Aminoacyl-coenzyme A synthesis catalyzed by a CoA ligase from Penicillium chrysogenum

Article abstract of DOI:10.1016/j.febslet.2011.02.018

Coenzyme A ligases play an important role in metabolism by catalyzing the activation of carboxylic acids. In this study we describe the synthesis of aminoacyl-coenzyme As (CoAs) catalyzed by a CoA ligase from Penicillium chrysogenum. The enzyme accepted medium-chain length fatty acids as the best substrates, but the proteinogenic amino acids l-phenylalanine and l-tyrosine, as well as the non-proteinogenic amino acids d-phenylalanine, d-tyrosine and (R)- and (S)-β-phenylalanine were also accepted. Of these amino acids, the highest activity was found for (R)-β-phenylalanine, forming (R)-β-phenylalanyl-CoA. Homology modeling suggested that alanine 312 is part of the active site cavity, and mutagenesis (A312G) yielded a variant that has an enhanced catalytic efficiency with β-phenylalanines and d-α-phenylalanine.

Full text of DOI:10.1016/j.febslet.2011.02.018

FEBS Letters 585 (2011) 893–898
journal homepage: www.FEBSLetters.org
Aminoacyl-coenzyme A synthesis catalyzed by a CoA ligase from
Penicillium chrysogenum
Martijn J. Koetsier a, Peter A. Jekel a, Hein J. Wijma a, Roel A.L. Bovenberg b, Dick B. Janssen a,
⇑
a Biochemical Laboratory, Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, Nijenborgh 4, Groningen NL-9747 AG, The Netherlands
b DSM Anti-Infectives, P.O. Box 425, Delft NL-2600 AK, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Coenzyme A ligases play an important role in metabolism by catalyzing the activation of carboxylic
acids. In this study we describe the synthesis of aminoacyl-coenzyme As (CoAs) catalyzed by a CoA
ligase from Penicillium chrysogenum. The enzyme accepted medium-chain length fatty acids as
Received 24 October 2010
Revised 13 February 2011
Accepted 14 February 2011
Available online 18 February 2011
the best substrates, but the proteinogenic amino acids
L-phenylalanine and L-tyrosine, as well as
the non-proteinogenic amino acids -phenylalanine, -tyrosine and (R)- and (S)-b-phenylalanine
D
D
were also accepted. Of these amino acids, the highest activity was found for (R)-b-phenylalanine,
forming (R)-b-phenylalanyl-CoA. Homology modeling suggested that alanine 312 is part of the active
site cavity, and mutagenesis (A312G) yielded a variant that has an enhanced catalytic efficiency with
b-phenylalanines and D-a-phenylalanine.
Ó 2011 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
Edited by Michael Ibba
Keywords:
Coenzyme A ligase
Aminoacyl-coenzyme A
b-Phenylalanine
Penicillium chrysogenum
1. Introduction
acid (C20). Despite this broad substrate range, amino acids have
not been described to be accepted as substrates.
Coenzyme A (CoA) ligases (or CoA synthetases) play an impor-
tant role in nature by catalyzing the activation of carboxylic acids
which allows for their further metabolism. For example, CoA
ligases activate fatty acids to initiate b-oxidation [1] and are
important in the biosynthesis of natural products such as lignin
[2] and penicillins [3] (Fig. 1). The mammalian fatty-acyl-CoA
ligases also function in the detoxification of diverse xenobiotic
compounds such as pesticides and drugs [4]. CoA ligases are
members of the superfamily of adenylate-forming enzymes and
activation proceeds in a two-step reaction (Fig. 2a). In the first half
reaction, the substrate reacts with adenosine 50-triphosphate (ATP)
to form an acyl-adenylate intermediate with the simultaneous
release of pyrophosphate. In the second half reaction, the adenylate
group is replaced by CoA and adenosine 50-monophosphate (AMP) is
released.
Activated amino acids play an important role in metabolic pro-
cesses such as protein and peptide synthesis, but the activation is
then not catalyzed by CoA ligases. In protein biosynthesis amino
acid activation is achieved by aminoacyl tRNA-synthetases,
whereas in the biosynthesis of small peptides, such as the antibi-
otic gramicidin S, amino acid activation is catalyzed by the adeny-
lation domains of non-ribosomal peptide synthetases (NRPSs) [5].
The aminoacyl tRNA synthetases and the adenylation domains of
the NRPSs are also members of the superfamily of adenylate-form-
ing enzymes. All these enzymes share formation of an adenylate
intermediate after the first reaction step (Fig. 2b). In the CoA
ligases this intermediate reacts with CoA to form the CoA-thioester
while in the tRNA synthetases and in the non-ribosomal peptide
synthetases the activated (a-amino)acyl group is transferred to
the 30 hydroxyl group of the tRNA molecule or the phosphopan-
theteine group of a peptidyl carrier domain, respectively.
The family of CoA ligases accepts a wide range of substrates. For
example, the group of fatty-acyl-CoA synthetases activates fatty
acids such as the short-chain acetate to the long-chain acid linoleic
A low level of aminoacyl-CoA ligase activity has previously been
described as a promiscuous activity of aminoacyl tRNA-syntheta-
ses [6] and of the adenylation domains of some NRPSs [7], suggest-
ing there might also be CoA ligases that are able to catalyze CoA
activation of amino acids. Aminoacyl-CoA ligase activity was also
suggested to be important for the epimerization of the side chain
of isopenicillin N in Penicillium chrysogenum [8], but the enzyme
Abbreviations: AMP, adenosine 50-monophosphate; ATP, adenosine 50-triphos-
phate; CoA, coenzyme A; HPLC, high-performance liquid chromatography; NRPS,
non-ribosomal peptide synthetase; PPi, inorganic pyrophosphate
⇑
responsible for CoA activation of isopenicillin
characterized.
N was not
Corresponding author. Fax: +31 (0)50 3634165.
E-mail address: d.b.janssen@rug.nl (D.B. Janssen).
0014-5793/$36.00 Ó 2011 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.febslet.2011.02.018

894
M.J. Koetsier et al. / FEBS Letters 585 (2011) 893–898
fatty acyl CoA Synthetase
O
O
O
n
OH
n
CoA
n
CoA
fatty acid beta-oxidation
H
N
S
phenylacetate-CoA ligase
OH
O
O
O
O
O
N
O
CoA
CoA
OH
penicillin G biosynthesis
O
coumarate-CoA ligase
R1
R3
Polymer
R1
R1
OH
HO
HO
R2
lignin biosynthesis
R2
R2
Fig. 1. Examples of CoA ligases in metabolic pathways. Fatty-acyl-CoA synthetases catalyze the first step in the b-oxidation of fatty acids. The activation of the penicillin G
side chain phenylacetic acid is catalyzed by phenylacetate-CoA ligase in Penicillium chrysogenum. CoA activation of coumarate derivatives in the biosynthesis of the complex
phenolic polymer lignin, in cells of higher plants.
half reaction 1
half reaction 2
ATP
CoA
O
O
O
O
O
O
A
B
AMP
CoA
OH
O
S
PPi
AMP
3'OH tRNA
ATP
AMP
AMP
tRNA
Ppant
OH
OH
O
O
O
S
NH2
NH2
NH2
PPi
AMP
O
O
O
Ppant pcp domain
ATP
NH2
NH2
NH2
PPi
AMP
adenylate intermediate
Fig. 2. Reactions catalyzed by members of the family of adenylate-forming enzymes. (A) phenylacetic acid activation catalyzed by a CoA ligase. (B) Ribosomal and non-
ribosomal activation of -phenylalanine.
L
The genome of P. chrysogenum contains a large number of genes
encoding CoA ligases [9], several of which have been characterized.
A cytosolic acetate-CoA synthetase was shown to have a broad
substrate range and was also able to catalyze the activation of side
chains that can be coupled to the b-lactam nucleus in vitro [10].
We have recently characterized a phenylacetyl-CoA ligase (PCL,
Pc22g14900) and an acyl-CoA ligase (ACLA, Pc22g20270) and
although these enzymes are involved in the side-chain activation
of penicillins and cephalosporins, respectively, they both are
acyl-CoA ligases with broad substrate specificities that show the
highest catalytic activities towards medium and long-chain fatty
acids. In the present work we describe the cloning and overexpres-
sion of a homologous and uncharacterized acyl-CoA ligase from P.
chrysogenum that has the ability to catalyze the synthesis of ami-
noacyl-CoAs.
2. Materials and methods
2.1. Materials
We used the CoA ligase, encoded by gene Pc21g20650, from P.
chrysogenum Wisconsin 54-1255 [9]. The gene was amplified from
a cDNA library [11] and subsequently cloned downstream of the
malE gene and in frame with a hexahistidine tag of vector pBAD-
MBP [12]. The resulting construct, encoding an MBP fusion protein,
was introduced in Escherichia coli using the pBAD vector (Invitro-
gen) and expressed as described previously [13]. The QuikChange
site-directed mutagenesis kit (Stratagene) was used to introduce
mutation A312G using primers Pc21g20650A312Gfw (50-CAAGA
TCATGTCTGCCGGCGCACCCCTTACTATT G-30, mutated codon is
underlined) and Pc21g20650A312Grv (50-CAATAGTAAGGGGTGC

M.J. Koetsier et al. / FEBS Letters 585 (2011) 893–898
895
GCCGGCAGACATGATCT TG-30, mutated triplet is underlined).
2.4. Homology modeling of aminoacyl-CoA ligase and docking
Production and purification of wild-type and mutant CoA ligase
were also done essentially as described earlier [13]. The amino
acids (S)- and (R)-b-phenylalanine were obtained from Peptech
Corporation. All other chemicals were from Sigma–Aldrich.
For homology model building and docking both online servers
(http://www.cbs.dtu.dk/services/CPHmodels/ and http://swissmodel.
expasy.org/) and YASARA software (www.yasara.org) were used
[15–17]. CPHmodels3.0 selected 3A9V as a template while Swiss
model selected 2D1R. YASARA made models based on PDB structures
of luciferase (2D1T, 2D1R, 2D1S, 3IEP, 1BA3 [18–20]), long-chain-
fatty-acid CoA ligase (3G7S), and coumarate-CoA ligase (3A9U,
3A9V, 3NI2 [21]) as templates. Docking was carried out with the YA-
SARA model using AutoDock4 [22] (350 separate runs). Figures were
prepared using Pymol software (www.pymol.org).
2.2. CoA ligase activity assay
CoA ligase activity was determined by following the substrate-
dependent formation of AMP by reversed-phase high-performance
liquid chromatography (HPLC) under essentially the same condi-
tions as described previously [13]. Standard reaction mixtures con-
tained 50 mM Tris–HCl, 200 mM NaCl, 3 mM ATP, 1.5 mM CoA,
5 mM MgCl2, 50
l
M to 100 mM of substrate, and 10 nM–2
lM
of purified recombinant enzyme in a total volume of 200
ll. Reac-
3. Results and discussion
tions were carried out at 30 °C and at pH 8.5. Substrate-indepen-
dent ATP hydrolysis by the CoA ligase was not observed
(kobs < 0.01 sꢀ1).
3.1. Formation of aminoacyl-coenzyme A
Apparent kinetic constants Km and kcat for CoA ligase substrates
were determined by varying the concentration of one substrate at
fixed concentrations of the other substrates. For determining the
apparent Km values with carboxylic acid substrates, fixed concen-
trations of 3 mM ATP, 1.5 mM CoA, and 5 mM Mg2+ were used.
Kinetic parameters for cosubstrates ATP and CoA were determined
using (R)-b-phenylalanine, fixed at 35 mM as the substrate.
The low levels of aminoacyl-CoA ligase activity observed in the
structurally related NRPSs and in tRNA synthetases suggest the
existence of CoA ligases that possess aminoacyl-CoA ligase activity.
In this study, the CoA ligase from P. chrysogenum encoded by the
gene Pc21g20650 was tested for the ability to catalyze the synthe-
sis of various aminoacyl-CoAs. The protein shares sequence simi-
larity to P. chrysogenum CoA ligases (PCL, ACLA) involved in
antibiotic synthesis that were previously characterized [3,13].
Analysis of CoA-dependent AMP production was used to deter-
mine which (amino) acids were accepted as substrates. We tested
2.3. LC–MS/MS identification of aminoacyl-CoAs
The formation of aminoacyl-CoAs was confirmed by electro-
spray mass spectrometry using a modified version of a described
protocol [14] in an LCQ Fleet Ion Trap mass spectrometer (Thermo
Scientific). Separation of reaction components was performed by
the proteinogenic amino acids and the amino acids
-Leu, -Ser, -Thr, -Met, -Tyr, -Phe, (R)- and (S)-norleucine,
(R)- and (S)-phenylglycine, (R)- and (S)-hydroxyphenylglycine and
(R)- and (S)-b-phenylalanine. Both -phenylalanine and -tyrosine
were accepted, and activity was also observed with -phenylala-
nine, -tyrosine and (R)- and (S)-b-phenylalanine. The enzyme also
had activity with the medium-chain fatty acids hexanoic, hepta-
noic, and octanoic acid, which are the preferred substrates of PCL
and ACLA.
D-Ala, D-Val,
D
D
D
D
D
D
L
L
LC using acidic eluent on a C18, 3.5-
lm column. Aminoacyl-CoAs
D
were eluted using a linear gradient (eluent A, 10:90 acetonitrile/
water containing 0.1% TFA and eluent B, 100% acetonitrile contain-
ing 0.1% TFA). Positive ion mode mass spectra were collected over a
scan range of m/z 200–1000.
D
Fig. 3. MS and MS/MS fragmentation of protonated (R)-b-phenylalanyl-CoA. The MS shows the parent ion in the single (m/z of 915) and double (m/z of 458) protonated form.
The MS/MS fragmentation (parent ion 915 m/z, collision energy 35 eV) shows the fragmented (R)-b-phenylalanylpantetheine molecule after a neutral loss of 507 (m/z of 408)
in the positive MS/MS mode. Similar results were obtained for the other isomers L-a-, D-a and (S)-b-phenylalanyl-CoA.

896
M.J. Koetsier et al. / FEBS Letters 585 (2011) 893–898
Table 1
was reported for the aminoacylation of CoA catalyzed by the
tRNA-synthetases [6,23]. However, in this case the aminoacyl-
CoA formation is a promiscuous activity with CoA replacing the
natural nucleophile of the second half reaction (Fig. 2), which
may explain the low turnover rates. Accordingly, the natural phen-
ylalanine activation reactions catalyzed by tRNA synthetases have
much higher observed turnover rates (Table 2). On the other hand,
compared to adenylation (A) domains of non-ribosomal peptide
synthetases (NRPS), the rates observed with the CoA ligase de-
scribed here appear higher [7,24], but it should be noted that the
low turnover number observed with the adenylation domain of
the gramicidin S synthetase initiation module (PheATE) is caused
by the slow release of the phenyladenylate. However, the affinity
of the Pc21g20650 CoA ligase for phenylalanine is very low com-
pared to that measured with the tRNA synthetases and the A do-
mains of the NRPS, which all have Km values in the micromolar
range [24–26]. Summarized, the current CoA ligase is an unspecific
acyl-CoA ligase accepting phenylalanine as a substrate, whereas
the tRNA synthetases and the NRPSs are very specific for their ami-
no acid substrate and tolerate CoA as an alternative nucleophile in
the second half reaction.
Steady-state kinetic parameters of (amino)acyl-CoA ligase activity of acyl-CoA ligase
(Pc21g20650) from P. chrysogenum.
Substrate
kcat (sꢀ1
)
Km (mM)
kcat/Km (mMꢀ1 sꢀ1
)
L
-
-
a
a
-Phenylalanine
-Tyrosine
0.37 0.13
0.007 0.004
1.4 0.2
0.06 0.03
0.25 0.02
3.0 0.2
3.3 0.4
3.9 1.1
17.4 2.22
10.8 0.05
7.3 0.2
44.7 2.8
3.5 1.2
83.0 6.4
1.6 0.8
33.9 0.1
3.5 0.4
0.78 0.01
2.0 0.4
(8.2 2.9) ꢁ 10ꢀ3
(2.0 1.4) ꢁ 10ꢀ3
(1.7 0.3) ꢁ 10ꢀ2
(3.8 2.7) ꢁ 10ꢀ2
(7.4 0.7) ꢁ 10ꢀ3
0.86 0.12
L
D
-
-
a
a
-Phenylalanine
-Tyrosine
D
(S)-b-Phenylalanine
(R)-b-Phenylalanine
ATP
CoA
Hexanoic acid
Heptanoic acid
Octanoic acid
1.63 0.21
0.94 0.06
0.15 0.01
(1.1 0.2) ꢁ 101
(1.2 0.1) ꢁ 101
(4.9 0.4) ꢁ 101
Table 2
Comparison of steady state kinetic parameters of phenylalanine activation catalyzed
by the CoA ligase, tRNA-synthetase and by the adenylation domain of non-ribosomal
peptide synthetases.
Substrate
kcat (sꢀ1
)
Km (mM) kcat/Km (mMꢀ1 sꢀ1
)
Reference
Acyl-CoA ligase
L
-
a
-Phenylalanine
-Phenylalanine
(R)-b-
Phenylalanine
0.37
1.4
3.0
44.7
83.0
3.5
0.0082
0.017
0.86
This work
This work
This work
3.2. Improved conversion of amino acids by mutagenesis of the
putative binding site
D-a
Phenylalanyl tRNA synthetase
Phenylalanine is a rather good substrate for the CoA ligase, but
the specificity for this substrate is low compared to activation of
fatty acids or compared to activation of (natural) amino acids cat-
alyzed by tRNA synthetases and NRPSs. This is caused by the rela-
tively low affinity for the amino acid (Table 2). To obtain further
insight in the causes of the selectivity of Pc21g20650 CoA ligase
with amino acids and carboxylic acids, we used homology model-
ing and site-directed mutagenesis.
L
-
-
a
a
-Phenylalaninea
240
1.72
0.0042
0.0103
57,000
177
[26]
[25]
L
-Phenylalanineb
A-domain PheATE NRPS Gramicidin S
L-
a
-Phenylalaninec
D-a
-Phenylalaninec
0.001
0.001
0.03
0.02
0.033
0.05
[24]
[24]
a
b
c
Phenylalanine-tRNA synthetase from Saccharomyces cerevisiae.
Phenylalanine-tRNA synthetase from Escherichia coli.
Kinetic parameters of the adenylation reaction. The low turnover number is due
to the slow release of phenylalanyl adenylate product.
Homology modeling was used to detect residues that may influ-
ence substrate selectivity. Three crystal structures of proteins that
have considerable sequence similarity to CoA ligase were used:
luciferase, long-chain fatty acid CoA ligase and coumarate-CoA li-
gase. A comparison between the homology models prepared by
CPHmodels3.0, SWISS-MODEL and YASARA (see Section 2) yielded
model structures with similar backbone folds (backbone RMSD
3.8–12.1 Å), with minor differences near the active site. The back-
bone RMSD within a range of 6 Å of the bound intermediates in
Fig. 4B is only 1.5–3.0 Å. Besides these relatively mild differences
in backbone positions there are more pronounced differences in
the predicted position of the side chains, as shown in Fig. 4A for
the predicted substrate binding site residues A312, Y242 and
H240. A YASARA predicted structure based on coumarate-CoA li-
gase (3A9V, sequence identity 33%) was selected for docking since
it had the highest score for structural quality. Automated docking
of the (R)-b-Phe-AMP intermediate resulted in a predicted binding
mode where the AMP tail is almost at the same position as that of a
co-crystallized intermediate in coumarate-CoA ligase (3NI2, yellow
in Fig. 4B). The (R)-b-Phe group of the docked (R)-b-Phe-AMP is at a
slightly different position as compared to the phenolic group in
3NI2. However, both in the crystallographic coumarate intermedi-
HPLC–MS/MS analysis was used to detect the formation of the
aminoacyl-CoA adducts from phenylalanines. The observed cleav-
age of ATP was coupled to the formation of phenylalanyl-CoA.
For both enantiomers of
a- and b-phenylalanine, the MS data
showed the presence of the parent ion in the single (m/z of 915)
and double (m/z of 458) protonated form (Fig. 3). The MS/MS frag-
mentation of protonated phenylalanyl-CoA showed the diffracted
phenylalanyl pantetheine molecule with a neutral loss of 507 (m/z
of 408) in the positive MS/MS mode.
The kinetic parameters of aminoacyl-CoA formation were deter-
mined by fitting the initial rate of AMP formation with the Michae-
lis–Menten equation (Table 1). The CoA ligase showed a preference
for the R-enantiomers of the substrates. As shown in Table 1, the
turnover number for (R)-D-phenylalanine, (R)-D-tyrosine and (R)-
b-phenylalanine was approximately 10-fold higher than with the
corresponding (S)-enantiomers. For b-phenylalanine the affinity
was also 10-fold higher for the (R)-enantiomer, resulting in the rel-
atively high catalytic efficiency observed for (R)-b-phenylalanine.
The measured apparent kcat values for the acylation of the ami-
no acids are at least 10-fold lower than for the fatty acids hexanoic,
heptanoic and octanoic acid (Table 1) and also the apparent Km val-
ues are higher for the amino acids. The kinetic parameters for the
fatty acids are very similar to what was found for PCL and ACLA
[3,13] suggesting that the CoA ligase is also a fatty acyl-CoA ligase.
The proteinogenic amino acids phenylalanine, tyrosine, alanine
and glycine were also tested as substrates for PCL and ACLA but
no activity was found [3, unpublished data] suggesting that the
ability to accept amino acids as a substrate is not a general feature
of the CoA ligases from P. chrysogenum.
ate and the docked the (R)-b-Phe intermediate, the C and Cb of the
a
activated substrate would be near the position of A312 in the
aligned structures. This residue is on a loop that protrudes into
the binding pocket of the intermediate. Thus, even though the
homology modeling did not produce atomistic detail, it suggested
that A312 is near the C or Cb atom of the bound activated
a
substrate.
To test the putative role of A312 in substrate binding and to
check the possibility of improving the activation of amino acids
through mutagenesis, we constructed mutant A312G, which was
expected to have a larger pocket. The mutation had clear effects
The apparent turnover number measured with CoA ligase
Pc21g20650 was several orders of magnitude higher than what

M.J. Koetsier et al. / FEBS Letters 585 (2011) 893–898
897
Fig. 4. Homology modeling and docking of substrate. (A) Superposition of three separately prepared homology models of P. chrysogenum CoA ligase. Protein backbones within
6 Å of the displayed binding pocket residues are shown as C atom traces. Light-blue, model produced by the CPH models server; purple, model produced by the Swiss-Model
a
server; green, top-ranked model produced by YASARA software. (B) Comparison of the top-ranked docked intermediate (light blue in green protein) with the position of the
intermediate observed in the crystal structure of coumarate-CoA ligase (yellow).
Table 3
Steady-state kinetic parameters of mutant CoA ligase A312G.
Substrate
kcat
Km
kcat/Km
(sꢀ1
)
r
(mM)
r
(mM sꢀ1
)
r
L
L
-
-
-
-
a
a
a
-Phenylalanine
-Tyrosine
-Phenylalanine
-Tyrosine
0.14 0.02
0.035 0.01
3.2 0.2
0.43 0.12
1.7 0.3
0.4
5
2
7
7
38.7 3.2
5.9 0.4
28.7 2.9
2.7 0.7
24.2 1.8
0.068 0.008
0.9
1.7
0.3
1.7
0.7
0.02
(3.6 0.6) ꢁ 10ꢀ3
0.4
3
6.5
4.2
9.5
4.0
(6 2) ꢁ 10ꢀ3
D
(1.1 0.2) ꢁ 10ꢀ1
(1.6 0.7) ꢁ 10ꢀ1
(7.0 1.4) ꢁ 10ꢀ2
3.4 0.6
D
a
(S)-b-Phenylalanine
(R)-b-Phenylalanine
0.23 0.03
0.08
r is the ratio of kinetic parameter of the mutant divided by the kinetic parameter for that substrate of the WT enzyme.
on the apparent kcat and Km with most substrates (Table 3), includ-
ing an increased kcat with - and -tyrosine and with (S)-b-phenyl-
observation that the A312G mutant has already fourfold improved
catalytic efficiency bodes well for further improvement by protein
engineering.
Nature appears to have evolved two major strategies to activate
carboxylic acids to thioesters. Most carboxylic acids are activated
by CoA ligases, but this reaction is uncommon for amino acids.
L
D
alanine. Most Km values were less affected, but a strong decrease
was observed with (R)-b-Phe, resulting in a net fourfold increase
of the selectivity constant with this substrate. The best substrate
for the mutant enzyme still is (R)-b-phenylalanine, but the largest
increase in activity was observed for (S)-b-phenylalanine (Table 3).
The fact that the A312G mutation enhances the conversion of var-
ious amino acids indicates that A312 is indeed located near the
binding site of the activated substrate, although subtle differences
cannot be explained in the absence of a crystallographic structure
of the enzyme.
CoA activation of
-aminoacyl thioesters which are less stable than normal acyl-
thioesters. This lower stability may be the reason that
a-amino acids would result in the formation of
a
a-
aminoacyl-CoAs are not used as intermediates in secondary
product biosynthesis. For protein synthesis, amino acids are acti-
vated as esters with tRNA. In the NRPSs the thioester-activated
amino acids remain covalently linked to the enzyme, preventing
hydrolysis of the energy rich intermediates. In contrast, b-
aminoacyl-CoAs are found as intermediates in b-alanine metab-
olism in Clostridium propionicum [28] and in taxol biosynthesis
[27], indicating that at least some aminoacyl-CoAs are stable
enough to occur in metabolic routes. Although these b-aminoacyl
3.3. Role of activated amino acids in natural product biosynthesis
To our knowledge this is the first paper describing a CoA ligase
that has considerable activity with aromatic proteinogenic amino
acids, as well as with b-phenylalanines. This raises the possibility
that CoA ligases are involved in some routes leading to the synthe-
sis of natural products that contain amino acid groups, or that CoA
ligases may be used in engineered routes towards natural prod-
ucts. For example, the activation of (R)-b-phenylalanine results in
the formation of (R)-b-phenylalanyl-CoA, which is an important
intermediate in taxol biosynthesis in Taxus brevifolia [27]. Taxol
is an important anti-cancer medicine. In one of the last steps of
taxol biosynthesis an acyltransferase couples the (R)-b-phenylala-
nyl side chain to the taxol nucleus. The acyltransferase requires a
CoA-activated side-chain precursor [27], but the enzyme responsi-
ble for the formation has not been identified. Although the cata-
lytic efficiency of the wild-type CoA ligase described here may be
insufficient for production of taxol in engineered systems, the
thioesters are more stable than their
the life-time of free -aminoacyl-CoAs has been shown to be in
the order of several minutes [29], sufficient to allow a role of
-aminoacyl-CoAs a role in biosynthetic routes.
In summary, in this paper we show that -aminoacyl-CoAs can
a-amino acid counterparts,
a
a
a
be formed in a reaction catalyzed by a CoA ligase from P. chrysog-
enum. This CoA ligase catalyzes the CoA activation (R)-b-
phenylalanine.
Acknowledgements
We thank Piet Wietzes for technical assistance with setting up
the LC–MS/MS experiments. This project was funded by the IBOS

898
M.J. Koetsier et al. / FEBS Letters 585 (2011) 893–898
J.M. (2010) The Penicillium chrysogenum aclA gene encodes a broad-substrate-
specificity acyl-coenzyme A ligase involved in activation of adipic acid, a side-
chain precursor for cephem antibiotics. Fungal Genet. Biol. 47, 33–42.
[14] Magnes, C., Sinner, F.M., Regittnig, W. and Pieber, T.R. (2005) LC/MS/MS
method for quantitative determination of long-chain fatty acyl-CoAs. Anal.
Chem. 77, 2889–2894.
[15] Arnold, K., Bordoli, L. and Schwede, T. (2006) The SWISS-MODEL Workspace: A
web-based environment for protein structure homology modelling.
Bioinformatics 22, 195–201.
(Integration of Biosynthesis and Organic Synthesis) Program of
Advanced Chemical Technologies for Sustainability (ACTS),
supported by the Dutch Ministry of Economic Affairs and The
Netherlands Organization for Scientific Research (NWO). Part of
this research was funded by NWO through an ECHO Grant.
References
[16] Nielsen, M., Lundegaard, C., Lund, O. and Petersen, T.N. (2010) CPHmodels-3.0
–
remote homology modeling using structure guided sequence profiles.
[1] Kunau, W.H., Dommes, V. and Schulz, H. (1995) Beta-oxidation of fatty acids in
mitochondria, peroxisomes, and bacteria: a century of continued progress.
Prog. Lipid. Res. 34, 267–342.
Nucleic Acids Res. 38, W576–581.
[17] Krieger, E., Joo, K., Lee, J., Lee, J., Raman, S., Thompson, J., Tyka, M., Baker, D. and
Karplus, K. (2009) Improving physical realism, stereochemistry, and side-
chain accuracy in homology modeling: four approaches that performed well in
CASP8. Proteins 77, 114–122.
[18] Nakatsu, T., Ichiyama, S., Hiratake, J., Saldanha, A., Kobashi, N., Sakata, K. and
Kato, H. (2006) Structural basis for the spectral difference in luciferase
bioluminescence. Nature 440, 372–376.
[2] Anterola, A.M. and Lewis, N.G. (2002) Trends in lignin modification:
a
comprehensive analysis of the effects of genetic manipulations/mutations on
lignification and vascular integrity. Phytochemistry 61, 221–294.
[3] Koetsier, M.J., Jekel, P.A., van den Berg, M.A., Bovenberg, R.A.L. and Janssen, D.B.
(2009) Characterization of
a phenylacetate-CoA ligase from Penicillium
chrysogenum. Biochem. J. 417, 467–476.
[19] Auld, D.S., Lovell, S., Thorne, N., Lea, W.A., Maloney, D.J., Shen, M., Rai, G.,
Battaile, K.P., Thomas, C.J., Simeonov, A., Hanzlik, R.P. and Inglese, J. (2010)
Molecular basis for the high-affinity binding and stabilization of firefly
luciferase by PTC124. Proc. Natl. Acad. Sci. USA 107, 4878–4883.
[20] Franks, N.P., Jenkins, A., Conti, E., Lieb, W.R. and Brick, P. (1998) Structural
basis for the inhibition of firefly luciferase by a general anesthetic. Biophys. J.
75, 2205–2211.
[21] Hu, Y., Gai, Y., Yin, L., Wang, X., Feng, C., Feng, L., Li, D., Jiang, X.N. and Wang,
D.C. (2010) Crystal structures of a Populus tomentosa 4-coumarate:CoA ligase
shed light on its enzymatic mechanisms. Plant Cell 22, 3093–3104.
[22] Morris, G.M., Goodsell, D.S., Haliday, R.S., Huey, R., Hart, W.E., Belew, R.K.
[4] Knights, K.M. and Drogemuller, C.J. (2000) Xenobiotic-CoA ligases: kinetic and
molecular characterization. Curr. Drug Metab. 1, 49–66.
[5] Conti, E., Stachelhaus, T., Marahiel, M.A. and Brick, P. (1997) Structural basis
for the activation of phenylalanine in the non-ribosomal biosynthesis of
gramicidin S. EMBO J. 16, 4174–4183.
[6] Jakubowski, H. (1998) Aminoacylation of coenzyme A and pantetheine by
aminoacyl-tRNA synthetases: possible link between noncoded and coded
peptide synthesis. Biochemistry 37, 5147–5153.
[7] Linne, U., Schäfer, A., Stubbs, M.T. and Marahiel, M.A. (2007) Aminoacyl-
coenzyme A synthesis catalyzed by adenylation domains. FEBS Lett. 581, 905–
910.
[8] Ullan, R.V., Casqueiro, J., Banuelos, O., Fernandez, F.J., Gutierrez, S. and Martin,
J.F. (2002) A novel epimerization system in fungal secondary metabolism
involved in the conversion of isopenicillin N into penicillin N in Acremonium
chrysogenum. J. Biol. Chem. 277, 46216–46225.
[9] Van den Berg, M.A., Albang, R., Albermann, K., Badger, J.H., Daran, J.M.,
Driessen, A.J.M., Garcia-Estrada, C., Fedorova, N.D., Harris, D.M., Heijne, W.H.,
Joardar, V., Kiel, J.A.K.W., Kovalchuk, A., Martín, J.F., Nierman, W.C., Nijland,
J.G., Pronk, J.T., Roubos, J.A., van der Klei, I.J., van Peij, N.N., Veenhuis, M., von
Döhren, H., Wagner, C., Wortman, J. and Bovenberg, R.A.L. (2008) Genome
sequencing and analysis of the filamentous fungus Penicillium chrysogenum.
Nat. Biotechnol. 26, 1161–1168.
[10] Martınez-Blanco, H., Reglero, A., Fernandez-Valverde, M., Ferrero, M.A.,
Moreno, M.A., Penalva, M.A. and Luengo, J.M. (1992) Isolation and
characterization of the acetyl-CoA synthetase from Penicillium chrysogenum:
involvement of this enzyme in the biosynthesis of penicillins. J. Biol. Chem.
267, 5474–5481.
[11] Kiel, J.A.K.W., van der Klei, I.J., van den Berg, M.A., Bovenberg, R.A.L. and
Veenhuis, M. (2005) Overproduction of a single protein, Pc-Pex11p, results in
2-fold enhanced penicillin production by Penicillium chrysogenum. Fungal
Genet. Biol. 42, 154–164.
and Olson, A.J. (1998) Automated docking using
a Lamarckian genetic
algorithm and empirical binding free energy function. J. Comput. Chem. 19,
1639–1662.
[23] Jakubowski, H. (2000) Amino acid selectivity in the aminoacylation of
coenzyme A and RNA minihelices by aminoacyl-tRNA synthetases. J. Biol.
Chem. 275, 34845–34848.
[24] Luo, L., Burkart, M.D., Stachelhaus, T. and Walsh, C.T. (2001) Substrate
recognition and selection by the initiation module PheATE of gramicidin S
synthetase. J. Am. Chem. Soc. 123, 11208–11218.
[25] Roy, H., Ling, J., Irnov, M. and Ibba, M. (2004) Post-transfer editing in vitro and
in vivo by the beta subunit of phenylalanyl-tRNA synthetase. EMBO J. 23,
4639–4648.
[26] Roy, H., Ling, J., Alfonzo, J. and Ibba, M. (2005) Loss of editing activity during
the evolution of mitochondrial phenylalanyl-tRNA synthetase. J. Biol. Chem.
280, 38186–38192.
[27] Walker, K., Fujisaki, S., Long, R. and Croteau, R. (2002) Molecular cloning and
heterologous expression of the C-13 phenylpropanoid side chain-CoA
acyltransferase that functions in Taxol biosynthesis. Proc. Natl. Acad. Sci.
USA 99, 12715–12720.
[28] Herrmann, G., Selmer, T., Jessen, H.J., Gokarn, R.R., Selifonova, O., Gort, S.J. and
Buckel, W. (2005) Two beta-alanyl-CoA:ammonia lyases in Clostridium
propionicum. FEBS J. 272, 813–821.
[29] Belshaw, P.J., Walsh, C.T. and Stachelhaus, T. (1999) Aminoacyl-CoAs as probes
of condensation domain selectivity in nonribosomal peptide synthesis.
Science 284, 486–489.
[12] Heuts, D.P.H.M., van Hellemond, E.W., Janssen, D.B. and Fraaije, M.W. (2007)
Discovery, characterization, and kinetic analysis of an alditol oxidase from
Streptomyces coelicolor. J. Biol. Chem. 282, 20283–20291.
[13] Koetsier, M.J., Gombert, A.K., Fekken, S., Bovenberg, R.A.L., van den Berg, M.A.,
Kiel, J.A.K.W., Jekel, P.A., Janssen, D.B., Pronk, J.T., van der Klei, I.J. and Daran,