Enzymatic conversion of unnatural amino acids by yeast D-amino acid oxidase

Article abstract of DOI:10.1002/adsc.200606188

Unnatural amino acids, particularly synthetic α-amino acids, are becoming crucial tools for modern drug discovery research. In particular, this application requires enantiomerically pure isomers. In this work we report on the resolution of racemic mixtures of the amino acids D,L-naphthylalanine and D,L-naphthylglycine by using a natural enzyme, D-amino acid oxidase from the yeast Rhodotorula gracilis. A significant improvement of the bioconversion is obtained using a single-point mutant enzyme designed by a rational approach. With this D-amino acid oxidase variant the complete resolution of all the unnatural amino acids tested was obtained: in this case, the bioconversion requires a shorter time and a lower amount of biocatalyst compared to the wild-type enzyme. The simultaneous production of the corresponding α-keto acid, a possible precursor of the amino acid in the L-form, improves the significance of the procedure.

Full text of DOI:10.1002/adsc.200606188

FULL PAPERS
DOI: 10.1002/adsc.200606188
Enzymatic Conversion of Unnatural Amino Acids by Yeast
d-Amino Acid Oxidase
Antonio Caligiuri,a Paola D’Arrigo,a Elena Rosini,b Davide Tessaro,a
Gianluca Molla,b Stefano Servi,a and Loredano Pollegionib,*
a
Dipartimento di C.M.I.C “G. Natta”, Politecnico di Milano, via Mancinelli 7, 20131 Milano, Italy
Fax : (+39)-02-2399-3080
b
Dipartimento di Biotecnologie e Scienze Molecolari, Università degli Studi dell’Insubria, via J.H. Dunant 3, 21100 Varese,
Italy
Phone (+39)-0332-421-506; fax: (+39)-0332-421-500; e-mail: loredano.pollegioni@uninsubria.it
Received: April 18, 2006; Accepted: July 28, 2006
Abstract: Unnatural amino acids, particularly syn- acid oxidase variant the complete resolution of all
thetic a-amino acids, are becoming crucial tools for the unnatural amino acids tested was obtained: in
modern drug discovery research.In particular, this
this case, the bioconversion requires a shorter time
application requires enantiomerically pure isomers. and a lower amount of biocatalyst compared to the
In this work we report on the resolution of racemic wild-type enzyme.The simultaneous production of
mixtures of the amino acids d,l-naphthylalanine and the corresponding a-keto acid, a possible precursor
d,l-naphthylglycine by using a natural enzyme, d- of the amino acid in the l-form, improves the signifi-
amino acid oxidase from the yeast Rhodotorula gra- cance of the procedure.
cilis.A significant improvement of the bioconversion
is obtained using a single-point mutant enzyme de- Keywords: d-amino acid oxidase; amino acids; bio-
signed by a rational approach.With this d-amino transformations; enzyme catalysis; rational design
Introduction
factor to give hydrogen peroxide (Figure 1).In partic-
ular, DAAO from the yeast Rhodotorula gracilis
The increasing interest of the pharmaceuticals and (RgDAAO) exhibits a very high turnover number,
fine chemicals industries in the production of drugs tight binding with the coenzyme FAD, a broad sub-
and intermediates is an incentive to devise new chiral strate specificity and it has an active site large enough
technologies and to improve enantioselective process- to accommodate even bulky substrates, for example,
es.[1] In fact, the worldwide sales of formulated chiral cephalosporin C (for a review see[4,5]).The redox reac-
drugs in single-enantiomer dosage forms is growing at tion of DAAO is exploited in a number of biotechno-
an annual rate of 13%.[2] On a commercial scale, the logical applications, for example, the two-step conver-
kinetic resolution approach allows one to separate sion of cephalosporin C into 7-aminocephalosporanic
enantiomers; in this context, the use of biocatalysts acid, to detect and quantify d-amino acids, to produce
has found widespread application and is rapidly be- a-keto acids from essential d-amino acids, and to re-
coming a cheap approach.[3] A key factor for the suc- solve racemic mixtures of d,l-amino acids.
cessful development of biocatalytic processes is the
In this work, we studied the enzymatic activity of
possibility to employ on a large-scale an inexpensive wild-type RgDAAO towards racemic mixtures of un-
enzyme with suitable properties (i.e. high activity, sta- natural aromatic amino acids (see Table 1) to achieve
bility and selectivity).Here we provide an example
the resolution of the corresponding l-amino acid
for the combination of biological tools for the produc- component.We also report on the use of rational
tion of optically pure unnatural l-amino acids. d- design as a strategy to obtain an enzyme that is in-
Amino acid oxidase (EC 1.4.3.3, DAAO) is a highly creasingly more suitable for this biotechnological ap-
stereoselective flavoenzyme which catalyses the dehy- plication.Non-proteinogenic amino acids that either
drogenation of the d-isomer of amino acids to give occur naturally or are chemically synthesised, and par-
the corresponding a-imino acids and, after subsequent ticularly synthetic a-amino acids, have played a signif-
hydrolysis, a-keto acids and ammonia.Oxygen, the
final redox acceptor, reoxidises the reduced FAD co- more, and due to their structural diversity and func-
icant role in the area of peptide research.[6] Further-
Adv. Synth. Catal. 2006, 348, 2183 – 2190
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
2183

Antonio Caligiuri et al.
FULL PAPERS
(e.g., l-DOPA), as well as components of numerous
therapeutically relevant compounds (e.g., d-2-naph-
thylalanine is found in the peptide drug Nafarelin).[8]
It is thus evident that unnatural amino acids are be-
coming indispensable tools in drug discovery efforts.
Results and Discussion
Kinetic Properties of Wild-Type RgDAAO on
Naphthylamino Acids
Table 1 summarises the apparent kinetic parameters
determined for wild-type RgDAAO towards unnatu-
ral amino acids 1–4 in comparison with the values ob-
tained for the reference substrate d-alanine (the cor-
responding Michaelis–Menten plots are reported in
Figure 2).The parameters were determined under
steady state conditions at 258C, pH 8.5, and at an
oxygen saturation of 21% (the assay mixture was
equilibrated with air).Wild-type RgDAAO shows a
higher substrate affinity for all naphthylamino acids
tested compared to d-alanine (in particular, signifi-
cantly lower Km values were determined for amino
acids 1–3).On the other hand, a appreciably higher
Figure 1. Reaction catalysed by d-amino acid oxidase.
specificity constant (as expressed by the ratio Vmax,app
/
tional versatility, they are widely used as chiral build-
Km,app) was only observed on d-2-naphthylalanine (1)
ing blocks and molecular scaffolds in constructing which is mainly due to a higher substrate affinity (see
chemical combinatorial libraries.Unnatural amino Table 1); the low Vmax,app values (up to ca. 600-fold
acids are of great value for the de novo design of pep- lower for d-2-naphthylglycine than for d-alanine) sig-
tides and proteins with a high propensity to fold with nificantly affect the specificity constant with the sub-
a predetermined secondary or tertiary structure[7] but strates 2–4.The Michaelis–Menten plot of the activity
are also valuable pharmaceuticals in their own right of wild-type RgDAAO on d-2-naphthylalanine (1)
Table 1. Comparison of the apparent steady-state kinetic parameters determined for wild-type and M213G RgDAAOs on
different d-amino acids.[a]
Substrate
Vmax,app [minÀ1]
Km,app [mM]
Vmax,app/Km,app [minÀ1 mmÀ1]
d-alanine
Wild-type
M213G
Wild-type
M213G
3900Æ200
770Æ95
0.9Æ0.1
4330
160
49100
19500
4.9Æ0.5
1965Æ80
1170Æ72
0.04Æ0.003
0.06Æ0.009
d-2-Naphthylalanine (1)[b]
d-1-Naphthylalanine (2)
Wild-type
M213G
125Æ8
0.04Æ0.007
0.03Æ0.003
3100
29000
870Æ23
Wild-type
M213G
6Æ0.1
0.01Æ0.001
0.03Æ0.002
600
1035
d-2-Naphthylglycine (3)
31Æ0.7
Wild-type
M213G
53Æ4
92Æ3
0.33Æ0.05
160
1840
d-1-Naphthylglycine (4)
0.05Æ0.006
[a]
Reaction conditions: the activity was determined using the oxygen consumption assay, at 258C, pH 8.5, and at air satura-
tion.The activity values were not modified by the use of the racemic d,l-mixture instead of the pure d-isomer.
Because of the substrate inhibition effect, the kinetic parameters on compound 1 were determined using the activity
values obtained in the 0–0.15 mM substrate concentration range.
[b]
2184
www.asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 2183 – 2190

Conversion of Unnatural Amino Acids by Yeast d-Amino Acid Oxidase
FULL PAPERS
(Figure 2A) shows a substrate inhibition effect[9] that
is evident at an amino acid concentration ꢀ0.1 mM.
On the other hand, a similar effect is not observed
using the compounds 2–4.The determination of the
kinetic parameters is not affected by the use of the re-
cemic mixture instead of the pure d-isomer, thus con-
firming that the l-isomer is not an inhibitor of the
DAAO reaction.[4,5] In conclusion, the investigation of
the substrate specificity of wild-type RgDAAO to-
wards these unnatural amino acids shows that d-2-
naphthylalanine (1) is the best substrate.
Bioconversions using Wild-Type RgDAAO
We investigated the ability of wild-type RgDAAO in
the resolution of racemic mixtures of unnatural aro-
matic amino acids 1–4 following the time course of
bioconversion using a stirred reactor and the enzyme
in the free form (Figure 3).The products of the bio-
conversion reaction were quantified by HPLC.In all
cases, a higher rate and total conversion were ob-
tained by bubbling air into the apparatus and in the
presence of catalase (data not shown).After 300 min
of reaction, the chromatograms show that over 90%
of the d-enantiomer has been consumed using sub-
strates 1, 2 and 4 (Figure 3A, B, D).Under optimized
conditions, the maximal conversion is reached using
different amounts of enzyme depending on the sub-
strate used (see legend of Figure 3 and Table 2).The
conversion of d-2-naphthylalanine (1) requires the
lowest enzyme concentration (40 U/L every hour), in
agreement with the higher specificity constant of
wild-type RgDAAO on this substrate (Table 1): a
total conversion of the d-isomer (ee >99%) is
reached in ca. 220 min.On the other hand, the con-
version of compound 3 is only ca. 60% even if a large
amount of wild-type RgDAAO has been used (ca.
6000 U/L).
Molecular Docking of Unnatural d-Amino Acids at
the Active Site of RgDAAO
At first, the mode of interaction of d-2-naphthylala-
nine (1), the best substrate among the unnatural d-
amino acids under investigation, at the active site of
RgDAAO was studied using the software AutoDock
3.0.[10] This program employs a Monte Carlo simulat-
ed annealing strategy where solutions (ligand-protein
complexes) possessing the lowest energy are the most
probable ones.The crystal structure of RgDAAO in
complex with d-trifluoroalanine (PDB entry 1c0L)
was used as the macromolecular target.[11] As shown
in Figure 4A and B, the d-2-naphthylalanine side
chain points towards the entrance of the active site of
RgDAAO engaging in strong hydrophobic and van
Figure 2. Michaelis–Menten plot of the activity values deter-
mined for wild-type (filled circles) and M213G (filled
squares) RgDAAOs on compound (1, A), (2, B), (3, C), and
(4, D).Bars indicate ÆSEM for at least three determina-
tions; where not shown, the error is smaller than the symbol
used.
Adv. Synth. Catal. 2006, 348, 2183 – 2190
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.asc.wiley-vch.de
2185

Antonio Caligiuri et al.
FULL PAPERS
der Waals interactions with the aromatic side chains
of F58 and Y238.Noteworthy, docking simulations
show that these interactions modify the correct posi-
tioning of the ligand at the active site by increasing
the distance between the N(5) position of the FAD
co-factor and the a-carbon of the substrate (Fig-
ure 4C).Anyway, and in agreement with the high spe-
cific activity of RgDAAO on d-2-naphthylalanine (1),
it is conceivable that random conformational motions
of protein side chains in solution allow the substrate
to reach a position suitable for oxidation.In fact, mo-
lecular modelling simulations show that minor move-
ments of the F58 and Y238 side chains could modify
the ligand positioning at the active site (not shown).
Interestingly, binding of the substrate (as well as of
the product) molecule between the side chain of these
two latter residues results in a partial hindrance of
the active site access (Figure 4B), giving the rationale
of the observed substrate inhibition effect (Fig-
ure 2A).
In order to understand the rationale of the low ac-
tivity measured for wild-type RgDAAO on the d-
naphthylglycine derivatives, a similar docking analysis
was performed using compounds 3 and 4 as ligands.
The results show that, although d-naphthylglycine
possesses a slightly shorter side chain, its major rigidi-
ty (due to the lack of a CH2 group) does not allow a
correct positioning of this substrate in the active site.
Modelling of d-2-naphthylglycine at the active site of
wild-type RgDAAO according to the positioning of
d-2-naphthylalanine (see above) clearly shows a clash
between the substrate side chain and the side chain of
the active site residue M213 (see Figure 4A).There-
fore, in an effort to improve the binding of compound
3 in the correct orientation at the active site of
RgDAAO, the M213 residue was mutated in glycine.
We have previously shown that the substrate specifici-
ty of yeast DAAO is modulated by the side chains of
residues belonging to the hydrophobic binding pocket
of the active site that interacted with the side chain of
the substrate.[12] Among these residues, M213 ap-
peared to be especially important and thus we have
been successful in attaining an enzyme active on
acidic d-amino acids using a structure-based rational
design approach (M213R mutant).[12] The substrate d-
2-naphthylglycine docked at the active site of the
model of the M213G mutant RgDAAO is depicted in
Figure 4D: it clearly shows that in this case the sub-
strate reaches the correct position for oxidation, that
Figure 3. Comparison of the time course of bioconversion of
compounds 1–4 in a stirred reactor using wild-type (filled
circles) and M213G (filled squares) RgDAAOs at 258C,
pH 8.5, and air bubbling. The arrows indicate the enzyme
additions. A: To 1. 2 mM of d,l-2-naphthylalanine is added
40 U/L (every hour) of wild-type RgDAAO or 20 U/L (at
the beginning of the bioconversion reaction) of M213G
RgDAAO. B: To 1. 2 mM d,l-1-naphthylalanine is added 105
U/L (every 60 min) and 20 U/L (every 40 min) of wild-type
and M213G RgDAAOs, respectively. C: To 0.75 mM d,l-2-
naphthylglycine is added 1500 U/L (every hour) of wild-type
RgDAAO or 100 U/L (at the beginning of the bioconver-
sion reaction) of M213G RgDAAO. D: To 1. 5 mM d,l-1-
naphthylglycine is added 200 and 100 U/L wild-type and
M213G RgDAAOs, respectively, every hour.
À
is, the a-C H bond of the substrate is in front of the
N(5) flavin position (at a distance <3.5 Š).
Purification and Properties of M213G RgDAAO
The M213G RgDAAO protein was expressed in E.
coli and purified using the protocol for the wild-type
2186
www.asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 2183 – 2190

Conversion of Unnatural Amino Acids by Yeast d-Amino Acid Oxidase
FULL PAPERS
Table 2. Comparison of the bioconversion results of racemic mixtures of unnatural amino acids (1–4) under optimized condi-
tions (and at 258C, air bubbling and pH 8.5).
Compound
Enzyme
form
Concentration
(d,l-form) [mM] of d-form [%]
Conversion yield Conditions
Enzyme concentra-
tion [U/L][a]
Time
[min]
40 every 60 min (160
total units)
20
d-2-Naphthyl-
alanine (1)
Wild-type 1.2
M213G 1.2
>99
>99
220
25
105 every 60 min
(735 total units)
20 every 40 min (60 100
total units)
d-1-Naphthyl-
alanine (2)
Wild-type 1.2
M213G 1.2
>96
>99
400
1500 every 60 min
(6000 total units)
100
d-2-Naphthylgly-
cine (3)
Wild-type 0.75
M213G 0.75
ꢁ60
>99
240
30
200 every 60 min
d-1-Naphthylgly-
cine (4)
Wild-type 1.5
M213G 1.5
>95
>99
(1000 total units)
100 every 60 min
(500 total units)
300
300
[a]
The enzymatic units are those determined on d-alanine as substrate.
RgDAAO containing an His-tag at the N-terminus.[13] enzyme (in terms of enzymatic units) and in a shorter
Starting from 10 L fermentation broth, 30 mg of pure time (see Table 2 and Figure 3).In particular, the con-
enzyme with a specific activity of 19.2 U/mg protein version of the d-isomer of compound 3 is complete in
were achieved: an overall purification yield of 67% only 30 min and using a 60-fold lower amount of
was obtained.The protein was about 95% pure and
yielded a single band at 40 kDa in SDS-PAGE.Like
enzyme than with wild-type RgDAAO (moreover,
with this latter enzyme form only a partial conversion
the wild-type RgDAAO, the mutant enzyme is puri- is obtained, Figure 3C).
fied as a holoenzyme (the E278nm/E455nm ratio is 8.5)
As an example of the possibility to employ
and is stable when stored at À208C for several RgDAAO on a preparative scale, the bioconversion
months.
of 1 mM d,l-2-naphthylalanine (60 mg total, 240 mL,
The maximal activity on the best substrates, d-ala- pH 8.5) was performed. By using 80 U/L of wild-type
nine and d-2-naphthylalanine, is significantly reduced RgDAAO or 20 U/L of M213G mutant, the produc-
in the mutant enzyme (Table 1).On the other hand,
tion of a-keto acid reached a plateau after 40 min and
an increase in Vmax,app is evident from compounds 2–4. 20 min, respectively, corresponding to a 50% amino
In the case of compounds 2 and 4, this change is also acid conversion.The residual amino acid was isolated
accompanied by an increased affinity for the substrate from the produced a-keto acid by standard ion ex-
and thus leads to a significant increase in the specifici- change chromatography on a Dowex column.In both
ty constant Vmax,app/Km,app value (up to 10-fold for d-1- cases the enantiomeric excess was >99.9% as judged
naphthylalanine).Interestingly, the substrate inhibi-
by chiral chromatography and the overall yield was
tion effect with compound 1 is less evident for the >90% (>27 mg of recovered product).
M213G mutant than for wild-type DAAO (see Fig-
ure 2A).
Conclusions
Bioconversions using M213G RgDAAO
In the present work, we provide an example for the
integration of an “in silico” analysis and a biological
The resolution of racemic mixtures of the unnatural tool such as site directed mutagenesis in the develop-
amino acids 1–4 was newly performed using the ment of resolution processes for the production of op-
M213G RgDAAO mutant and the same experimental tically pure unnatural amino acids.The assay of wild-
conditions used with the wild-type enzyme (see type RgDAAO on the unnatural d-amino acids 1–4
above).In all cases, the complete resolution of the
d,l-mixtures is obtained using a lower amount of RgDAAO is active on all the compounds tested and
confirms its broad substrate specificity.[4] In fact,
Adv. Synth. Catal. 2006, 348, 2183 – 2190
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.asc.wiley-vch.de
2187

Antonio Caligiuri et al.
FULL PAPERS
shows a higher specificity constant with naphthylala-
nine (compounds 1, 2), than with naphthylglycine de-
rivatives (compounds 3, 4) (Table 1), probably be-
cause of the higher flexibility of the substrate side
chain.In particular, compound 1 is the best substrate
for the wild-type enzyme: the high specificity constant
on this compound is mainly due to a higher substrate
affinity (Table 1).On the other hand, using the wild-
type RgDAAO only a partial resolution of d,l-2-
naphthylglycine has been obtained, and for com-
pounds 2 and 4 the complete conversion of the d-
isomer required the addition of a huge amount of
enzyme.
Models of RgDAAO-d-naphthylamino acid com-
plexes were simulated by AutoDock software and
showed that the low activity observed with naphthyl-
glycine derivatives is largely due to the steric hin-
drance of the side chain of M213.The substitution of
M213 to glycine does not result in gross perturbation
of the FAD microenvironment but affected the kinet-
ic properties of the enzyme.This substitution allowed
a better interaction with compounds 2–4, resulting in
an increase in maximal activity and/or substrate affini-
ty (see Table 1).Such a change in substrate specificity
allows a complete resolution of the racemic mixtures
of all the four unnatural d-amino acids using a signifi-
cant lower amount of enzyme (see Table 2 and
Figure 3).
In conclusion, the high activity on naphthylamino
acids and the high yield of d-isomer conversion ob-
served with the M213G RgDAAO mutant appear to
be a major advantage of the process we have success-
fully developed on a laboratory scale.The overall
conversion is not a simple kinetic resolution: in fact
the a-keto acid produced is the substrate for an
amino transferase or an amino acid dehydrogenase.[14]
If enzymes with l-specificity are applied, the two con-
vergent steps (oxidation of the d-enantiomer followed
Figure 4. A: Proposed model of the positions of substrates
1 and 3 at the active site of RgDAAO as predicted by dock-
ing analysis using the program AutoDock 3.0 (the position
of substrate 3 was modelled based on the orientation ob-
tained for compound 1).This software combines a rapid,
grid-based method for evaluation of the ligand-protein inter-
action energies with a Monte Carlo simulated annealing
search algorithm for optimal conformation of ligands.[10,20] B:
The same complex as in A for substrate 1 seen from outside
the active site; protein surface is depicted in cyan, F58 and
Y238 in red and M213 in ochre.The bulk naphthyl group of
the substrate is shown in green using van der Waals repre-
sentation. C: Alternative docking solutions at the active site
of RgDAAO highlighting the interaction between the naph-
thyl moiety of 1 and the aromatic side chains of F58 and
Y238. D: The model of M213G DAAO mutant was devel-
oped using Swiss PDB Viewer program; the docking of the
d-2-naphthylglycine (3) at the active site was achieved as re-
ported above (see legend of Figure 4A).
2188
www.asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 2183 – 2190

Conversion of Unnatural Amino Acids by Yeast d-Amino Acid Oxidase
FULL PAPERS
was reached.The white precipitate was collected by filtra-
by reductive amination of the produced a-keto acid
to the l-amino acid) should produce a deracemisation
of the starting amino acid.Moreover it has been re-
cently shown that the in situ chemical reduction of
the initially formed imine can also converge to a final
complete conversion into one single enantiomer.[15]
tion and dried under vacuum to give d,l-1-naphthylglycine;
1
yield: 0.4 g (47%); H NMR (DMSO-d6): d=4.98 (s, 1H),
7.46–7.59 (m, 4H), 7.85–7.95 (m, 2H), 8.31 (d, 1H).
Activity Assay and Kinetic Measurements
DAAO activity was assayed with an oxygen electrode at air
saturation (0.253 mM O2) and 258C, using 28 mM d-alanine
as substrate in 75 mM sodium pyrophosphate buffer, pH 8.5.
One DAAO unit is defined as the amount of enzyme that
converts 1 mmol of d-amino acid per minute at 258C.The ki-
netic parameters of the DAAO reaction with different
amino acids (both d- and d,l-forms) were determined in
100 mM sodium pyrophosphate buffer, pH 8.5, and at 258C.
The initial reaction rates were used to calculate the kinetic
parameters employing the KaleidaGraph software (Synergy
Software, PA, U.S.A.). The pure d-isomer of compound 1
and 2 was purchased from Bachem and Sigma Aldrich, re-
spectively; the D-isomer of compounds 3 and 4 was isolated
by chiral HPLC chromatography as detailed below but
using the Crownpak CR SemiPrep column (10”150 mm,
5 mm).
Experimental Section
Enzyme Expression and Purification
The M213G mutant was generated by site-directed muta-
genesis (QuikChange kit, Stratagene) using as template the
cDNA coding for His-tagged wild-type DAAO.[13] The muta-
tion was introduced using the following 37-mer mutagenic
oligonucleotides M213G-up: GCAAGCGATGCACGGGA-
GACTCGTCCGACCCCGCTTC;
GAAGCGGGGTCGGACGAGTCTCCCGTG-
M213G-dn:
CATCGCTTGC.The introduction of the desired mutation
was confirmed by automated DNA sequencing.Enzymatic
DNA modifications were carried out according to the manu-
facturerꢁs instructions and essentially as described.[16] Re-
combinant wild-type and M213G RgDAAOs were expressed
Bioconversion
in BL21(DE3)pLysS E. coli cells using the pT7-HisDAAO
N
expression vector and purified by HiTrap Chelating chroma-
tography (GE Healthcare Bio-sciences).[13] The best expres-
sion of the mutant enzyme was obtained by growing the
cells at saturation, inducing the protein expression with
0.8 mM IPTG and collecting the cells after 3 h of growth at
308C.The pure enzymes possess a specific activity on d-ala-
nine of 110 and 19.2 U/mg protein for wild-type and M213G
RgDAAO, respectively.The enzyme concentration was de-
termined spectrophotometrically by using an extinction co-
efficient of 12.6 mMÀ1 cmÀ1.
The time course of bioconversion was determined by high
performance liquid chromatography (HPLC) assay.The re-
action mixture contained (in a final volume of 35 mL)
100 mm sodium phosphate buffer, pH 8.5, 0.5–1.5 mM d,l-
amino acid [highest solubility was: (1) 1.20 mM, (2)
1.20 mM, (3) 0.75 mM, (4) 1.50 mM] and different amounts
of wild-type (160–6000 U/L) or M213G (20–500 U/L)
RgDAAO and catalase (from Sigma Aldrich, 280 U/liter).
Reaction mixtures were incubated at 258C and aliquots
were drawn at different times (up to 500 min).Samples
were separated by HPLC on a Zorbax SB-Aq (150”
4.6 mm, 5 mm; Agilent Technologies) using TFA 0.1%/ace-
tonitrile (70/30), flow 1 mLminÀ1 for 1–2 and TFA 0.1%/
acetonitrile (80/20), flow 1.1 mLminÀ1 for 3–4 , with benzoic
acid as internal standard (UV detector at 254 nm).Enantio-
meric purity of the final products was determined by means
of chiral HPLC on a Crownpak CR+ column (4”150 mm,
5 mm, Chiral Technologies) using aqueous perchloric acid/
methanol (90/10) (except where stated otherwise).The fol-
lowing Rt values have been determined: compound 1 d=
31.6 min, l=41.5 min (pH 1.5, flow=1 mLminÀ1); com-
Preparation of Compounds 1–4
d,l-2-Naphthylalanine (rac-1): To a suspension of 5 g of
monoethyl (2-naphthylmethyl) acetamidomalonate[17] in
100 mL water-ethanol (1:1), 1 mL of 12 N HCl was added
and the suspension was refluxed for 12 h under magnetic
stirring.The reaction was monitored by TLC (AcOEt:
i-PrOH:AcOH, 2:2:1).After this time 10 N NaOH was
added dropwise under stirring until a pH of about 6.2 was
reached.The white precipitate was collected by filtration
and dried under vacuum to afford d,l-2-naphthylalanine;
pound
2
d=33.2 min, l=40.5 min (pH 1.5, flow=
1
1 mLminÀ1); compound 3 d=13.4 min, l=28.2 min (pH 2.0,
flow=1.2 mLminÀ1); compound 4 d=10.4 min, l=20.3 min
(aqueous perchloric acid, pH 2.0, flow=1.0 mLminÀ1).The
isolation of the residual l-amino acid from the produced a-
keto acid during a preparative bioconversion was performed
by standard ion exchange chromatography on a Dowex 50”
8 column (the bound l-amino acid was eluted with 1M am-
monia).
yield: 2.3 g (70%). H NMR (DMSO-d6, TFA-d4): d=3.30
(m, 2H), 4.23 (s, 1H), 7.4–7.56 ( (m, 3H), 7.75–7.92 (m,
4H), 8.37 (br s, 2H).
d,l-1-Naphthylalanine (rac-2): Obtained from Sigma Al-
drich.
d,l-2-Naphthylglycine (rac-3): Prepared according to the
literature.[18]
d,l-1-Naphthylglycine (rac-4): In a 50-mL flask, 20 mL of
a 20% sodium hydroxide solution were mixed with 5 mL of
[19]
acetone.1 g of 1-naphthylglycine amide
was added and
Molecular Modelling Studies
the solution was brought to reflux under magnetic stirring.
The reaction was monitored by TLC (AcOEt:i-PrOH:
AcOH, 2:2:1).After 24 h the mixture was cooled to room
temperature, and treated with 20 mL of water.8 mL of 37%
HCl were added under stirring at 08C until a pH of ca. 6.3
Automated ligand docking was performed by AutoDock 3.0,
a suite of automated docking tools based on a Monte Carlo
simulated annealing approach.[20] The d-enantiomers of the
compounds 1 and 3 were prepared applying a mirror sym-
Adv. Synth. Catal. 2006, 348, 2183 – 2190
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.asc.wiley-vch.de
2189

Antonio Caligiuri et al.
FULL PAPERS
metry to the coordinates of the corresponding l-forms,
downloaded from the macromolecular structure database of
EMBL-EBI (http://www.ebi.ac.uk/msd-srv/chempdb/cgi-bin/
cgi.pl). Swiss PDBviewer 3.7 was used for protein three-di-
mensional structure visualisation and for the preparation of
the model of the M213G mutant RgDAAO.[21]
[8] P.P.Taylor, D.P.Pantaleone, R.F.Senkpeil , I.G.Fo-
theringham, Trends in Biotechnology 1998, 16, 412–
418.
[9] W.W. Cleland, Contemporary enzyme kinetics and
mechanism, (Ed.: D. L. Purich), Academic Press, New
York, 1983, pp.253–266.
[10] G.M. Morris, D.S. Goodsell, R.S. Halliday, W.E.
Huey, R.K. Belew, A.S. Olson, J. Comp. Chem. 1998,
19, 1639–1662.
[11] S.Umhau, L.Pollegioni, G.Molla, K.Diederichs, W.
Acknowledgements
Welte, MS. .Pilone, S.Ghisla,
Proc. Natl. Acad. Sci.
U.S.A. 2000, 97, 12463–12468.
[12] S.Sacchi, S.Lorenzi, G.Molla, M.S.Pilone, C.Rosset-
ti, L.Pollegioni, J. Biol. Chem. 2002, 277, 27510–27516.
[13] S.Fantinato, L.Pollegioni, MS..Pilone,
This work was supported by a grant from FAR to LP and
GM and from Fondazione Cariplo to LP. Financial support
from COFIN-MURST to SS is acknowledged.
Enzyme
Microb. Technol. 2001, 29, 407–412.
[14] a) HS..Bae, SG. .Lee, SP..Hong, MS..Kwak, N.
References
Esaki, K.Soda, M.H.Sung,
1999, 6, 241–247; b) S.P.Crump, J.D.Rozzell, in.
catalytic production ofamino acids and derivatives ,
(Eds.: J.D.Rozell, F.Wagner), Wiley, New York, 1992,
Mol. Cat. B: Enzymatic
Bio-
[1] a) M.K. OꢁBrien, B. Vanasse, Curr. Opin. Drugs Dis-
covery Dev. 2000, 3, 793–806; b) A.J.J. Straathof, S.
Panke, A.Schmid, Curr. Opin. Biotechnol. 2002, 13,
548–556; c) H.-U. Blaser, Chem. Commun. 2003, 7,
293–296.
pp.43–58; c) T.Li, A.B.Kootstra, I.G.Fotheringham,
Org. Proc. Res. Dev. 2002, 6, 533–538.
[15] a) K.Soda, T.Oikawa, K.Yokoigawa,
J. Mol. Cat.
[2] a) Chem. Eng. News 2002, 78, 45–57; b) W.Leuchten-
B:Enzymatic, 2001, 11, 149–153; b) I.Fotheringham, I.
berger, K.Huthmacher, K.Drauz,
Applied Microb.
Archer, R.Carr, R.Speight, NJ..Turner,
Biochem.
Biotechnol. 2005, 69, 1–8.
[3] a) B.Schulze, M.G.Wubbolts,
Soc. Trans. 2006, 34, 287–290.
[16] J.Sambrook, E.F.Fritsch, T.Maniatis,
Curr. Opin. Biotechnol.
Molecular clon-
1999, 10, 609–615; b) A.Schmid, JS..Dordick, B.
ing: a laboratory manual, 2nd edn., Cold Spring
Harbor Laboratory Press, Cold Spring Harbor, NY.
1989.
Hauer, A.Kiener, M.Wubbolts, B.Witholt,
2001, 409, 258–268.
Nature
[4] M.S.Pilone, Cell. Mol. Life Sci. 2000, 57, 1732–1747.
[17] A.Berger, M.Smolarsky, N.Kurn, H.R.Bosshard,
J.
[5] M.S.Pilone, M.S., L.Pollegioni,
Biocatal. Biotransfor-
Org. Chem. 1973, 38, 457–460.
mation 2002, 20, 145–159.
[6] S.James, Chemistry Today 2003, 3, 65–68.
[7] a) J.P. Schneider, J.W. Kelly,
[18] S. Kukolja, S.E. Draheim, J.L. Pfeil, R.D. Cooper,
B.J.Graves, R.E.Holmes, D.A.Neel, G.W.Huffman,
J.A.Webber, M.D.Kinnick, R.T.Vasileff, B.J.Foster,
J. Med. Chem. 1985, 28, 1886–1896.
Chem. Rev. 1995, 95,
2169–2187; b) G.Tuchscherer, M.Mutter, Pure Appl.
Chem. 1996, 68, 2153–2162; c) M.H.J. Cordes, A.R.
Davidson, R.T.Sauer, Curr. Opin. Struct. Biol. 1996, 6,
3–10; d) VJ..Hruby, G.Li, C.Haskell-Luevano, M.
Shenderovich, Biopolymers 1997, 43, 219–266; e) R.N.
Patel, Adv. Appl. Microbiol. 2000, 47, 33–78; f) W.
[19] H.L.van Lingen, J.K.W.van de Mortel, K.F.W.Hek-
king, F.L.van Delft, T.Sonke, F.P.J.T.Rutjes,
Eur. J.
Org. Chem. 2003, 317–324.
[20] D.S. Goodsell, G.M. Morris, A.J. Olson,
J. Mol. Rec-
ognit. 1996, 9, 1–5.
Leuchtenberger, K.Huthmacher, K.Drauz,
Applied
[21] N.Guex, M.C.Peitsch, Electrophoresis 1997, 18, 2714–
Microb. Biotechnol. 2005, 69, 1–8.
2723.
2190
www.asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 2183 – 2190