Engineered dehydrogenase biocatalysts for non-natural amino acids: Efficient isolation of the d-enantiomer from racemic mixtures

Article abstract of DOI:10.1039/b809147a

With a view to their use in the kinetic resolution of racemic non-natural amino acids, five variants of the enzyme l-phenylalanine dehydrogenase, the wild-type enzyme from Bacillus sphaericus and four active-site mutants, have been tested with a range of amino acids. In each case, the rates of reaction with 0.2 mM l-amino acid and with the racemic mixture at 0.4 mM were compared, so that the starting concentration of the active substrate was kept constant. Although the d-amino acids are not substrates, they were inhibitory in all cases. The extent of inhibition, however, varied greatly from compound to compound and among the mutants. With the N145L mutant and dl 4-O-Me-Phe, the equimolar d-enantiomer gave 83.2% inhibition, and with the wild-type enzyme there was 86.7% inhibition with racemic norleucine. By contrast, with these same substrates the N145V mutant showed less than 9% and 24% inhibition respectively. The N145A mutant was selected for use with dl-4-Cl-Phe. The pH was decreased from the enzyme's optimum of 10.4 to 9.5 to minimise breakdown of the coenzyme NAD⁺, and the coenzyme was recycled by molecular oxygen with the assistance of a commercial diaphorase. Reaction on a 200 μmole scale in 20 ml ethanolamine HCl buffer, pH 9.5, with 25 μg N145A enzyme and 100 μg diaphorase, was monitored by chiral HPLC. The l-isomer was removed to an extent of >99% after 40 h, with the d-isomer peak undiminished. The pure d-isomer was isolated from the reaction mixture in 85% overall yield after ion-exchange chromatography.

Full text of DOI:10.1039/b809147a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Engineered dehydrogenase biocatalysts for non-natural amino acids:
efficient isolation of the D-enantiomer from racemic mixtures
Francesca Paradisi,*a Philip A. Conway,a,c Anita R. Maguireb and Paul C. Engel*c
Received 29th May 2008, Accepted 7th July 2008
First published as an Advance Article on the web 5th August 2008
DOI: 10.1039/b809147a
With a view to their use in the kinetic resolution of racemic non-natural amino acids, five variants of
the enzyme L-phenylalanine dehydrogenase, the wild-type enzyme from Bacillus sphaericus and four
active-site mutants, have been tested with a range of amino acids. In each case, the rates of reaction with
0.2 mM L-amino acid and with the racemic mixture at 0.4 mM were compared, so that the starting
concentration of the active substrate was kept constant. Although the D-amino acids are not substrates,
they were inhibitory in all cases. The extent of inhibition, however, varied greatly from compound to
compound and among the mutants. With the N145L mutant and DL 4-O-Me-Phe, the equimolar
D-enantiomer gave 83.2% inhibition, and with the wild-type enzyme there was 86.7% inhibition with
racemic norleucine. By contrast, with these same substrates the N145V mutant showed less than 9%
and 24% inhibition respectively. The N145A mutant was selected for use with DL-4-Cl-Phe. The pH was
decreased from the enzyme’s optimum of 10.4 to 9.5 to minimise breakdown of the coenzyme NAD+,
and the coenzyme was recycled by molecular oxygen with the assistance of a commercial diaphorase.
Reaction on a 200 lmole scale in 20 ml ethanolamine HCl buffer, pH 9.5, with 25 lg N145A enzyme
and 100 lg diaphorase, was monitored by chiral HPLC. The L-isomer was removed to an extent of
>99% after 40 h, with the D-isomer peak undiminished. The pure D-isomer was isolated from the
reaction mixture in 85% overall yield after ion-exchange chromatography.
has been described,8 utilising the enantioselectivity of penicillin-G
Introduction
acylase in the cleavage of N-phenacetyl amino acid derivatives.
Its application has so far been reported only for biological amino
acids, and apart from the requirement for prior derivatisation, it is
also relevant that very variable ee values were reported, depending
on the target amino acid chosen. An alternative enzymatic route to
D-amino acids is offered by the recently reported D-specific amino
acid dehydrogenase.9
Our strategy is focused on the creation of new biocatalysts
through site-directed mutagenesis to shift and broaden the
substrate specificity of natural enzymes. We have previously
reported the successful application of mutants of phenylalanine
dehydrogenase from Bacillus sphaericus in the synthesis of non-
natural L-amino acids starting from the corresponding 2-oxo-
acids.10,11 Given the stringent enantioselectivity of PheDHs, these
catalysts may also be used for the efficient kinetic resolution of the
D- and L- enantiomers. Without any requirement for derivatisation,
a single biocatalyst thus offers a route to both enantiomers at high
ee. In this paper, we report the selectivity of four PheDH mutants in
dealing with racemic mixtures of non-natural amino acids and an
example of kinetic resolution involving recycling of the cofactor.
The synthesis of amino acids in pure L- and D- forms is a major
interest in the pharmaceutical field. Amino acids can be used as
chiral synthons in multistep reactions, being easy to derivatise at
the carboxylic moiety as well as at the amino group. The chiral
amino acid can then induce chirality in a subsequent reaction with
a non-chiral substrate, conferring a specific spatial geometry on
the new molecule, and, where new stereocentres are introduced,
their configuration can be influenced.1 Natural L-amino acids
are easy to obtain from biological sources, but the production
of the corresponding D- forms is more challenging. Non-natural
amino acids (e.g. tert-leucine, phenylglycine) are in demand for the
synthesis of novel drugs,2,3 and, through combinatorial chemistry,
libraries of new compounds may be created. In this last context,
neither enantiomer is usually readily available, and the resolution
of racemic mixtures4 or asymmetric synthesis5 are the only options.
Enzymatic approaches are limited by the lack of enzymes that
would naturally accept non-natural amino acids as substrates.
However, a few industrial processes have optimised the biocatalytic
synthesis of optically pure non-natural amino acids, such as the
method for L- (or D-) tert-leucine using leucine dehydrogenase,6,7
which shows low but useable activity with the non-natural sub-
strate. A kinetic resolution method for preparing D-amino acids
Results and discussion
aUCD School of Chemistry and Chemical Biology, University College Dublin,
Belfield, Dublin 4, Ireland. E-mail: francesca.paradisi@ucd.ie; Fax: +353 1
716 2501; Tel: +353 1 716 2967
Tables 1 and 2 present data under standard assay conditions for
the oxidative deamination reaction of wild-type phenylalanine
dehydrogenase (WT PheDH) and mutated variants with a series of
derivatives of L-phenylalanine (Phe) substituted at position 4 (1–4
and the corresponding D-enantiomers (7–10), and aliphatic amino
acids (L-norleucine 5, L-cyclohexylalanine 6, and the racemate
bAnalytical and Biological Chemistry Research Facility, Department of
Chemistry and School of Pharmacy, University College, Cork, Ireland
cUCD School of Biomolecular and Biomedical Science, Conway Institute,
University College Dublin, Belfield, Dublin 4, Ireland, paul.engel@ucd.ie
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3611–3615 | 3611
©

View Article Online
Table 1 Activity of WT PheDH and mutants with 0.2 mM amino acid substrate. Assays were performed at pH 10.4 in Gly–NaOH buffer by following
the production of NADH spectrophotometrically at 340 nm. The table shows specific activities (in brackets) with pure L-amino acids, and also relative
activities (first number) as a percentage of the activity with L-Phe in each case
Amino acids
WT
N145A
N145L
N145V
N145I
L-Phe
4-F-L-Phe (1)
4-Cl-L-Phe (2)
4-Me-L-Phe (3)
4-OMe-L-Phe (4)
L-Norleucine (5)
L-Cyclohexyl-alanine (6)
100 (7.5 U mg−1
52 (3.9 U mg−1
11 (0.8 U mg−1
5 (0.4 U mg−1
3 (0.2 U mg−1
)
)
)
)
)
100 (35 U mg−1
)
100 (17.5 U mg−1
172 (30.1 U mg−1
204 (35.7 U mg−1
124 (21.7 U mg−1
)
)
)
)
100 (20 U mg−1
265 (53 U mg−1
)
)
100 (16.4 U mg−1
168 (27.6 U mg−1
130 (21.3 U mg−1
69 (11.3 U mg−1
)
)
)
)
129 (45.2 U mg−1
123 (43.1 U mg−1
82 (28.7 U mg−1
105 (36.8 U mg−1
)
)
)
)
)
)
112 (22.4 U mg−1
67 (13.4 U mg−1
146 (29.2 U mg−1
)
)
)
20 (3.5 U mg−1
)
14 (2.3 U mg−1
)
5.9 (0.4 U mg−1
9.8 (0.7 U mg−1
)
)
9.4 (3.3 U mg−1
17.3 (6.1 U mg−1
20.6 (3.6 U mg−1
13.7 (2.4 U mg−1
)
)
25 (5 U mg−1
)
14.6 (2.4 U mg−1
12.2 (2 U mg−1
)
32.5 (6.5 U mg−1
)
)
Table 2 Activity of WT PheDH and mutants with racemic amino acids. Each amino acid was present at 0.4 mM concentration i.e. 0.2 mM of each
enantiomer. The first number in each case gives the activity as a percentage of that with the corresponding pure L-amino acid at 0.2 mM (Table 1).
Reactions were performed at pH 10.4 in Gly–NaOH buffer by following the production of NADH at 340 nm spectrophotometrically. The racemic form
of 6 could not be obtained and therefore it has not been included in the table
Amino acids
WT
N145A
N145L
N145V
N145I
DL-Phe
77 (5.8 U mg−1
)
41 (14.4 U mg−1
60 (27.1 U mg−1
50 (21.6 U mg−1
70 (21.1 U mg−1
65 (23.9 U mg−1
)
)
)
)
)
52 (11.2 U mg−1
54 (16.3 U mg−1
29 (10.4 U mg−1
)
)
)
64 (12.8 U mg−1
52 (27.6 U mg−1
)
)
39 (6.4 U mg−1
)
4-F-DL-Phe (7)
4-Cl-DL-Phe (8)
4-Me-DL-Phe (9)
4-OMe-DL-Phe (10)
DL-Norleucine (11)
77 (3 U mg−1
)
52 (14.4 U mg−1
)
83 (0.7 U mg−1
66 (0.3 U mg−1
95 (0.2 U mg−1
)
)
)
43 (9.6 U mg−1
54 (7.2 U mg−1
)
)
26 (5.3 U mg−1
36 (4.1 U mg−1
60 (1.4 U mg−1
)
)
)
31 (6.7 U mg−1
17 (0.6 U mg−1
)
)
91 (26.6 U mg−1
)
)
13.3 (0.1 U mg−1
)
58 (1.9 U mg−1
)
72.2 (2.6 U mg−1
)
76.7 (3.8 U mg−1
68.3 (1.6 U mg−1
)
DL-norleucine 11). These reactions are exceptionally sensitive to
pH and the experiments are strictly performed at pH 10.4 and at
25 ◦C. For each of the amino acids, an activity of at least 5 U mg−1
(L-norleucine with N145V) was measured with one or more of
the mutated enzymes, and for several substrates activities up to
10-fold higher than this were obtained (e.g. 53 U mg−1 for 1 and
36.8 U mg−1 for 4). Owing to the low concentration of the sub-
strates (0.2 mM in case of pure L-amino acids and 0.4 mM for the
racemic mixtures), consistent through the tables and imposed by
the low water solubility of some of the amino acids, the measured
activities are certainly sub-optimal in some cases. For example, the
specific activity of WT PheDH with L-phenylalanine under these
conditions is only 7.5 U mg−1, much lower than that of the mutants,
but this is explained by the much higher Km of WT PheDH
for L-phenylalanine,11 and clearly, for this particular enzyme
protein, much higher activity would be achievable by increasing the
substrate concentrations. It interesting to note, however, that our
WT enzyme behaves as expected with L-norleucine, for example,
even at this low concentration of substrate; the retained activity
of 5.9%, compared to L-phenylalanine at the same concentration,
correlates quite well with that reported by Asano et al.12 whereby
the same enzyme shows a retention of activity of 3.9% at 10 mM
concentration.
The data in Table 2, reporting reaction rates measured with the
amino acids added at double the concentration but as racemic
mixtures (0.2 mM each enantiomer), clearly show that the D-
enantiomer is invariably inhibitory (all the relative activities are
referred to the figure for the corresponding pure L-amino acid),
in some cases quite potently so. Thus, for example, with 0.2 mM
L-4-OMe-Phe and N145L, the rate with an equal mixture of the
D-enantiomer is 6 times lower than the rate with the pure L-form,
and the unmutated enzyme (WT) shows over 7-fold inhibition by
the D-enantiomer of norleucine. This is an important factor to
consider in any kinetic resolution, since, as the reaction proceeds,
competitive inhibition inevitably becomes more severe. For each
compound tested, the range of inhibition seen with the five
different biocatalysts was considerable. Thus with 4-Cl-Phe, the
inhibition ranged from 17% (WT) to 74% (N145I) and for 4-
OMe-Phe from 4.8% (WT) to 83.2% (N145L). With Phe and
with nearly all the 4-substituted analogues tested, WT PheDH
stands out as showing the least inhibition by the D-enantiomer
(with the 4-Me analogue, both N145A (30.3%) and WT PheDH
(34%) show much less inhibition than the other mutants). Thus
the unmutated enzyme shows the least tendency to accept the D-
enantiomer, even as an unreactive inhibitory ligand. This pattern
is sharply reversed in the case of the aliphatic substrate, for which
3612 | Org. Biomol. Chem., 2008, 6, 3611–3615
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Scheme 1 The DL-4-Cl-phenylalanine is subjected to enzymatic deamination. The L-enantiomer is oxidised to the corresponding oxo-acid while the
D-enantiomer does not react. The NAD+ cofactor is reduced to NADH in this process and is reoxidised by molecular oxygen in a coupled reaction
catalysed by a commercial diaphorase (from Enzolve Technologies, Dublin).
WT PheDH shows much higher inhibition by the unreactive
enantiomer than any of the mutant enzymes. This contrasting
pattern of behaviour in WT PheDH presumably reflects the fact
that it has been evolutionarily perfected for a perfect fit to the
aromatic substrate. PheDH from B. sphaericus, either in its wild-
type form or generated mutants, has not been previously tested
either with racemic amino acids, or with pure D-enantiomers.
However, Asano et al. reported a study on PheDH from B. badius
(82% similarity with our enzyme),13 in which they tested the
enzyme with a series of substrates and compared activities with
L-Phe as a standard. B. badius PheDH retained 34% activity with
DL-4-F-phenylalanine (7), in very good agreement with our results
where the WT retains a 40% activity with 7 when compared with
L-Phe.
colour change (blue and colourless respectively for oxidised and
reduced) (Scheme 2).
Synthesis of D-4-Cl-phenylalanine
To establish whether these biocatalysts could be used in the
racemic resolution of amino acids, one substrate among those
tested was selected as an example for further investigations
(compound 8). Using the mutant N145A PheDH, chosen both
for its higher activity with that substrate and relatively moderate
inhibition (activity of 21.4 U mg−1 even in the presence of the
equimolar D-enantiomer), we optimised conditions for a successful
preparation of D-4-Cl-phenylalanine starting from the cheaper DL-
4-Cl-phenylalanine. In view of the expense of the cofactor NAD+,
a coupled recycling system was introduced (Scheme 1), allowing
the use of catalytic amounts of this reactant.
Recycling systems reported in the literature mainly utilise
NADH oxidases (NOX) and molecular oxygen to re-oxidise the
cofactor,7 yielding H2O or H2O2. Unfortunately, the optimal pH
range for these enzymes in terms of both stability and optimal
activity is between pH 5 and 7, quite unsuitable for our experiment
in view of the high pH optimum of PheDH. The diaphorase used
here is much more stable at alkaline pH values.
Scheme 2 Oxidoreduction mechanism of 2,6-dichlorophenolindophenol.
A further aspect to consider was the instability of NAD+ at
high pH over extended periods. NAD+ was therefore incubated at
different pH values and periodically assayed spectrophotometri-
cally with alcohol dehydrogenase from Saccharomyces cerevisiae
(see Experimental). In view of the results (not shown), the pH
was lowered from 10.4, found to be optimal for activity in the
experiments reported above, to 9.5, thus considerably decreasing
cofactor breakdown. The overall kinetic resolution was followed
by chiral HPLC10,11 with the pure enantiomers and 2-oxo-acid
separately injected as standards. The reaction went to completion
(99%) after 40 hours (Fig. 1) without any decline in the peak for the
D-enantiomer, confirming that only the L-enantiomer is recognised
by the enzyme as a substrate.
The pure D-4-Cl-phenylalanine was recovered from the crude
reaction mixture by ion-exchange chromatography with a final
yield of 85%.
To maximise the concentration of dissolved oxygen, necessary
for the reoxidation of NADH, the reaction was performed in a
large vessel with constant orbital shaking. To further facilitate
reaction with oxygen, 0.1 mg of 2,6-dichlorophenolindophenol
(DCPIP) was added to the mixture as a mediator. DCPIP is a
readily autoxidisable dye with an oxidoreduction characterised by
A second experiment was carried out under the same conditions
(substrate and enzyme concentrations) utilising WT PheDH as
biocatalyst. It can be noted that, while the mutant N145A PheDH
is certainly more active than the unmodified enzyme with the
substrate DL-4-Cl-Phe (21.6 U mg−1 and 0.7 U mg−1 respectively),
the WT PheDH is less inhibited by the D-amino acid (12.5%
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3611–3615 | 3613
©

View Article Online
Fig. 1 Chromatogram of the crude reaction mixture after 40 hours. The oxoacid elutes after 3 min (arrow) and the peak at 8.1 min corresponds to
the D-enantiomer. The L-enantiomer would be expected at ∼6.7 min (arrow), as visible in the inset graph, which shows a close-up of the trace for the
DL-starting material.
inhibition versus 50% for the mutant). The reaction progressed
much more slowly, as expected, reaching only 35% conversion
after 40 hours.
catalyse the reaction with non-natural substrates, its much lower
activity requires a correspondingly much higher ratio of enzyme
to substrate for the reaction time to be comparable with that of the
mutant. The fact that the WT enzyme seems to be less inhibited
by the D-amino acids does not adequately compensate for the low
activity. Finally, it can be envisaged that the separated oxoacid
could be fed straight back into the reaction under the synthetic
conditions described by Paradisi et al.11 to regenerate the separate,
pure L-enantiomer, thus completing the resolution of the racemate.
The method described here appears to be both robust and versatile.
Conclusion
PheDH mutants have proved to be versatile biocatalysts able to
accommodate non-natural substrates. The strict enantioselectivity
of the unmutated enzyme with the natural substrate, Phe, has been
fully retained, but even though the D-enantiomer is never a sub-
strate for the biocatalyst, it can nevertheless be a strong inhibitor.
The extent of inhibition (at a fixed substrate concentration) by the
D-enantiomer varies across the substrates and mutants, making the
optimal pairing of biocatalyst and substrate an important choice.
With other enzymatic systems such as lipases and hydrolases,
this phenomenon is less common: while product inhibition is
well investigated, no direct inhibition studies of the less reactive
enantiomer have been reported.14 D-Amino acid inhibition is, on
the other hand, a well exploited phenomenon with peptidases,
and novel drugs aimed at inhibition of proteases often do contain
D-amino acids.15,16 In our opinion, the inhibition observed with
PheDH mutants can be explained by the broadened substrate
selectivity introduced with the mutations: the side chain of an
amino acid and the a-carboxyl group are the key binding elements,
but even though both enantiomers can be accommodated in the
active pocket, correct orientation of the amino group and the
proton to be removed, and therefore catalysis, only occurs with the
L-isomer. The D-enantiomer effectively competes with the L- for the
active site in a reversible manner, and, since it is not removed by the
catalytic reaction, over time, the non-reactive molecule produces
a somewhat similar effect to the (inhibitory) accumulation of a
product. The results obtained here are generic, in the sense that
the system could readily be employed in the same way with any
suitable pairing of biocatalyst and racemic amino acid mixture
e.g. from the results in Tables 1 and 2. The biocatalyst efficiently
removes 100% of the L-enantiomer, leaving a pure peak of the D-
stereoisomer. Although WT PheDH could in principle be used to
Experimental
All the amino acids were obtained from Bachem (Switzerland).
Laboratory chemicals were reagent-grade and used without
further purification. HPLC-grade solvents were used for HPLC
separations. NAD+ (grade II) was purchased from Roche. The
wild-type PheDH and the mutant enzymes N145A/V/L/I in
which the Asp residue in position 145 in the active pocket of
the biocatalyst had been replaced with Ala/Val/Leu/i-Leu were
over-expressed in E. coli TG1 cells and purified as described
elsewhere.17,18 The enzymes were stored as precipitates in 60%
ammonium sulfate at 4 ◦C and desalted before use on PD-10
desalting columns from Amersham Biosciences.
Evaluation of enzyme activity—UV assays
Each amino acid (0.2 mM or 0.4 mM) was dissolved to form a
component of a reaction mixture containing KCl (100 mM) and
Gly–NaOH buffer (50 mM). The pH was adjusted to 10.4 by
adding a suitable amount of NaOH. In view of the instability
of NAD+ at pH 10.4, the coenzyme was prepared as a stock
solution in water (125 mM) and 20 lL was added to the
reaction mixture immediately before starting each experiment
(final concentration of NAD+ 2.5 mM). A total volume of 1 mL
of reaction mixture was incubated at 25 ◦C. The reaction was
followed spectrophotometrically at 340 nm over 1 minute after
3614 | Org. Biomol. Chem., 2008, 6, 3611–3615
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
adding an appropriate amount of enzyme to achieve an optimally
measurable reaction rate (between 0.01–0.03 Abs units per min).
Each reaction was carried out in duplicate and the average rate is
reported.
2 A. S. Bommarius, M. Schwarm, K. Stingl, M. Kottenhahn, K.
Huthmacher and K. Drauz, Tetrahedron: Asymmetry, 1995, 6, 2851–
2888.
3 J. J. Lynch III, P. Honore, D. J. Anderson, W. H. Bunnelle, K. H.
Mortell, C. Zhong, C. L. Wade, C. Z. Zhu, H. Xu, K. C. Marsh, C. H
Lee, M. F. Jarvis and M. Gopalakrishnan, Pain, 2006, 125, 136–142.
4 R. M. Kellogg, J. W. Nieuwenhuijzen, K. Pouwer, T. R. Vries, Q. B.
Broxterman, R. F. P. Grimbergen, B. Kaptein, R. M. La Crois, E. de
Wever, K. Zwaagstra and A. C. Van Der Laan, Synthesis, 2003, 10,
1626–1638.
Resolution of 4-Cl-phenylalanine
The reaction was performed at pH 9.5 (ethanolamine–HCl, 20 mL,
50 mM) and at room temperature (15–20 ◦C) in a vessel with
a large surface area and shaking or stirring to maximise the
exchange with molecular oxygen. The reaction mixture was 5–
10 mM in DL-Cl-phenylalanine (9), 1 mM in NAD+, 100 mM in
KCl and contained 0.1 mg DCPIP (dichlorophenol indophenol).
25 lg of the biocatalyst (N145A PheDH or WT PheDH) and
0.1 mg diaphorase were added to initiate the reaction, which
was monitored by chiral HPLC (CHIROBIOTIC T column,
eluent 80 : 20 MeOH–H2O 1 mL min−1) as reported previously.10
Upon completion, the reaction mixture was adsorbed onto an
ion exchange resin, Dowex 50WX8-200. The resin was washed
with distilled water, and eluted with 25% NH4OH to give D-4-Cl-
phenylalanine (78% 15.6 mg). 1H NMR (500 MHz, D2O, NaOD)
d ppm 2.66 (dd J = 13.56, 7.28 Hz, 1H), 2.79 (dd J = 13.53,
5.50 Hz, 1H), 3.31 (t, J = 6.38 Hz 1H), 7.06 (d, J = 8.12 Hz,
2Ar-H), 7.19 (d, J = 8.14 Hz, 2Ar-H).
5 V. Constantinou-Kokotou, V. Magrioti, T. Markidis and G. Kokotos,
J. Pept. Res., 2001, 58, 325–331.
6 M. R. Kula and C. Wandrey, Methods Enzymol., 1988, 136, 34–45.
7 W. Hummel, M. Kuzu and B. Geueke, Org. Lett., 2003, 5, 3649–3650.
8 C. Carboni, H. G. T. Kierkels, L. Gardossi, K. Tamiola, D. B. Janssen
and P. J. L. M. Quaedflieg, Tetrahedron: Asymmetry, 2006, 17, 245–251.
9 K. Vedha-Peters, M. Gunawardana, J. D. Rozzell and S. J. Novick,
J. Am. Chem. Soc., 2006, 128, 10923–10929.
10 P. Busca, F. Paradisi, E. Moynihan, A. R. Maguire and P. C. Engel,
Org. Biomol. Chem., 2004, 2, 2684–2691.
11 F. Paradisi, S. Collins, A. R. Maguire and P. C. Engel, J. Biotechnol.,
2007, 128, 408–411.
12 Y. Asano, A. Nakazawa and K. Endo, J. Biol. Chem., 1987, 262, 10346–
10354.
13 Y. Asano, A. Nakazawa, K. Endo, Y. Hibino, M. Ohomori, N. Numao
and K. Kondo, Eur. J. Biochem., 1987, 168, 153–159.
14 I. S. Ng and S. W. Tsai, Biotechnol. Bioeng., 2005, 89, 88–95.
15 D. C Spellmeyer, S. Brown, G. B. Stauber, H. M. Geysen and R. Valerio,
Bioorg. Med. Chem. Lett., 1993, 3, 1253–1256.
16 L. N. Jungheim, T. A. Shepherd, A. J. Baxter, J. Burgess, S. D. Hatch,
P. Lubbehusen, M. Wiskerchen and M. A. Muesing, J. Med. Chem.,
1996, 39, 96–108.
17 S. Y Seah, K. L. Britton, D. W. Rice, Y. Asano and P. C. Engel,
Biochemistry, 2002, 41, 11390–11397.
18 S. Y. Seah, K. L. Britton, P. J. Baker, D. W. Rice, Y. Asano and P. C.
Engel, FEBS Lett., 1995, 370, 93–96.
References
1 F. Paradisi, G. Porzi and S. Sandri, Tetrahedron: Asymmetry, 2001, 12,
3319–3324.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3611–3615 | 3615
©