ω-Transaminase-catalyzed asymmetric synthesis of unnatural amino acids using isopropylamine as an amino donor

Article abstract of DOI:10.1039/c3ob40495a

Isopropylamine is an ideal amino donor for reductive amination of carbonyl compounds by ω-transaminase (ω-TA) owing to its cheapness and high volatility of a ketone product. Here we developed asymmetric synthesis of unnatural amino acids via ω-TA-catalyzed amino group transfer between α-keto acids and isopropylamine.

Full text of DOI:10.1039/c3ob40495a

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
ω-Transaminase-catalyzed asymmetric synthesis of
unnatural amino acids using isopropylamine as an
amino donor†
Cite this: Org. Biomol. Chem., 2013, 11,
6929
Eul-Soo Park, Joo-Young Dong and Jong-Shik Shin*
Received 11th March 2013,
Accepted 2nd July 2013
Isopropylamine is an ideal amino donor for reductive amination of carbonyl compounds by ω-transami-
nase (ω-TA) owing to its cheapness and high volatility of a ketone product. Here we developed asymmetric
synthesis of unnatural amino acids via ω-TA-catalyzed amino group transfer between α-keto acids and
isopropylamine.
DOI: 10.1039/c3ob40495a
www.rsc.org/obc
L-glutamic acid was used as an amino donor for a primary
transaminase and the reaction equilibrium was shifted by
Introduction
Nonproteinogenic amino acids, including L-α-amino acids carry- simultaneously running an ornithine transaminase reaction
ing an unnatural side chain and D-α-amino acids, are gaining which led to spontaneous cyclization of a resulting keto acid
ever-growing importance as chiral building blocks for various product.14
pharmaceuticals, because they provide enhanced enzymatic
Contrary to the equilibrium constant (Keq) close to unity for
and pharmacodynamic stability as well as structural diversity the reactions catalyzed by α-transaminase (α-TA) such as TTA
that cannot be mimicked by natural counterparts.1–4 For and DATA,12,16 transaminations between primary amines and
example, L-homoalanine (L-1a) is pharmaceutically important, α-keto acids catalyzed by ω-transaminase (ω-TA) are energeti-
serving as a key intermediate for production of levetiracetam cally favorable.17 For example, Keq of the transamination
and brivaracetam (antiepileptic drugs),5 and ethambutol (an between 2b and α-methylbenzylamine (3a) was reported to
antituberculosis drug).6
be 1130.17 Therefore, the transaminations using primary
A
number of chemocatalytic7–10 and biocatalytic2,3,11 amines as an amino donor permit thermodynamically un-
methods have been developed to enable cost-effective syn- restricted asymmetric amination of keto acids (Scheme 1).18
thesis of the unnatural amino acids. Industrial production of
L-1a has been established by reductive amination of 2-oxobuty-
ric acid (2a) using tyrosine transaminase (TTA).2,12–14 To over-
come the unfavorable reaction equilibrium, L-aspartate was
employed as an amino donor because the resulting keto acid
product (i.e. oxaloacetic acid) underwent spontaneous decar-
boxylation to pyruvic acid (2b). However, this reaction led to
product contamination with L-alanine (L-1b) because TTA is
reactive toward 2b as well as 2a.2,12–14 To reduce the side
product formation, the TTA reaction was coupled with aceto-
lactate synthase capable of converting 2b to acetolactic acid.15
The same strategy was applied to production of D-1a by repla-
cing TTA with D-amino acid transaminase (DATA).2,12 As
an alternative to addressing the unfavorable equilibrium,
Department of Biotechnology, Yonsei University, 262 Seongsanno, Seodaemun-Gu,
120-749 Seoul, South Korea. E-mail: enzymo@yonsei.ac.kr; Fax: +82-2-362-7265;
Tel: +82-2-2123-5884
†Electronic supplementary information (ESI) available: Materials, cloning of
ARmutTA gene, purification of enzymes, HPLC analysis and structural characteri-
zation. See DOI: 10.1039/c3ob40495a
Scheme 1 Asymmetric reductive amination of α-keto acids using primary
amines as an amino donor to produce enantiopure amino acids.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 6929–6933 | 6929

View Article Online
Paper
Organic & Biomolecular Chemistry
We previously demonstrated asymmetric synthesis of L-1a two (S)-selective ω-TAs showed reactivities of 3b higher than
using benzylamine (3b) as an amino donor.19 However, a (S)-3a whereas the two (R)-selective ω-TAs showed reactivities of
serious drawback of this method was a high amino acceptor 3b much lower than (R)-3a. PDTA and ARTA showed modest
reactivity of the resulting deamination product, i.e. benz- reactivities toward alkylamines (i.e. 3c–d). In contrast, OATA
aldehyde (4b), which engendered a drastic reduction in the net showed substantial reactivities toward the alkylamines and the
reaction rate because of a high reverse reaction rate. For reactivity of 3c was 43% relative to (S)-3a. As expected, ARmutTA
example, (S)-selective ω-TA from Paracoccus denitrificans displayed higher reactivity toward 3c than its parental ω-TA
showed 84% reactivity of 4b relative to 2b (i.e. a typical amino (i.e. ARTA) did. Based on the results, OATA and ARmutTA turned
acceptor for most ω-TAs).19 Therefore, a cheap amino donor, out to be suitable for asymmetric synthesis of L- and D-amino
whose deamination product is either non-reactive or easily acids, respectively, using 3c as an effective amino donor.
removable, is highly demanded to implement preparative
asymmetric synthesis of unnatural amino acids using ω-TAs.
To ensure whether 3c is a better amino donor than 3b, we
compared the reaction progress of the amination of 2a cata-
In this regard, isopropylamine (3c) is an ideal amino donor lyzed by OATA using the two amines as amino donors (Fig. 1).
for ω-TA because its deamination product, i.e. acetone (4c), is a Despite the 6-fold higher reactivity of 3b than 3c, the two reac-
non-reactive amino acceptor and easy to remove owing to high tions showed similar conversions up to 30 min and then the
volatility.20 In addition, 3c is much cheaper than 3b. These reaction with 3b became even slower than the one with 3c.
beneficial features of 3c as an amino donor have recently After a 4 h reaction, conversion reached >99% with 3c whereas
prompted intense research efforts to asymmetric transfer of an 3b permitted 86% conversion. This result clearly corroborates
amino group from 3c to prochiral ketones for preparation of that product inhibition caused by a high amino acceptor reac-
chiral amines.20–23 However, to the best of our knowledge, 3c tivity of 4b (35% relative to 2b)28 is highly detrimental to
has not yet been exploited as an amino donor for asymmetric efficient amination of keto acids.
amination of α-keto acids although this approach is promising
It is known that most ω-TAs possess severe steric con-
for cost-effective synthesis of unnatural amino acids including straints in the small binding pocket which precludes entry of a
L-1a and D-1a. In this study, we sought to develop ω-TA-cata- substituent larger than an ethyl group.20,24,27 To clarify the
lyzed synthesis of enantiopure unnatural amino acids from range of amino acids producible by OATA and ARmutTA, we
α-keto acids using 3c as an amino donor.
examined the substrate specificity of the two ω-TAs toward
α-keto acids (Table 2). As expected, OATA showed substantial
reactivities only toward α-keto acids carrying substituents no
larger than an ethyl group, i.e. five α-keto acids (2a–e) among
the sixteen tested. It was intriguing that the steric constraint of
a parental enzyme toward α-keto acids was found to be con-
served in ARmutTA in spite of the multiple mutations intro-
duced to accommodate bulky substituents of arylalkyl
ketones.20 However, the enlarged binding pocket of ARmutTA
permitted improved reactivities toward α-keto acids carrying a
linear alkyl substituent (i.e. 2j, 2m–n).29 It is notable that 2n
showed a reactivity even higher than 2j and 2m.
Results and discussion
Although 3c is an ideal amino donor, most ω-TAs do not
exhibit substantial reactivity toward 3c.24,25 For example, two
typical ω-TAs, i.e. (S)-selective ω-TA from Vibrio fluvialis JS1724
and (R)-selective ω-TA from Arthrobacter sp. (ARTA),25 were
known to be non-reactive toward 3c. Therefore, we set out to
search for ω-TAs that could utilize 3c as an effective amino
donor. To this end, we examined the amino donor reactivities
of four primary amines (two arylalkylamines and two alkyl-
amines) toward two (S)-selective ω-TAs cloned in our group from
P. denitrificans (PDTA)19 and Ochrobactrum anthropi (OATA),26,27
and (R)-selective ARTA25 and its variant (ARmutTA)20 engineered
for reductive amination of bulky ketones by 3c (Table 1). The
For asymmetric synthesis of unnatural amino acids, we
decided to move forward with α-keto acids showing relative
Table 1 Amino donor reactivities of primary amines toward ω-TAs
Relative reactivitya (%)
Amines
PDTA
OATA
ARTA
ARmutTA
3ab
3b
100
128
7
100
259
43
100
3
2
100
28
8
3c
3dc
n.r.d
23
2
14
a Relative reactivity represents the initial reaction rate (i.e. conversion
<15%) normalized by that of (S)- or (R)-3a. Reaction conditions were
amine (20 mM except 3d) and 2b (20 mM). b Enantiopure (S)- and (R)-
3a (20 mM) were used for (S)- and (R)-selective ω-TAs, respectively.
c Rac-3d (40 mM) was used for the reactivity measurements. d n.r.:
not reactive (i.e. relative reactivity <1%).
Fig. 1 Time-course monitoring of the reductive amination of 2a using 3b or 3c
as an amino donor. Reaction conditions were 2a (100 mM), amine (150 mM)
and OATA (5 U mL−1).
6930 | Org. Biomol. Chem., 2013, 11, 6929–6933
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 2 Amino acceptor specificities of OATA and ARmutTA toward α-keto acids
Relative reactivitya (%)
α-Keto acids
OATA
ARmutTA
2a
2b
2c
2d
2e
2f
2g
2h
2i
2j
2k
2l
2m
2n
2o
2p
13
33
100
36
5
6
n.r.
n.r.
n.r.
n.r.
2
n.r.
n.r.
2
100
135
43
Scheme 2 Production of both enantiomers of 1a from L-5 and 3c.
14
n.r.b
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
n.r.
8
n.r.
n.r.
a Relative reactivity represents the initial reaction rate normalized by
that of 2b. Reaction conditions: α-keto acid (20 mM) and (S)- or (R)-3a
(20 mM). b n.r.: not reactive.
Fig. 2 Asymmetric synthesis of L- and D-1a from L-5 and 3c. Reaction conditions
were L-5 (100 mM), 3c (200 mM), TD (5 U mL−1) and ω-TA (20 U mL−1) in a 1 mL
reaction mixture.
reactivities higher than 5%, i.e. 2a and 2d with OATA, and
2a–b, 2d–e and 2n with ARmutTA (Table 3). Reductive amination
of the α-keto acids (100 mM) by 3c (1.5 molar equiv.) led to
efficient synthesis of unnatural amino acids with >98% conver- Abarelix (antineoplastic)30 and D-serine (D-1e) is a precursor of
sion and excellent enantiopurity (>99.9%) within 2 h except for D-cycloserine (cognition enhancer).31 Besides, D-fluoroalanine
2d. Compared with the other α-keto acids, amination of 2d (D-1d) is recognized as an antibiotic owing to its ability to in-
showed exceptionally slow reaction progress with both ω-TAs. activate bacterial alanine racemase.32
We suspected that either 2d or its amination product (i.e. 1d)
Among the α-keto acids used in Table 3, 2a is readily acces-
might induce irreversible enzyme inactivation, because we sible from a cheap precursor, i.e. L-threonine (L-5), using threo-
observed formation of enzyme aggregates during both ami- nine deaminase (TD) as previously demonstrated.19 L-5 is one
nation reactions. Indeed, it turned out that 1d induced such of the cheap natural amino acids mass-produced by microbial
detrimental enzyme inactivation, because the enzyme precipi- fermentation.33 Therefore, we investigated whether both enan-
tate was not observed in the incubation solution containing tiomers of 1a could be prepared from L-5 and 3c by coupling
either 2d or 3c. Therefore, for efficient amination of 2d, a the ω-TA reactions with TD cloned from Escherichia coli
method to enhance enzyme stability such as enzyme immobi- (Scheme 2). In this approach, the carbon skeleton and the
lization needs to be employed.
amino group of L- and D-1a come from L-5 and 3c, respectively.
The unnatural amino acids listed in Table 3 are of indus- The TD/OATA coupled reaction using L-5 (100 mM) and 3c (2
trial importance. For example, in addition to L-1a of pharma- equiv.) was completed at 60 min, leading to 99% conversion
ceutical importance,5 D-alanine (D-1b) is a building block of yield and >99.9% ee of produced L-1a (Fig. 2). To produce D-1a,
OATA was substituted by ARmutTA. Under the same substrate
conditions, the TD/ARmutTA reaction resulted in production of
Table 3 Asymmetric synthesis of unnatural amino acids from keto acids using
3c as an amino donora
D-1a with 98% conversion yield and >99.9% ee at 90 min. The
longer reaction time required to complete the TD/ARmutTA
reaction presumably resulted from the lower relative reactivity
of 3c toward ARmutTA than that toward OATA. These results
demonstrate that both enantiomers of 1a of pharmaceutical
importance can be readily prepared from cheap substrates.
We also carried out preparative-scale synthesis of L-1a and
D-1a via the TD/ω-TA strategy in a 50 mL reaction mixture
charged with L-5 (1.79 g, 15 mmol) and 3c (1.94 mL,
22.5 mmol). Conversions over 99% were attained within 12
and 24 h for the TD/OATA and TD/ARmutTA reactions, res-
pectively. Purification and structural characterization of the
desired amino acid products were performed, leading to
Reaction
time (h)
Conversionb
(%)
Product
(% ee)
Substrate
ω-TA
2a
2d
2a
2b
2d
2e
2n
OATA
OATA
ARmutTA
ARmutTA
ARmutTA
ARmutTA
ARmutTA
0.5
10
0.5
0.2
10
2
0.5
99
79
99
98
86
99
99
L-1a (>99.9)
L-1d (>99.9)
D-1a (>99.9)
D-1b (>99.9)
D-1d (>99.9)
D-1e (>99.9)
D-1n (>99.9)
a Reaction conditions: 0.5 mL reaction mixture containing keto acid
(100 mM), 3c (150 mM) and ω-TA (50 U mL−1). b Conversions were
based on consumption of the keto acid substrate.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 6929–6933 | 6931

View Article Online
Paper
Organic & Biomolecular Chemistry
recovery of pure L-1a (1.20 g, 77.6% isolation yield, >99.9% ee) 0.7 and 0.2 U mL−1 for PDTA, OATA and (R)-selective ω-TAs,
and D-1a (1.21 g, 78.2% yield, >99.9% ee).
respectively. Reactions were allowed for 10 min and the initial
rates (i.e. conversion >15%) were measured by analyzing pro-
duced L- or D-1b. Amino acceptor reactivities were measured
with (S)- or (R)-3a (20 mM), α-keto acid (20 mM) and ω-TA (0.1
and 0.05 U mL−1 for OATA and ARmutTA, respectively). The
acetophenone produced was analyzed for the initial rate
measurements.
Conclusions
In this work, we demonstrated efficient biocatalytic asym-
metric synthesis of unnatural amino acids from corresponding
keto acids using 3c as a cosubstrate. Unlike the α-TA reactions,
the ω-TA approach using the cheap achiral amino donor is
thermodynamically favorable. This eliminates the need for
additional enzymes to shift unfavorable equilibrium or to
prevent side product accumulation, which has been compul-
sory in the previous α-TA-based approaches.2,12–15 Moreover,
when it comes to the synthesis of D-amino acids, a serious
drawback of the α-TA approach is the requirement of an
expensive D-amino acid as an amino donor (e.g. D-aspartic acid
for DATA reactions).2,12 However, to expand our approach to
asymmetric synthesis of structurally diverse unnatural amino
acids, the steric constraint in the substrate binding pocket of
ω-TAs should be addressed. Although more than twenty ω-TAs
have been reported, no naturally occurring enzyme capable of
accepting α-keto acids carrying substituents bulkier than an
ethyl group has been identified.34,35 However, as demonstrated
by ARmutTA capable of accepting bulky linear aliphatic groups
of α-keto acids, directed evolution combined with in silico
modeling20 may create an ω-TA variant of which the substrate
binding pocket is redesigned in a way to accept diverse bulky
substituents.
Recently, Seo et al. reported deracemization of rac-1a into
L-1a using ω-TA from V. fluvialis JS17 coupled with D-amino acid
oxidase.36 Because they used 3b as an amino donor for ω-TA,
the severe product inhibition by 4b had to be mitigated using
a biphasic reaction system to extract the inhibitory 4b.
However, the biphasic system is often detrimental to enzyme
stability. Moreover, hydrogen peroxide generated from the
oxidase reaction usually leads to severe enzyme inactivation
without the aid of catalase. In contrast, the TD/ω-TA approach
using L-5 and 3c lacks such inhibitory products and affords
production of both enantiomers of 1a.
Asymmetric synthesis of unnatural α-amino acids
Time-course monitoring of the reductive amination of 2a
using 3b or 3c as an amino donor was performed at 100 mM
2a, 150 mM amine, 0.5 mM PLP and 5 U mL−1 OATA in 50 mM
phosphate buffer (pH 7). The reaction volume was 500 μL and
the reaction mixture was incubated at 37 °C. Aliquots of the
reaction mixture (10 μL) were taken at predetermined reaction
times and mixed with 60 μL acetonitrile to stop the reaction.
The reaction mixtures were subjected to HPLC analysis of 2a to
measure conversions.
Asymmetric synthesis of unnatural amino acids (1 mL reac-
tion volume) was carried out with 100 mM α-keto acid (2a–b,
2d–e and 2n), 150 mM 3c, 0.5 mM PLP and 50 U mL−1 ω-TA in
50 mM phosphate buffer (pH 7) at 37 °C. Keto acids and
amino acids were analyzed for measurements of conversion
and enantiomeric excess, respectively.
TD/ω-TA coupled reactions
The coupled reaction for synthesis of L- and D-1a (1 mL reac-
tion volume) was performed with 100 mM L-5, 200 mM 3c,
0.5 mM PLP, 5 U mL−1 TD and 20 U mL−1 ω-TA in 50 mM
phosphate buffer (pH 7) at 37 °C. OATA and ARmutTA were
employed for preparation of L- and D-1a, respectively. 1a was
analyzed for measurements of conversion yield and enantio-
meric excess.
Preparative-scale synthesis of L- and D-1a was performed in
a 50 mL reaction mixture containing 300 mM L-5, 450 mM 3c,
0.5 mM PLP, 5 U mL−1 TD and 10 U mL−1 ω-TA (OATA or
ARmutTA) under magnetic stirring at 37 °C. When the conver-
sion exceeded 99%, the reaction mixture was subjected to
product isolation.
Purification and structural characterization of L- and D-1a
Experimental
The pH of the reaction mixture was adjusted to 1.0 by adding
5 N HCl for protein precipitation. The reaction mixture was fil-
tered through a glass-fritted filter funnel to remove protein
precipitate. The filtrate solution was loaded on a glass column
packed with Dowex 50WX8 cation-exchange resin (40 g), fol-
lowed by washing with 0.1 N HCl (120 mL) and water
(120 mL), and then elution with 145 mL of 10% ammonia
solution. The elution fractions were pooled and evaporated at
30 °C and 0.1 bar. The resulting solids were washed with EtOH
Enzyme assay
Typical enzyme assays were carried out at 37 °C and 50 mM
phosphate buffer (pH 7). One unit of ω-TA activity was defined
as the enzyme amount that catalyzed formation of 1 µmol of
acetophenone in 1 min at 20 mM 2b and 20 mM (S)- or (R)-3a.
One unit of TD was defined as the enzyme amount producing
1 µmol of 2a in 1 min at 50 mM L-5.
Measurement of amino donor and acceptor reactivity
(50 mL) and then oven-dried overnight. The purified L- and
1
Amino donor reactivities were measured with 2b (20 mM) and D-1a were structurally characterized by H NMR, 13C NMR, IR,
amine (20 mM except 3d) in 50 mM phosphate buffer (pH 7). elemental analysis and LC/MS, leading to confirmation of the
The enzyme concentrations in the reaction mixtures were 0.5, recovery of the desired pure products (see ESI†).
6932 | Org. Biomol. Chem., 2013, 11, 6929–6933
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
18 E. García-Urdiales, I. Alfonso and V. Gotor, Chem. Rev.,
2011, 111, PR110–PR180.
Acknowledgements
This work was supported by the Advanced Biomass R&D 19 E. Park, M. Kim and J. S. Shin, Adv. Synth. Catal., 2010,
Center (ABC-2010-0029737) and the Basic Science Research 352, 3391–3398.
Program (2010-0024448) through the National Research Foun- 20 C. K. Savile, J. M. Janey, E. C. Mundorff, J. C. Moore,
dation of Korea funded by the Ministry of Education, Science
and Technology.
S. Tam, W. R. Jarvis, J. C. Colbeck, A. Krebber, F. J. Fleitz,
J. Brands, P. N. Devine, G. W. Huisman and G. J. Hughes,
Science, 2010, 329, 305–309.
21 K. E. Cassimjee, C. Branneby, V. Abedi, A. Wells and
P. Berglund, Chem. Commun., 2010, 46, 5569–5571.
22 M. E. B. Smith, B. H. Chen, E. G. Hibbert, U. Kaulmann,
K. Smithies, J. L. Galman, F. Baganz, P. A. Dalby, H. C. Hailes,
G. J. Lye, J. M. Ward, J. M. Woodley and M. Micheletti, Org.
Process Res. Dev., 2010, 14, 99–107.
23 F. G. Mutti, C. S. Fuchs, D. Pressnitz, J. H. Sattler and
W. Kroutil, Adv. Synth. Catal., 2011, 353, 3227–3233.
24 J. S. Shin and B. G. Kim, J. Org. Chem., 2002, 67,
2848–2853.
Notes and references
1 J. S. Ma, Chimica Oggi, 2003, 21, 65–68.
2 M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Keßeler,
R. Stürmer and T. Zelinski, Angew. Chem., Int. Ed., 2004, 43,
788–824.
3 A. S. Bommarius, M. Schwarm and K. Drauz, Chimia, 2001,
55, 50–59.
4 S. Servi, D. Tessaro and G. Pedrocchi-Fantoni, Coord. Chem.
Rev., 2008, 252, 715–726.
5 M. Sasa, J. Pharm. Sci., 2006, 100, 487–494.
25 A. Iwasaki, K. Matsumoto, J. Hasegawa and Y. Yasohara,
Appl. Microbiol. Biotechnol., 2012, 93, 1563–1573.
6 W. A. Nugent and J. E. Feaster, Synth. Commun., 1998, 28, 26 M. S. Malik, E. S. Park and J. S. Shin, Green Chem., 2012,
1617–1623.
14, 2137–2140.
7 L. D. Tran and O. Daugulis, Angew. Chem., Int. Ed., 2012, 27 E. S. Park, M. Kim and J. S. Shin, Appl. Microbiol. Bio-
51, 5188–5191. technol., 2012, 93, 2425–2435.
8 S. J. Zuend, M. P. Coughlin, M. P. Lalonde and 28 Relative reactivity was calculated by comparing the initial
E. N. Jacobsen, Nature, 2009, 461, 968–970.
9 S. Bera, D. Mondal, M. Singh and R. K. Kale, Tetrahedron,
2013, 69, 969–1011.
reaction rates measured at 100 mM rac-3a and 50 mM 2b
or 4b using OATA.
29 ARTA showed less than 1% reactivities of 2j and 2m–n
relative to 2b.
10 J. A. Ma, Angew. Chem., Int. Ed., 2003, 42, 4290–4299.
11 A. S. Bommarius, M. Schwarm and K. Drauz, J. Mol. Catal. 30 K. Bellmann-Sickert and A. G. Beck-Sickinger, Trends
B: Enzym., 1998, 5, 1–11.
Pharmacol. Sci., 2010, 31, 434–441.
12 P. P. Taylor, D. P. Pantaleone, R. F. Senkpeil and 31 A. Pittaluga, R. Pattarini and M. Raiteri, Eur. J. Pharmacol.,
I. G. Fotheringham, Trends Biotechnol., 1998, 16, 412–418. 1995, 272, 203–209.
13 D. J. Ager, T. Li, D. P. Pantaleone, R. F. Senkpeil, 32 L. P. B. Gonçalves, O. A. C. Antunes and E. G. Oestreicher,
P. P. Taylor and I. G. Fotheringham, J. Mol. Catal. B:
Enzym., 2001, 11, 199–205.
14 T. Li, A. B. Kootstra and I. G. Fotheringham, Org. Process
Res. Dev., 2002, 6, 533–538.
15 I. G. Fotheringham, N. Grinter, D. P. Pantaleone, R. F. Senkpeil
and P. P. Taylor, Bioorg. Med. Chem., 1999, 7, 2209–2213.
16 D. J. Ager and I. G. Fotheringham, Curr. Opin. Drug Dis-
covery Dev., 2001, 4, 800–807.
17 J. S. Shin and B. G. Kim, Biotechnol. Bioeng., 1998, 60,
534–540.
Org. Process Res. Dev., 2006, 10, 673–677.
33 W. Leuchtenberger, K. Huthmacher and K. Drauz, Appl.
Microbiol. Biotechnol., 2005, 69, 1–8.
34 D. Koszelewski, K. Tauber, K. Faber and W. Kroutil, Trends
Biotechnol., 2010, 28, 324–332.
35 M. S. Malik, E. S. Park and J. S. Shin, Appl. Microbiol. Bio-
technol., 2012, 94, 1163–1171.
36 Y. M. Seo, S. Mathew, H. S. Bea, Y. H. Khang, S. H. Lee,
B. G. Kim and H. Yun, Org. Biomol. Chem., 2012, 10,
2482–2485.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 6929–6933 | 6933