CAL-B catalyzed desymmetrization of 3-alkylglutarate: "olefin effect" and asymmetric synthesis of pregabalin

Article abstract of DOI:10.1039/c3ob40311d

CAL-B catalyzed desymmetrization of prochiral 3-alkylglutaric acid diesters was performed to prepare optically active 3-alkylglutaric acid monoesters bearing various alkyl substituents, including methyl, ethyl, propyl and allyl groups. Allyl esters showed

Full text of DOI:10.1039/c3ob40311d

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
CAL-B catalyzed desymmetrization of 3-alkylglutarate:
“olefin effect” and asymmetric synthesis of pregabalin†
Cite this: DOI: 10.1039/c3ob40311d
Jae-Hoon Jung,a Doo-Ha Yoon,a Philjun Kang,b Won Koo Lee,*b Heesung Euma and
Hyun-Joon Ha*a
CAL-B catalyzed desymmetrization of prochiral 3-alkylglutaric acid diesters was performed to prepare
optically active 3-alkylglutaric acid monoesters bearing various alkyl substituents, including methyl, ethyl,
propyl and allyl groups. Allyl esters showed far better stereoselectivity among the alkyl esters, suggesting
possible π–π interactions between the olefin of the substrate and the Trp104 or His224 side chains at the
enzyme active site. Based on this reaction, the synthesis of (S)-(+)-3-aminomethyl-5-methylhexanoic acid
(pregabalin) was achieved with a 70% overall yield.
Received 12th February 2013,
Accepted 28th February 2013
DOI: 10.1039/c3ob40311d
www.rsc.org/obc
3-((alkyloxycarbonyl)methyl)-5-methylhexanoic acid, which is a
Introduction
synthetic precursor for pregabalin®,6 a lipophilic GABA (γ-amino-
Obtaining an optically pure isomer that is free from other butyric acid) analogue for treating several central nervous
stereoisomers is strongly demanded, particularly in the system disorders including epilepsy, neuropathic pain, anxiety
pharmaceutical industry,1 which drives the development of and social phobia.7 Though there is a success story5 of desym-
various catalysts including enzymes. Among many industrially metrization of diisopropyl 3-isobutylglutarate by porcine liver
applicable enzymes, lipases are the most useful to catalyze esterase with 70% ee a more stereoselective method is still
diverse reactions based on the carbonyl functionality of needed. In this study, enzymatic desymmetrization of 3-alkyl-
carboxylic acids and their derivatives.2 Many different types of glutaric acid diesters was undertaken to achieve a practical syn-
lipases are available from various sources including mamma- thetic route to prepare pregabalin. We describe one of the
lian cells and diverse microorganisms. Furthermore, most most efficient synthetic routes for pregabalin through lipase-
lipases are tolerant of a wide spectrum of substrates and reac- mediated desymmetrization of 3-alkylglutaric acid diesters. In
tion media. However, there is a significant limitation in addition, a systematic study of the enzymatic desymmetriza-
obtaining optically pure compounds from lipase-mediated re- tion of the 3-alkylglutaric acid diesters provides valuable infor-
solution of racemates, and only 50% reaction yields can be mation regarding the active site binding mode, which is
obtained. To overcome this limitation, racemization of the dependent not only on the alkyl group at C3 but also the alkyl
unreacted isomer with a strong base or a metal complex is a group at the ester. Many structurally diverse organic molecules
possible alternative.3 Desymmetrization is another way to have been utilized as lipase substrates for reactions, including
produce an optically pure reaction product with a 100% theor- hydrolysis, esterification and ammoniolysis, some of which
etical yield in the absence of a racemizer.4 Various chemical provide quite clear pictures of the binding mode near the
and enzymatic methods have been reported for desymmetriza- active site residues combined with their crystalline structures.2
tion to yield optically pure products. However, few studies have However, there are still many uncertainties predicting the reac-
reported the chemical and enzymatic desymmetrization of tivity and stereoselectivity of certain substrates due to enzyme
3-alkylglutaric acid diesters to obtain chiral 3-alkylglutaric acid mobility and remote interactions. Throughout this study, we
monoesters, which are good starting materials for biologically found an “olefin effect”, in which the diallyl esters among
important molecules.5 One of the representative examples is other alkyl esters showed far better stereoselectivity with possi-
ble remote interactions between the substrate and the enzyme
active site.
aDepartment of Chemistry and Protein Research Center for Bio-Industry,
Hankuk University of Foreign Studies, Yongin, Kyunggi-Do 449-719, Korea.
E-mail: hjha@hufs.ac.kr; Tel: +82-31-330-4369
CAL-B catalyzed desymmetrization of
3-alkylglutarates
bDepartment of Chemistry, Sogang University, Seoul 121-742, Korea.
E-mail: wonkoo@sogang.ac.kr; Tel: +82-2-705-8449
†Electronic supplementary information (ESI) available: Preparation of sub-
The first step focused on enzymatic desymmetrization of pro-
chiral 3-isobutylglutaric acid, its diesters, and anhydrides with
strates. NMR spectra. HPLC data for ee values. Further results of molecular
dynamics simulation. See DOI: 10.1039/c3ob40311d
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
some commercial lipases including CAL-B (an immobilized discriminate one enantiomer from the other by the specific
lipase B from Candida antartica), porcine pancreatic, Pseudo- interactions between the substrates and the enzyme. Propyl
monas cepacia, Candida cylindracea, Aspergillus niger, Rhizopus and allyl groups have the same carbon-chain length, but the
niveus, Rhizopus arrhizus, wheat germ, and porcine liver ester- allyl group has olefin at the end rather than a saturated alkane
ase. Enzymatic desymmetrization of 3-isobutylglutaric acid seen in the propyl group. This “olefin effect” was further investi-
and its anhydride with alcohols in toluene did not yield the gated with 3-propyl (2) and 3-methylglutarates (3). Dipropyl
corresponding monoester products over 24 hours. Thus, we (2c) and diallyl (2d) 3-propylglutarates were reacted with the
switched our attention to the selective hydrolysis of prochiral CAL-B enzyme under all of the same conditions. A larger differ-
dialkyl 3-isobutylglutarate prepared from esterification of the ence in stereoselectivity was observed in the dipropyl (2c) and
corresponding diacids. Fortunately, two enzymes, CAL-B and diallyl (2d) 3-propylglutarates, with ee values of 12 and 72%,
porcine liver esterase, were active in the hydrolytic reaction of respectively (entries 6 and 7). The chemically similar esters
diethyl 3-isobutylglutarate (1b), and CAL-B was slightly better such as dipropargyl (2e) and dicyclopropylmethyl (2f) 3-propyl-
than porcine liver esterase. We thereby decided to further glutarates showed poor stereoselectivities of 35 and 15% ee
study the CAL-B enzyme to determine how efficiently the compared to those of allylesters (2d) (entries 8 and 9). These
desymmetrization proceeds and how good the stereoselectivity results highlight the “olefin effect” that was not observed
would be by changing the alkyl group of prochiral dialkyl 3-iso- either in the alkynyl group with two π-bonds or in the confor-
butylglutarate (1).
mationally less mobile cyclopropylmethyl group. Hydrolytic
The typical reaction was carried out with the enzyme in reactions of dipropyl (3c) and diallyl (3d) 3-methylglutarates
phosphate buffer at pH 7 with a pH adjustment throughout revealed ee values of 11 and 15% which were quite poor, but
the reaction using 0.25 M NaOH (Table 1). After completion of there was still some “olefin effect” (entries 10 and 11).
the hydrolytic reaction in 5.1 h, the reaction product, a mono-
ethyl ester of 3-isobutylglutaric acid (4b), was obtained at a lyzed 4a, the carboxylic acid was converted to the Cbz-
quantitative yield of 76% enantiomeric excess (ee) (entry 2). protected amine 7 followed by hydrolysis to yield amino acid 8.
To determine the absolute stereochemistry of the hydro-
Prochiral dimethyl (1a) and dipropyl 3-isobutylglutarates The [α]2D5 value was +4.0°, indicating that the absolute stereo-
(1c) were prepared and reacted with CAL-B to obtain their chemistry of the amino acid was 3R.8 Therefore, the desymme-
corresponding monoesters (4a and 4c) in quantitative yields trization reaction of prochiral dialkyl 3-isobutylglutarates
(entries 1 and 3). As expected, the stereochemical outcomes proceeded to give the monoester as a product with a 3S con-
were also quite similar to the ethyl case with 75 and 50% ee, figuration (Scheme 1).
respectively.
The diallyl esters (1d) showed astonishingly improved
stereoselectivity of 93% ee without sacrificing the reaction
yield, which was the average of quadruplicates with different
reaction scales from several hundred milligrams up to several
grams (entry 4). However, the di-i-butyl ester (1e) substrate,
bearing one additional carbon chain on the allyl or propyl
group, was quite inert to the hydrolytic reaction after 36 hours
(entry 5). The difference in the ee between the propyl and allyl
Scheme 1 Stereochemical identification of the 3-isobutylglutaric acid mono-
ester hydrolytic product (4a).
esters at 50 and 93% was equivalent to 1.197 kcal mol−1 to
Table 1 CAL-B catalyzed desymmetrization of prochiral dialkyl 3-isobutylglutarate (1–3)
Entry
S.M.
R
R′
Conversion (%)
Reaction time (h)
Product (ee)
1
2
3
4
5
6
7
8
1a
1b
1c
1d
1e
2c
2d
2e
2f
i-Bu
i-Bu
i-Bu
i-Bu
i-Bu
Pr
Pr
Pr
Pr
Me
Me
Me
Et
Pr
Allyl
i-Bu
Pr
Allyl
Propargyl
100
100
100
100
5
100
100
100
100
100
100
5.0
5.1
15.5
15.6
36.0
12.7
5.7
1.4
3.0
0.4
0.2
4a (75)
4b (76)
4c (50)
4d (93)
—
5c (12)
5d (72)
5e (35)
5f (15)
6c (11)
6d (15)
9
10
11
Cyclopropylmethyl
Pr
Allyl
3c
3d
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 2 The tetrahedral intermediate Td1 shows the schematic details of active
site residues including Trp104 where the allyl group binds. Active site details
feature the hydrogen bonds essential for catalysis (red dotted lines D1–D6) and
the distances from the olefinic carbon at the substrate to the nearest arylic
carbon in the Trp104 side chain (D7) and to the nearest N3 of the imidazole
ring attached at the catalytic His224 (D8).
Fig. 1 Tetrahedral intermediates (Td1) of the selected substrates with different
configurations during individual 15 ns molecular dynamics simulations.
Molecular dynamics study
Then, we determined the role of this olefin by directly compar-
ing the diallyl and dipropyl 3-isobutylglutarate during binding
at the active site of the CAL-B enzyme with a 15 ns molecular
dynamics (MD) simulation based on the first tetrahedral inter-
mediate Td1 for serine esterase-catalyzed hydrolysis. Four
different tetrahedral intermediates (Td1) of the selected sub-
strates were generated with different configurations such as
Fast-1d, Slow-1d, Fast-1c, and Slow-1d (Fig. 1).
Fig. 3 Results of the molecular dynamics simulation between the fast and
slow-reacting substrates (Fast-1d vs. Slow-1d and Fast-1c vs. Slow-1c) based on
the difference in the T1 torsional angle.
(D7 and D8) and the T1 torsional angle around the active
The tetrahedral intermediate Td1 must form six catalytically carbonyl group (Fig. 2). The D7 descriptor represents the dis-
essential hydrogen bonds (D1–D6), containing three for the tance between the olefinic carbon in the substrate and the
oxyanion hole (D1–D3) and another three for the catalytic triad nearest arylic carbon in the Trp104 side chain which is posi-
(D4–D6) with lipase, as shown in Fig. 2 with the representative tioned at the nearby active site. Another descriptor D8 shows
substrate 1d.
the distance between the olefinic carbon in the substrate and
The formation and collapse of the tetrahedral intermediate the nearest N3 of the imidazole ring attached at the catalytic
Td1 dominated the enzymatic reaction suggesting that Td1 His224. We observed that the more selective substrates (1d vs.
resembled the transition state of the catalytic reaction.9 There- 1c) showed a larger difference in the T1 torsional angle, i.e. the
fore, the origin of the kinetic resolution by lipase-catalyzed torsional angle difference between Fast-1d and Slow-1d was
hydrolysis should be studied by inspecting Td1 through MD much bigger than the same angle difference between Fast-1c
simulations. The six essential hydrogen bond interactions and Slow-1c.
between the substrate and the active site in the tetrahedral
Moreover, the T1s torsional angles for Slow-1d, Fast-1c and
intermediate Td1 were investigated for each of four substrates Slow-1c gradually approached about 60° of unreactive geo-
(Fast-1d, Slow-1d, Fast-1c and Slow-1c). They were evaluated by metry, whereas only Fast-1d went to 0° as a favorable confor-
comparison with known criteria10,11 including the atomic dis- mer (Fig. 3).10
tance (N–O or O–O < 3.1 Å)12 and the bond angle (N–H–O or
The results of the MD study for the descriptors D7 and D8
O–H–O > 120°) (Fig. 2). All essential hydrogen bonds (D1–D6) showed the π–π interactions between the olefin of the substrate
generated from the four different substrates were sufficiently and the Trp104 and His224 side chain at the active site (Fig. 4
effective to proceed without a notable difference to explain the and Table 2). The average distance D7 for the Fast-1d between
selectivity observed in the allyl ester.
the olefin in the substrate and the aryl group in Trp104 was
Further investigation into the transition state featured new 4.20 Å, which was shorter than that of the Slow-1d substrate at
descriptors including the non-bonding interaction distances 4.63 Å. A similar difference was observed for the Fast-1d and
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 2 Pregabalin synthesis.
Pregabalin synthesis
The hydrolytic reaction product was utilized to prepare (S)-(+)-
3-aminomethyl-5-methylhexanoic acid (10) (pregabalin). Con-
version of the ester (4d) to an amide (9) was achieved in a
highly efficient manner without racemization and no interfer-
ence of carboxylic acid with Mg3N2 in methanol.16 Amide 9
was recrystallized twice in diethyl ether to produce an enantio-
merically pure product in 93% yield. This was the synthetic
precursor used for preparing pregabalin,5 which is adminis-
tered to treat several diseases related to the central nervous
system (Scheme 2).7
Fig. 4 (a) The three-dimensional Td1 structure of Fast-1d at the enzyme active
site. Three different functional groups including the olefin in the substrate,
His224 and Trp104 at the active site were close together and expressed by the
surface filling model. (b) A comparison of the three-dimensional Td1 structure of
Fast-1d (brown colored molecule) vs. Slow-1d (grey colored molecule) at the
active sites. Green solid lines show the strong interactions among three nearby
groups including the olefin in the substrate and His224 and Trp104 at the active
site of Fast-1d. Green dotted lines show the same but weak interactions of
Slow-1d.
Table 2 Fast-1d and Slow-1d distances D7 and D8 generated from the mole-
cular dynamics simulation snapshots
Conclusions
D7
D8
In conclusion, we examined the “olefin effect” on the CAL-B
catalyzed-hydrolytic reaction of prochiral diallyl 3-alkylgluta-
rate. We also successfully synthesized pregabalin from the
hydrolytic product 3-((allyloxy-carbonyl)methyl)-5-methylhexa-
noic acid (4d) through two consecutive steps including trans-
formation of the ester to the corresponding amide and a
Hoffman rearrangement with a 70% overall yield.
Fast-1d
Slow-1d
Fast-1d
Slow-1d
Average (Å)
4.20
300
122
41
4.63
300
21
3.94
300
207
69
4.24
300
99
Total snapshots (15 ns)
Within criteriaa
a
%
7
33
a Distance criteria were less than 4.0 Å between the olefinic carbon in
the substrate and the nearest arylic carbon in the side chain of Trp104
(D7) and the nearest N3 of the imidazole ring attached at the catalytic
His224 (D8).
Experimental
The CAL-B, CHIRALZYME® L-2, c.-f., lyo, was purchased from
Roche. The high performance liquid chromatography (HPLC)
analysis was performed using a YL 9100 HPLC system with a
Slow-1d substrates in D8 (3.94 and 4.24 Å, respectively), which
was the distance between the olefin in the substrate and the
nearest N3 of the imidazole ring attached at the catalytic
His224. The effective distances D7 and D8 for the nonbonding
interaction were less than 4.0 Å,13 and a large difference was
observed in the snapshots of each enantiomer during the
15 ns. Among 300 cases, 122 (41% for D7) and 207 (69% for
D8) snapshots were effective for Fast-1d, compared to only 21
(7% for D7) and 99 (33% for D8) snapshots for Slow-1d. The
olefin in the case of Fast-1d drove the favorable conformation
change in Td1 with a big discrimination against Slow-1d,
which we called the “olefin effect”.14
Undoubtedly, the experimental and modeling results
corroborate the “olefin effect” between the substrate and the
enzyme,15 which was the crucial factor increasing the enantio-
selectivity on the CAL-B catalyzed desymmetrization of
prochiral 3-alkylglutaric acid diallylesters over other esters.
1
UV/VIS detector. H NMR and 13C NMR spectra were obtained
using a Varian 200 (200 MHz for 1H and 50.3 MHz for 13C)
spectrometer and deuterated solvents (CDCl3 + 0.03% TMS,
CD3OD; Euriso-top®). Optical rotation was measured using a
polarimeter (Rudolph Autopol IV digital polarimeter). Low and
high resolution mass spectrometry (MS) analyses were per-
formed using a Macromass ZQ4000 LC/MS system and an AB
Sciex 4800 Plus MALDI TOF/TOF™ with a 2,5-dihydroxy-
benzoic acid (DHB) matrix was used to prepare samples for
MS. Data were obtained in the reflector positive mode with
internal standards for calibration, respectively.
Analytical methods
After CAL-B-mediated hydrolysis, a chiral HPLC method was
used to analyze the mono-ester products obtained. The mono-
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
ester products 4a–6d were analyzed with a Chiralpak® IA 2.05–2.37 (m, 3H), 3.07–3.34 (m, 2H), 5.03 (s, 1H), 5.10 (s, 2H),
(4.6 mm × 250 mm) using IPA–hexane–TFA 1 : 99 : 0.1, 0.3 mL 7.35 (m, 5H); 13C NMR (50 MHz, CD3OD): δ = 176.83, 159.08,
min−1 (5c) or 0.7 mL min−1 (5d), IPA–hexane 2 : 98, 0.6 mL 138.45, 129.42, 128.90, 128.71, 67.37, 45.47, 42.56, 38.28,
(6c–6d) and 4a–4e, 5e–5f were analyzed by Chiralcel® OD-RH 34.95, 26.34, 23.05; HRMS-MALDI (m/z): [M + Na]+ calcd for
(4.6 mm × 150 mm) using IPA–ACN–TFA 97 : 3 : 0.1, 0.4 mL C16H23NaNO4, 316.1519; found, 316.1524.
min−1, H2O–ACN 80 : 20, 0.3–0.8 ml min−1 and detection at
210–215 nm.
(R)-3-(2-Amino-2-oxoethyl)-5-methylhexanoic acid 9
General procedure for CAL-B mediated hydrolysis
Mg3N2 (0.155 g, 1.53 mmol, 5 equiv.) was added to a solu-
tion of (S)-3-(2-(allyloxy)-2-oxoethyl)-5-methylhexanoic acid 4d
(70.1 mg, 0.31 mmol, 1 equiv.) in MeOH (1.3 mL) at room
temperature. The mixture was warmed to 80 °C and stirred for
16 h. The reaction mixture was acidified to pH 3–4 with 6 N
HCl. The aqueous layer was extracted with EtOAc (3 × 30 mL).
The organic layer was dried over MgSO4, filtered, and concen-
trated in vacuo. The crude compound was purified by recrystal-
lization in Et2O (20 mL) to furnish title compound 9 (53.9 mg,
Dialkyl 3-alkylpentanedioate (0.65 mmol) was diluted in phos-
phate buffer solution (pH 7) at 25 °C. When the pH had stabil-
ized, CAL-B (180 mg) was added to the solution. pH was
maintained with NaOH (0.25 M, 2.6 mL, 0.65 mmol). The reac-
tion mixture was diluted with EtOAc (10 mL) and filtered. The
aqueous layer was extracted with EtOAc (2 × 20 mL). The
organic layer extracts were dried over MgSO4, filtered, and con-
centrated in vacuo. The % ee value of the given compound as a
yellow oil (99%) was measured using the chiral HPLC method.
1
93%) as a white solid. H NMR (200 MHz, DMSO-d6): δ = 0.80
(d, J = 6.4 Hz, 6H), 1.07 (t, J = 6.4 Hz, 2H), 1.57 (m, 1H),
1.97–2.22 (m, 5H), 6.37 (s, 1H), 6.77 (s, 1H); 13C NMR (50 MHz,
DMSO-d6): δ = 174.7, 174.6, 43.8, 40.4, 30.4, 25.2, 23.3, 23.3;
MS (m/z): [M + Na]+ calcd for C9H17NO3Na, 210.11; found,
209.88; Mp 131–135 °C.
(R)-Methyl 3-((benzyloxycarbonylamino)methyl)-5-methyl-
hexanoate 7
TEA (0.072 mL, 0.52 mmol, 1 equiv.) and DPPA (0.11 mL,
0.52 mmol, 1 equiv.) were added to a solution of (S)-3-
(2-methoxy-2-oxoethyl)-5-methylhexanoic acid 4a (104.3 mg,
0.52 mmol, 1 equiv.) in toluene (1.3 mL) at room temperature.
The solution was warmed to 90 °C and stirred for 2 h. Benzyl
alcohol (0.053 mL, 0.52 mmol, 1 equiv.) was added, and the
reaction mixture was refluxed overnight. The reaction mixture
Molecular modeling
Preparation of the MD simulation
was washed with a solution of 1% NaNO2 and 1.5% NaHCO3 Proteins were prepared using Discovery Studio 2.5 (Accelrys
in H2O (2 × 100 mL), followed by H2O (100 mL). The organic Inc., San Diego, CA, USA). The starting crystal structure of
layer was dried over MgSO4, filtered, and concentrated CAL-B (PDB ID: 1TCA) was taken from the Protein Data Bank.17
in vacuo. The crude compound was purified by column All simulations were performed using the CHARMm force-
chromatography on silica gel (9 : 1 Hex–EtOAc) to furnish the field.18 The missing atoms and residues in the structure were
title compound 7 (141 mg, 88%) as a colorless oil. [α]2D5 +3.84° added at pH 7.0. The partial charges of the tetrahedral inter-
(c1.51, EtOH); 1H NMR (200 MHz, CDCl3): δ = 0.89 (m, 6H), mediate were set according to the Momany–Rone calculation
1.10–1.25 (m, 2H), 1.58–1.64 (m, 1H), 2.13–2.31 (m, 3H), in this and subsequent steps.19 This includes three stages of
3.08–3.32 (m, 2H), 3.65 (s, 3H), 4.93 (s, 1H), 5.10 (s, 2H), 7.35 minimization: (1) minimization of just hydrogen atoms while
(m, 5H); 13C NMR (50 MHz, CD3OD): δ = 171.78, 155.64, all heavy atoms are fixed; (2) minimization of all hydrogens
135.05, 126.05, 125.54, 125.39, 63.98, 48.59, 42.14, 39.22, and side chain atoms while all backbone atoms are fixed and
34.71, 31.62, 22.97, 19.69; HRMS-MALDI (m/z): [M + Na]+ calcd (3) minimization of all atoms. The water molecules from the
for C17H25NaNO4, 330.1676; found, 330.1674.
X-ray structure and hydrogen in the Ser105Oγ side chain were
removed, and Ser105Oγ was covalently bound to the tetra-
hedral hemiacetal carbon and the catalytic His224 residue was
protonated. The first gentle minimization of enzyme was
(R)-3-((Benzyloxycarbonylamino)methyl)-5-methylhexanoic
acid 8
NaOH (25 mg, 0.63 mmol, 2 equiv.) was added to a solution briefly performed with the Smart Minimizer algorithm com-
of (R)-methyl 3-((benzyloxycarbonylamino)methyl)-5-methyl- posed of the steepest descent and conjugate gradient with
hexanoate 7 (97 mg, 0.32 mmol, 1 equiv.) in MeOH–H2O (8 : 1, 5000 steps set to 0.001 kcal mol−1 Å of RMS gradient during
2 ml) at room temperature. The reaction mixture was stirred generalized Born solvation. Next, Fast-1d, one of the fast-react-
for 4 h, and then MeOH was evaporated in vacuo. The reaction ing substrates as a reference ligand, was covalently bound to
mixture in H2O was acidified with 3 N HCl to pH 3–4, and Ser105 and the transition state structure (Td1) maintained.
extracted with DCM (3 × 30 mL). The organic layer was dried Further calculations were carried out for Slow-1d, Fast-1c, and
over MgSO4, filtered and concentrated in vacuo. The crude Slow-1c. The ligand was minimized in the same manner
compound was purified by column chromatography on silica by keeping the enzyme fixed. The system was resolved with
gel (1 : 1 Hex–EtOAc) to furnish title compound 8 (74 mg, 80%) several water molecules in an explicit spherical boundary with
1
as a yellow oil. [α]2D5 +4.00° (c1.20, EtOH); H NMR (200 MHz, a harmonic restraint set to a 20 Å spherical radius from the
CDCl3): δ = 0.89 (m, 6H), 1.13–1.33 (m, 2H), 1.60–1.66 (m, 1H), mass center to save calculation time.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
MD simulation
J. Org. Chem., 2011, 76, 5709–5718; (d) G. Obame,
H. Pellissier, N. Vanthuyne, J. B. Bongui and G. Audran,
Tetrahedron Lett., 2011, 52, 1082–1085; (e) N. Ríos-Lombar-
día, V. Gotor-Fernández and V. Gotor, J. Org. Chem., 2011,
76, 811–819; (f) C. Manzuna Sapu, J. E. Bäckvall and
J. Deska, Angew. Chem., Int. Ed., 2011, 50, 9731–9734.
5 M. S. Hoekstra, D. M. Sobieray, M. A. Schwindt,
T. A. Mulhern, T. M. Grote, B. K. Huckabee,
V. S. Hendrickson, L. C. Franklin, E. J. Granger and
G. L. Karrick, Org. Process Res. Dev., 1997, 1, 26–38.
6 (a) M. J. Burk, P. D. de Koning, T. M. Grote, M. S. Hoekstra,
G. Hoge, R. A. Jennings, W. S. Kissel, T. V. Le, I. C. Lennon,
T. A. Mulhern, J. A. Ramsden and R. A. Wade, J. Org. Chem.,
2003, 68, 5731–5734; (b) G. M. Sammis and E. N. Jacobsen,
J. Am. Chem. Soc., 2003, 125, 4442–4443; (c) T. Ok, A. Jeon,
J. Lee, J. H. Lim, C. S. Hong and H. S. Lee, J. Org. Chem.,
2007, 72, 7390–7393; (d) C. A. Martinez, S. Hu, Y. Dumond,
J. Tao, P. Kelleher and L. Tully, Org. Process Res. Dev., 2008,
12, 392–398; (e) Y. Chen, X. Li and R. Cheng,
Chin. J. Org. Chem., 2011, 31, 1582–1594; (f) J. M. Liu,
X. Wang and Z. M. Ge, Tetrahedron, 2011, 67, 636–640;
(g) H. Mukherjee and C. A. Martinez, ACS Catal., 2011, 1,
1010–1013.
To minimize all substrates, a standard dynamics cascade was
performed by a series of two-stage minimizations, heating,
equilibrium, and production by adopting a 12 Å non-bound
spherical cut-off, using the isothermal-isochoric ensemble
(NVT), and a distance-dependent dielectric implicit solvent
model with a dielectric constant set to 1.0. The first minimiz-
ation was performed with 5000 steps using the steepest
descent algorithm with a 0.001 of RMS gradient. The second
minimization introduced 50 000 steps with the conjugate gra-
dient and a 0.0001 RMS gradient. Next, the complex was
heated with 10 000 steps to a 250 K initial temperature and a
300 K target temperature, and equilibrated with 200 000 steps
using a 0.001 time step. Finally, 15 000 000 production stage
steps with a 0.001 ps time step yielded the final product gener-
ated from 300 snapshots. Only potential or total energy was
used to confirm whether the generated conformers were stable
or not during MD. All snapshots for each enantiomer were
analyzed with focus on the distance and the angle between
heteroatoms.
Acknowledgements
7 (a) I. Selak, Curr. Opin. Invest. Drugs, 2001, 2, 828–834;
(b) B. A. Lauria-Horner and R. B. Pohl, Expert Opin. Invest.
Drugs, 2003, 12, 663–672.
8 Z. Hameršak, I. Stipetić and A. Avdagić, Tetrahedron:
Asymmetry, 2007, 18, 1481–1485.
This study was supported by the GRRC program of Gyeonggi
province (GRRC HUFS-2012-B03) and by the National Research
Foundation of Korea (NRF-2011-0012379).
9 K. Nishizawa, Y. Ohgami, N. Matsuo, H. Kisida and
H. Hirohara, J. Chem. Soc., Perkin Trans. 2, 1997, 1293–
1298.
Notes and references
1 (a) V. Gotor-Fernández, R. Brieva and V. Gotor, J. Mol. 10 (a) J. Nyhlén, B. Martín-Matute, A. G. Sandström, M. Bocola
Catal., B, 2006, 40, 111–120; (b) R. N. Patel, Coord. Chem.
Rev., 2008, 252, 659–701; (c) R. N. Patel, ACS Catal., 2011, 1,
1056–1074; (d) C. M. Clouthier and J. N. Pelletier, Chem.
Soc. Rev., 2012, 41, 1585–1605.
and J.-E. Bäckvall, ChemBioChem, 2008, 9, 1968–1974;
(b) J. H. Park, H. J. Ha, W. K. Lee, T. Généreux-Vincent and
R. J. Kazlauskas, ChemBioChem, 2009, 10, 2213–2222.
11 I. J. Colton, D. L. T. Yin, P. Grochulski and R. J. Kazlauskas,
Adv. Synth. Catal., 2011, 353, 2529–2544.
2 (a) U. T. Bornscheuer and R. J. Kazlauskas, Hydrolases in
Organic Synthesis: Regio- and Stereoselective Biotransform- 12 M. Cygler, P. Grocbulski, R. J. Kazlauskas, J. D. Schrag,
ations, Wiley-VCH, Weinheim, 2nd edn, 2006, pp. 124–127;
(b) K. Faber, Biotransformations in Organic Chemistry,
F. Bouthillier, B. Rubin, A. N. Serreqi and A. K. Gupta,
J. Am. Chem. Soc., 1994, 116, 3180–3186.
Springer, Berlin, 5th edn, 2004, pp. 29–176; (c) V. Gotor, 13 R. O. Gould, A. M. Gray, P. Taylor and M. D. Walkinshaw,
I. Alfonso and E. García-Urdiales, Asymmetric Organic Syn- J. Am. Chem. Soc., 1985, 107, 5921–5927.
thesis with Enzymes, Wiley-VCH, Weinheim, 1st edn, 2008, 14 (a) E. A. Meyer, R. K. Castellano and F. Diederich, Angew.
pp. 133–169.
Chem., Int. Ed., 2003, 42, 1210–1250; (b) A. T. Macias and
A. D. MacKerell Jr., J. Comput. Chem., 2005, 26, 1452–1463;
(c) L. M. Salonen, M. Ellermann and F. Diederich, Angew.
Chem., Int. Ed., 2011, 50, 4808–4842.
3 (a) O. Pàmies and J.-E. Bäckvall, Chem. Rev., 2003, 103,
3247–3261; (b) B. Martín-Matute and J.-E. Bäckvall, Curr.
Opin. Chem. Biol., 2007, 11, 226–232; (c) J. H. Lee, K. Han,
M.-J. Kim and J. Park, Eur. J. Org. Chem., 2010, 999–1015; 15 (a) H. Yang, E. Henke and U. T. Bornscheuer, J. Org. Chem.,
(d) H. Pellissier, Tetrahedron, 2011, 67, 3769–3802;
(e) Y. Kim, J. Park and M.-J. Kim, ChemCatChem, 2011, 3,
271–277.
1999, 64, 1709–1712; (b) K. Thodi, E. Barbayianni,
I. Fotakopoulou, U. T. Bornscheuer, V. Constantinou-
Kokotou, P. Moutevelis-Minakakis and G. Kokotos, J. Mol.
Catal., B, 2009, 61, 241–246.
4 Reviews: (a) E. Schoffers, A. Golebiowski and
C. R. Johnson, Tetrahedron, 1996, 52, 3769–3826; 16 G. E. Veitch, K. L. Bridgwood and S. V. Ley, Org. Lett., 2008,
(b) E. García-Urdiales, I. Alfonso and V. Gotor, Chem. Rev., 10, 3623–3625.
2005, 105, 313–354; Recent examples: (c) N. Ríos- 17 J. Uppenburg, M. T. Hansen, S. Patkar and T. A. Jones,
Lombardía, E. Busto, V. Gotor-Fernández and V. Gotor,
Structure, 1994, 2, 293–308.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
18 A. D. MacKerell Jr, D. Bashford, M. Bellott, R. L. Dunbrack
Jr, J. D. Evanseck, M. J. Field, S. Fischer, J. Gao, H. Guo,
S. Ha, D. Joseph-McCarthy, L. Kuchnir, K. Kuczera,
J. C. Smith, R. Stote, J. Straub, M. Watanabe,
J. Wiórkiewicz-Kuczera, D. Yin and M. Karplus, J. Phys.
Chem. B, 1998, 102, 3586–3616.
F. T. K. Lau, C. Mattos, S. Michnick, T. Ngo, D. T. Nguyen, 19 F. A. Momany and R. Rone, J. Comput. Chem., 1992, 13,
B. Prodhom, W. E. Reiher III, B. Roux, M. Schlenkrich,
888–900.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.