A structure-based mechanism for benzalacetone synthase from Rheum palmatum

Article abstract of DOI:10.1073/pnas.0909982107

Benzalacetone synthase (BAS), a plant-specific type III polyketide synthase (PKS), catalyzes a one-step decarboxylative condensation of malonyl-CoA and 4-coumaroyl-CoA to produce the diketide benzalacetone. We solved the crystal structures of both the wild-type and chalcone-producing I207L/L208F mutant of Rheum palmatum BAS at 1.8 A resolution. In addition, we solved the crystal structure of the wild-type enzyme, in which a monoketide coumarate intermediate is covalently bound to the catalytic cysteine residue, at 1.6 A resolution. This is the first direct evidence that type III PKS utilizes the cysteine as the nucleophile and as the attachment site for the polyketide intermediate. The crystal structures revealed that BAS utilizes an alternative, novel activesite pocket for locking the aromatic moiety of the coumarate, instead of the chalcone synthase's coumaroyl-binding pocket, which is lost in the active-site of the wild-type enzyme and restored in the I207L/L208F mutant. Furthermore, the crystal structures indicated the presence of a putative nucleophilic water molecule which forms hydrogen bond networks with the Cys-His-Asn catalytic triad. This suggested that BAS employs novel catalytic machinery for the thioester bond cleavage of the enzyme-bound diketide intermediate and the final decarboxylation reaction to produce benzalacetone. These findings provided a structural basis for the functional diversity of the type III PKS enzymes.

Full text of DOI:10.1073/pnas.0909982107

A structure-based mechanism for benzalacetone
synthase from Rheum palmatum
Hiroyuki Moritaa, Yoshihiko Shimokawab, Michikazu Tanioc, Ryohei Katod, Hiroshi Noguchib, Shigetoshi Sugiod,1
Toshiyuki Kohnoc,1, and Ikuro Abea,1
,
aGraduate School of Pharmaceutical Sciences, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan;bSchool of Pharmaceutical Science,
University of Shizuoka, 52-1 Yada, Shizuoka 422-8526, Japan;cMitsubishi Kagaku Institute of Life Sciences, 11 Minamiooya, Machida, Tokyo 194-8511,
Japan; anddInnovation Center Yokohama, Mitsubishi Chemical Corporation, 1000 Kamoshida, Aoba, Yokohama, Kanagawa 227-8502, Japan.
Edited by Robert M. Stroud, University of California, San Francisco, CA, and approved November 18, 2009 (received for review September 1, 2009)
Benzalacetone synthase (BAS), a plant-specific type III polyketide
synthase (PKS), catalyzes a one-step decarboxylative condensation
of malonyl-CoA and 4-coumaroyl-CoA to produce the diketide
benzalacetone. We solved the crystal structures of both the
wild-type and chalcone-producing I207L/L208F mutant of Rheum
palmatum BAS at 1.8 Å resolution. In addition, we solved the crys-
tal structure of the wild-type enzyme, in which a monoketide
coumarate intermediate is covalently bound to the catalytic cys-
teine residue, at 1.6 Å resolution. This is the first direct evidence
that type III PKS utilizes the cysteine as the nucleophile and as
the attachment site for the polyketide intermediate. The crystal
structures revealed that BAS utilizes an alternative, novel active-
site pocket for locking the aromatic moiety of the coumarate, in-
stead of the chalcone synthase’s coumaroyl-binding pocket, which
is lost in the active-site of the wild-type enzyme and restored in the
I207L/L208F mutant. Furthermore, the crystal structures indicated
the presence of a putative nucleophilic water molecule which
forms hydrogen bond networks with the Cys-His-Asn catalytic
triad. This suggested that BAS employs novel catalytic machinery
for the thioester bond cleavage of the enzyme-bound diketide
intermediate and the final decarboxylation reaction to produce
benzalacetone. These findings provided a structural basis for the
functional diversity of the type III PKS enzymes.
the decarboxylation of malonyl-CoA by maintaining the orienta-
tions of substrates and intermediates during the sequential con-
densation reactions (5). Indeed, the BAS I207L/L208F mutant
(numbering in BAS) restored the chalcone-forming activity, sup-
porting the hypothesis that the replacement of Phe208 in BAS
(corresponding to Phe215 in M. sativa CHS) accounts for the
interruption of the polyketide chain elongation at the diketide
stage (13, 14).
We now present the crystal structures of both the wild-type and
chalcone-producing I207L/L208F mutant of R. palmatum BAS at
1.8 Å resolution. In addition, we solved the crystal structure of the
wild-type enzyme, in which a monoketide coumarate intermedi-
ate is covalently bound to the catalytic cysteine residue, at 1.6 Å
resolution. This is direct evidence that a type III PKS utilizes the
cysteine as the nucleophile and as the attachment site for the
polyketide intermediate. The crystal structures revealed that
BAS utilizes an alternative active-site pocket for locking the aro-
matic moiety of the coumarate, instead of the CHS’s “coumaroyl-
binding pocket”, and indicated the presence of a putative nucleo-
philic water molecule that forms hydrogen bond networks with
the Cys-His-Asn catalytic triad. These findings have led to a pro-
posal for a unique mechanism of the enzyme reaction and pro-
vided a structural basis for the functional diversity of the type III
PKS enzymes.
biosynthesis ∣ enzyme ∣ polyketide
Results and Discussion
Overall Structure of BAS. The homodimeric apo-structure of BAS
consists of residues 8–383 of monomer A, and residues 8–382 of
monomer B (Fig. S3). Both monomers are nearly identical to
each other, with root-mean-square deviations (RMSDs) of
0.44 Å. Each monomer binds the other monomer with a twofold
axis, thereby forming a biologically active, symmetric dimer.
Upon dimerization, each monomer buries 2200 Å2 of its surface,
and Met130 protrudes into the other monomer by the formation
of a cis-peptide bond between Met130 and Pro131 on a loop, to
complete the wall of an active-site cavity in each monomer. The
catalytic triad of Cys157, His296 and Asn329 is buried deep with-
in each monomer, and sits at the intersection of a traditional
16 Å-long CoA-binding tunnel and a large internal cavity, in a
location and orientation very similar to those of the other plant
type III PKSs (5–7, 10). The CoA-binding tunnel is connected to
the protein surface, thus facilitating the entrance of the substrate
into the catalytic center. The overall structures of BAS are highly
homologous to those of the previously reported plant type III
enzalacetone synthase (BAS), a member of the plant-specific
B
chalcone synthase (CHS) superfamily of type III polyketide
synthases (PKSs) (1–3), catalyzes the one-step decarboxylative
condensation of 4-coumaroyl-CoA with malonyl-CoA to produce
a diketide benzalacetone, 4-(4-hydroxyphenyl)but-3-en-2-one
(Fig. 1A and Fig. S1) (4). BAS is a crucial enzyme in the biosyn-
thesis of the C6-C4 moiety of biologically active phenylbutanoids
such as the antiinflammatory glucoside lindleyin in rhubarb, and
raspberry ketone, the characteristic aroma of raspberry fruit. In
contrast, the typical type III PKSs catalyze iterative
condensations of malonyl-CoA with a CoA-linked starter mol-
ecule (Fig. S1) (1–3). For example, CHS and stilbene synthase
(STS), sharing ≈70% amino acid sequence identity with BAS,
catalyze sequential condensations of 4-coumaroyl-CoA and three
molecules of malonyl-CoA to produce the tetraketides naringen-
in chalcone and resveratrol, respectively. Recent crystallographic
analyses of the type III PKSs have revealed that the functional
diversity of the CHS-superfamily enzymes is principally derived
from the small modifications of the active-site architecture
(5–12).
Author contributions: H.M. and I.A. designed research; H.M., Y.S., and I.A. performed
research; H.M., M.T., R.K., H.N., S.S., T.K., and I.A. analyzed data; H.M. and I.A. wrote
the paper.
We previously reported that the diketide-forming activity of
Rheum palmatum BAS is attributed to the characteristic substi-
tution of the active-site Phe215; the conserved CHS residues
214LF are uniquely replaced by IL in R. palmatum BAS (number-
ing in Medicago sativa CHS, Fig. S2) (13). The conformationally
flexible Phe215, located at the junction between the active-site
cavity and the CoA-binding tunnel, is absolutely conserved in
all of the known type III PKSs, and is considered to facilitate
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1To whom correspondence may be addressed. E-mail: ssugio@rc.m-kagaku.co.jp, tkohno@
mitils.jp or abei@mol.f.u-tokyo.ac.jp.
This article contains supporting information online at www.pnas.org/cgi/content/full/
0909982107/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.0909982107
PNAS ∣ January 12, 2010 ∣ vol. 107 ∣ no. 2 ∣ 669–673

Fig. 1. Enzyme reaction and active-site structure of R. palmatum BAS. (A) Proposed mechanism for the formation of benzalacetone by BAS. The active-site
structures of wild-type BAS apo (B), M. sativa CHS (C), BAS I207L/L208F mutant (D), and wild-type BAS with the covalently bound coumarate (E). The coumarate
and naringenin molecules are shown as blue and green stick models, respectively. The water molecule and the hydrogen bonds are indicated with light-blue
spheres and dotted lines, respectively. The Fo-Fc density map of the monoketide intermediate covalently bound to the catalytic Cys157 (F) in monomer A, and in
monomer B of wild-type BAS, countered at 2.0 s (G).
PKSs, including M. sativa CHS (5) and Pinus sylvestris STS (7),
with RMSDs of 0.66 Å and 0.73 Å, respectively, for the Cα-atoms
(Fig. S3).
(−112, 6), is shifted by a ϕ angle of −61° and a ψ angle of
þ48° toward Leu208 in BAS. A similar situation has also been
found in the crystal structure of resveratrol-forming P. sylvestris
STS (7), which will be discussed later. The conformational dif-
ferences of Leu208 and Ser331 cause the loss of the CHS’s
coumaroyl-binding pocket (5) from the BAS active-site cavity.
As a result, the total cavity volume (350 Å3) of the active site
of BAS is much smaller than that ofM. sativa CHS (750 Å3)
(Fig. 2A and B), suggesting that the steric contraction leads to
the shorter diketide-forming activity of BAS (Fig. 2D and E).
Phe258 in R. palmatum BAS and the corresponding Phe265 in
M. sativa CHS are located at the entrance of the active-site, as in
the case of the above mentioned Leu208 in BAS and Phe215 in
CHS. The backbone torsion angle of Phe258 (−131, 118), in com-
parison with that of Phe265 (−123, 131), is slightly shifted by a ϕ
angle of þ8° and a ψ angle of þ13°, and the aromatic moiety of
Phe258 protrudes more toward Leu208, as compared to that of
Phe265 of M. sativa CHS, and forms a hydrophobic interaction
Active-Site Architecture of BAS. As previously reported (13), the
benzalacetone-forming activity of R. palmatum BAS is attributed
to the characteristic substitution of the conserved active-site
Phe215 (numbering in M. sativa CHS) with leucine. Indeed, a
comparison of the active-site structure of R. palmatum BAS with
that of M. sativa CHS revealed that both the location and angle of
the planar residue Phe215 of M. sativa CHS are well conserved in
Leu208 of BAS, but the Cδ carbon of the Leu208 side-chain
protrudes into the CHS’s active-site cavity, thereby causing a
substantial contraction of the active-site cavity of BAS (Fig. 1B
and C). In addition, in R. palmatum BAS, the terminal hydroxyl
group of Ser331 is isometric with that of the corresponding
Ser338 in M. sativa CHS. The serine residue neighboring the cat-
alytic cysteine is considered to be crucial for modulation of the
catalytic activity (7). In fact, we previously reported that the BAS with the side-chain of Leu208 (Fig. 1B and C, and Fig. S4).
S331 V mutant exhibited a twofold increase in the kcat∕KM value
for the benzalacetone-producing activity, which suggested that
the residue is important for providing steric guidance for the
diketide formation reaction (14). In the BAS crystal structure,
the hydroxyl group of Ser331 rotates by nearly 120°, thereby
blocking the entrance of the CHS’s coumaroyl-binding pocket
(5). Moreover, a structural comparison with M. sativa CHS re-
vealed significant backbone changes in the loops corresponding
to residues 123–128 in R. palmatum BAS and the residues 131–
135 in M. sativa CHS, respectively (RMSDs of 1.00 Å for the
Cα-atoms, Fig. 1B and C). This movement is caused by various
neighboring amino acid substitutions at the loop, most notably
the replacement of Val98 in M. sativa CHS with the bulkier
side-chain Gln in BAS, with the result that the backbone torsion
angle of Leu125 (−51, −42), as compared with that of Thr132
Although positional differences of this gatekeeper phenylalanine
are often observed in the type III PKSs (7), this conformational
change appears to be caused by the F208L substitution in BAS.
The total area of the active-site entrance of BAS is thus ≈34 Å2,
which is twice as large as that of M. sativa CHS (17 Å2) (Fig. S4).
A similar widening of the active-site entrance has also been re-
ported for the structure of the M. sativa F215S CHS mutant,
which accepts the bulky N-methylanthraniloyl-CoA as a starter
substrate (15). Noel and coworkers (15) have proposed that
the F215S substitution opens the space at the cavity entrance
to accommodate the methylamine moiety of N-methylanthrani-
loyl-CoA and facilitates the positioning of the thioester-carbonyl
moiety next to the Cys-His-Asn catalytic triad. These observations
may also account for the previous report that R. palmatum BAS
readily accepts N-methylanthraniloyl-CoA as a starter substrate
670
∣
www.pnas.org/cgi/doi/10.1073/pnas.0909982107
Morita et al.

Fig. 2. Surface and schematic representations of the active-site cavities of wild-type BAS (A and D), M. sativa CHS (B and E) , and BAS I207L/L208F mutant
(C and F). The bottoms of the “coumaroyl-binding pocket” are highlighted as purple surfaces. The covalently bound coumarate, the water molecules and the
hydrogen bonds are colored as in Fig. 1.
to efficiently produce N-methyl-4-hydroxy-2(1H)-quinolone after
The phenolic OH of the coumarate forms a hydrogen bond with
the backbone carbonyl oxygen of Ser257 (Fig. 1E). This is in sharp
contrast with the active-site structure ofM. sativa CHS in which
the coumaroyl moiety of naringenin is locked into the coumaroyl-
binding pocket, with its phenolic OH forming a hydrogen bond
with the backbone carbonyl oxygen of Gly216, and with the car-
bonyl oxygen of naringenin interacting with the hydroxyl of
Thr197 (Fig. 1C) (5). These observations suggest that BAS uti-
lizes the unique active-site pocket to lock the coumaroyl moiety
for the production of benzalacetone, and the chain elongation is
interrupted at the diketide stage because of the steric contraction
of the active-site cavity (Fig. 2D).
condensation with one molecule of malonyl-CoA (16).
Restoration of Chalcone-Forming Activity in the BAS I207L/L208F
Mutant. The X-ray crystal structure of the BAS I207L/L208F
mutant clearly demonstrated that both the location and orien-
tation of the residues lining the active-site cavity in the mutant
are almost perfectly conserved with those in the wild-type
enzyme (Fig. 1B and D), whereas those of Ile207, Leu208,
Phe258 and Ser331 in the mutant were sterically altered to those
of CHS (Fig. 1C and D). Notably, the conformational changes at
Phe208 and Ser331 in the mutant are caused by the replacement
of Leu208 with the bulkier and planar Phe, which restored the
coumaroyl-binding pocket within the active-site of the mutant
(Fig. 2B and C). Thus, the Ile207L/Leu208F substitutions open
a gate to the buried coumaroyl-binding pocket, thereby increasing
the polyketide chain elongation up to three condensations with
malonyl-CoA, and leading to the recovery of the chalcone-
forming activity in the mutant (Fig. 2E and F). It is remarkable
that the functional diversity of the type III PKS evolved from the
simple steric modulation of a single, chemically inert, residue
lining the active-site cavity.
Mechanism of Thioester Bond Cleavage and Decarboxylation. Consid-
ering the conservation of the overall folding and the catalytic
triad, it is conceivable that BAS utilizes similar catalytic machin-
ery as that used by CHS for the initiation of the enzyme reaction
and the decarboxylative Claisen-type condensation of malonyl-
CoA (2, 3). Here, the most interesting points are the catalytic
processes of the thioester bond cleavage of the enzyme-bound
diketide intermediate and the final decarboxylation reaction that
produces the C6-C4 benzalacetone.
It should be noted that the crystal structure of BAS revealed
slight conformational changes in the loop between residues 123–
128, as compared to those in CHS and STS (Fig. S5). Recent
structural analyses ofP. sylvestris STS demonstrated that the back-
bone change of the loop leads to the subtle displacement of
Thr132, resulting in the rearrangement of the so-called “aldol-
switch” hydrogen bond network, including Thr132 and a nucleo-
philic water molecule neighboring the catalytic Cys164 at the
active-site center (numbering in M. sativa CHS, Fig. S5) (7). Noel
and Schröder (7) thus proposed that the internal aldol cyclization
and sequential decarboxylation reactions in STS are principally
triggered by the cleavage of the thioester bond of the enzyme-
bound intermediate, by the nucleophilic attack of the activated
water molecule. In contrast, careful examination of the BAS
crystal structure precluded the presence of such hydrogen bond
networks, since Thr132 in STS is replaced with the hydrophobic
Leu in BAS (Fig. 1B and E, and Fig. S5). Instead, the structure
clearly demonstrated the presence of a distinct, activated water
molecule which forms hydrogen bond networks with the Cys-His-
Asn catalytic triad (Fig. 1B and E, and Fig. S5). Interestingly, the
Crystal Structure of Wild-Type BAS with Covalently Bound Coumarate.
To gain further insights into the polyketide formation machinery,
we soaked the apo crystals of BAS with the sole initial substrate,
4-coumaroyl-CoA (see Methods for details). The crystal structure
unambiguously revealed that a monoketide intermediate (i.e.,
4-coumaroyl thioester) is covalently bound to the SH group of
the catalytic Cys157 in each monomer, which had never been
observed in any type III PKS structures (Fig. 1F and G, and
Fig. S3). This is direct evidence that the type III PKS enzyme
utilizes the cysteine residue as the nucleophile and as the attach-
ment site for the polyketide intermediate. The coumarate occu-
pies the CoA-binding tunnel in monomer A, while it lies in the
active-site cavity in monomer B (Fig. S3). To accommodate the
coumarate in the active-site, the SH group of Cys190 must rotate
by about 93°, as compared with that of the BAS apo structure
(Fig. 1B and E).
Amazingly, the aromatic moiety of the coumarate in monomer
B is locked into an alternative, unique active-site pocket, which
protrudes into the “floor” of the conventional CHS active site.
Morita et al.
PNAS
∣
January 12, 2010
∣
vol. 107
∣
no. 2
∣
671

Fig. 3. Proposed mechanism for the BAS enzyme reaction. The active-site structures of the apo (A) and the coumarate-bound BAS (B), respectively. Three-
dimensional model for the diketide intermediate covalently bound to the catalytic cysteine. The putative nucleophilic water molecule is shown as a light-blue
sphere. The dotted lines indicate the hydrogen bond (C). Proposed mechanism for the enzyme catalyzed hydrolysis and decarboxylation reaction (D).
water molecule is also sterically conserved in the structure of the
BAS I207L/L208F mutant (Fig. 1E and Fig. S4).
acid is released from ORAS, as in the case of the β-keto acid
in BAS.
Since the chalcone-forming I207L/L208F mutant still produces
benzalacetone (13), both the wild-type and mutant enzymes could
possibly maintain the same catalytic machinery for the
formation of benzalacetone. Furthermore, as described above,
the BAS S331V mutant, in which the corresponding Ser338 in
the aldol-switch hydrogen bond network is substituted with the
hydrophobic Val, exhibited a twofold increase in the kcat∕KM
value for the benzalacetone-producing activity (14). In addition,
we have recently successfully detected the 4-coumaroyl β-keto
acid, 5-(4-hydroxyphenyl)-3-oxopent-4-enoic acid, in the reaction
products of BAS (Fig. S6). All of these observations strongly sug-
gest that BAS utilizes alternative and unique catalytic machinery,
which is distinct from that of STS, for the thioester bond cleavage
of the enzyme-bound intermediate and the final decarboxylation
reaction, and that the β-keto acid produced by the nucleophilic
attack of the water molecule, presumably activated by His296,
subsequently undergoes decarboxylation to yield the C6-C4
benzalacetone (Fig. 3D). Thus, we propose that the decarbox-
ylation of the β-keto acid proceeds via proton abstraction by
His296, reactivated by the Cys157 thiolate, and formation of
an enolate anion presumably stabilized by the His296-Asn329
oxyanion hole, just as in the case of the decarboxylation of
malonyl-CoA (2). Finally, tautomerization to the keto form pro-
duces benzalacetone and restores the Cys157-His296 thiolate-
imidazolium ion pair (Fig. 3D).
In conclusion, the crystal structures revealed that BAS utilizes
an alternative, unique active-site pocket for locking the aromatic
moiety of the coumarate, instead of the coumaroyl-binding
pocket, which is lost in the active site of the wild-type enzyme
and restored in the I207L/L208F mutant. Furthermore, the crys-
tal structures indicated the presence of a putative nucleophilic
water molecule which forms hydrogen bond networks with the
Cys-His-Asn catalytic triad. This suggested that BAS employs
unique catalytic machinery for the thioester bond cleavage of
the enzyme-bound diketide intermediate and the final decarbox-
ylation reaction to produce benzalacetone. These findings pro-
vide a structural basis for the functional diversity of the type
III PKS enzymes.
Methods
Chemicals. 4-coumaroyl-CoA was synthesized as described previously. Oligo-
nucleotides were obtained from Hokkaido System Science Co., Ltd. Standard
chemicals were obtained from Sigma-Aldrich and Hampton Research.
Expression and Purification. Wild-type BAS was expressed in E. coli M15 and
purified as described previously (18). The L207L/L208F mutant expression
plasmid was constructed with a QuikChange Site-Directed Mutagenesis Kit
(Stratagene), according to the manufacturer’s protocol, by using 50-
CCATGATAGGCCAA GCATTATTCGGCGATGGGGCTGC-30 as the sense primer,
50-GCAGCCCCATCGCC GAATAATGCTTGGCCTATCATGG-30 as the antisense
primer, and the wild-type BAS expression plasmid as the template. The
mutant protein was expressed and purified by the same procedure used
for the wild-type BAS, and was concentrated to 20 mg∕mL in 20 mM
HEPES-NaOH (pH7.5) buffer, containing 100 mM NaCl and 2 mM DTT.
Recently, 20-oxoalkylresorcylic acid synthase (ORAS), which
catalyzes the condensation of four molecules of malonyl-CoA
with a long-chain acyl-CoA thioester (C16, C18 and C20) as a
starter substrate to produce the pentaketide alkylresorcylic acid
and resorcinol, has been characterized from Neurospora crassa
(Fig. S1) (12, 17). The stable resorcinol acid is considered to
be an intermediate which is further enzymatically converted into
the final alkylresorcinol product by a decarboxylation reaction. In
spite of the similar catalytic process to that of STS, the crys-
tallographic analysis of ORAS revealed that, unlike STS, the
enzyme structure does not share the aldol switch in the active-site
cavity (12). In contrast, our structural analyses revealed that the
location of the putative nucleophilic water molecule in BAS is
nearly identical to that of the sulfinic oxygen of the oxidized cat-
alytic cysteine in ORAS (Fig. S7). Therefore, it is quite probable
that the sulfinic oxygen in ORAS can be regarded as an analog of
the nucleophilic water molecule in BAS, and that the resorcinol
Crystallization. Crystallization of the wild-type BAS was performed as pre-
viously reported (18). The I207L/L208F mutant crystals were also obtained
by the same crystallization methods as those used for the wild-type BAS. Both
crystals were independently transferred to a reservoir solution with 20% (v/v)
glycerol as a cryoprotectant, and were then flash-cooled at 100 K in a nitro-
gen-gas stream. For the monoketide intermediate-complexed crystal of the
wild type, a single crystal, as described above, was captured in a nylon loop
and transferred to the reservoir solution containing
2
mM
4-couamroyl-CoA. After an incubation at 20° for 2 d, the crystal was trans-
ferred to the same cryoprotectant as that used for the wild-type and mutant
apo crystals, except that it also contained 4-coumaroyl-CoA.
Structure Determination. Both X-ray diffraction datasets from the wild-type
BAS crystals were collected at 100 K at BL24XU of SPring-8 (wavelength,
0.82656 Å), by using a Rigaku R-AXIS V imaging plate. The X-ray diffraction
672
∣
www.pnas.org/cgi/doi/10.1073/pnas.0909982107
Morita et al.

dataset from the mutant crystal was collected at 100 K at BL41XU of SPring-8
(wavelength, 1.00000 Å), by using an ADSC Quantum 210 CCD detector. All
data were indexed, integrated, and scaled with the HKL2000 program (19).
The initial phases of the wild-type apo structure were determined by
molecular replacement by using the BAS structure model generated by
the SWISS-MODEL package (http://expasy.ch/spdpv/) based on the crystal
structure of M. sativa CHS [Protein Data Bank (PDB) entry 1BQ6], as a search
model. The molecular replacement was performed with the program Crystal-
lography and NMR System (CNS) (20). Crystallographic refinement and model
building were performed with CNS and XTALVIEW (21), respectively. Each re-
finement cycle was followed by model building using the σA-weighted 2Fo-Fc
and Fo-Fc electron density maps. The water molecules were automatically
placed into the difference electron density maps with XTALVIEW, and were
retained or rejected on the basis of geometric criteria as well as their refined
B-factors. After several rounds of model building and refinement, the final
model was obtained. The other structures were solved by the same procedure
as used for the model refinement of the BAS wild-type apo structure, except
for the use of either the entire or I204L/L208F-substituted final models of the
wild-type apo structure as the search model in the molecular replacement
methods. Both the 2Fo-Fc and Fo-Fc maps indicated the presence of a portion
of the monoketide intermediate covalently bound to the catalytic cysteine of
each monomer in the intermediate-complexed structure, and the intermedi-
ate manually fits into the visible electron density. Each model consists of re-
sides 8–383 of monomer A, and residues 8–382 of monomer B. The qualities
of the final models were assessed with PROCHECK (22). Details of the data
collection, processing, and structure refinement are summarized in Table S1.
The cavity volume and the active-site entrance area were calculated by the
program CASTP (http://cast.engr.uic.edu/cast/). All crystallographic figures
were prepared with PyMOL (DeLano Scientific, http://www.pymol.org).
volume of 500 μL of 100 mM Tris-HCl buffer (pH 8.0) and 1 mM EDTA. Incuba-
tions were performed at 30 °C for 1 hr, and were stopped by the addition of
50 μL of 20% HCl. The products were then extracted with 3 mL of ethyl
acetate. The products were separated by reverse-phase HPLC (JASCO 880)
on a TSK-gel ODS-80Ts column (4.6 ÅE 150 mm, TOSOH), at a flow rate of
0.8 mL∕ min. Gradient elution was performed with H2O and MeOH, both
containing 0.1% TFA: 0–5 min, 30% MeOH; 5–17 min, linear gradient from
30 to 60% MeOH; 17–25 min, 60% MeOH; 25–27 min, linear gradient from 60
to 70% MeOH. Elutions were monitored by a multichannel UV detector
(MULTI 340, JASCO) at 280 nm. UV spectra (198–400 nm) were recorded every
0.4 s. Online LC-ESIMS spectra were measured with an Agilent Technologies
series 1100 HPLC coupled to a Bruker Daltonics esquire4000 ion-trap mass
spectrometer fitted with an ESI source. HPLC separations were performed
under the same conditions as described above. The ESI capillary temperature
and the capillary voltage were 320 °C and 4.0 V, respectively. The tube lens
offset was set at 20.0 V. All spectra were obtained in the positive mode over a
mass range of m∕z 50–500, and at a range of one scan every 0.2 s. The
collision gas was helium, and the relative collision energy scale was set at
30.0% (1.5 eV).
Spectroscopic Data for (E)-5-(4-Hydroxyphenyl)-3-Oxopent-4-Enoic Acid
(4-Coumaroyl Diketide β-Keto Acid). UV λmax 283 nm; LC-ESIMS: MS, m∕z 207
½M þ Hꢀþ, MS/MS (precursor ion at m∕z 207), m∕z 147 ½M þ H-CO2-CH2ꢀþ; HRMS
(FAB) found for ½C11H11O4ꢀþ 207.0668, calculated value 207.0657.
ACKNOWLEDGMENTS. This work was supported in part by a Grant-in-Aid for
Scientific Research from the Ministry of Education, Culture, Sports, Science
and Technology, Japan (I.A. and H.M.), by grants from The Naito Foundation
(I.A.) and Takeda Science Foundation (H.M.), and from the National Project
on Protein Structural and Functional Analyses (S.S. and T.K.).
Enzyme Reaction. The reaction mixture contained 54 μM of 4-coumaroyl-CoA,
108 μM of malonyl-CoA, and 20 μg of the purified wild-type enzyme in a final
1. Schröder J (1999) Comprehensive Natural Products Chemistry, (Elsevier, Oxford), Vol 1,
pp 749–771.
12. Rubin-Pitel SB, et al. (2008) Distinct structural elements dictate the specificity of the
type III pentaketide synthase from Neurospora crassa. Chem Biol, 15:1079–1090.
2. Austin MB, Noel JP (2003) The chalcone synthase superfamily of type III polyketide
synthases. Nat Prod Rep, 20:79–110.
3. Morita H, Abe I, Noguchi H (2009) Comprehensive Natural Products Chemistry (Elsevier,
Oxford, in press).
4. Abe I, Takahashi Y, Morita H, Noguchi H (2001) Benzalacetone synthase. A novel
polyketide synthase that plays a crucial role in the biosynthesis of phenylbutanones
in Rheum palmatum. Eur J Biochem, 268:3354–3359.
5. Ferrer JL, Jez JM, Bowman ME, Dixon RA, Noel JP (1999) Structure of chalcone synthase
and the molecular basis of plant polyketide biosynthesis. Nat Struct Biol, 6:775–784.
6. Jez JM, et al. (2000) Structural control of polyketide formation in plant-specific
polyketide synthases. Chem Biol, 7:919–930.
13. Abe I, Sano Y, Takahashi Y, Noguchi
H (2003) Site-directed mutagenesis of
benzalacetone synthase. The role of the Phe215 in plant type III polyketide synthases.
J Biol Chem, 278:25218–25226.
14. Abe T, et al. (2007) Structure function analysis of benzalacetone synthase from Rheum
palmatum. Bioorg Med Chem Lett, 17:3161–3166.
15. Jez JM, Bowman ME, Noel JP (2002) Expanding the biosynthetic repertoire of plant
type III polyketide synthases by altering starter molecule specificity. Proc Natl Acad
Sci USA, 99:5319–5324.
16. Abe I, Abe T, Wanibuchi K, Noguchi H (2006) Enzymatic formation of quinolone
alkaloids by a plant type III polyketide synthase. Org Lett, 8:6063–6065.
17. Funa N, Awakawa T, Horinouchi S (2007) Pentaketide resorcylic acid synthesis by type
III polyketide synthase from Neurospora crassa. J Biol Chem, 282:14476–14481.
18. Morita H, et al. (2008) Crystallization and preliminary crystallographic analysis of a
plant type III polyketide synthase that produces benzalacetone. Acta Crystallogr Sect
F Struct Biol Cryst Commun, 64:304–306.
19. Otwinowski Z, Minor W (1997) Processing of x-ray diffraction data collected in oscilla-
tion mode. Methods Enzymol, 276:307–326.
20. Brunger AT, et al. (1998) Crystallography & NMR system: A new software suite for
macromolecular structure determination. Acta Crystallogr
54:905–921.
21. McRee DE (1992) A visual protein crystallographic software system for X11/Xview. J
Mol Graphics Modell, 10:44–46.
7. Austin MB, Bowman ME, Ferrer JL, Schröder J, Noel JP (2004) An aldol switch dis-
covered in stilbene synthases mediates cyclization specificity of type III polyketide
synthases. Chem Biol, 11:1179–1164.
8. Austin MB, et al. (2004) Crystal structure of a bacterial type III polyketide synthase and
enzymatic control of reactive polyketide intermediates.
J
Biol Chem,
279:45162–45174.
9. Sankaranarayanan R, et al. (2004) A novel tunnel in mycobacterial type III polyketide
synthase reveals the structural basis for generating diverse metabolites. Nat Struct Mol
Biol, 11:894–900.
10. Morita H, et al. (2007) Structural insight into chain-length control and product
specificity of pentaketide chromone synthase from Aloe arborescens. Chem Biol,
14:359–369.
D Biol Crystallogr,
22. Laskowski RA, MacArthur MW, Moss DS, Thornton JM (1993) PROCHECK: a program to
check the stereochemical quality of protein structures. J Appl Crystallogr, 26:283–291.
11. Goyal A, et al. (2008) Structural insights into biosynthesis of resorcinolic lipids by a type
III polyketide synthase in Neurospora crassa. J Struct Biol, 162:411–421.
Morita et al.
PNAS
∣
January 12, 2010
∣
vol. 107
∣
no. 2
∣
673