Asymmetric synthesis of l-tert-leucine and l-3-hydroxyadamantylglycine using branched chain aminotransferase

Article abstract of DOI:10.1016/j.molcatb.2010.05.014

l-Tert-leucine (Tle) and l-3-hydroxyadamantylglycine of (HAG) are important intermediates for a variety of pharmaceutical classes. They were asymmetrically produced from corresponding keto acids using branched-chain aminotransferase (BCAT) with l-glutamate (Glu) as an amino donor. For the production of l-Tle and l-HAG, BCAT from Enterobacter sp. TL3 (BCATen) and BCAT from Escherichia coli K12 (ilvE, newly named as BCATes) were used, respectively. In our current study, we characterized the basic properties of BCATen and BCATes such as substrate specificity, enantioselectivity, and kinetic parameters. The activities of BCATen and BCATes were inhibited severely by α-ketoglutarate which is a deaminated product of l-Glu. In the presence of 10 mM α-ketoglutarate, both enzymes activities were reduced up to 80%. In order to overcome product inhibition by α-ketoglutarate and the problem of equilibrium of the transamination reaction, coupling reactions were carried out with l-glutamate dehydrogenase (GDH)/formate dehydrogenase (FDH) and AspAT. The coupling reaction dramatically increased the yields of both target compounds. 135 mM of l-Tle (>99% ee) was produced from 150 mM corresponding keto acid in BCATen/GDH/FDH coupling reaction with 90% conversion. In addition, 90.5 mM l-HAG (>99% ee) was produced from 100 mM corresponding keto acid in BCATes/AspAT coupling reaction using recombinant whole-cells.

Full text of DOI:10.1016/j.molcatb.2010.05.014

Journal of Molecular Catalysis B: Enzymatic 66 (2010) 228–233
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Asymmetric synthesis of l-tert-leucine and l-3-hydroxyadamantylglycine using
branched chain aminotransferase
Eun young Honga,1, Minho Chaa,1, Hyungdon Yun , Byung-Gee Kima,∗
b
a
School of Chemical and Biological Engineering, Seoul National University, Seoul #151-742, Republic of Korea
School of Biotechnology, Yeungnam University, Gyeongsan, Gyeongbuk # 712-749, Republic of Korea
b
a r t i c l e i n f o
a b s t r a c t
Article history:
l-Tert-leucine (Tle) and l-3-hydroxyadamantylglycine of (HAG) are important intermediates for a vari-
ety of pharmaceutical classes. They were asymmetrically produced from corresponding keto acids using
branched-chain aminotransferase (BCAT) with l-glutamate (Glu) as an amino donor. For the production
of l-Tle and l-HAG, BCAT from Enterobacter sp. TL3 (BCATen) and BCAT from Escherichia coli K12 (ilvE,
newly named as BCATes) were used, respectively. In our current study, we characterized the basic prop-
erties of BCATen and BCATes such as substrate specificity, enantioselectivity, and kinetic parameters.
The activities of BCATen and BCATes were inhibited severely by ␣-ketoglutarate which is a deaminated
product of l-Glu. In the presence of 10 mM ␣-ketoglutarate, both enzymes activities were reduced up
to 80%. In order to overcome product inhibition by ␣-ketoglutarate and the problem of equilibrium
of the transamination reaction, coupling reactions were carried out with l-glutamate dehydrogenase
Received 14 January 2010
Received in revised form 27 May 2010
Accepted 28 May 2010
Available online 4 June 2010
Keywords:
l-Tert-leucine
l-3-Hydroxyadamantylglycine
Branched-chain aminotransferase (BCAT)
Nonproteinogenic amino acid
(GDH)/formate dehydrogenase (FDH) and AspAT. The coupling reaction dramatically increased the yields
of both target compounds. 135 mM of l-Tle (>99% ee) was produced from 150 mM corresponding keto acid
in BCATen/GDH/FDH coupling reaction with 90% conversion. In addition, 90.5 mM l-HAG (>99% ee) was
produced from 100 mM corresponding keto acid in BCATes/AspAT coupling reaction using recombinant
whole-cells.
©
2010 Elsevier B.V. All rights reserved.
1
. Introduction
be through degradation of GIP and the degradation of GLP-1. Thus,
Saxagliptin is now under development by Bristol-Myers Squibb for
treatment of type 2 diabetes [5,6].
Nonproteinogenic amino acids are important intermediates for
a variety of pharmaceutical classes, most notably in peptidomimet-
ics with antiviral and ontological applications [1]. Because of the
great usage and versatility of these compounds, many optical active
nonproteinogenic amino acids have been the targets for asym-
metric synthesis and chemical or enzymatic resolution [2]. Among
them, l-tert-leucine (Tle) and l-3-hydroxyadamantylglycine (2-(3-
hydroxy-1-adamantyl)-(2S)-amino ethanoic acid: HAG) are very
important precursors for the synthesis of pharmaceuticals (Fig. 1).
l-Tle itself is a component of several pharmaceutical develop-
ment projects as tumor-fighting agents or HIV protease inhibitors
against several diseases such as tumors, rheumatic arthritis and
AIDS [3]. l-HAG can be used as a precursor for the synthesis of
Saxagliptin which is a dipeptidyl peptidase IV inhibitor [4]. Dipep-
tidyl peptidase-IV’s role in blood glucose regulation is thought to
There are numerous methods for the synthesis of l-Tle includ-
ing chemical resolution, stereoselective synthesis and enzymatic
approaches [7–10]. Kragl el al. [7] reported continuous production
of l-Tle in series of two enzyme–membrane reactors by reduc-
tive amination of trimethylpyruvate with leucine dehydrogenase.
Recently, Penicillin G acylase (PGA) from Kluyvera citrophila immo-
bilized on Amberzyml was used for enantioselective hydrolysis of
N-phenylacetylated-dl-tert-leucine to produce l-Tle [8]. Among
them, the Degussa AG (now Evonik AG)-mediated process, uti-
lizing leucine dehydrogenase-catalyzed reductive amination of
trimethylpyruvate, has been successfully operated on a large scale
[8]. In case of synthesizing l-HAG compound, it was originally pre-
pared using an asymmetric Stecker amino acid synthesis [4]. After
this research, approaches to enzymatically synthesize Saxagliptin’s
key intermediate l-HAG have been studied using modified form
of a recombinant phenylalanine dehydrogenase cloned from Ther-
moactinomyces intermedius and expressed in Pichia pastoris or
Escherchia coli [11].
∗ Corresponding author at: Laboratory of Molecular Biotechnology and Bioma-
terials, School of Chemical and Biological Engineering, Seoul National University,
Seoul, Republic of Korea. Tel.: +82 2 880 6774; fax: +82 2 874 1206.
E-mail address: byungkim@snu.ac.kr (B.-G. Kim).
Transaminases (TAs) play important roles in amino acid
metabolism and are ubiquitous in microorganisms and eukaryotic
cells. TAs have been studied extensively, due to its potential to
1
These authors equally contributed to this work.
1
381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.05.014

E.y. Hong et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 228–233
229
1
mL of 50 mM phosphate buffer. After centrifugation, cells were
added to 0.5 mL of reaction solution composed of 50 mM phosphate
buffer (pH 8.0), 10 mM trimethylpyruvate and 50 mM l-glutamate
◦
(
Glu) at 37 C, and the specific activity and enantioselectivity were
measured.
2.3. Genetic analysis
Fig. 1. The target unnatural amino acids by trasaminase reaction. (A) Tert-leucine
Tle); (B) l-3-hydroxyadamantylglycine (2-(3-hydroxy-1-adamantyl)-(2S)-amino
(
DNA manipulations, including preparation of plasmids, restric-
ethanoic acid: HAG).
tion enzyme digestion and ligation, transformation of E. coli were
followed the methods by Sambrook, etc. Multiple alignments of
the amino acid sequences among the homologous proteins were
performed with CLUSTAL multiple sequence alignment program.
produce chiral amino acids and amines in biosystems, Till now
various examples of using the transaminases for the produc-
tion of natural and non-natural d- and l-amino acids as well as
chiral amines have been published [12–15]. Despite their many
advantages such as broad substrate specificity and high enantios-
electivity, their industrial use have been limited mainly due to a
low equilibrium constant of the TA reaction and/or severe inhi-
bition by the product [16]. To over-come these problems, many
attempts have been done to shift the equilibrium by stripping out
or removing the coproduct [17,18].
2.4. Preparation of enzymes
To express BCATen without excessive flanking parts, the coding
region of BCATen was amplified from genomic DNA of Enterobacter
ꢀ
sp. by PCR using P1 (5 -ATCATGGAATTCATGACCACGAAGAAAGCT-
ꢀ
ꢀ
ꢀ
3
) and P2 (5 -AAAAAACTCGAGTTATTGATTAACTTGATCTAACCA-3 )
primers. To express C-terminal fusion BCATen with a His-tagged
polypeptide, the coding region of BCATen was amplified by
Among them, the most widely studied transaminase is
␣-transaminase, such as aspartate transaminase, tyrosine transam-
ꢀ
ꢀ
PCR using P3 (5 -ATATATGGATCCATGACCACGAAGAAAGCT-3 ) and
inase, branched-chain amino acid transaminase (BCAT), and
valine–alanine transaminase of E. coli K-12 [19–22]. Although
branched-chain amino acid transaminase shows reactivity towards
various aliphatic amino acids such as leucine and iso-leucine, few
attempts have been undertaken to produce l-Tle and HAG from
the corresponding keto acids using transaminase. In order to pro-
duce l-Tle, we isolated Enterobacter sp. TL3 showing BCAT activity
toward l-Tle from soil samples using enrichment culture method
ꢀ
ꢀ
P4 (5 -ATATATCTCGAGTTGATTAACTTGATC-3 ) primers. The PCR
products were digested with restriction enzymes (EcoRI/XhoI,
or BamHI/XhoI site), and inserted into the vector pET24ma
[
24]. Since DNA sequence of BCATes is highly homologues
to that of BCATen, the same primer sets are used to con-
struct plasmids (pET24ma) for the expression of BCATes from E.
coli K-12 (accession number U00096). AspAT gene was ampli-
ꢀ
ꢀ
fied with P5 (5 -ATATATGGATCCATGTTTGAGAACATTACC-3 ) and
[
23]. The present study illustrates the isolation of Enterobacter sp.
ꢀ
ꢀ
P6 (5 -ATATATCTCGAGTTACAGCACTGCCACAAT-3 ) primers using
genomic DNA of E. coli K12 as template, and was inserted into
pET23b (+) containing BamHI/XhoI restriction sites.
Recombinant E. coli was grown at 37 C in 1 L LB broth contain-
ing 50 ␮g mL of appropriate antibiotic. When the OD600 of the
TL3, molecular cloning, and expression of the gene encoding the
new aminotransferase (BCATen). Using the cloned BCAT from E. coli
(
BCATes) and BCATen, enantiopure l-HAG and l-Tle were produced
◦
in quantitative yield and high enantiomeric purity (ee > 99%).
−1
culture reached 0.5, 1 mM IPTG was added to the culture broth.
After a 5 h induction, the cells were harvested. The harvested cells
2
. Materials and methods
were suspended in 50 mL of 10 mM Tris–HCl (pH 7.2) containing
2.1. Material
ꢀ
2
0 ␮M pyridoxal 5 -phosphate, 2 mM EDTA, 1 mM PMSF, and 0.01%
(
v/v) 2-mercaptoethanol and ultrasonically disrupted for 25 min
Trimethylpyruvate (TMP) was obtained from the CKD Research
◦
at 4 C. The samples were centrifuged (17,000 × g, 20 min), and
Institute (Cheonan, Korea). Keto acid 2-(3-hydroxy-1-adamantyl)-
-oxoethanoic acid (HAOE) was purchased from Hi@Tech
Chongqing, China). Racemic l-3-hydroxyadamantylglycine
and l-3-hydroxyadamantylglycine was obtained from the
Samchully Pharm. Co. (Gyeonggi-do, Korea). Formate dehy-
drogenase (FDH) and l-glutamate dehydrogenase (GLDH),
the supernatants were dialyzed against 10 mM Tris–HCl buffer
2
(
ꢀ
(pH 7.2) containing 20 ␮M pyridoxal 5 -phosphate, 0.2 mM EDTA,
1
␮M PMSF, and 0.001% (v/v) 2-mercaptoethanol. The dialyzed cell
◦
extract was stored at −70 C for further study.
The His -tagged fusion protein was purified with Ni-NTA
6
◦
agarose resin from Quiagen (Hilden, Germany) at 4 C. The crude
isopropyl-␤-d-thiogalactopyranoside
(IPTG),
2,3,4,6-tetra-O-
extract was passed directly over a column containing 10 mL of
Ni-NTA agarose resin. After the column was washed with 50 mL
of phosphate buffer (pH 8.0) containing 20 mM imidazole, the C-
terminal His -tagged BCATs was eluted with 20 mL of phosphate
buffer (pH 8.0) containing 200 mM imidazole buffer. The elution
acetyl-d-glucopyranosyl isothiocyanate (GITC) and all other
chemicals were purchased from Aldrich chemical Co. (St Louis,
MO). HPLC grade water and acetonitrile were obtained from
Duksan chemical Co. (Ansan, Korea).
6
solution containing partially purified His -tagged BCAT was con-
6
2.2. Enrichment culture
centrated by ultra-filtration using Centriplus YM-30 (Millipore,
Bedford, MA) with molecular mass cut-off of 10 kDa.
A spoonful of soil samples collected from various domestic sites
were incubated in 50 mL of Luria–Bertani (LB) broth medium for
h. After the incubation, 0.1 mL of culture broth was transferred
2.5. Enzyme assays
1
to 10 mL minimal medium containing 10 mM l-Tle as sole nitrogen
The activity assay of crude-extract was performed in 50 mM
source [23]. Enrichment culture was carried out by repeated dilu-
phosphate buffer (pH 8.0) containing 10 mM keto acid, 10 mM l-
Glu at 37 C, and the product was measured by HPLC. One unit of
the enzyme activity represents 1 ␮mol of product formation per
minute in 10 mM l-Glu.
Determination of the kinetic parameters was performed in
200 ␮L assay solution (50 mM phosphate buffer (pH 8.0)) contain-
◦
tion of culture broth with fresh minimal medium (20–30-folds) at
◦
3
7 C. After the enrichment, the diluted culture broth was spread
out on the same minimal agar plates. Each colony on the plates
was inoculated into 4 mL LB media. After the over-night incubation
o
at 37 C, the harvested cells by centrifugation were washed with

2
30
E.y. Hong et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 228–233
ing the purified enzyme and 10 mM l-Glu. The concentration of
keto acid (amino acceptor) was varied from 0.6 to 20 mM. BCAT
was added to the reaction mixture to initiate the reaction. After
the two consensus amino acid sequences to clone the gene.
ꢀ
ꢀ
ꢀ
Pr1f (5 -CTGCACTAYGGIMMGIMGRTITTYGAAGG-3 ) and Pr2r (5 -
ꢀ
AYYGGIGTRAYTWCIGCCGC-IGTWCC-3 ) were synthesized, where
3
0 min incubation, the reaction was stopped by heating the mix-
I, R, S, and Y indicate inosine, A or G, C or G, and C or T, respec-
tively. PCR was performed using genomic DNA of Enterobacter sp.
TL3 as a template. When Prlf/Pr2r was used as primers, about
700 bp gene fragments were obtained. This size matches to the
expected size. The amplified fragments were cloned in the pGEM-T
vector and sequenced. BLAST results showed that the cloned frag-
ments had high homology to bacterial BCATs, especially showing
86% identity of DNA sequence with E. coli. In order to clone the
BCATen gene, PCR primers were designed based upon nucleotide
sequences of BCATes from E. coli due to highly homologous DNA
◦
ture at 96 C for 15 min. Initial reaction rate was determined by
analysis of the production of l-HAG and l-Tle.
Quantitative analysis and enantiomeric purities of l-HAG
and l-Tle were performed using Waters HPLC system with 3.9
(
radius) × 150 mm C1 Symmetry reverse phase column (Waters,
8
Milford, USA) at 254 nm. In order to analyze chiral compounds, the
racemic acids were derivatized with GITC [25].
Separation of derivatized HAG enantiomers was achieved with
an elution of water/acetonitrile/trifluoacetic acid (70/30/0.1, v/v%)
at a flow rate 1.0 mL min
−1
◦
ꢀ
at 25 C. Retention times for l-HAG
sequences; prf (5 -ATCATGGAATTCA-TGACCACGAAGAAAGCT) and
ꢀ
derivative and d-HAG derivative were 9.7 and 11.3 min, respec-
tively. Separation of derivatized Tle enantiomers was achieved with
an elution of water/acetonitrile/trifluoacetic acid (65/35/0.1, v/v%)
at a flow rate 1.0 mL min at 25 C. Retention times for l-Tle deriva-
tive and d-Tle derivative were 4.8 and 5.2 min, respectively.
For the whole cell reaction of L-HAG production, the prepared
recombinant E. coli BL21 over-expressing BCATes (6 mg of wet
cell weight) was added to the 0.5 mL reaction mixture consisting
prr (5 -CTCGAGTTATTGATTAACTTGATCTAACCA) were designed.
918 bp fragment was obtained from genomic DNA of Enterobac-
ter sp. TL3 by PCR with primers (prf and prr). This DNA sequence
encodes a protein of 309 amino acid residues and identity with
E. coli was 87% when the nucleotide sequences were compared
(Fig. 2).
−1
◦
3.3. Over-expression and purification of the BCATs in E. coli BL21
of 100 mM HAOE, 500 mM l-Glu, 300 mM l-Asp, 20 ␮M PLP and
◦
2
00 mM phosphate buffer (pH 8.0) at 37 C. Reaction products were
The recombinant BCATen and C-terminal His-tagged BCATen
were over-expressed in E. coli BL21. The relative quantities of
the recombinant proteins were almost equal in SDS-PAGE (data
not shown). The specific activities for l-Tle using crude extract
of the native enzyme and the His-tagged BcATen were 17 and
quantitatively analyzed by HPLC after specific time intervals.
3
. Results
−
1
1
0 U mg , respectively. The recombinant E. coli BL21 expressing
3
.1. Screening of microorganisms producing optically pure l-Tle
BCATen showed 45-fold higher activity compared to the wild-type
Enterobacter sp. TL3. The specific activities of the recombinant E.
from trimethylpyruvate with l-Glu
−
1
coli over-expressing His -tagged BCATes were 8.2 U mg . The puri-
6
After several rounds of enrichment culture, a bacteria (TL3) later
identified as Enterobacter sp. by 16s rRNA analysis (Korea Research
Institute of Bioscience and Biotechnology, Korea) showed the high-
est reaction rate and l-specific stereo-selectivity. When whole cell
reaction was carried out with 10 mM TMP and 20 mM l-Glu, the
enantiomeric excess of the produced l-Tle was above 99%. To inves-
tigate whether or not the enzyme activity for producing l-Tle is
really an aminotransferase, the enzyme activity was examined in
the presence of typical inhibitors of PLP-dependent enzyme using
the crude extract of Enterobacter sp. TL3. It was strongly inhib-
ited by gabaculine, hydroxylamine and aminooxyacetic acid [26].
In addition, almost equivalent amounts of l-Tle were produced
according to the consumed l-Glu. Therefore, the enzyme involved
in the enzymatic production of l-TLe was confirmed to be a kind of
a BCAT.
fied BCATen and BCATes gave a single protein band on SDS-PAGE,
respectively.
3.4. Characterization of the his-tagged BCATs from E. coli BL21
The enzyme activity versus the pH was determined within pH
range of 6.0–9.0 in the presence of 10 mM trimethylpyruvate and
0 mM l-glutamate. Three kinds of reaction buffer, 100 mM citrate
pH 6.0), 100 mM potassium phosphate (pH 6.0–8.0), and 100 mM
1
(
boric acid (pH 8.0–9.0) were used for the enzyme assay. The pH-
activity profile of the enzyme for BCATen and BCATes was both bell-
shaped, showing a maximum value at pH 7.5. The enzyme activities
at pH 6.0 and 9.0 were 58 and 78% of that at pH 7.5, respectively.
The amino acceptor specificity of the BCAT from E. coli and Enter-
obacter sp. was examined in the presence of 10 mM l-Glu, and
the enzyme showed high activity toward various aliphatic keto
acids such as 4-methyl-2-oxovaleric acid (keto form of leucine),
3.2. Cloning and DNA sequencing of BCATen
3
-methyl-2-oxobutyric acid (keto form of valine) and 3-methyl-2-
Enterobacter sp. TL3 belongs to gammaproteobacter in pro-
oxovaleric acid (keto form of iso-leucine). This result corresponds
to the enzyme activity that had been reported [27]. Synthesizing
our target unnatural amino acids, trimethylpyruvate; TMP (keto
form of Tle) and 2-(3-hydroxy-1-adamantyl)-2-oxoethanoic acid:
HAOE (keto form of 3-hydroxyadamantylglycine) also showed high
activity (Table 1). Whereas, aromatic ␣-keto acid like phenylpyru-
vate was inert. The amino donor specificity of the enzyme was
examined in the presence of 10 mM TMP and 10 mM HAOE.
Among Glu, Asp and Ala, l-Glu was the most reactive amino
donor.
In comparison between BCAT from E. coli and Enterobacter
sp., BCATen from Enterobacter sp. showed slightly higher activity
toward branched chain amino acids including TMP. Thus, BCATen
was chosen for further studies for the production of l-Tle. On the
other hand, the result indicates that BCATes from E. coli is more suit-
able for the synthesis of bulky amino acid, l-HAG, therefore, BCATes
teobacteria. In order to clone the gene coding the BCATen,
the degenerative primers were designed from highly con-
served regions. We collected 13 amino acid sequences of
BCAT from gammaproteobacter using Gen-Bank Database (E. coli
K12, Haemophilus influenzae Rd, Pseudomonas putida KT2440,
Salmonella enterica subsp., Shigella flexneri 2a str, Shewanella
oneidensis MR-1, Salmonella typhimurium LT2, Xanthomonas
axonopodis pv, Xanthomonas campestris pv, Xylella fastidiosa,
Yersinia pestis, Pseudomonas aeruginosa). The multiple alignments
of the selected bacterial BCAT amino acid sequences revealed
two highly conserved regions; LHYGT(Q,M)S(G,E,Q)V(I,C)FEG
(
30th–39th amino acids based on BCATes from E. coli) and
GTAAE(V)I(V)TP (257th–264th amino acids based on BCATes from
E. coli) which were long enough for designing the degenera-
tive primers. Degenerate PCR primers were designed based upon

E.y. Hong et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 228–233
231
Fig. 2. DNA and amino acid sequences of BCATen from Entrerobacter sp. The boxes indicate the consensus amino acid sequences used to design the degenerate primers.
was selected for further studies for the asymmetric synthesis of
at 200 mM TMP (Fig. 3). BCATes also showed substrate inhibition
l-HAG.
by HAOE. The enzyme activity was maximal at 80 mM and then
decreased as the concentration increased. Judging from the sub-
strate inhibition by keto acids, the control of concentrations of the
keto acids in the reaction mixture would be a key factor maintaining
the maximum aminotransferase activity of these enzymes.
3.5. Substrate and product inhibition of the enzymes for
asymmetric synthesis
The activities of BCATen and BCATes were inhibited severely
by ␣-ketoglutarate (deaminated product of l-Glu). In the pres-
ence of 10 mM ␣-ketoglutarate, only 20% of the enzyme activity
remained (Fig. 3B). This result indicates that the removal of the
Since the reaction rate of aminotransferase is greatly influenced
by the product and substrate concentrations, product and substrate
inhibitions of BcATen and BCATes were examined. The substrate
inhibitions of the BCATen and BCATes by l-Glu were determined in
the presence of 20 mM TMP and HAOE, respectively. Both enzymes
did not show substrate inhibition by l-Glu up to 200 mM. The sub-
strate inhibition of BCATen by TMP was not observed until 20 mM.
However, as the concentration of TMP exceeded 20 mM, the initial
rate decreased, and about 44% of the initial rate at 20 mM was lost
␣-ketoglutarate would be essential for the successful enzymatic
production of l-Tle and l-HAG using BCATs. Indeed, when enzyme
reaction was carried out in 1 mL of 100 mM phosphate buffer
(
(
pH 8.0) containing 20 or 50 mM TMP, 100 mM l-Glu and BCATen
7 U mL ), the amounts of produced l-Tle did not exceed 17 mM.
−
1
Table 1
Kinetic parameters of BCATs from E. coli and Enterobacter sp.
BCAT (E. coli)
BCAT (Enterobacter sp.)
kcat (s−1)
kcat/Km (s mM
−1
−1
)
Km (mM)
kcat (s
−1
)
kcat/Km (s mM
−1
−1
)
Km (mM)
4
3
3
-Methyl-2-oxovaleric acid
-Methyl-2-oxobutyric acid
-Methyl-2-oxovaleric acid
0.42
1.64
0.99
0.09
7.38
6.61
16.84
13.75
2.20
15.74
10.27
13.89
24.44
0.56
0.97
0.78
0.63
0.06
9.52
17.32
12.45
12.15
2.59
17.85
15.96
19.29
43.16
0.3
Trimethylpyruvate
-(3-Hydroxy-1-adamantyl)-2-oxoethanoic acid
2
4.11
2.90

2
32
E.y. Hong et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 228–233
to plateau with 75% conversion. Taylor et al. [2] briefly mentioned
that one of the byproduct, pyruvate, accumulation is often undesir-
able because it can undergo a competitive transamination to form
l-alanine, and the presence of this additional byproduct can effect
the reaction.
In case of GLDH coupling system, formate dehydrogenase (FDH)
was used to regenerate NAD/NADH. Here ammonium formate sup-
plies ammonium for GLDH and formate for FDH. The favorable
equilibrium of FDH for NADH derives the overall reaction toward
the production of l-Tle. Ammonium formate dehydrogenase takes
the ammonium formate and releases carbon dioxide and the amine
source. NADH can be regenerated on this account. Therefore, addi-
tional feeding of amine source to avoid biproduct formation by
dehydrogenase itself is unnecessary. To minimize substrate inhibi-
tion by TMP (Fig. 3), fed-batch reaction was carried out by feeding
−
1
5
0 mM TMP. BCATen (7 U mL ) was added to 10 mL of 100 mM
phosphate buffer (pH 8.0) containing 50 mM of TMP, 30 mM of
−
1
l-Glu, 200 mM ammonium formate, GLDH (15 U mL ) and FDH
(
−
1
20 U mL ). The solid-state substrate corresponding to 50 mM TMP
was added to the reaction mixture intermittently and the course
of reaction was monitored using HPLC analysis. The yield of l-Tle
reached up to 90% with 99% ee in 150 mM high substrate concen-
tration conditions (Fig. 5).
3.7. Asymmetric synthesis of l-3-hydroxyadamantylglycine in
BCATes/AspAT coupling system
For the asymmetric synthesis of l-3-hydroxyadamantylglycine,
−
1
the purified enzyme (0.7 U mL BACTes) was added to 0.2 mL of
1
1
00 mM phosphate buffer (pH 8.0) containing 20 mM of the HAOE,
00 mM l-Glu, 60 mM l-Asp at 37 C for 15 h. 10 mM of HAOE
◦
was converted into 12.2 mM l-HAG with 61.2% conversion. When
−
1
0
.7 U mL BACTes and crude extract of AspAT was added in this
Fig. 3. Substrates and product inhibition of enzymes (A). Substrate inhibition of
BCATen by TMP (ꢀ) and substrate inhibition of BCATes by HAOE (᭹). For inhibi-
tion by TMP against the BCATes, reaction mixture was comprised of as follows: 7 U
BCATen, 10 mM l-glutamate, the specific concentrations of TMP, and 100 mM phos-
phate buffer (pH 8.0). For inhibitions by HAOE against BCATes, reaction mixtures
were comprised of as follows: 0.7 U BCATes, 10 mM l-glutamate, the specific con-
centrations of HAOE, and 100 mM phosphate buffer (pH 8.0). 100% corresponds to
condition, the coupling system gave 20 mM l-HAG with 85.6% con-
version, which is 1.4-folds higher conversion than BCTAes single
enzyme reaction.
−
1
0
.21 mM min reaction rate; (B) ␣-ketoglutarate inhibition of BCATen and BCATes.
Reaction mixtures were comprised of as follows: 0.7 U BCATes or 0.7 U BCATen,
0 mM l-glutamate, the specific concentrations of ␣-keto glutarate, and 100 mM
1
−
1
phosphate buffer (pH 8.0). 100% corresponds to 0.07 mM min reaction rate for
both enzymes.
3.6. Asymmetric synthesis of tert-leucine with BCATen using
coupling system
To overcome product inhibition of ␣-ketoglutarate, coupling
reaction with GLDH or AspAT was applied to the BCAT reaction
(
Fig. 4). In BCAT/AspAT coupling system, keto-acid acceptor (TMP)
is converted into l-Tle using BCAT with l-Glu as the amino donor.
The corresponding ␣-ketoglutarate is then converted back to l-
Glu using l-aspartic acid (l-Asp) as the amino donor by AspAT,
accompanied by the formation of oxaloacetic acid. Therefore, the
inhibitory ␣-ketoglutarate can be maintained at low concentra-
tion in the reaction mixture. In the BCAT/AspAT coupling reaction,
enzyme reaction was carried out in 100 mM phosphate buffer (pH
8
.0) containing 50 mM of TMP, 100 mM of l-Glu, 200 mM l-Asp,
−1
−1
BCATen (7 U mL ) and AspAT (15 U mL ). 37.5 mM l-Tle was pro-
duced with 75% yield. Although the reaction yield increased about
2
.2-fold, compared to previous results without coupling system
(
17 mM l-Tle was produced), still the yield (75%) was not high
enough. We examined inhibitory effects on the BCATen reaction
by pyruvate. The inhibitory effect by pyruvate was negligible up
to 150 mM of pyruvate. It is not clear why reaction profile reach
Fig. 4. The reaction schemes of enzyme coupling reaction. (A) BCAT/GLDH coupling
reaction; (B) BCAT/AspAT coupling reaction.

E.y. Hong et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 228–233
233
In addition, E. coli may also maintain inhibitory by-products at low
concentration by intracellular-metabolism.
4. Conclusion
In this study, the basic properties of BCATen from Enterobac-
ter sp. TL3 which is isolated by selective enrichment from soil
samples were characterized in detail. This study also success-
fully demonstrated the production of l-TLe and l-HAG using
BCAT/AspAT coupling and BCAT/GLDH/FDH coupling reactions
which are designed to overcome product inhibition of the enzymes
by ␣-ketoglutarate as well as the equilibrium problem of the
transamination reaction. We were able to achieve 90% conver-
sion with 99% ee in 150 mM high substrate concentration condition
by BCATen/GLDH/FDH coupling system. In the synthesis of l-HAG,
The AspTA/BcAT coupling system using two recombinant whole
cells achieved 90.5% conversion of 100 mM substrate with >99%
ee. In this study, coupling system was proven to be successful for
increasing the bioconversion of the transamianse reaction, giv-
ing almost full conversion for the synthesis of both l-Tle and
l-HAG.
Fig. 5. The production of l-Tle using BCATen/GLDH coupling reaction. The reaction
started with the reaction mixture containing 100 mM phosphate buffer (pH 8.0),
5
0 mM of TMP, 30 mM of l-Glu, 200 mM ammonium formate, 7 U BCATen, 15 U GDH
and 20 U FDH in 10 mL. The arrows indicate when solid TMP (50 mM) was added.
Acknowledgements
This research was supported by a grant from the National
Research Laboratory Program of the Korea Science and Engineering
Foundation (Grant No. ROA-2007-000-10007-0), and Seoul R&BD
Program (KU080657).
References
[
[
1] T. Li, A.B. Kootstra, I.G. Fotheringham, Org. Process Res. Dev. 6 (2002) 533–538.
2] P.P. Taylor, D.P. Pantaleone, R.F. Senkpeil, I.G. Fotheringham, Trends Biotechnol.
16 (1998) 412–418.
[
[
3] P. Lehr, A. Billich, B. Charpiot, P. Ettmayer, D. Scholz, B. Rosenwirth, H. Gstach,
J. Med. Chem. 39 (1996) 2060–2067.
4] D.J. Augeri, J.A. Robl, D.A. Betebenner, D.R. Magnin, A. Khanna, J.G. Robertson,
A. Wang, L.M. Simpkins, P. Taunk, Q. Huang, S.P. Han, B. Abboa-Offei, M. Cap, L.
Xin, L. Tao, E. Tozzo, G.E. Welzel, D.M. Egan, J. Marcinkeviciene, S.Y. Chang, S.A.
Biller, M.S. Kirby, R.A. Parker, L.G. Hamann, J. Med. Chem. 48 (2005) 5025–5037.
5] B. Gallwitz, Treat. Endocrinol. 4 (2005) 361–370.
[
[
Fig. 6. The asymmetric synthesis of l-HAG using whole cells reaction. The reaction
6] L.L. Nielsen, Drug Discov. Today 10 (2005) 703–710.
mixture (0.5 mL) contained 200 mM phosphate buffer (pH 8.0), 100 mM of HAOE,
[7] U. Kragl, W. Kruse, W. Hummel, C. Wandrey, Biotechnol. Bioeng. 52 (1996)
309–319.
[8] S.L. Liu, Q.X. Song, D.Z. Wei, Y.W. Zhang, X.D. Wang, Prep. Biochem. Biotechnol.
36 (2006) 235–241.
[9] Y. Zhang, T. Sammakia, Org. Lett. 6 (2004) 3139–3141.
10] U. Kragl, T. Dwars, Trends Biotechnol. 19 (2001) 442–449.
11] R.L. Hanson, S.L. Goldberg, D.B. Brozozwski, R.N. Patel, Adv. Synth. Catal. 349
−
1
5
00 mM l-Glu and 300 mM l-Asp using whole cell harboring BCAT (0.7 U mL ).
−1
Course of the conversion when only BCATes (0.7 U mL ) was used (᭹). Course of
the conversion when whole cell harboring BCATes (0.7 U mL ) and AspTA (12 mg
−
1
−1
wet cell weight mL ) in E. coli BL21 was separately added (ꢀ).
[
[
(
2007) 1369–1378.
The use of whole cells as the biocatalyst in an enzymatic pro-
cess has the advantage of simple and economical preparation as
expenses for cell lysis and/or purification are saved. For the whole
cell reaction, freshly prepared whole-cells were added to reaction
mixture containing 100 mM of HAOE, 500 mM l-Glu and 300 mM l-
[
12] D.J. Ager, I.G. Fotheringham, Curr. Opin. Drug Discov. Dev. 4 (2001) 800–807.
[13] B.K. Cho, J.H. Seo, T.W. Kang, B.G. Kim, Biotechnol. Bioeng. 83 (2003) 226–
234.
14] J.S. Shin, B.G. Kim, Biotechnol. Lett. 31 (2009) 1595–1599.
15] B.Y. Hwang, B.K. Cho, H. Yun, K. Koteshwar, B.G. Kim, J. Mol. Catal. B: Enzym.
[
[
3
7 (2005) 27–55.
◦
Asp at 37 C. The recombinant E. coli expressing BCATes (12 mg wet
[16] A. Iwasaki, Y. Yamada, N. Kizaki, Y. Ikenaka, J. Hasegawa, Appl. Microbiol.
Biotechnol. 69 (2006) 499–505.
17] D. Koszelewski, I. Lavandera, D. Clay, G.M. Guebitz, D. Rozzell, W. Kroutil,
Angew. Chem. Int. Ed. Engl. 47 (2008) 9337–9340.
[18] U.T. Bornscheuer, ChemCatChem (2009) 42–51.
[19] M.J. Benecky, R.A. Copeland, R.P. Rava, R. Feldhaus, R.D. Scott, C.M. Metzler, D.E.
Metzler, T.G. Spiro, J. Biol. Chem. 260 (1985) 11671–11678.
20] I.G. Fotheringham, S.A. Dacey, P.P. Taylor, T.J. Smith, M.G. Hunter, M.E. Finlay,
S.B. Primrose, D.M. Parker, R.M. Edwards, Biochem. J. 234 (1986) 593–604.
21] E.S. Venos, M.H. Knodel, C.L. Radford, B.J. Berger, BMC Microbiol. 4 (2004) 39.
22] J.O. Falkinham 3rd, Mol. Gen. Genet. 176 (1979) 147–149.
23] H. Yun, B.Y. Hwang, J.H. Lee, B.G. Kim, Appl. Environ. Microbiol. 71 (2005)
4220–4224.
24] J.H. Lee, S.W. Chung, H.J. Lee, K.S. Jang, S.G. Lee, B.G. Kim, Bioprocess. Biosyst.
Eng. 33 (2010) 71–78.
25] A. Ota, S. Ito, K. Yamamoto, Y. Kawashima, J. Chromatogr. 626 (1992) 187–196.
−1
−1
cell weight mL , which is corresponding to 0.7 U mL ) gave 41%
[
conversion yield. When reaction was carried out using E. coli BL21
−
1
expressing BCATes (12 mg wet cell weight mL ) and E. coli BL21
−
1
expressing AspTA (12 mg wet cell weight mL ), 90.5 mM l-HAG
was produced with >99% ee, which was 2.2-fold higher than that of
using BCAT alone (Fig. 6). When the same reaction conditions were
executed using the crude extract of the enzymes in same units,
[
[
[
[
1
00 mM of HAOE was converted into 55.4 mM l-HAG with 55.4%
conversion in case of only using BACTes, and the coupling system
gave 78.5 mM l-HAG with 78.5% conversion. Although the substrate
keto acid could be converted to hydroxy acid by the ketoreduc-
tase existing in the whole cell or the product could be metabolized,
the whole-cell appeared to be less sensitive to the product inhi-
bitions exerted by ␣-ketoglutrate and/or by-products because the
enzyme was indirectly exposed to high substrate concentrations.
[
[
[26] H. Yun, S. Lim, B.K. Cho, B.G. Kim, Appl. Environ. Microbiol. 70 (2004)
2529–2534.
27] K. Inoue, S. Kuramitsu, K. Aki, Y. Watanabe, T. Takagi, M. Nishigai, A. Ikai, H.
Kagamiyama, J. Biochem. 104 (1988) 777–784.
[