteristic behaviour. It
turned out that (Novozym 435)ꢀcatalyzed transesterification of (±)ꢀ3 in chloroform proceeded
quite successfully, nevertheless in the case of methanolysis of (±)ꢀ4a, activity of this enzyme
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3 2 2
was completely lost in waterꢀimmiscible chlorinated coꢀsolvents (CHCl or CH Cl ), as the
process did not start even after a day. Finally, acetonitrile was found to be the solvent of
choice for lipaseꢀmediated methanolysis of (±)ꢀ4a, where other solvents offered decreased
rates and worse enantioselectivities.
Table 3. Solvent screening for the enantioselective methanolysis of (±)ꢀ4a in the presence of Novozym 435
under kinetically controlled conditions
a
b
c
d
e
Entry
Solvent (logP)
1,4ꢀDioxane (ꢀ0.31)
CH CN (0.17)
t [d]
10
8
10
1
1
1
14
Conv. [%] ee
s
[%]/(config.) ee
97/(S)
95/(S)
p
[%]/(config.)
66/(R)
73/(R)
E
f
1
2
3
4
5
6
7
60f
20
23
21
N.D.
N.D.
N.D.
10
3
57f
54
Acetone (0.20)
THF (0.40)
89/(S)
75/(R)
g
h
h
h
h
h
0
N.D.
N.D.
N.D.
N.D.
g
h
h
CHCl
3
(1.67)
0
N.D.
g
h
h
CH Cl (1.01)
0
N.D.
2
2
f
PhCH
3
(2.52)
63
92/(S)
54/(R)
a
Conditions: (±)ꢀ4a 100 mg, lipase 20 mg, solvent 1 mL, methanol 114 mg, 0.15 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
b
Logarithm of the partition coefficient of a given solvent in nꢀoctanolꢀwater system according to ChemBioDraw Ultra 13.0 software
indications.
c
Determined by chiral HPLC analysis by using a Chiralcel ODꢀH column.
Determined by chiral HPLC analysis of corresponding alcohol obtained after derivatization of alcohol (R)ꢀ(ꢀ)ꢀ3 into the corresponding
d
2
acetate (R)ꢀ(ꢀ)ꢀ4a, which was performed by addition of DMAP and Ac O (5 equiv) since direct analysis of (R)ꢀ(ꢀ)ꢀ3 with Chiralcel ODꢀH
column was unsatisfactory.
e
f
42
Calculated according to Chen et al., using the equation: E = {ln[(1 – conv.)(1 – ee
Based on GC, for confirmation the % conversion was calculated from the enantiomeric excess of the unreacted ester (ee
ee ) according to the formula conv. = ee /(ee + ee ).
Based on gas chromatography (GC).
Not determined.
s s
)]}/{ln[(1 – conv.)(1 + ee )]}.
s
) and the product
(
p
s
s
p
g
h
2.5. Lipaseꢀcatalyzed KR of (±)ꢀ4aꢀb – alcoholysis vs hydrolysis
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Adverse results of (±)ꢀ4a kinetic resolution forced us to explore an alternative asymmetric
approach towards synthesis of optically active proxyphylline. In living organisms lipases
catalyze hydrolysis of higher fatty acid esters of glycerol, thus fulfill an essential function in
metabolism of lipids (e.g. fats and oils) and lipoproteins. As lipases are natural ‘workhorses’
of such chemical transformations in the cells, hence it seems to be reasonable to increase the
lipophilicity of proxyphylline (±)ꢀ3 by replacing acetate moiety with longer fatty acyl chains.
We envisaged that using butyrate (±)ꢀ4b or decanoate (±)ꢀ4c derivatives, which closer mimic
natural substrates, the reactions might occur with enhanced enantioselectivity. Unfortunately,
as it soon turned out, monitoring of enzymatic reaction with proxyphylline decanoate (±)ꢀ4c
was impossible since GC and HPLC analyses have both failed due to nonꢀvolatility of (±)ꢀ4c
in the recommended temperature ranges limits (<260 °C) for the available HPꢀ50+ semipolar
column and nonꢀresolvability of (±)ꢀ4c enantiomers on a Chiralcel ODꢀH column,
respectively. These led us to cease further experiments with (±)ꢀ4c and hence, we focused on
the alcoholysis of acetate (±)ꢀ4a and butanoate (±)ꢀ4b. Additionally, the enzymatic KR was
examined in waterꢀmiscible (monophasic aqueousꢀorganic) solvent systems composed of
acetonitrile and water or commercially available 1 M TrisꢀHCl Buffer, respectively (Table 4).
The influence of acyl group nature on the course of the ester methanolysis/hydrolysis reaction
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
was examined using 10 equiv of the appropriate acyl acceptor reagent (methanol, H
water which is present in buffer solution) under catalysis of Novozym 435 suspended in
CH CN as coꢀsolvent, and stirred at 25 °C using a magnetic stirrer (500 rpm). The courses of
2
O or
3
all the enzymatic reactions were monitored by gas chromatography (GC), arrested
deliberately after 8 days, all of the lipaseꢀcatalytically resolved products were
chromatographically purified, the hydroxyl group of (R)ꢀ(ꢀ)ꢀ3 properly derivatized, and only
then their final evolution were subsequently followed by chiral HPLC to compare efficiency
of particular methodology. At first glance, the size of acyl group in the examined esters (±)ꢀ
ACS Paragon Plus Environment

Page 15 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
4aꢀb strongly influence the reaction enantioselectivity. The results presented in Table 4
indicate that KR of butanoate (±)ꢀ4b proceeded with a considerable improvement in the
reaction enantioselectivity (E up to 260), furnishing in case of both types of hydrolytic
attempts (Table 4, entry 5 and 6) very good results in terms of enantiomeric excess of
unreacted ester (S)ꢀ(+)ꢀ4b (96ꢀ98% ee) as well as the released alcohol (R)ꢀ(ꢀ)ꢀ3 (94ꢀ97% ee).
In turn, although (Novozym 435)ꢀcatalyzed methanolysis of (±)ꢀ4b was significantly less
selective (E=34), it turned out that enantioenrichment was remarkably high as the ester (S)ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(+)ꢀ4b was obtained with an excellent enantiomeric excess (>99% ee). Moreover, a
comprehensive understanding of the kinetics of lipaseꢀcatalyzed reactions prompt us that the
potential of methanolysis of (±)ꢀ4b has not been fully evaluated, and an attractive optical
purity could be obtained for the opposite enantiomer (R)ꢀ(ꢀ)ꢀ3 as well. This is fully
42
understandable especially since one realizes that Chen’s logarithmic equation demands very
similar levels of conversion, so that the results of Eꢀfactor may be reliably compared. As a
single experiment cannot be decisive for full estimation of enzymatic performance as well as
its potential applicability, it was therefore desirable to compare two types of lipaseꢀmediated
transformations – hydrolysis of (±)ꢀ4b in TrisꢀHCl Buffer as the most enantioselective
method, and methanolysis of (±)ꢀ4b as the one, which led to the highest enantiomeric excess
of unreacted ester (S)ꢀ(+)ꢀ4b.
Table 4. Examination of (Novozym 435)ꢀcatalyzed KR conditions for the enantioselective methanolysis
and hydrolysis of (±)ꢀ4a and (±)ꢀ4b suspended in CH CN after 8 days, respectively
3
a
b
c
d
Entry
Compound
(±)ꢀ4a
Nucleophile
Conv. [%]
ee [%]/(config.) eep [%]/(config.)
E
s
e
1
2
3
4
5
6
MeOH
57
41
62
57
51
50
95/(S)
64/(S)
52/(S)
>99/(S)
98/(S)
96/(S)
73/(R)
94/(R)
84/(R)
74/(R)
94/(R)
97/(R)
23
63
8
34
149
260
f
2
H O
TrisꢀHCl Buffer
g
h
(±)ꢀ4b
MeOH
i
H O
2
j
TrisꢀHCl Buffer
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
s
Based on GC, for confirmation the % conversion was calculated from the enantiomeric excess of the unreacted ester (ee ) and the product
(
ee
p s s p
) according to the formula conv. = ee /(ee + ee ).
b
Determined by chiral HPLC analysis by using a Chiralcel ODꢀH column.
Determined by chiral HPLC analysis of corresponding alcohol obtained after derivatization of alcohol (R)ꢀ(ꢀ)ꢀ3 into the corresponding
c
2
acetate (R)ꢀ(ꢀ)ꢀ4a, which was performed by addition of DMAP and Ac O (5 equiv) since direct analysis of (R)ꢀ(ꢀ)ꢀ3 with Chiralcel ODꢀH
column was unsatisfactory.
d
e
f
42
Calculated according to Chen et al., using the equation: E = {ln[(1 – conv.)(1 – ee
Conditions: (±)ꢀ4a 100 mg, lipase 20 mg, CH CN 1 mL, MeOH 114 mg, 0.15 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
Conditions: (±)ꢀ4a 100 mg, lipase 20 mg, CH CN 1 mL, H O 64 mg, 0.64 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
Conditions: (±)ꢀ4a 100 mg, lipase 20 mg, CH CN 1 mL, 1 M TrisꢀHCl Buffer, pH 7.5 64 mg, 0.64 mL (10 equiv), 25 ºC, 500 rpm
s s
)]}/{ln[(1 – conv.)(1 + ee )]}.
3
3
2
g
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(magnetic stirrer).
h
Conditions: (±)ꢀ4b 100 mg, lipase 20 mg, CH
3
CN 1 mL, MeOH 104 mg, 0.13 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
CN 1 mL, H O 58 mg, 0.58 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
Conditions: (±)ꢀ4b 100 mg, lipase 20 mg, CH CN 1 mL, 1 M TrisꢀHCl Buffer, pH 7.5 58 mg, 0.58 mL (10 equiv), 25 ºC, 500 rpm
(magnetic stirrer).
i
Conditions: (±)ꢀ4b 100 mg, lipase 20 mg, CH
3
2
j
3
Stimulated by the above results, in the next part of our study, we have examined time
dependency on the rate and enantioselectivity of butyrate (±)ꢀ4b methanolysis/hydrolysis
under kinetic resolution conditions described in the previous section. For that reason, the
conversion degrees of enzymatic reactions were monitored by chiral HPLC over a wide range
of reaction times and at various intervals. From the data collected in Table 5 it become
obvious that the methanolysis reaction proceeded more enantioselectively (up to E=584) to
give both unreacted substrate (S)ꢀ(+)ꢀ4b and product (R)ꢀ(ꢀ)ꢀ3 in enantiomerically pure form
(
>99% ee) depending on the moment of process arresting. Another clear advantage of lipaseꢀ
catalyzed methanolysis over hydrolysis method is easiness of methanol removal by
evaporation under reduced pressure or by the extraction, so there is no need to use special
water removal procedures.
Table 5. Examination of (Novozym 435)ꢀcatalyzed KR time courses for the enantioselective methanolysis
3
and hydrolysis of (±)ꢀ4b suspended in CH CN, respectively
a
b
c
p
d
Entry
Nucleophile
MeOH
t [h] Conv. [%]
ee
s
[%]/(config.)
84/(S)
ee
[%]/(config.)
>99/(R)
>99/(R)
96/(R)
92/(R)
90/(R)
88/(R)
98/(R)
98/(R)
98/(R)
E
e
1
2
3
4
5
6
7
8
9
4
8
46
47
48
52
52
53
30
39
45
46
46
47
532
584
152
126
99
88/(S)
24
48
72
96
24
48
72
96
120
144
90/(S)
>99/(S)
>99/(S)
>99/(S)
43/(S)
63/(S)
80/(S)
84/(S)
85/(S)
87/(S)
82
f
TrisꢀHCl Buffer
151
190
244
264
270
283
10
11
12
98/(R)
98/(R)
98/(R)
ACS Paragon Plus Environment

Page 17 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
a
s
Based on GC, for confirmation the % conversion was calculated from the enantiomeric excess of the unreacted ester (ee ) and the
product (ee
p
) according to the formula conv. = ee
Determined by chiral HPLC analysis by using a Chiralcel ODꢀH column.
Determined by chiral HPLC analysis of corresponding alcohol obtained after derivatization of alcohol (R)ꢀ(ꢀ)ꢀ3 into the
s s p
/(ee + ee ).
b
c
2
corresponding acetate (R)ꢀ(ꢀ)ꢀ4a, which was performed by addition of DMAP and Ac O (5 equiv) since direct analysis of (R)ꢀ(ꢀ)ꢀ3
with Chiralcel ODꢀH column was unsatisfactory.
d
e
f
42
Calculated according to Chen et al., using the equation: E = {ln[(1 – conv.)(1 – ee
Conditions: (±)ꢀ4b 100 mg, lipase 20 mg, CH CN 1 mL, MeOH 104 mg, 0.13 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
CN 1 mL, 1 M TrisꢀHCl Buffer, pH 7.5 58 mg, 0.58 mL (10 equiv), 25 ºC, 500 rpm
s s
)]}/{ln[(1 – conv.)(1 + ee )]}.
3
Conditions: (±)ꢀ4b 100 mg, lipase 20 mg, CH
3
(
magnetic stirrer).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analysis of data from kinetic experiments also highlighted that the reaction rates differ
substantially in each case. It was noticed that the rate of (Novozym 435)ꢀmediated
methanolysis of (±)ꢀ4b was profoundly enhanced compared to hydrolysis, as it reaches ca.
50% conversion after approximately 24 hours, while hydrolysis catalyzed by the same lipase
could not achieve this level even after elongation the reaction time over 6 days. Comparing
both enantioselective KR courses one can see very characteristic tendency. Although
hydrolytic attempt excludes the possibility of achieving enantiopure alcohol (R)ꢀ(ꢀ)ꢀ3 during
the examined timeꢀperiod, the liberated alcohol has always been afforded with excellent
enantiomeric excesses of about 98% (Table 5, entries 7ꢀ12). In turn, an enantioenrichment
increment of the slower reacting enantiomer (S)ꢀ(+)ꢀ4b was negligible, and it finally reached
barely 87% ee. In the case of enantioselective hydrolysis of (±)ꢀ4b it was observed that the
conversion increased to a certain level (45ꢀ47%) after which there was no significant change
(Table 5, entries 9ꢀ12). Note also that it was crucial to follow the kinetics of both lipaseꢀ
catalyzed reactions very carefully, as it allowed us to determine the appropriate moment to
terminate biotransformation of (±)ꢀ4b in order to obtain products of desired enantiomeric
purity. Undoubtedly, the experiments revealed classic kinetics of the process that means when
lipaseꢀmediated methanolysis process does not exceed 47% conversion it is in favor of
enantiopure alcohol (R)ꢀ(ꢀ)ꢀ3 formation (Table 5, entry 1 and 2), while when reaching the
conversion above 52% (Table 5, entries 4ꢀ6) it is beneficial for obtaining unreacted ester (S)ꢀ
(
+)ꢀ4b as a single enantiomer (>99%). It is noteworthy that maximum enantiomeric excess of
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the product (92%) and substrate (>99%) from the same attempt was obtained when the
reaction was arrested at conversion of ca. 52% after 2 days of proceeding (Table 5, entry 4).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2.6. Preparativeꢀscale lipaseꢀcatalyzed KR of (±)ꢀ4b via methanolysis
Gratifyingly, an inꢀdepth analysis performed on biotransformation reaction conditions
allowed us to obtain strategy fully compatible for the employed xanthineꢀlike derivative (±)ꢀ
4b. In the course of our investigation, we found that enzymeꢀpromoted methanolysis
unequivocally turned to be the method of choice, as it led to accomplishments of both
resolved enantiomers of proxyphylline in high enantiomeric purity, depending on the time and
conversion rate. As we developed reliable and robust catalytic system at analytical scale, next
we have turned to lipaseꢀcatalyzed methanolysis of (±)ꢀ4b under the conditions of kinetic
resolution at preparative scale. The commercial CALꢀB in immobilized form (Novozym 435)
was employed as the catalyst. Crucial advantage of the aboveꢀselected lipase preparation is its
easy handling and simple isolation of the resolved products. The upꢀscaling investigations
were carried out with 1, 2 and 5 grams of butanoate (±)ꢀ4b respectively using the same
optimized reaction procedure as in the analytical scale studies, but obviously with
proportionally enlarged reaction stoichiometry (Table 6). Unfortunately, we have found that
when performing enzymatic processes in a bigger scale, different reaction times were required
to obtain both enantiomers of proxyphylline with comparable % ee. The reason of such
behavior is unclear for us because we have not met any drawbacks with upꢀscaling until now.
This suggest that such misleading is committed mostly at the stage of the process monitoring.
As we mentioned earlier, during this synthesis we could not optimized HPLC analytic method
ACS Paragon Plus Environment

Page 19 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
for reliable tracing of the reaction progress directly from the crude mixture. In this instance,
the differences in experimental attempts may be due to the used GC analysis, which is
definitely saddled with some error and might appear to be the main source of inaccuracies.
Nevertheless, from the results summarized in Table 6, it is evident that preparative scale
methanolysis of (±)ꢀ4b provided access to both enantiomers of proxyphilline reaching a very
high (94ꢀ97% ee) to excellent enantiomeric enrichment (98ꢀ100% ee).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 6. Gramꢀ and multigramꢀscale enantioselective methanolysis of (±)ꢀ4b catalyzed by Novozym 435
a
b
e
f
D
c
e
f
d
Entry
Scale
t [d]
1
Conv. [%]
ee
s
[%]/Yield (%)/[α]
ee
p
[%]/Yield (%)/[α]
D
E
g
1
2
3
4
5
6
1 g
47
51
48
51
50
53
87/95/+42.96 (c 1.35)
98/89/ꢀ54.00 (c 1. 00)
94/90/ꢀ50.00 (c 1. 00)
97/93/ꢀ52.38 (c 1. 05)
95/89/ꢀ51.43 (c 1. 05)
96/91/ꢀ48.00 (c 1.05)
89/94/ꢀ47.14 (c 1.00)
283
170
192
206
207
90
3
>99/96/+50.36 (c 1.40)
88/93/+36.00 (c 1.50)
>99/94/+45.29 (c 1.70)
97/90/+41.35 (c 1.85)
>99/93/+48.89 (c 1.80)
h
2 g
1
3
i
5 g
2
6
a
Based on GC, for confirmation the % conversion was calculated from the enantiomeric excess of the unreacted ester (ee
ee ) according to the formula conv. = ee /(ee + ee ).
Determined by chiral HPLC analysis by using a Chiralcel ODꢀH column.
s
) and the product
(
p
s
s
p
b
c
Determined by chiral HPLC analysis of corresponding alcohol obtained after derivatization of alcohol (R)ꢀ(ꢀ)ꢀ3 into the corresponding
2
acetate (R)ꢀ(ꢀ)ꢀ4a, which was performed by addition of DMAP and Ac O (5 equiv) since direct analysis of (R)ꢀ(ꢀ)ꢀ3 with Chiralcel ODꢀH
column was unsatisfactory.
d
42
Calculated according to Chen et al., using the equation: E = {ln[(1 – conv.)(1 – ee
s s
)]}/{ln[(1 – conv.)(1 + ee )]}.
e
This value indicates isolated yield after purification step, and is calculated on the basis of the theoretical number of moles arising from
conversion rate, relative to theoretical amount, i.e., when 50% conversion is reached, up to the half of the acetate could be obtained.
f
Specific rotation, c solution in chloroform, T=27.5 °C.
g
Conditions: (±)ꢀ4b 1 g, lipase 200 mg, CH
3
3
CN 10 mL, MeOH 1.04 g, 1.3 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
CN 20 mL, MeOH 2.08 g, 2.6 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
h
i
Conditions: (±)ꢀ4b 2 g, lipase 400 mg, CH
3
Conditions: (±)ꢀ4b 5 g, lipase 1 g, CH CN 50 mL, MeOH 5.20 g, 6.6 mL (10 equiv), 25 ºC, 500 rpm (magnetic stirrer).
In response to 1 gram reactions, Novozym 435 exhibited excellent enantioselection
(E=170ꢀ283) in the preparative methanolysis of substrate (±)ꢀ4b providing access to the ester
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(S)ꢀ(+)ꢀ4b with >99% ee and alcohol (R)ꢀ(ꢀ)ꢀ3 with 98% ee (Table 6, entries 1 and 2)
depending on time of the process termination and thus the conversion. In the case of 2 gram
scale reactions one can see that the results strongly correlated with the previous 1 gram scale
attempts (Table 6, entries 1 and 2 vs entries 3 and 4). Hence, if the reaction was arrested after
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
24 h of proceeding Novozym 435 acted with a high stereopreference in the formation of
alcohol (R)ꢀ(ꢀ)ꢀ3 (97% ee) (Table 6, entry 3), whereas the reaction terminated after 3 days
yielded unreacted ester (S)ꢀ(+)ꢀ4b of >99% ee (Table 6, entry 4). Finally we scaledꢀup the
reaction to 5 grams of racemic substrate (±)ꢀ4b reaching a similar to the presented
enantiomeric excess values, but after twice as long reaction times. The lower reaction rate
might be due to lower level of enzyme activity regarding the mass transfer coefficient, which
in this case mostly depends on the relevant type of the employed reactor. To reach higher
conversions in shorter reaction times, the amount of lipase should be doubled and/or the
biotransformation of (±)ꢀ4b carried out at higher temperatures. However, industry sector
prefers simple reaction systems in which enzyme loading does not exceed 20% (w/w) in
relation to substrate due to costꢀeffectiveness, and elevated temperatures are not willingly
applied because of solvent volatility problems, decrease in lipase stereopreference and the
limitations in the catalyst reusability. As the preparative scale reactions allowed us to obtain
enantiomerically pure ester (S)ꢀ(+)ꢀ4b (>99% ee) for acceptable kinetics, and showed potency
to be amenable to a largeꢀscale execution, the effects of enzyme concentration and
temperature on the activity and enantioselectivity of Novozym 435 for KR of (±)ꢀ4b was
intentionally not studied. Examination of the data included in Table 6 reveals that scaling of
the reaction up to 1ꢀ, 2ꢀ and subsequently up to 5 grams of the substrate (±)ꢀ4b gave almost
the same isolated yields of recovered ester (S)ꢀ(+)ꢀ4b ranging from 90% to 96%, and the
formed alcohol (R)ꢀ(ꢀ)ꢀ3 ranging from 89% to 93% with high to excellent enantiomeric excess
(87ꢀ100% ee) for both resolved enantiomers. In addition, the reusability of the biocatalyst was
ACS Paragon Plus Environment

Page 21 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
tested by conducting the same experiments with the spent catalyst, and we found that
Novozym 435 did not show any significant activity loss and erosion of ee even after 3 reuses.
This can be recognized as crucial advantage since the lipase was exposed on long and nobble
contact with stirring bar, which can mechanically crush the immobilization support, resulting
in detachment of the enzyme from the carrier as well as on contact with methanol during the
reaction, and with chloroform while the filtration and washing procedures were performed.
Interestingly, we have found that the resultant alcohol (R)ꢀ(ꢀ)ꢀ3 was precipitated as a
crystalline solid during the progress of the reaction. The phenomenon of selective solubility of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
the alcohol and butanoate in the chosen medium (CH CN) potentially increases the utility and
practical application of the used synthetic route, as the cumbersome chromatography
purification step might be fully excluded after KR procedure. This can be successfully used
both in liquidꢀliquid extractive membrane reactors and/or immobilized enzyme biocatalytic
membrane reactors (BMRs) commonly used in industry. The experimental validity of this
feature will be verified more deeply in our laboratory in a due course by conducting
enzymatic reaction on a larger scale.
2.7. Determination of the lipase stereopreference
Lastly, the absolute configurations of both separated enantiomers [(S)ꢀ(+)ꢀ4b and (R)ꢀ(ꢀ)ꢀ3]
were determined by comparison of their optical rotation signs with the data described in the
literature. The optical rotation measured for obtained by us samples of enantiomerically pure
27.5
27.5
D
butyric ester (S)ꢀ(+)ꢀ4b (>99% ee) varied from [α]D = +45.29 (c 1.70, CHCl ) to [α]
=
3
+50.36 (c 1.40, CHCl
3
) depending on the source of their origin (experiment). In turn, the
optical rotation determined for the most enantiomerically enriched alcohol (R)ꢀ(ꢀ)ꢀ3 gave
2
D
7.5
[
α]
3
= ꢀ54.00 (c 1.00, CHCl ) at 98% ee, which is in accordance with the value given in
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
0
literature [α]D = ꢀ63.80 (c 0.42, CHCl ) at >99% ee for the compound obtained from
3
39
enantiomerically pure propylene oxide under basic conditions. In light of these findings it is
clear that the employed Novozym 435 exhibit (R)ꢀstereopreference toward racemic ester (±)ꢀ
4
b since in all of the attempted enzymatic assays it predominantly catalyzed transformation of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the (R)ꢀenantiomer into the corresponding optically active alcohol (R)ꢀ(ꢀ)ꢀ3 leaving (S)ꢀester
almost unreacted. It is reasonable to assume that Novozym 435 manifests the same
enantiopreference with (±)ꢀ3 in transesterification reactions, giving the complementary results
to lipaseꢀmediated methanolysis in terms of stereochemistry of the resolved enantiomers.
2.8. Docking
There is clearly an urgent need for development of more effective strategies for synthetic
methods, hence in depth knowledge of a catalytic behavior of the enzymes as natural
asymmetric selectors toward chiral molecules is beneficial. Limited attention has been
devoted so far to gain a deeper understanding of the molecular basis of the lipaseꢀcatalyzed
44
processes. Therefore an employment of molecular modeling techniques using docking tools,
molecular dynamics (MD) procedures, and quantum mechanics/molecular mechanics
(QM/MM) calculations to gain insight into the mechanistic details of lipase catalytic
properties should allow rational design of substrates (substrate engineering), reaction
conditions (medium engineering), and enzyme (protein engineering) for improved reaction
kinetics, yields, and higher enantioselectivity.
In order to rationalize the observed enantioselectivity of (CALꢀB)ꢀcatalyzed
methanolysis of (±)ꢀ4b, we applied an enzymeꢀsubstrate docking protocol using nonꢀ
45
commercial AutoDock Vina software. According to the wellꢀknown mechanism of action of
46
serine hydrolases, the employed immobilized CALꢀB (Novozym 435) catalyzes the
methanolysis of (±)ꢀ4b in the way depicted in Fig. 1. In general, the CALꢀB activeꢀsite
ACS Paragon Plus Environment

Page 23 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
consists of a catalytic triad of Ser105, His224 and Asp187 in which serine is the key
aminoacid engaged in biotransformations of xenobiotic substrates, and the remaining two
residues are responsible for its nucleophilic activation. The mechanism of (CALꢀB)ꢀcatalyzed
methanolysis of racemic butanoate (±)ꢀ4b involves two steps during which two nonꢀcovalent
enzyme–substrate complexes [namely, Michaelis complexes (MCCs)] and tetrahedral
intermediates (TIs) are formed. In the first step of the reaction, the activated nucleophilic
hydroxyl group of the Ser105 attacks the acylꢀcarbonyl group of the substrate (±)ꢀ4b, thus
affording the acylꢀenzyme intermediate (acylation step). The formation and collapse of the
first tetrahedral intermediate (TIꢀ1) is the rateꢀdetermining step of a whole catalytic process.
In the second step, the acylꢀgroup of the acylated enzyme (AcCALꢀB) reacts with
nucleophilic methanol (deacylation step) to form the second tetrahedral intermediate (TIꢀ2),
and finally to liberate the enzyme closing the catalytic cycle. In addition, both TIs are
stabilized by NH and OH functions in the soꢀcalled oxyanion hole of the enzyme, constituted
by the Gln106 and Thr40 residues (see Fig. 1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Reaction mechanism (catalytic cycle) of the (CALꢀB)ꢀcatalyzed methanolysis of (±)ꢀ4b.
Moreover, the binding side of CALꢀB is composed of two hydrophobic pockets: the
large pocket is lined by the respective set of Ile189, Val190, Val154, Leu140, Leu144,
Asp134 and Gln157 residues, while the medium pocket is crowded by Trp104, and Leu278–
Ala287 helix. The architecture of these pockets force substrate (±)ꢀ4b to accommodate within
the activeꢀsite of CALꢀB in two binding modes, namely binding mode I (Fig. 2 A) and
binding mode II (Fig. 2 B), thereby determining the orientation of the substrate molecule
toward catalytic triad and the oxyanion hole residues. Because some part of more bulky
substrates may extend toward the entrance of the mediumꢀsize hydrophobic pocket, the
interactions with the solvent in the particular binding mode is obviously relevant as well. For
the formation of the enzymeꢀsubstrate active complex, the proper location of the substrate
ACS Paragon Plus Environment

Page 25 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
acylꢀgroup against hydroxyl group of the catalytic Ser105 as well as the appropriate
interatomic distances, including the catalytic intramolecular hydrogen bonding network
between Asp187 and His224 residues, which lead to abstraction of acidic proton from
hydroxyl group of Ser105, and thus facilitate the nucleophilic attack (Fig. 2 C magenta
dashes), must be maintained. Bearing in mind all of the above mentioned requirements, we
have taken into account three major criteria when employing docking studies: (i) the distance
of the carbon atom of carbonyl group of alkyl side chain of (±)ꢀ4b to the oxygen atom of OH
group of the catalytic serine (Fig. 2 C orange dashes), (ii) hydrogen bond interactions
between the acetyl oxygen of (±)ꢀ4b and the residues of the oxyanion hole (Fig. 2 C green
dashes), (iii) steric clashes with the enzyme and (iv) other intermolecular polar interactions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Characteristic binding modes of racemic proxyphylline butanoate (±)ꢀ4b (yellow sticks) in the CALꢀB
(PDB ID: 1TCA) binding pocket. The activeꢀsite geometries of CALꢀB with (±)ꢀ4b docked in binding mode I
(A) and in binding mode II (B). The binding pocket of CALꢀB is constituted by a medium hydrophobic pocket
(orange sticks) above the catalytic triad AspꢀHisꢀSer (green sticks) and a large pocket (cyan sticks) below it (A
and B). (C) The most important interatomic distances for the catalytic process between substrate (±)ꢀ4b, catalytic
triad and the oxyanion hole residues (gray sticks) of the CALꢀB are indicated by magenta, green, and orange
dashed lines, respectively. The figures were generated using PyMOL.
To disclose the structural factors responsible for higher reactivity of (R)ꢀester than its
counterpart in the lipaseꢀcatalyzed methanolysis reactions of (±)ꢀ4b, the two enantiomers of
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
proxyphylline butanoate [(S)ꢀ(+)ꢀ4b and (R)ꢀ(ꢀ)ꢀ4b] were docked separately in the CALꢀB
4
7
structure taken from the Protein Data Bank (PDB), with the code 1TCA. Before docking
was performed, the target structure of CALꢀB was appropriately prepared by deleting all the
water molecules (including those that were present in the catalytic cavity), and by adding
polar hydrogen atoms (see Experimental Section). When both enantiomers of (±)ꢀ4b were
docked independently into the CALꢀB active site, 9 of the most energetically favorable
binding modes for the ligandꢀprotein complexes for each substrate molecule were generated.
The results of binding affinity energies (kcal/mol) of the two ligands with the CALꢀB enzyme
are shown in Tables S2 and S3, which are placed in the Supporting Information. Next step
provided a detail analysis of all the unproductive and productive poses (a stable
conformations that enable the generation of an acylꢀenzyme reactive complexes) selected on
the basis of the docking conformational search.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Concerning the two examined enantiomers of substrate (±)ꢀ4b, it was clear that only
the Rꢀenantiomer displayed a consistent orientation within the CALꢀB binding site to obtain
near attack conformations (NACs) leading to productive complexes (R1 and R2). As shown in
Fig. 3, the fastꢀreacting (R)ꢀenantiomer is embraced inside the hydrophilic pockets of CALꢀB
in position corresponding to the binding mode II postulated above (see Fig. 2 B). In this case,
the large substituent (1,3ꢀdimethylxanthine group) was oriented toward the activeꢀsite
entrance surrounded by hydrophobic Val154, Leu144, Ala141, Leu140 and Ile189 residues,
meanwhile the acyl moiety was positioned in the stereospecificity acidꢀbinding pocket
surrounded by Trp104 and Gln157. All the hydrogen bonds required for catalysis were
spatially arranged, and the distance between Ser105 residue and the carbonyl function of Rꢀ
ester was suitable for effective nucleophilic attack. Nevertheless, the selected docking
conformations of (R)ꢀenantiomer (R1 and R2) illustrated in Fig. 3 are not optimal for lipaseꢀ
catalysis as the distances between carbonyl atom in acyl moiety and the oxygen atom of the
ACS Paragon Plus Environment

Page 27 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
catalytic serine exceed >4 Å (blue dashed lines), and it is too far for rapid nucleophilic attack
to occur. Importantly, this can explain slow lipaseꢀmediated reaction of the employed racemic
substrate (±)ꢀ4b. Moreover, the productive conformers of fastꢀreacting enantiomer are
efficiently stabilized by polar interactions (magenta dashed lines) between carbonyl oxygen
atom of Rꢀester and the oxyanion hole Thr40 residue as well as the hydrogen bonds formed
between carboxylic group of Ile189 and oxygen atom (R1) or nitrogen atom (R2) of xanthine
ring, respectively.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In sharp contrast to the Rꢀester, all the complexes of the slowꢀreacting (S)ꢀenantiomer
remains at the outer region of the active site, being oriented in space mainly in the binding
mode I (see Supporting Information), which makes the nucleophilic attack of Ser105
unfeasible. In the case of the nonpreferential Sꢀester, the docking experiment revealed that
acyl moiety of the slowꢀreacting enantiomer is sterically accommodated in the pocket
surrounded by the subsequent residues Ile189, Ala141, Leu140, and Gln157. This stems
mainly from van der Waals (vdW) interactions with those surrounding hydrophobic residues
as well as stabilization of the ligandꢀprotein complexes by electrostatic interactions with the
nearest neighboring polar atom of the hydrophilic residues Thr40, Gln157, and Ser105. The
formation of two hydrogen bonds between unsubstituted nitrogen atom of the xanthine moiety
and the hydroxy side chains of Thr40 and Ser105 residues (S1) pulled the acyl moiety of Sꢀ
ester far away from the catalytic triade and the oxyanion hole. Moreover, the analysis of the
second nonpreferential complex of Sꢀester (S2) revealed that CALꢀB could not accommodate
(S)ꢀenantiomer in a catalytically active configuration as easily as (R)ꢀenantiomer owing to
very strong quaternary electrostatic interaction occurred. In this context, the carbonyl group of
xanthine ring is stabilized through Coulombic interactions with the amide side chain of
Gln157 as well as intermolecular hydrogen bonding with Thr40 and Ser105, respectively. The
additional hydrogen bond between the carbonyl oxygen of the ligand acyl moiety and the
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 52
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
carboxylic group of Ile189 residue is believed to maintain (S)ꢀester in its ‘uncatalytical
accommodation’. The aboveꢀmentioned hydrogen bonding network is considered to play an
important role in Sꢀligand binding, and thus might impede access to the carbonyl group of the
resolved substrate (±)ꢀ4b in the case of the slowꢀreacting (S)ꢀenantiomer.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Predominant conformations of (R)ꢀ(ꢀ)ꢀ4b (yellow sticks) (R1ꢀR2) and (S)ꢀ(+)ꢀ4b (lightorange sticks)
(S1ꢀS2) in CALꢀB (PDB ID: 1TCA) activeꢀsite. The residues constituting the catalytic triad (Asp187ꢀHis224ꢀ
Ser105) of CALꢀB are shown in green sticks representation. The oxyanionic Thr40 residue is shown in cyan. The
rest of the most significant residues contributing to the stabilization of (±)ꢀ4b enantiomers by polar interactions
(magenta dashed lines) and by CHꢀCH van der Waals (vdW) interactions are shown in gray sticks representation.
Nitrogen atoms are presented with blue colour, and the oxygen atoms with red colour. The overall enzyme
structure is shown as a gray cartoon diagram. The mutual distances between the amino acid residues and ligands
atoms are given in Ångström. The different orientation of the enantiomers toward Ser105 and Thr40 is clearly
visible. Only the fastꢀreacting (R)ꢀenantiomer (R1 and R2) is near Ser105 (4.3ꢀ4.4 Å, the distances are indicated
ACS Paragon Plus Environment

Page 29 of 52
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
by blue dashed lines) and additionally stabilized through the formation of hydrogen bonds (magenta dashes) with
Thr40 residue (3.2 Å), and Ile189 (3.2ꢀ3.5 Å).
In summary, the docking results unambiguously revealed that asymmetric
environment in the lipase activeꢀsite force stereoselective methanolysis of one of the (±)ꢀ4b
enantiomers since the carbonyl group of the alkyl side chain is situated definitely much closer
to the catalytic residues in the Rꢀester complex than in the Sꢀester counterpart. This
conformational difference suggests a faster transformation of the Rꢀester over Sꢀester what is
certainly correlated to the high Eꢀvalues obtained experimentally. Another evident
discriminating factor illustrated in Fig. 3 is related to the interactions between docked
substrate enantiomers and the aminoꢀacid residues of the oxyanion hole. One can see that for
the slowꢀreacting enantiomer, all of the generated conformers are not able to place their
potential oxyanion species in the oxyanion hole, with a consequent energy destabilization as
compared to the fastꢀreacting counterpart, where stabilizing hydrogen bonds exists
particularly between the oxyanions and Thr40 residue. Based on the results of the molecular
docking simulation, we have proved that this method can be successfully used to predict
stereopreference of lipases used in biotransformations, and thus it could be helpful in
determination of absolute configuration of newly resolved chiral compounds, for which
stereochemistry can not be simply assigned by XRD or NMR methodology.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CONCLUSION
In summary, first lipaseꢀcatalyzed ecoꢀfriendly, facile and scalable approach toward
preparation of the singleꢀenantiomer drug proxyphilline was developed. This procedure
includes simple kinetic resolution strategy using immobilized lipase B from Candida
antarctica (Novozym 435). Various enzymatic systems were employed in the critical stage of
ACS Paragon Plus Environment