Immobilization of ω-transaminases by encapsulation in a sol-gel/celite matrix

Article abstract of DOI:10.1016/j.molcatb.2009.12.001

Commercially available ω-transaminases ω-TA-117, -113, and Vibrio fluvialis (Vf-AT) have been immobilized in a sol-gel matrix. Improved results were obtained by employing Celite 545 as additive. The immobilized ω-transaminases ω-TA-117, -113, and V. fluvialis (Vf-AT) were tested in the kinetic resolution of α-chiral primary amines. In contrast to the free enzyme ω-TA-117, the sol-gel/celite immobilized enzyme showed activity even at pH 11. Recycling of the sol-gel/Celite 545 immobilized ω-transaminase ω-TA-117 was performed over five reaction cycles without any substantial loss in enantioselectivity and conversion. Finally, the immobilized ω-TA 117 was employed in a one-pot two-step deracemization of rac-mexiletine and rac-4-phenyl-2-butylamine, two pharmacologically relevant amines. The corresponding optically pure (S)-amines were obtained in up to 95% isolated yield (>99% ee).

Full text of DOI:10.1016/j.molcatb.2009.12.001

Journal of Molecular Catalysis B: Enzymatic 63 (2010) 39–44
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Immobilization of ␻-transaminases by encapsulation in a sol–gel/celite matrix
∗
Dominik Koszelewski, Nicole Müller, Joerg H. Schrittwieser, Kurt Faber, Wolfgang Kroutil
Research Centre Applied Biocatalysis, c/o Department of Chemistry, Organic and Bioorganic Chemistry, University of Graz, Heinrichstrasse 28, A-8010 Graz, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 October 2009
Received in revised form
Commercially available ␻-transaminases ␻-TA-117, -113, and Vibrio fluvialis (Vf-AT) have been immo-
bilized in a sol–gel matrix. Improved results were obtained by employing Celite 545 as additive. The
immobilized ␻-transaminases ␻-TA-117, -113, and V. fluvialis (Vf-AT) were tested in the kinetic resolu-
tion of ␣-chiral primary amines. In contrast to the free enzyme ␻-TA-117, the sol–gel/celite immobilized
enzyme showed activity even at pH 11. Recycling of the sol–gel/Celite 545 immobilized ␻-transaminase
2
7 November 2009
Accepted 1 December 2009
Available online 5 December 2009
␻
-TA-117 was performed over five reaction cycles without any substantial loss in enantioselectivity
and conversion. Finally, the immobilized ␻-TA 117 was employed in a one-pot two-step deracem-
ization of rac-mexiletine and rac-4-phenyl-2-butylamine, two pharmacologically relevant amines. The
corresponding optically pure (S)-amines were obtained in up to 95% isolated yield (>99% ee).
© 2009 Elsevier B.V. All rights reserved.
Keywords:
Sol–gel
Transaminase
Kinetic resolution
Immobilization
Deracemization
1
. Introduction
sol–gel materials offer unique intrinsic properties, such as high
surface to volume ratio, large surface area and porosity. Further-
Aminotransferases or transaminases (TAs) (EC 2.6.1.X) are pyri-
more, nanoporous materials are usually nontoxic, inert, chemically
and thermally stable, so they are applicable where biocompatibil-
ity and thermal stability is required. Due to the low temperature
processing, the sol–gel technology represents a useful methods to
immobilize sensitive biomolecules [32]. The sol–gel encapsulation
in silica matrix has been found as an excellent method to pre-
pare high-performance enzymes, since the immobilized catalysts
typically showed improved resistance to thermal and chemical
denaturation, as well as enhanced storage and operation stability
[33].
ꢀ
doxal 5 -phosphate (PLP) dependent enzymes which are ubiquitous
in nature being responsible for transferring amino groups [1].
Recently, ␻-transaminases have gained increased attention due to
their potential for the preparation of chiral amines [2–9], which are
needed in optically pure form e.g. by the pharmaceutical industry
[
6,10–15].
For expensive enzyme preparation recycling of the biocat-
alyst is compulsory to run economic viable processes [16,17].
Recycling/reusing enzymes can be accomplished for instance by
immobilization of the enzyme [18–20]. Immobilized enzymes are
considered to be more stable with wide potential applications
ranging from chemical synthesis to biotechnology and medicines
2. Experimental
[
21–24]. To the best of our knowledge, ␻-transaminases have only
been immobilized by covalent attachment to different supports
25] but not by sol–gel entrapment. Sol–gel entrapment is a partic-
2.1. Materials and instruments
[
The
amine
rac-1-(2,6-methoxy)-2-aminopropane
(rac-
®
ularly easy and effective method to immobilize enzymes [26,27],
since it involves the simple to perform transition of a liquid ‘sol’
into a solid ‘gel’ phase. Various mechanisms underlying sol–gel
processes have been the subject of several books and reviews
mexiletine), rac-4-phenyl-2-butylamine, Tween 80 , Celite
545, tetramethyl orthosilicate (TMOS), trimethoxypropylsilane
(TMOPS), isobutyltrimethoxysilane (iBTMOS), poly(vinyl alcohol)
(PVA), sodium fluoride, 2-propanol, glucose as well as solvents
were commercially available from Sigma–Aldrich (Vienna, Austria)
and were used as received unless otherwise stated. The solvents
were of analytical grade. ␻-Transaminases ␻-TA 117, ␻-TA-113,
␻-TA-114, and from Vibrio fluvialis as well as d-glucose dehy-
drogenase (475 U/mL, order no. 29.10) and NAD-specific formate
dehydrogenase (200 U/mL, order no. FDH-101) were obtained from
Codexis® Inc. l-Lactate dehydrogenase from bovine heart (Type III,
ammonium sulfate suspension, ≥500 U/mg protein, one unit will
ꢀ
ꢀꢀ
[
28–30]. Generally, the silane precursor, R Si(OR ) undergoes acid-
3
or base-catalyzed hydrolysis and cross-linking condensation with
formation of an SiO2 matrix in which the enzyme is encapsu-
lated [31]. Regarding to porous structure in nanometer dimensions,
∗ Corresponding author. Fax: +43 0316 380 9840.
E-mail address: wolfgang.kroutil@uni-graz.at (W. Kroutil).
1
381-1177/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2009.12.001

4
0
D. Koszelewski et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 39–44
◦
◦
reduce 1.0 ␮mol of pyruvate to l-lactate per min at pH 7.5 at 37 C,
catalog no. L2625) and l-alanine dehydrogenase from Bacillus
subtilis [buffered aqueous glycerol solution, 30 U/mg protein
at 20 C, followed by evaporating of the excess of anhydride under
reduced pressure. After adding a fresh portion of ethyl acetate
(600 ␮L) the ee of the derivatized compound was measured by GC
on a chiral phase using hydrogen as carrier gas.
(
Lowry), one unit will convert 1.0 ␮mol of l-alanine to pyruvate
◦
◦
and NH per min at pH 10.0 at 25 C, catalog no. A7653] were pur-
GC program parameters: injector 220 C; flow 12.5 psi; temper-
3
◦ ◦ ◦
chased from Sigma–Aldrich. Optical rotations were measured on a
ature program 100 C/hold 2 min; 130 C/rate 1 C per min/hold
1
◦
◦
PerkinElmer Polarimeter 341 in a 1 mL cuvette of 10 cm length. H
5 min; 170 C/rate 20 C per min/hold 5 min. Retention times: (S)-1a
38.8 min, (R)-1a 40.1 min; (S)-1b 26.5 min, (R)-1b 28.6 min; (S)-1c
31.0 min, (R)-1c 30.1 min.
13
and C NMR were recorded on a Bruker 300 MHz spectrometer at
00 and 75 MHz, respectively, using TMS as internal standard. The
3
conversion of amine was measured by gas chromatography using
Agilent GC-MS 5975C that was equipped with a coated DB-1701
DF 0.25 column (30 m × 0.25 mm, Agilent Technologies). All ee
values were analyzed by using a Varian GC 3900 that was equipped
with a coated CP Chirasil-Dex CB DF 0.25 column (25 m × 0.32 mm,
Varian, Inc.).
2
.5. Enzyme recycling
The recycling of the sol–gel/Celite 545 immobilized ␻-TA 117
(50 mg) was studied for the kinetic resolution of rac-4-phenyl-
◦
2
-butylamine (1a) at 30 C. The experiments were performed
according to Section 2.3. Each reaction was terminated after 24 h by
filtration. Collected immobilized enzyme was washed with distilled
water (3 × 5 mL) and used in the successive cycle.
2
.2. Sol–gel entrapment
Sol–gel entrapment of ␻-transaminases was performed accord-
ing to method described by Reetz et al. with slight modifications
34]. A commercial ␻-transaminase (lyophilizate) (100 mg) was
dissolved in TRIS/HCl buffer (100 mM, pH 7.5, 400 ␮L, 1 mM pyri-
doxal 5 -phosphate). Celite 545 (50 mg) or/and Tween 80 (60 mg)
were employed as additives if required and the mixture was vig-
orously shaken with a vortex-mixer. Then aqueous PVA (4% w/v,
2
.6. Temperature studies
[
The temperature study was performed employing sol–gel/Celite
45 immobilized ␻-TA 117 545 (50 mg) for the kinetic resolution
ꢀ
®
5
of rac-4-phenyl-2-butylamine (1a) at varied temperature. The con-
version was determined according to Section 2.3.
1
00 ␮L), aqueous sodium fluoride (1 M, 50 ␮L) and isopropyl alco-
hol (50 ␮L) were added, and the mixture was again vigorously
shaken with a vortex-mixer. Alkylsilane (2.5 mmol) and TMOS
2.7. pH studies
(
0.5 mmol, 74 ␮L) were added and the mixture agitated for 20 s, fol-
The pH study was performed employing sol–gel/Celite 545
lowed by drying over night at room temperature in opened tubes.
The gel was washed with distilled water (20 mL); during this step a
spatula was used to crush the gel. In case of immobilized ␻-TA-117,
immobilized ␻-TA 117 (50 mg) and rac-4-phenyl-2-butylamine
1a) as substrate. Depending on the pH three different buffers
were used: citrate buffer (pH 4, 100 mM), phosphate buffer (pH 7,
00 mM) and borate buffer (pH 9 and 11, 100 mM). The conversion
was determined according to Section 2.3.
(
2
-propanol (10 mL) was used before water to wash the immo-
bilizate as described by Reetz et al. Thereafter the immobilized
-transaminase was dried exposed to air at room temperature.
1
␻
The enzyme concentration in the washing liquid was estimated
by the Bradford’s method [35]. The amount of immobilized pro-
tein was estimated to be between 20% and 35% depending on the
experiment.
2
(
.8. Preparation of optically pure (R)-4-phenyl-2-butylamine
1a) using immobilized ␻-TA from V. fluvialis via kinetic
resolution
In a plastic screw tube (50 mL) rac-4-phenyl-2-butylamine (1a)
100 mg, 0.67 mmol, 33 mM) was suspended in phosphate buffer
20 mL, 100 mM, pH 7.0, 1 mM PLP). Sodium pyruvate (74 mg,
2
.3. Transaminase assay
(
(
The free (4 mg) and immobilized ␻-transaminase (50 mg) was
◦
0.67 mmol, 33 mM), and the sol–gel/Celite 545 immobilized ␻-TA
from V. fluvialis (300 mg) were added and the reaction was shaken
performed at 30 C for 24 h in sodium phosphate buffer (100 mM,
pH 7 with 50 mM sodium pyruvate) containing pyridoxal 5 -
ꢀ
◦
at 30 C (600 rpm) for 24 h. The pH of the mixture was adjusted
phosphate monohydrate (1 mM) in 2 mL eppendorf tube. The
reaction mixture contained 50 mM of rac-amine. The conversion
to ketone was followed by GC chromatography. The reaction was
stopped by adding aqueous NaOH (200 ␮L, 10N), followed by
extraction with ethyl acetate (600 ␮L, twice). The organic phase
was dried (Na SO ) and the conversion was measured by GC using
to pH 1 employing aqueous HCl (5 M), and the remaining ketone
was extracted five times with dichloromethane (5 × 10 mL). After
the extraction, no ketone was detectable anymore in the resid-
ual aqueous phase. The pH was adjusted to pH 12, and the amine
was extracted four times with dichloromethane (4 × 10 mL). The
solvent of the combined extracts was evaporated and (R)-4-phenyl-
2
4
nitrogen as carrier gas.
2
-butylamine (R)-1a was obtained with 48% yield (48 mg) at 51%
20
of conversion; >99% ee; [˛]D −6.55, c 1.0, CHCl , [lit. 6.40, c 0.47,
2.4. Determination of enantiomeric excess
3
1
CHCl for (S)-enantiomer, 98% ee) [36]; H NMR (300 MHz) ı 1.10
3
(
2
d, 3H, J = 6.70 Hz), 1.32 (s, 2H), 1.50-1.75 (m, 2H), 2.54-2.73 (m, 2H),
.85-2.95 (m, 1H), 7.10-7.50 (m, 5H).
The enantiomeric excess of amines 1a and 1b was analyzed by
gas chromatography on a chiral phase after derivatization to the
acetoamide derivatives, which was performed by adding DMAP and
2
0-fold excess of acetic acid anhydride. The derivatization reaction
2.8.1. Synthesis of optically pure (S)-4-phenyl-2-butylamine (1a)
using immobilized ␻-TA 117 via kinetic resolution
◦
was performed in ethyl acetate for 3 h at 40 C, followed by adding
distilled water to hydrolyze the excess of acetic acid anhydride.
After drying the organic phase (Na SO ) the ee of the derivatized
compound was measured by GC on a chiral phase. Amine 1c was
analyzed after derivatization to trifluoroacetamide, which was per-
formed by adding a 20-fold excess of trifluoroacetic acid anhydride.
The derivatization reaction was performed in ethyl acetate for 3 h
In a screw tube (50 mL) rac-4-phenyl-2-butylamine (1a)
(100 mg, 0.67 mmol, 33 mM) was suspended in phosphate buffer
(20 mL, 100 mM, pH 7.0, 1 mM PLP). Sodium pyruvate (74 mg,
0.67 mmol, 33 mM), and the sol–gel/Celite 545/Tween 80® immo-
bilized ␻-TA-117 (250 mg) were added and the reaction was shaken
2
4
◦
at 30 C (600 rpm) for 24 h. After work-up as described above (S)-

D. Koszelewski et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 39–44
41
1
a was obtained with 43% yield (43 mg); >99% ee; [˛]D20 7.1, c 1.0,
CHCl , [lit. 6.40, c 0.47, CHCl for (S)-enantiomer, 98% ee] [36].
(600 ␮L). The organic phase was dried (Na2SO4) and analyzed by
GC chromatography. The product (S)-1c was obtained with 98% of
conversion and >99% ee.
3
3
2.8.2. Synthesis of optically pure (S)-phenylethylamine (1b) using
immobilized ␻-TA 117 via kinetic resolution.
2.10. Preparative deracemization of rac-phenyl-2-butylamine
In a screw tube (50 mL) rac-phenylethylamine (1b) (100 mg,
(1a)
0
1
4
.82 mmol, 41 mM) was suspended in phosphate buffer (20 mL,
00 mM, pH 7.0, 1 mM PLP). Sodium pyruvate (90 mg, 0.82 mmol,
1 mM), and the sol–gel/Celite 545 immobilized ␻-TA-117
In a screw tube (50 mL) rac-1a (50 mg, 0.28 mmol) was sus-
pended in phosphate buffer (20 mL, 100 mM, pH 7.0, 1 mM PLP).
Sodium pyruvate (31 mg, 0.28 mmol) and the sol–gel/Celite 545
immobilized ␻-transaminase TA-117 (200 mg) were added and the
◦
(
250 mg) were added and the reaction was shaken at 30 C
(
600 rpm) for 24 h. After work-up as described above (S)-1b was
20
◦
obtained with 41% yield (41 mg); >99% ee; [˛]D −42.1, c 1.0, ben-
zene. The absolute configuration was assigned by comparison of
elution order on GC and co-injection with commercially available
reference material.
reaction was shaken at 30 C (300 rpm). After the kinetic resolution
(24 h) immobilized enzyme was separated by filtration and lactate
+
dehydrogenase from bovine heart (100 ␮L), 1 mM NAD , glucose
(42 mM), glucose dehydrogenase (100 ␮L) were added and shaken
◦
for 1 h at 30 C. l-Alanine (5 equiv., 100 mM) and the second ␻-
◦
2.9. Deracemization of rac-4-phenyl-2-butylamine (1a) and
transaminase TA-114 (30 mg) were added and shaken at 30 C. After
rac-mexiletine (1c) via one-pot two-step procedure on an
analytical scale
24 h, the pH of the mixture was adjusted to pH 1 employing aque-
ous HCl (5 M), and the remaining ketone was extracted five times
with dichloromethane (5 × 10 mL). After the extraction, no ketone
was detectable anymore in the residual aqueous phase. The pH was
adjusted to pH 12, and the amine 1a was extracted four times with
dichloromethane (4 × 10 mL). The solvent of the combined extracts
of (S)-1a was evaporated and (S)-1a was obtained with 88% yield
(44 mg); >99% ee; [˛]D20 6.8 c 1.0, chloroform.
2
(
.9.1. Deracemization of rac-4-phenyl-2-butylamine (1a) to the
S)-enantiomer
The reaction was performed at 30 C for 24 h in sodium phos-
◦
ꢀ
phate buffer (100 mM, pH 7) containing pyridoxal 5 -phosphate
monohydrate (1 mM) in a 2 mL eppendorf tube. The reaction buffer
(
(
1 mL) was mixed with sol–gel/Celite 545 immobilized ␻-TA-117
50 mg) and sodium pyruvate (5.5 mg, 0.05 mmol, 50 mM). The
2.11. Preparative deracemization of rac-mexiletine (1c)
reaction mixture contained 50 mM of the corresponding amine 1a.
After the kinetic resolution (24 h) immobilized enzyme was sep-
In a screw tube (50 mL) rac-mexiletine (1c) (50 mg, 0.28 mmol,
14 mM) was suspended in phosphate buffer (20 mL, 100 mM, pH
7.0, 1 mM PLP). Sodium pyruvate (77 mg, 0.70 mmol, 14 mM), and
the immobilized ␻-ATA-117 (200 mg) were added and the reac-
arated by filtration and lactate dehydrogenase from bovine heart
+
(
20 ␮L, 220 U), NAD (1 mM), glucose (27 mg, 0.15 mmol, 150 mM),
glucose dehydrogenase (20 ␮L, 10 U) were added (to remove pyru-
◦
◦
vate from the first step) and shaken for 1 h at 30 C. l-Alanine (5
tion was shaken at 30 C (600 rpm) for 24 h. After the kinetic
+
equiv., 250 mM) and the second ␻-transaminase TA-114 (6 mg)
resolution (24 h) lactate dehydrogenase (50 ␮L), NAD (1 mM), glu-
◦
were added and shaken at 30 C. After 24 h aqueous NaOH (200 ␮L,
0N) and ethyl acetate (600 ␮L) were added. The organic phase was
dried (Na SO ) and analyzed by GC chromatography. The product
cose (42 mM), and glucose dehydrogenase (50 ␮L) were added (to
◦
1
remove pyruvate from the first step) and shaken for 1 h at 30 C. l-
Alanine (5 equiv., 70 mM), and the second ␻-transaminase TA-113
2
4
◦
(
S)-1a was obtained with 90% of conversion and >99% ee.
(30 mg) were added and shaken at 30 C for 24 h. After work-up as
described above (S)-mexiletine (1c) was obtained with 94% yield
(47 mg) at 99% of conversion; >99% ee; [˛]D20 +3.50, c 1.0, CHCl3,
[lit. +3.0, c 1.05, CHCl3 for (S)-enantiomer) [36]; H NMR (300 MHz)
2
.9.2. Deracemization of rac-mexiletine (1c) to yield (S)-1c
1
employing LDH for pyruvate removal in the second step
The kinetic resolution (24 h) of rac-1c (50 mM) was performed
as described above but employing sol–gel/Celite 545 immobilized
ı 1.19 (d, J = 6.6 Hz, 3H), 1.71 (bs, 2H), 2,30 (s, 6H), 3.32-3.45 (m, 1H),
3.52-3.65 (m, 1H), 3.67 (dd, J = 4.0, 10.2 Hz, 1H), 6.83-6.95 (m, 1H),
7.05 (d, J = 7.5 Hz, 2H); 13C NMR (75 MHz) ı 16.3, 19.8, 47.3, 78.3,
123.8, 128.9, 130.8, 155.5.
␻
-TA-117 (50 mg). The first step of the deracemization sequence
was stopped by removal of the immobilized enzyme by filtration
followed by the addition of lactate dehydrogenase (20 ␮L, 220 U),
NAD+ (1 mM), glucose (27 mg, 0.15 mmol, 150 mM), and glucose
dehydrogenase (20 ␮L, 10 U) (to remove pyruvate from the first
3. Results and discussion
◦
step) and shaken for 1 h at 30 C. l-Alanine (5 equiv., 250 mM), and
For the sol–gel immobilization of the transaminases we
expected that the type of silane precursor has an impact
on the activity of the encapsulated ␻-transaminase; there-
fore two different alkylsilanes were tested in a first approach
[trimethoxypropylsilane (TMOPS) and isobutyltrimethoxysilane
(iBTMOS)]. Procedures for sol–gel matrix immobilization have
been optimized recently [34] e.g. for optimal reagent ratios
(TMOPS/TMOS = 5:1) [37] and have been employed as previously
reported. The precursor iBTMOS led to a catalyst preparation with
very low residual activity (less then 5%, data not shown). On the
other hand TMOPS was found to be a suitable xerogel precursor and
used for the further studies. For the immobilization procedure pyri-
second ␻-transaminase TA-113 (6 mg) were added and shaken at
◦
3
0 C. After 24 h aqueous NaOH (200 ␮L, 10N) and ethyl acetate
(
600 ␮L) were added. The organic phase was dried (Na SO ) and
2 4
analyzed by GC chromatography. The product (S)-1c was obtained
with >99% of conversion and >99% ee.
2.9.3. Deracemization of rac-mexiletine (1c) to yield (S)-1c
employing l-AlaDH for l-alanine recycling in the second step
The kinetic resolution (24 h) was performed as described in
2
.8.2. After removal of the enzyme by filtration l-alanine dehy-
+
drogenase (20 ␮L, 8 U), NAD (1 mM), ammonium formate (9 mg,
ꢀ
0.15 mmol, 150 mM), and formate dehydrogenase (20 ␮L, 4 U) were
doxal 5 -phosphate (PLP, 1 mM) was added, since it has been shown
added (to transform remaining pyruvate from the first step to l-
alanine) and shaken for 1 h at 30 C. Additional l-alanine (5 equiv.,
previously that PLP was needed to stabilize ␻-transaminase dur-
ing immobilization [25]. Immobilized transaminases were tested
for the kinetic resolution of racemic pharmacologically relevant
amines (Scheme 1). For instance, optically pure 4-phenyl-2-
butylamine (1a) is a precursor of the antihypertensive dilevalol [38]
◦
2
50 mM), and the second ␻-transaminase TA-113 (6 mg) were
◦
added and shaken at 30 C. The reaction was stopped after 24 h
by the addition of aqueous NaOH (200 ␮L, 10N) and ethyl acetate

4
2
D. Koszelewski et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 39–44
Table 1
Kinetic resolution employing immobilized enzyme preparations.
Entry Substrate ␻-TA
Additivesa
c (%)b ee (%)c
Ed
1
2
3
4
5
6
7
8
9
0
1
9
0
1
3
4
5
6
5
2
3
4
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1b
1b
1b
1b
1b
1c
1c
1c
1c
1c
Vf-TA
113
117
Vf-TA
113
None
None
None
8
2
30
8
<1
48
54
16
n.d.
n.d.
43 (S)
n.d.
n.d.
98 (S)
>99 (R) >100
18 (R)
2 (R)
n.d.
–
–
–
–
–
Celite 545
Celite 545
Celite 545
Celite 545
Celite 545
117
>100
e
Vf-TA
e
113
–
–
–
Vf-TA
113
117
Vf-TA
Vf-TA
Vf-TA
117
Celite 545, Tween 80® 36
®
1
1
1
2
2
1
1
1
2
2
2
2
2
Celite 545, Tween 80
<1
Celite 545, Tween 80® 58
>99 (S) >100
None
19
18
2
53
61
11 (R)
42 (R)
n.d.
>99 (S) >100
>99 (S) >100
>99 (S) >100
15 (R)
5 (R)
39 (S)
–
–
–
Celite 545
Celite 545, Tween 80®
None
117
Celite 545
Scheme 1. Kinetic resolution of rac-amines catalyzed by immobilized ␻-
transaminases.
117
Celite 545, Tween 80® 57
e
Vf-TA
Celite 545
Celite 545
None
16
16
25
58
–
–
–
e
113
117
117
117
and mexiletine (1c) is used to treat arrhythmias within the heart
as well as for anesthesia [39–41].
First immobilization experiments employing ␻-TAs 113, 117
and V. fluvialis (Vf-TA) led only to a reasonable conversion in the
case of employing ␻-TA 117 for the kinetic resolution of rac-
Celite 545
>99 (S) >100
97 (S) >100
Celite 545, Tween 80® 47
a
b
c
Celite 545 (50 mg) or/and Tween 80® (60 mg).
Determined by GC, conversion after 24 h, 50 mM amine concentration.
Determined by GC on a chiral phase.
Enantioselectivity calculated from c and ee [45].
2-Propanol avoided during immobilization and work-up. n.d. not determined
1
a (Table 1, entries 1–3). In a next step sol–gel immobilization
d
e
was performed in the presence of Celite 545 (50 mg) as a porous
solid support. Celite 545 is described as a highly porous diatoma-
due to too low conversion.
ceous earth composed of silica (SiO )/metal oxides [42], and is
2
frequently employed for enzyme immobilization due to its chem-
ical inertness and unique interconnected pore structure [42–44].
Again, only in the case of ␻-TA 117 an improvement of conversion
was observed (entries 4-6). Reasonable active immobilized enzyme
preparations for the two other transaminases Vf-TA and ␻-TA 113
were only obtained when omitting the 2-propanol treatment dur-
ing immobilization by the sol–gel process as well as work-up of
the immobilizate (entries 7-8). Obviously the 2-propanol treatment
during immobilization did not affect ␻-TA 117 but ␻-TA 113 and
Vf-TA.
®
the moderate improvements obtained with Tween 80 for ␻-TA
®
117 did not justify the further usage of Tween 80 ; thus, ␻-TA 117
was used in the further studies as a sol–gel/Celite 545 immobilized
preparation without Tween 80 . Testing the possibility to recycle
the sol–gel/Celite 545 immobilized ␻-TA 117, the enzyme prepara-
tion was employed for a series of kinetic resolutions of rac-1a. The
gel was separated and washed after each cycle before reusing it.
The preparation could successfully be employed in eight cycles: the
enantioselectivity remained constant (E > 100) over all cycles and
the ␻-TA 117-immobilizate retained 78% of its initial conversion
after eight repetitive uses (Fig. 1).
In a further experiment the conversion was measured at varied
temperature (Fig. 2) as well as varied pH values (Fig. 3).
The apparent temperature optimum for the free enzyme was
at 30 C while the one for the immobilized enzyme was between
40 C and 50 C. The shift of temperature optimum was postulated
previously to result from a multipoint contact through hydro-
gen bonding as well as ionic and hydrophobic interactions of the
enzyme with the matrix [34]. It could be speculated that the ␻-
transaminase is conformationally arrested in the matrix leading to
®
Addition of Tween 80® as a further additive during immobiliza-
tion effected mainly Vf-TA leading to a slightly better conversion
than without (entry 9). The ␻-ATA 117 immobilized by the sol–gel
process in the presence of Celite 545 led to good conversion in the
®
absence as well as presence of Tween 80 . In both cases protein
◦
loading yielded up to 35% during immobilization. Employing the
best preparations (S)-amine 1a (entries 6, 11) as well as (R)-amine
◦
◦
1
a (entry 7) were obtained with excellent enantioselectivity for the
kinetic resolution (>100). The immobilized ␻-TA 117 proved also
to be best suited for the two other substrates 1b and 1c.
®
The addition of Tween 80 resulted in an immobilized catalyst
®
◦
which was not as easy to handle as without Tween 80 . Additionally
good activity even at higher temperature (60 C).
Fig. 1. Recycling of the sol–gel/Celite 545 immobilized ␻-TA 117 for the kinetic
resolution of rac-1a.
Fig. 2. Conversion of native (gray) and sol–gel/Celite 545 immobilized ␻-TA 117
(white) at varied temperature.

D. Koszelewski et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 39–44
43
Scheme 2. Deracemization of rac-amines 1a and 1c via a one-pot two-step synthesis procedure.
employing the immobilized enzyme was combined with stereose-
lective amination using either lactate dehydrogenase (system i) or
alanine dehydrogenase (system ii) to shift the equilibrium in the
second step to the product side. This sequence enabled an efficient
one-pot two-step deracemization leading to optically pure (S)-
amines 1a and 1c with 90% and up to >99% conversion respectively.
Having optimized process conditions, a preparative transformation
of 50 mg racemic 4-phenyl-2-butylamine (1a) at 14 mM substrate
concentration yielded (S)-1c at 96% conversion after 48 h with
>
99% ee and 89% isolated yield. Additionally, 50 mg of racemic 1-
(2,6-dimethylphenoxy)-2-propanamine (1c) was transformed into
optically pure amine (S)-1c with complete conversion and >99% ee
and 95% isolated yield.
Fig. 3. Relative conversion of native (gray) and the sol–gel/Celite 545 immobilized
␻
-TA 117 (white) at varied pH values.
4. Conclusion
Even more pronounced was the apparent stabilization effect of
Immobilization of ␻-transaminases in sol–gel matrix was stud-
the sol–gel/Celite 545 matrix when testing the conversion at varied
pH (Fig. 3).
ied. The ␻-TA 117 immobilized with Celite 545 as an additive
retained its activity over a broader range of temperature and pH
compared to the native enzyme. The enzyme could be recycled
eight times with only moderate decrease of conversion for each
cycle. Finally, the ␻-TA 117 was employed in a two-step deracem-
ization protocol, whereby the enzyme used in the first step could
simply be removed by centrifugation.
The pH range of the immobilized ␻-TA 117 was broader than
the one of the native enzyme; e.g. at pH 9.0 77% conversion was
achieved and at pH 11.0 39% conversion while the free enzyme
was almost not active anymore at pH 9.0. Thus, the immobilization
method preserved the enzyme activity over a broader pH range.
Again this result could probably be attributed to the stabilization
within the sol–gel matrix. The optimum pH of the immobilized
enzyme was found to be similar to the one of native enzyme (pH
Acknowledgments
7
.0).
Finally, the ␻-TA 117 and Vf-TA immobilized with Celite 545
Financial support by the Österreichische Forschungsförderungs-
gesellschaft (FFG) and the Province of Styria are gratefully
acknowledged.
were used in a preparative kinetic resolution of 100 mg of rac-
amines 1a and 1b leading to enantiomerically pure (S)-enantiomer
with 43% and 41% yield respectively and (R)-1a with 48% yield. In all
cases the ee was >99%.Recently, we have published a deracemiza-
tion concept for prim-amines based on a two-step one-pot process
which consists of (i) a kinetic resolution and (ii) a stereoselective
amination (Scheme 2) [46]. Substantial limitation of the presented
method was a possible interference of the employed (R)- and (S)-
References
[
1] A.E. Braunstein, E.E. Snell, Transaminases, Wiley, New York, 1985, pp. 2–35.
[2] M. Höhne, U.T. Bornscheuer, ChemCatChem 1 (2009) 42–51.
[
[
3] N.J. Turner, R. Carr, in: R.N. Patel (Ed.), Biocatalysis in the Pharmaceutical and
Biotechnology Industries, CRC Press, Boca Raton, 2007, pp. 743–755.
4] S. Buchholz, H. Gröger, in: R.N. Patel (Ed.), Biocatalysis in the Pharmaceutical
and Biotechnology Industries, CRC Press, Boca Raton, 2007, pp. 829–847.
␻
-transaminases between the two steps. The problem was solved
by a heat treatment after the first step to inactivate the enzyme
employed in the first step. Thus after the kinetic resolution, the
sample was kept at 75 C for 30 min, before the enzyme required
for the second step was added. This method led to complete denat-
uration and loss of the enzyme which significantly increased the
overall cost of this process due to the high price of transaminases.
To improve this process immobilized (R)-transaminase ␻-TA-
[5] J.D. Stewart, Curr. Opin. Chem. Biol. 5 (2001) 120–129.
6] B.-Y. Hwang, B.-K. Cho, H. Yun, K. Koteshwar, B.-G. Kim, J. Mol. Catal. B: Enzym.
7 (2005) 47–55.
[7] D.J. Ager, T. Li, D.P. Pantaleone, R.F. Senkpeil, P.P. Taylor, I.G. Fotheringham, J.
Mol. Catal. B: Enzym. 11 (2001) 199–205.
[
3
◦
[
8] D. Koszelewski, I. Lavandera, D. Clay, G.M. Guebitz, D. Rozzell, W. Kroutil,
Angew. Chem. Int. Ed. 47 (2008) 9337–9340.
[
9] D. Koszelewski, I. Lavandera, D. Clay, D. Rozzell, W. Kroutil, Adv. Synth. Catal.
350 (2008) 2761–2766.
[
[
10] U. Kaulmann, K. Smithies, M.E.B. Smith, H.C. Hailes, J.M. Ward, Enzyme Microb.
Technol. 41 (2007) 628–637.
11] H. Yun, S. Lim, B.-K. Cho, B.-G. Kim, Appl. Environ. Microbiol. 70 (2004)
2529–2534.
1
17 was employed in the first step of this sequence. The catalyst
preparation could be separated easily by filtration or centrifugation
after the first step – the kinetic resolution. The kinetic resolution

4
4
D. Koszelewski et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 39–44
[
[
12] R.L. Hanson, B.L. Davis, Y. Chen, S.L. Goldberg, W.L. Parker, T.P. Tully, M.A. Mon-
tana, R.N. Patel, Adv. Synth. Catal. 350 (2008) 1367–1375.
13] M. Höhne, S. Kühl, K. Robins, U.T. Bornscheuer, ChemBioChem 9 (2008)
[30] R. Gupta, N.K. Chaudhury, Biosens. Bioelectron. 22 (2007) 2387–2399.
[31] C.J. Brinker, G.W. Scherer, Sol–Gel Science: The Physics and Chemistry of
Sol–Gel Processing, Academic Press, Boston, 1990.
3
63–365.
[32] M.T. Reetz, Adv. Mater. (Weinheim, Germany) 9 (1997) 943–954.
[33] D. Pirozzi, E. Fanelli, A. Aronne, P. Pernice, A. Mingione, J. Mol. Catal. B: Enzym.
59 (2009) 116–120.
[
[
14] M.D. Truppo, N.J. Turner, J.D. Rozzell, Chem. Commun. (2009) 2127–2129.
15] M.D. Truppo, J.D. Rozzell, J.C. Moore, N.J. Turner, Org. Biomol. Chem. 7 (2009)
3
95–398.
[34] M.T. Reetz, P. Tielmann, W. Wiesenhofer, W. Konen, A. Zonta, Adv. Synth. Catal.
345 (2003) 717–728.
[
[
16] K.M. Koeller, C.H. Wong, Nature 409 (2001) 232–240.
17] C.-H. Wong, G.M. Whitesides, Enzymes in Synthethic Organic Chemistry, Perg-
amon Press, Oxford, 1994.
18] G.F. Bickerstaff, Immobilization of Enzymes and Cells: Methods in Biotechnol-
[35] M. Bradford, Anal. Biochem. 72 (1976) 248–254.
[36] J. González-Sabín, V. Gotor, F. Rebolledo, Tetrahedron: Asym. 13 (2002)
1315–1320.
[
ogy, vol. 1, Humana Press, Totowa, 1997.
[37] D. Koszelewski, A. Redzej, R. Ostaszewski, J. Mol. Catal. B: Enzym. 47 (2007)
51–57.
[38] J.E. Clifton, I. Collins, P. Hallett, D. Hartley, L.H.C. Lunts, P.D. Wicks, J. Med. Chem.
25 (1982) 670–679.
[
[
[
19] L. Cao, Carrier-Bound Immobilized Enzymes, Wiley-VCH, Weinheim, 2005.
20] U. Hanefeld, L. Gardossi, E. Magner, Chem. Soc. Rev. 38 (2009) 453–468.
21] K. Drauz, H. Waldman (Eds.), Enzyme Catalysis in Organic Synthesis, vol. 1–3,
2
nd ed., VCH, Weinheim, 2002.
[39] H. Koppe, K. Zeile, W. Kummer, H. Stahle, P. Danneberg, U.S. Patent 3,659,019,
1979.
[
[
[
22] C. Mateo, J.M. Palomo, G. Fernandez-Lorente, J.M. Guisan, R. Fernandez-
Lafuente, Enzyme Microbiol. Technol. 40 (2007) 1451–1463.
23] A. Liese, C. Wandrey, Industrial Biotransformations, Wiley-VCH, Weinheim,
[40] M.M. Cavalluzzi, A. Catalano, C. Bruno, A. Lovece, A. Carocci, F. Corbo, C. Fran-
chini, G. Lentini, V. Tortorella, Tetrahedron: Asym. 18 (2007) 2409–2417.
[41] E. Kalso, M.R. Tramer, H.J. McQuay, R.A. Moore, Eur. J. Pain 2 (1998) 3–14.
[42] S.-F. Chang, S.-W. Chang, Y.-H. Yen, C.-J. Shieh, Appl. Clay Sci. 37 (2007) 67–
73.
2
000.
24] U.T. Bornscheuer, in: H.J. Rehm, G. Reed, A. Pühler, P.J.W. Stadler, D.R. Kelly
Eds.), Industrial Biotransformations. Biotechnology–Series, vol. 8b, Wiley-
(
VCH, Weinheim, 2000, pp. 277–294.
25] S.-S. Yi, C. Lee, J. Kim, D. Kyung, B.-G. Kim, Y.-S. Lee, Process Biochem. 42 (2007)
[43] H.-X. Wang, H. Wu, C.-T. Ho, X.-C. Weng, Food Chem. 97 (2006) 661–665.
[44] W.A.M. Alloue, J. Destain, T.E. Medjoub, H. Ghalfi, P. Kabran, P. Thonart, Appl.
Biochem. Biotechnol. 150 (2008) 51–61.
[
8
95–898.
[
[
[
[
26] A.C. Pierre, Biocatal. Biotransform. 22 (2004) 145–170.
[45] C.-S. Chen, Y. Fujimoto, G. Girdaukas, C.J. Sih, J. Am. Chem. Soc. 104 (1982)
7294–7299.
[46] D. Koszelewski, D. Clay, J.D. Rozzell, W. Kroutil, Eur. J. Org. Chem. (2009)
2289–2292.
27] D. Avnir, T. Coradin, O. Lev, J. Livage, J. Mater. Chem. 16 (2006) 1013–1030.
28] D. Avnir, S. Braun, O. Lev, M. Ottolenghi, Chem. Mater. 6 (1994) 1605–1614.
29] I. Gill, Chem. Mater. 13 (2001) 3404–3421.