Nonproteinogenic α-amino acid preparation using equilibrium shifted transamination

Article abstract of DOI:10.1021/op025518x

Microbial α-transaminases such as tyrosine aminotransferase (TAT) and branched chain aminotransferase (BCAT) of Escherichia coli, are useful as industrial biocatalysts to prepare nonproteinogenic L-amino acids from α-keto acids and an amino donor. However, they typically yield only 50% product when L-glutamic acid, the preferred amino donor, is used due to accumulated 2-ketoglutaric acid. Accordingly, methods have been sought to increase the reaction yield by the recycle or removal of the keto acid by-product. In this report, we have investigated the biocatalytic coupling of δ-transamination with α-transamination to recycle 2-ketoglutaric acid, and thereby increase the yield of aminotransferase reaction products. Ornithine δ-aminotransferase (OAT) catalyses the reversible transfer of the δ-amino group of L-ornithine to 2-ketoglutaric acid forming L-glutamic acid semialdehyde and L-glutamic acid. The cyclisation of L-glutamic acid semialdehyde to form Δ¹-pyrroline-5-carboxylate under physiological conditions, favours the reaction in the direction of L-glutamic acid formation. The Bacillus subtilis rocD gene encoding OAT was cloned and produced at high levels in E. coli. Combined cell extracts of separate E. coli strains overproducing OAT and E. coli tyrosine aminotransferase enabled the synthesis of L-2-aminobutyrate from 2-ketobutyric acid to reach a yield of 92% compared to 50% achievable by TAT alone. Similarly, combined extracts of strains overproducing OAT and E. coli branched-chain amino acid aminotransferase synthesised L-tert-leucine from trimethylpyruvic acid with a 73% yield compared to 31% with BCAT alone. The use of OAT as a general biocatalytic tool to achieve high yields in aminotransferase reactions is discussed.

Full text of DOI:10.1021/op025518x

Organic Process Research & Development 2002, 6, 533−538
Nonproteinogenic r-Amino Acid Preparation Using Equilibrium Shifted
Transamination
†
‡
Tao Li, Anna B. Kootstra, and Ian G. Fotheringham*
Great Lakes Fine Chemicals, 601 East Kensington Rd, Mt. Prospect, Illinois 60056, U.S.A.
Abstract:
enzymatic bioprocesses to achieve the enantiospecific syn-
Microbial r-transaminases such as tyrosine aminotransferase
thesis of this family of compounds. These bioprocesses
include kinetic and dynamic kinetic resolution of race-
(
TAT) and branched chain aminotransferase (BCAT) of Es-
cherichia coli, are useful as industrial biocatalysts to prepare
nonproteinogenic -amino acids from r-keto acids and an amino
donor. However, they typically yield only 50% product when
-glutamic acid, the preferred amino donor, is used due to
7
,19
4,9,28
mates
and enantio-specific syntheses.
Amongst the
L
latter, aminotransferases have been extensively studied in
the large-scale preparation of nonproteinogenic L- or D-R-
1
,9,25
L
amino acids
as these enzymes typically display many
accumulated 2-ketoglutaric acid. Accordingly, methods have
been sought to increase the reaction yield by the recycle or
removal of the keto acid by-product. In this report, we have
investigated the biocatalytic coupling of δ-transamination with
r-transamination to recycle 2-ketoglutaric acid, and thereby
increase the yield of aminotransferase reaction products.
Ornithine δ-aminotransferase (OAT) catalyses the reversible
desirable features of industrial biocatalysts, including high
turnover, relaxed substrate specificity, and no requirement
for external cofactor recycle.
(1) Ager, D. J.; Fotheringham, I. G.; Laneman, S. A.; Pantaleone, D. P.; Taylor,
P. P. Chim. Oggi 1997, 15, 11-14.
(
(
2) Bearne, S. L.; Wolfenden, R. Biochemistry 1995, 34, 11515-11520.
3) Bewley, M. C.; Lott, J. S.; Baker, E. N.; Patchett, M. L. FEBS Lett. 1996,
386, 215-218.
(
4) Bommarius, A. S.; Drauz, K.; Groeger, U.; Wandrey, C. Chirality Ind.
transfer of the δ-amino group of
acid forming -glutamic acid semialdehyde and
The cyclisation of
-glutamic acid semialdehyde to form ∆1-
pyrroline-5-carboxylate under physiological conditions, favours
the reaction in the direction of -glutamic acid formation. The
L
-ornithine to 2-ketoglutaric
1992, 371-397.
L
L-glutamic acid.
(
5) Borsuk, P.; Dzikowska, A.; Empel, J.; Grzelak, A.; Grzeskowiak, R.;
Weglenski, P. Acta Biochim. Pol. 1999, 46, 391-403.
L
(
(
6) Bradford, M. M. Anal. Biochem. 1976,72, 248-254.
7) Chibata, I.; Tosa, T.; Shibatani, T. Chirality Ind. 1992, 351-370.
L
(8) Transaminases; Christen, P., Metzler, D. E., Eds.; Wiley: New York, 1985.
(
9) Crump, S. P.; Rozzell, J. D. Biocatalytic Production of Amino Acids by
Transamination. In Biocatalytic Production of Amino Acids and DeriVatiVes;
Rozzell, J. D., Wagner, F., Eds.; Wiley: New York, 1992; pp 43-58.
Bacillus subtilis rocD gene encoding OAT was cloned and
produced at high levels in E. coli. Combined cell extracts of
separate E. coli strains overproducing OAT and E. coli tyrosine
(10) Cunin, R.; Glandsorff, N.; Pierard, A.; Stalon, V. Microbiol. ReV. 1986,
0, 314-352.
5
aminotransferase enabled the synthesis of
from 2-ketobutyric acid to reach a yield of 92% compared to
0% achievable by TAT alone. Similarly, combined extracts
of strains overproducing OAT and E. coli branched-chain
amino acid aminotransferase synthesised -tert-leucine from
L-2-aminobutyrate
(
(
(
11) Dzikowska, A.; Swianiewicz, M.; Talarczyk, A.; Wisniewska, M.; Goras,
M.; Scazzocchio, C.; Weglenski, P. Curr. Genet. 1999, 35, 118-126.
12) Fotheringham, I. G.; Bledig, S. A.; Taylor, P. P. J. Bacteriol. 1998, 180,
5
4319-4323.
13) Fotheringham, I. G.; Dacey, S. A.; Taylor, P. P.; Smith, T. J.; Hunter, M.
G.; Finlay, M. E.; Primrose, S. B.; Parker, D. M.; Edwards, R. M. B. Journal
L
1986, 234, 593-604.
trimethylpyruvic acid with a 73% yield compared to 31% with
BCAT alone. The use of OAT as a general biocatalytic tool to
achieve high yields in aminotransferase reactions is discussed.
(
(
(
14) Fotheringham, I. G.; Grinter, N.; Pantaleone, D. P.; Senkpeil, R. F.; Taylor,
P. P. Bioorg. Med. Chem. 1999, 7, 2209-2213.
15) Gardan, R.; Rapoport, G.; Debarbouille, M. J. Mol. Biol. 1995, 249, 843-
856.
16) Herrmann, K. M., Somerville, R. L., Eds. Amino Acids: Biosynthesis and
Genetic Regulation; Advanced Book Program; Addison Wesley: Reading,
MA, 1983.
(
17) Jhee, K.-H.; Yoshimura, T.; Esaki, N.; Yonaha, K.; Soda, K. J. Biochem.
Introduction
(Tokyo) 1995, 118, 101-108.
Nonproteinogenic amino acids are important intermediates
for a variety of pharmaceutical classes, most notably in
peptidomimetics with antiviral and oncological applications.
They are increasingly required at large scale and with high
enantiomeric purity as demand for single-enantiomer drugs
continues to grow. The stereoselectivity inherent in natural
biochemistry has prompted the development of a variety of
(18) Jones, B. N.; Paabo, S.; Stein, S. J. Liq. Chromatogr. 1981, 4, 565-586.
(
19) Kamphuis, J.; Boesten, W. H. J.; Kapten, B.; Hermes, H. F. M.; Sonke,
T.; Broxterman, Q. B.; Van Den Tweel, W. J. J.; Schoemaker, H. E.
Chirality Ind. 1992, 187-208.
(20) Kenklies, J.; Ziehn, R.; Fritsche, K.; Pich, A.; Andreesen, J. R. Microbiology
(
Reading, U.K.) 1999, 145, 819-826.
(
21) Laemmli, U. K. Nature 1970, 227, 680-685.
(22) Medina, V.; Pontarollo, R.; Glaeske, D.; Tabel, H.; Goldie, H. J. Bacteriol.
990, 172, 7151-7156.
23) Messenguy, F.; Vierendeels, F.; Scherens, B.; Dubois, E. J. Bacteriol. 2000,
82, 3158-3164.
1
(
1
*
Corresponding author: Ian G. Fotheringham, Department of Chemistry,
University of Edinburgh, King’s Buildings, West Mains Rd, Edinburgh EH9
JJ, UK. Telephone: 0131-650-7747. Fax: 0131-650-4743. E-mail:
(24) Miller, C. A.; Tucker, W. T.; Meacock, P. A.; Gustafsson, P.; Cohen, S.
N. Gene 1983, 24, 309-315.
(25) Pantaleone, D. P.; Geller, A. M.; Taylor, P. P. J. Mol. Catal. B: Enzym.
2001, 11, 795-803.
3
ian.fotheringham@ed.ac.uk.
†
Present address: Biotransformation, Schering Plough Research Institute,
(26) Remaut, E.; Stanssens, P.; Fiers, W. Gene 1981, 15, 81-93.
(27) Roosens, N. H. C. J.; Thu, T. T.; Iskandar, H. M.; Jacobs, M. Plant Physiol.
1998, 117, 263-271.
U-13-3000, 1011 Morris Ave., Union, NJ 07083.
‡
Present address: Albany Molecular Research, 601 East Kensington Rd.,
Mt. Prospect, IL 60056.
(28) Stirling, D. I. Chirality Ind. 1992, 209-222.
1
0.1021/op025518x CCC: $22.00 © 2002 American Chemical Society
Vol. 6, No. 4, 2002 / Organic Process Research & Development
•
533
Published on Web 05/21/2002

Figure 1. General r-aminotransferase reaction.
Figure 2. Coupled reactions involving an r-aminotransferase and ornithine-δ-aminotransferase. 1: r-Keto acid. 2: L-Amino acid,
1
3:
L-Glutamic acid. 4: r-Ketoglutaric acid. 5: L-Ornithine., 6: L-Glutamate-γ-semialdehyde. 7: ∆ -Pyrroline-5-carboxylic acid.
Aminotransferases catalyse the pyridoxal 5′-phosphate-
undesired L-alanine as a by-product and to employ the
generally preferred amino donor for this enzyme class.
L-Ornithine δ-aminotransferase (OAT) [EC 2.6.1.13] cataly-
ses the pyridoxal 5′-phosphate-dependent, reversible tran-
samination of the δ-amino group from L-ornithine to
R-ketoglutarate, producing L-glutamate-γ-semialdehyde and
L-glutamic acid. The formation of L-glutamic acid from
R-ketoglutarate is strongly favoured by the spontaneous
dependent reversible transfer of an amino and keto group
between an amino acid and keto acid substrate pair yielding
new amino acid and keto acid products as shown in Figure
. In E. coli and most other bacteria studied, aminotrans-
ferases are involved in the biosynthesis of many R-amino
1
acids from their corresponding R-keto acids with L-glutamic
acid serving as the preferred amino donor.8
,9,16
Due to the
1
relaxed substrate specificity of microbial aminotransferases
such as the tyrosine aminotransferase (TAT) and branched-
chain aminotransferase (BCAT) of E. coli, these enzymes
are also efficient in the enantiospecific preparation of
nonproteinogenic R-amino acids such as L-homophenylala-
nine, L-2-aminobutyrate and L-tert-leucine from their corre-
sponding R-keto acids.9 However, the reversible transam-
ination reaction typically results in a 50% conversion of
substrates to products. This low yield and corresponding low
product purity is not feasible for economical large-scale
production and has prompted efforts to overcome this
limitation of an otherwise attractive biocatalytic process.
Principally, this has been addressed by the use of
aspartate, rather than glutamate, as the amino donor since
aspartate yields the unstable â-keto acid product oxaloacetate
cyclisation of L-glutamate-γ-semialdehyde to form ∆ -
2
27
pyrroline-5-carboxylate, a precursor of L-proline in plants
and a product of arginine catabolism in microorganisms.1
0,16
Ornithine aminotransferases have been found in a number
11,15,17,20,23,35
of microbial isolates, many of them Bacilli.
This
report describes the use of OAT, produced from the Bacillus
subtilis rocD gene cloned in E. coli, in a coupled transami-
nation, as shown in Figure 2, to recycle R-ketoglutarate
generated in R-aminotransferase reactions in order to increase
the production of specific nonproteinogenic amino acids.
,31
Results
Cloning and Expression of Aminotransferase Genes.
Plasmid pTL17 carries the cloned rocD gene of B. subtilis
1
5
1
68 encoding the 44 kD OAT protein. Cell extracts
1
6
instead of the relatively stable R-ketoglutarate. Rapid
decarboxylation of oxaloacetate to form pyruvate under
biotransformation conditions favours aspartate consumption,
prepared following induction of rocD expression and analy-
sed by SDS-PAGE indicated a new 44 kD band. The
9
thus increasing product yield significantly beyond 50%. This
(29) Stoker, N. G.; Fairweather, N. F.; Spratt, B. G. Gene 1982, 18, 335-341.
(30) Takechi, M.; Kanda, M.; Hori, K.; Kurotsu, T.; Saito, Y. J. Biochem. (Tokyo)
is not a complete solution as residual pyruvate is often
transaminated to L-alanine which also compromises product
yield and purity. Additionally many aminotransferases do
1994, 116, 955-959.
(31) Taylor, P. P.; Pantaleone, D. P.; Senkpeil, R. F.; Fotheringham, I. G. Trends
Biotechnol. 1998, 16, 412-418.
(32) Tuchman, M.; Rajagopal, B. S.; McCann, M. T.; Malamy, M. H. Appl.
16
not accept aspartate as amino donor. Although multienzyme
EnViron. Microbiol. 1997, 63, 33-8.
(33) Twigg, A. J.; Sherratt, D. Nature (London) 1980, 283, 216-18.
9
,14
systems have been devised to address these problems, an
alternative approach which enables a versatile high-yielding
transamination from L-glutamate is preferable to eliminate
(
(
34) Yang, J.; Pittard, J. J. Bacteriol 1987, 169, 4710-5.
35) Yasuda, M.; Tanizawa, K.; Misono, H.; Toyama, S.; Soda, K. J. Bacteriol.
1981, 148, 43-50.
534
•
Vol. 6, No. 4, 2002 / Organic Process Research & Development

Figure 3. Influence of OAT on the yield of L-2-aminobutyrate (L-2-aba) synthesised from 2-ketobutyric acid by TAT. Reaction 1
() contained TAT in extract of W3110/pME64. Reactions 2 (9) and 3 (2) additionally contained OAT in cell extract of W3110/
(
pTL17 and 25 or 50 mM ornithine (Ort), respectively.
intensity of the band represented approximately 40% of the
releasable cell protein. The aminotransferase specific activity
of the crude extract was 17 units/mg protein. Plasmid pME64
carries the cloned E. coli K12 tyrB gene13 encoding the
aromatic aminotransferase. Whole cells of strain W3110/
pME64 have previously been shown to efficiently catalyse
acid. Samples at 3, 5, and 16 h contained L-2-aminobutyrate
at 24.93, 27.96, and 24.95 mM, respectively. The OAT
transamination was incorporated into the other two reactions.
Reaction 2 contained 3 mg of the soluble protein from
W3110/pTL17 and 25 mM of L-ornithine in addition to the
components of reaction 1. Samples of reaction 2 at 3, 5 and
16 h contained L-2-aminobutyrate at 30.71, 34.74, and 37.97
mM, respectively. In reaction 3 the concentration of ornithine
was 50 mM, while everything else remained the same as
reaction 2. Samples of reaction 3 at 3, 5 and 16 h contained
L-2-aminobutyrate at 34.8, 45.96, and 43.79 mM, respec-
tively.
Reaction 1 appeared to have reached equilibrium follow-
ing incubation for 3 h. The final yield of reaction 1 was
49.9%, the equilibrium of the reversible TAT transamination.
The yields of reaction 2 and 3 after 16 h incubation were
75.9 and 87.6%, respectively. Reaction 3 showed a 92%
conversion of 2-ketobutyric acid to L-2-aminobutyrate after
5 h. When coupled with OAT reaction, there was a further
increase in product yield as the amount of ornithine was
increased.
14
the L-2-aminobutyrate/L-glutamate transamination reaction.
Expression of the tyrB gene from its native promoter on
plasmid pAT153,33 is deregulated by the repressor titration
3
4
effect of the very high copy number plasmid vector.
Plasmid pIF328 carries the cloned ilVE gene encoding the
E. coli K12 branched-chain aminotransferase (BCAT) which
has previously been shown to catalyse the L-tert-leucine/L-
glutamate transamination reaction.31 The ilVE gene is ex-
2
2
pressed from the E. coli K12 pckA promoter region
introduced immediately upstream on the vector.
Biosynthesis of L-2-Aminobutyric Acid. The synthesis
of L-2-aminobutyrate was catalyzed by tyrosine aminotrans-
ferase using the crude cell extract of W3110/pME64. The
effect on the reaction yield of coupling this reaction to
ornithine δ-transamination was investigated in three reactions
as shown in Figure 3. The reactions were sampled over the
course of 16 h for amino acid analysis by HPLC. Reaction
Biosynthesis of L-tert-Leucine. The synthesis of L-tert-
leucine was catalyzed by the branched-chain aminotrans-
ferase from the crude extract of W3110/pIF328. The effect
of coupling to the OAT reaction was also investigated in
three reactions similar to those for L-2-aminobutyrate prepa-
1
is the R-transamination by TAT in the absence of OAT
which contained 3.7 mg of the soluble protein from W3110/
pME64, 50 mM 2-ketobutyric acid and 50 mM L-glutamic
Vol. 6, No. 4, 2002 / Organic Process Research & Development
•
535

Figure 4. Influence of OAT on the yield of L-tert-leucine synthesised from trimethylpyruvic acid (TMPA) by TAT. Reaction 1 (()
contained TAT in extract of W3110/pIF328. Reactions 2 (9) and 3 (2) additionally contained OAT in cell extract of W3110/pTL17
and 25 or 50 mM ornithine (Ort), respectively.
ration as shown in Figure 4. Reaction 1 included 3.7 mg of
the extracted soluble protein from W3110/pIF328, 40 mM
trimethylpyruvic acid (TMPA) and 40 mM L-glutamic acid.
Samples at 3, 5, and 16 h contained L-tert-leucine at 5.72,
Similarly the yield of L-tert-leucine produced by the E. coli
branched-chain aminotransferase increased from 31 to 73%
when OAT was included. Since neither 2-ketobutyric acid
nor trimethylpyruvic acid is a substrate for OAT, the results
in this report demonstrate that OAT can efficiently regenerate
L-glutamate as the amino donor for the R-aminotransferase
reaction. The improved reaction yields were due to the
reduced reversibility of the OAT reaction under conditions
where L-ornithine is transaminated to L-glutamate semial-
12.53, and 12.92 mM, respectively. Reaction 2 contained 3
mg of soluble protein from W3110/pTL17 and 25 mM
L-ornithine in addition to everything in reaction 1. Samples
of reaction 2 at 3, 5 and 16 h contained L-tert-leucine at 9.2,
20.21, and 27.79 mM, respectively. In reaction 3 the
1
concentration of ornithine was 50 mM, while everything else
remained the same as reaction 2. Samples of reaction 3 at
dehyde and spontaneously cyclised to ∆ -pyrroline-5-car-
boxylate. In previous reports of preparative aminotransferase
reactions the use of aspartic acid as the amino donor has
been the only general strategy to enable the reaction to
progress beyond the normal conversion of 50%. However,
this often leads to accumulation of L-alanine which is difficult
to remove and decreases the overall yield by creating an
alternate pathway converting L-aspartate directly to L-alanine.
Conversely, ornithine aminotransferase creates no alternate
pathway, and the removal of residual L-ornithine and
L-glutamate by precipitation or chromatographic approaches
is facilitated by their charged nature. Selective crystallisation
3
2
,5 and 16 h contained L-tert-leucine at 7.71, 21.72, and
9.19 mM, respectively. Reaction 1 did not progress further
after 5 h incubation at 37 °C with a yield of 31.3% of L-tert-
Leucine. Reactions 2 and 3 did not stop at 5 h. Yields for
these two reactions at 16 h reached 69.5 and 73%, respec-
tively.
Discussion
The use of B. subtilis ornithine aminotransferase (OAT)
to couple L-ornithine ω-transamination to L-glutamate R-tran-
samination enabled the latter reaction to greatly exceed the
typical 50% product yield in the biosynthesis of two high-
value amino acids. The yield of L-2-aminobutyrate produced
by E. coli tyrosine aminotransferase increased from 50 to
1
or precipitation can be used as appropriate to remove ∆ -
pyrroline-5-carboxylate on a case-specific basis. Future work
will examine the optimal conditions for this bioprocess,
including the minimum catalytic amount of L-glutamate
required, operation at very high substrate concentrations, and
92% when OAT was included in the biotransformation.
536
•
Vol. 6, No. 4, 2002 / Organic Process Research & Development

the coordinated maximal expression of the enzymes. Ad-
ditionally further cost reduction could be achieved by
replacing L-ornithine with L-lysine, a very low-cost feedstock
amino acid whose transamination product also generates a
cyclic imine, or by incorporation of arginase to produce
ornithine directly from arginine.3,5,32 The coupling of R- and
ω-transamination offers a general method to enhance the
overall applicability of these enzymes as biocatalysts in the
preparation of unnatural amino acids at high enantiomeric
purity. Since none of the enzymes used in this approach
require active cellular metabolism for prolonged use as a
biocatalyst, the process is appropriate for immobilisation of
crude enzyme or whole cells. In this way, the production
and handling of the enzymes can be greatly simplified, and
recovery and recycling of the biocatalysts are improved,
thereby reducing costs and minimising the effect on down-
stream processing and product isolation. The high expression
of the OAT enzyme from the cloned rocD gene of B. subtilis
again illustrates the growing potential to construct multien-
zyme biotransformations in recombinant strains of E. coli
for scalable production of valuable compounds.
generated by SphI and BspEI digestion of a fragment
prepared by PCR from pLG338 as the template and the
following oligonucleotides as primers: 5′-GACGCATG-
CACCATTCCTTGCGGCGGCG-3′ and 5′-GACTCCG-
GAGGCAAATCGCTGAATATTCC-3′. The resulting plas-
mid was named pIF328. Plasmid pTL17 was constructed as
follows. The rocD gene was cloned from B. subtilis strain
1
68 by PCR with the following pair of primers: 5′-
GACGGATCCACTATGACAGCTTTATCTAAATCCAAA-
and 5′-CAGGGTACCTCATTATGCGTTTCGCAG-
3
′
CACGTG-3′. The 1.2 kb DNA from the PCR was then
cleaved with BamHI and KpnI and ligated to the 4.8 kb
14
vector fragment of pPOT3 obtained from similar digestion.
The resulting plasmid was named as pTL17. In this plasmid,
the transcription of rocD gene is controlled by an upstream
l P
R
promoter. This promoter is regulated by a temperature-
26
sensitive cI857 repressor carried on the same plasmid.
Preparation of Enzymes for Biotransformation. Each
plasmid was used to transform separate E. coli K12 W3110
(ATCC 2735) hosts for enzyme expression. Transformation
was carried out by electroporation using a Bio-Rad Gene
Pulser. Cell cultures carrying pME64, or pIF328 were
prepared by inoculating 50 mL of LB medium with a single
colony from an LB plate and were incubated with shaking
at 37 °C overnight. Antibiotics, where appropriate, were used
at the concentrations of 100 mg/mL ampicillin, 40 mg/mL
kanamycin for solid and liquid media. Overnight cultures
were then used to inoculate 1 L of LB plus antibiotics. They
were grown in a 4-L flask with agitation at 37 °C until the
OD600 reached 1.0. The cells were then recovered by
centrifugation at 10000g for 5 min. The culture of W3110/
pTL17 was prepared by inoculating 50 mL LB containing
Experimental Section
General DNA Manipulation. General DNA handling
including PCR, restriction analysis, DNA recovery, chro-
mosomal and plasmid DNA preparations, and E. coli
transformation were carried out using standard methods as
12
described previously. Restriction enzymes and DNA ligase
were purchased from New England Biolabs (Beverly, MA).
PCR was carried out using ClonTech HF PCR kits (Palo
Alto, CA). Molecular biology grade reagents were purchased
from Sigma-Aldrich (Milwaukee, WI).
Plasmid Construction. Plasmid pME64 was constructed
as described.13 Plasmid pIF328 was constructed as follows.
The ilVE gene encoding the branched-chain aminotransferase
was cloned from E. coli K12 chromosomal DNA by PCR
using the following pair of oligonucleotide primers:
1
0 mg/mL chloramphenicol with a single colony from an
overnight LB agar/chloramphenicol plate. This culture was
incubated with shaking at 30 °C overnight. The overnight
culture was used to inoculate 1 L of LB plus chloramphenicol
in a 4-L flask. The culture was grown to OD600 ) 0.6-0.8,
at which point OAT expression was induced by raising the
incubator temperature to 42 °C. The incubation at 42 °C was
continued for a further 2 h before the cells were recovered
by centrifugation. To prepare enzyme extracts for biotrans-
formation, the cells were broken with a French press at a
5
′-CGCGGATCCACTATGACCACGAAGAAAGCT-
GATTACATTTGG-3′ and 5′-CAGCGTGCATGCTTAT-
TGATTAACTTGATCTAACCA-3′. The amplified PCR
product was cleaved with BamHI and SphI, and ligated to
the 3.9 kb fragment resulting from similar cleavage of of
12
pIF306. The resulting plasmid was named pIF307. Plasmid
pIF307 was cleaved with EcoRI and PstI and the 4.1 kb band
was isolated. This was ligated to a 982 bp fragment
containing the kanamycin resistance gene from similarly
2
pressure of 1000 lb/in , and the insoluble particles were
removed by centrifugation at 10000g × 30 min.
29
Enzyme and Protein Assays. Protein concentration was
cleaved plasmid pLG338. The resulting plasmid was named
pIF312. This plasmid was cleaved with EcoRI and BamHI.
The vector fragment of 4.97 kb was isolated and ligated to
a 0.27 kb DNA fragment containing the pckA promoter of
E. coli K12. This fragment was generated by EcoRI/BamHI
double digestion of an E. coli chromosomal fragment
amplified using PCR with the following oligonucleotides as
primers: 5′-GACGAATTCACTTTACCGGTTGAATTTGC-
6
determined using the Bradford assay. Aminotransferase
expression levels were analyzed by SDS-PAGE of crude cell
2
1
extracts using 12% polyacrylamide. The activity of OAT
was determined using a standard aminotransferase assay.30
Amino Acid Preparation via Coupled Transamina-
tions. All reactions were conducted in 50 mM phosphate,
pH 8.5 at 37 °C, in a total volume of 2 mL in 5-mL capped
tubes with agitation. After 2, 4, and 5 h of incubation the
pH was checked and adjusted to pH 8.5 with 1 N NaOH.
Samples of 100 mL were taken at 3, 5, and 16 h, centrifuged
to remove cells, and diluted as appropriate for amino acid
analysis by quantitative HPLC.
3′ and 5′-GACGGATCCTCCTTAGCCAATATGTATTGCC-
3′. The resulting plasmid, named pIF313, was cleaved with
SphI and BspEI and the 4.1 kb vector fragment isolated. This
fragment was then ligated to a 0.97 kb DNA fragment
containing the par locus24 of pLG338 This fragment was
Vol. 6, No. 4, 2002 / Organic Process Research & Development
•
537

HPLC Analysis of Amino Acids. Amino acid samples
and standards were derivatised with o-phthaldialdehyde 18
before analysis. The amino acid derivatives were separated
on a C18 column (5 mm, 250 mm × 4.6 mm) by a 2-100%
acetonitrile gradient in phosphate buffer (15 mM, pH 6.2)
at the flow rate of 1.5 mL/min.
Received for review February 11, 2002.
OP025518X
538
•
Vol. 6, No. 4, 2002 / Organic Process Research & Development