Structure-based engineering of a plant type III polyketide synthase: Formation of an unnatural nonaketide naphthopyrone

Article abstract of DOI:10.1021/ja070375l

Pentaketide chromone synthase (PCS) from Aloe arborescens is a novel plant-specific type III polyketide synthase (PKS) that produces 5,7-dihydroxy-2-methylchromone from five molecules of malonyl-CoA. On the basis of the crystal structures of wild-type and M207G mutant PCS, the F80A/Y82A/M207G triple mutant was constructed and shown to produce an unnatural novel nonaketide naphthopyrone by sequential condensations of nine molecules of malonyl-CoA. This is the first demonstration of the formation of a nonaketide by the structurally simple type III PKS. A homology model predicted that the active-site cavity volume of the triple mutant is increased to 4 times that of the wild-type PCS.

Full text of DOI:10.1021/ja070375l

Published on Web 04/17/2007
Structure-Based Engineering of a Plant Type III Polyketide
Synthase: Formation of an Unnatural Nonaketide
Naphthopyrone
Ikuro Abe,*, Hiroyuki Morita, Satoshi Oguro, Hisashi Noma,
†,¶
§
†
†
†
‡
‡
†
Kiyofumi Wanibuchi, Nobuo Kawahara, Yukihiro Goda, Hiroshi Noguchi, and
,
§
Toshiyuki Kohno*
Contribution from the School of Pharmaceutical Sciences, UniVersity of Shizuoka, Shizuoka
4
22-8526, Japan, PRESTO, Japan Science and Technology Agency, Kawaguchi, Saitama
3
32-0012, Japan, National Institute of Health Sciences (NIHS), Setagaya, Tokyo 158-8501,
Japan, and Mitsubishi Kagaku Institute of Life Sciences (MITILS),
Machida, Tokyo 194-8511, Japan
Received January 18, 2007; E-mail: abei@ys7.u-shizuoka-ken.ac.jp; tkohno@mitils.jp
Abstract: Pentaketide chromone synthase (PCS) from Aloe arborescens is a novel plant-specific type III
polyketide synthase (PKS) that produces 5,7-dihydroxy-2-methylchromone from five molecules of malonyl-
CoA. On the basis of the crystal structures of wild-type and M207G mutant PCS, the F80A/Y82A/M207G
triple mutant was constructed and shown to produce an unnatural novel nonaketide naphthopyrone by
sequential condensations of nine molecules of malonyl-CoA. This is the first demonstration of the formation
of a nonaketide by the structurally simple type III PKS. A homology model predicted that the active-site
cavity volume of the triple mutant is increased to 4 times that of the wild-type PCS.
(Figure 1B).2,3 The pentaketide-forming PCS was thus function-
ally transformed into an octaketide synthase by the simple steric
Introduction
Pentaketide chromone synthase (PCS) from Aloe arborescens
Liliaceae) is a novel plant-specific chalcone synthase (CHS)
EC 2.3.1.74) superfamily type III polyketide synthase (PKS)
modulation of the chemically inert single residue lining the
active-site cavity. Indeed, recently solved crystal structures of
wild-type and M207G mutant PCS revealed that the large-to-
small substitution dramatically increase the volume of the
polyketide elongation tunnel by opening a gate to newly found
hidden pockets behind the active site of the enzyme (Figure
(
(
that catalyzes sequential decarboxylative condensation of five
molecules of malonyl-CoA (1) to produce 5,7-dihydroxy-2-
methylchromone (2), which is a biosynthetic precursor of the
anti-asthmatic furochromones including khellin and visnagin
4
3A,B). To further manipulate the PCS enzyme reaction, here
Figure 1A).1 A. arborescens PCS shares 50-60% amino acid
,2
we extended the polyketide elongation tunnel of the M207G
mutant by simultaneously replacing two aromatic residues,
Phe80 and Tyr82, forming the bottom of the novel pocket, with
Ala. A homology model based on the crystal structure of the
PCS M207G mutant predicted that the active-site cavity volume
(
sequence identity with those of other type III PKSs of plant
origin. In PCS, CHS’s conserved active-site Thr197, Gly256,
and Ser338 (numbering in M. satiVa CHS), sterically altered in
a number of divergent type III PKSs, are uniquely replaced with
Met, Leu, and Val, respectively (Figure 2). Remarkably,
substitution of the Met207 of PCS (corresponding to the CHS’s
Thr197) with Gly yielded a mutant that efficiently catalyzes
successive condensation of eight molecules of malonyl-CoA to
produce octaketides, SEK4 (3) and SEK4b (4), the products of
the minimal type II (subunit type) PKS for the benzoisochro-
manequinone actinorhodin (act from Streptomyces coelicolor)
3
of the F80A/Y82A/M207G triple mutant (1031 Å ) is increased
3
to 4 times that of the wild-type PCS (247 Å ) (Figure 3).
Experimental Section
14
Chemicals. [2- C]Malonyl-CoA (48 mCi/mmol) was purchased
from Moravek Biochemicals (California). Authentic samples of SEK4/
2
,3
SEK4b and aloesone were obtained in our previous works.
Site-Directed Mutagenesis. Aloe arborescens PCS F80A/Y82A/
M207G mutant was constructed with the QuickChange Site-Directed
†
University of Shizuoka.
PRESTO, Japan Science and Technology Agency.
NIHS.
MITILS.
¶
(
3) (a) Abe, I.; Oguro, S.; Utsumi, Y.; Sano, Y.; Noguchi, H. J. Am. Chem.
Soc. 2005, 127, 12709-12716. (b) Abe, I.; Watanabe, T.; Lou, W.; Noguchi,
H. FEBS J. 2006, 273, 208-218. (c) Abe, I.; Watanabe, T.; Morita, H.;
Kohno, T.; Noguchi, H. Org. Lett. 2006, 8, 499-502. (d) Abe, I. In ACS
Symposium Series 955; American Chemical Society: Washington, DC,
2007; pp 109-127.
‡
§
(
1) For recent reviews on the CHS superfamily type III PKSs, see: (a) Schr o¨ der,
J. In ComprehensiVe Natural Products Chemistry; Elsevier: Oxford, 1999;
Vol. 2, pp 749-771. (b) Austin, M. B.; Noel, J. P. Nat. Prod. Rep. 2003,
2
0, 79-110.
(4) (a) Morita, H.; Kondo, S.; Abe, T.; Noguchi, H.; Sugio, S.; Abe, I.; Kohno,
T. Acta Crystallogr. 2006, F62, 899-901. (b) Morita, H.; Kondo, S.; Oguro,
S.; Noguchi, H.; Sugio, S.; Abe, I.; Kohno, T. Chem. Biol. 2007, in press.
(
2) Abe, I.; Utsumi, Y.; Oguro, S.; Morita, H.; Sano, Y.; Noguchi, H. J. Am.
Chem. Soc. 2005, 127, 1362-1363.
5976
9
J. AM. CHEM. SOC. 2007, 129, 5976-5980
10.1021/ja070375l CCC: $37.00 © 2007 American Chemical Society

Structure-Based Engineering of Type III Polyketide Synthase
A R T I C L E S
Figure 1. Enzyme reaction of (A) wild-type PCS, (B) PCS M207G mutant, and (C) PCS F80A/Y82A/M207G mutant. HPLC elution profile of the reaction
products of the triple mutant.
Mutagenesis Kit (Stratagene), according to manufacturer’s protocol,
using a pair of primers (mutated codons are underlined): sense 5′-
GGGAAGCGTTAGCCAACGCTGACGAGGAGTTC-3′ and anti-
sense as a sense primer, 5′-GAACTCCTCGTCAGCGTTGGCG-
TAACGCTTCCC-3′, and the PCS M207G expression plasmid as the
template.
and the protein was subsequently eluted using 50-200 mM NaCl linear
gradient. Finally, the mutant PCS protein solution was further purified
to homogeneity by Superdex 200HR (10/100GL) (Amersham) and
concentrated to 10 mg/mL in 20 mM HEPES/NaOH (pH 7.0) containing
100 mM NaCl and 2 mM DTT.
Enzyme Reaction. The standard reaction mixture contained 216 µM
of malonyl-CoA and 10 µg of enzyme in a final volume of 500 µL of
100 mM potassium phosphate buffer, pH 6.0. Incubations were carried
out at 30 °C for 1 h. The products were extracted with 1000 µL of
ethyl acetate and analyzed by reverse-phase HPLC on a TSK-gel ODS-
80TsQA column (2.0 × 150 mm) (TOSOH, Tokyo, Japan) with a flow
Enzyme Expression and Purification. After confirmation of the
sequence, protein expression, extraction, and initial purification by
glutathione Sepharose 4B affinity chromatography (Amersham) were
carried out as previously described by PreScission protease digestion
4
to remove the GST tag. The resultant PCS mutant protein thus contains
three additional residues (GPG) at the N-terminal flanking region
derived from the PreScission protease recognition sequence. After
affinity purification and the GST tag removal, the protein solution
containing the mutant PCS was diluted 5-fold with 50 mM HEPES/
NaOH buffer (pH 7.0) containing 5% glycerol and 2 mM DTT and
then was applied to a Resource-Q column (Amersham). The column
was washed with the HEPES/NaOH buffer containing 50 mM NaCl,
2
rate of 0.8 mL/min. Gradient elution was performed with H O and
MeOH, both containing 0.1% TFA: 0-5 min, 30% MeOH; 5-17 min,
30-60% MeOH; 17-25 min, 60% MeOH; 25-27 min, 60-70%
MeOH; 27-35 min, 70% MeOH; 35-40 min, 70-100% MeOH.
Retention time: SEK4 (15.3 min), SEK4b (16.3 min), and the novel
nonaketide (21.1 min). On-line LC-ESIMS spectra were measured with
a Agilent Technologies HPLC 1100 series coupled to a Bruker Daltonics
J. AM. CHEM. SOC.
9
VOL. 129, NO. 18, 2007 5977

A R T I C L E S
Abe et al.
Figure 2. Comparison of the amino acid sequences of Aloe arborescens PCS and other CHS superfamily type III PKSs. M.s CHS, Medicago satiVa CHS;
A.h STS, Arachis hypogaea stilbene synthase; G.h 2PS, Gerbera hybrida 2PS; R.p ALS, Rheum palmatum ALS; A.a OKS, A. arborescens octaketide
synthase; A.a PCS, A. arborescens PCS. The critical active-site residue 197 (in pink), the catalytic triad (Cys164, His303, and Asn336) (in red), and the
residues lining the active site (Phe215, Gly256, F265, and Ser338) (in blue) are marked with # (numbering in M. satiVa CHS), and residues for the CoA
binding with +. Phe80 and Tyr82 of A. arborescens PCS are also marked (in pink).
Figure 3. Comparison of the active-site cavities of (A) wild-type PCS (PDB code 2D3M), (B) PCS M207G mutant (PDB code 2D52), and (C) PCS
F80A/Y82A/M207G mutant (homology model). The residues lining the cavity (blue and orange) are shown with the catalytic triad (yellow). The bottoms
of the pockets of the mutants are highlighted in light blue. The cavity volumes are calculated to be 247, 649, and 1031 Å, respectively.
esquire4000 ion trap mass spectrometer fitted with an ESI source. HPLC
separations were carried out under the same conditions as described
above. The ESI capillary temperature and 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 both negative and positive mode, over a mass
range of m/z 50-600, 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%
Homology Modeling. The model of A. arborescens PCS F80A/
Y82A/M207G mutant was generated with the CPHmodels 2.0 package
5
(http://www.cbs.dtu.dk/services/CPHmodels/) based on the crystal
4
structure of the PCS M207G mutant (PDB code: 2D52). The model
quality was checked using PROCHECK.6 The cavity volume was
calculated on the basis of the residues composing the active-site cavity
wall, by using CASTP (http://cast.engr.uic.edu/cast/).
(
1.5 eV).
Results and Discussion
For a large-scale reaction, 24 mg of malonyl-CoA was incubated
with 10 mg of the enzyme in 40 mL of 100 mM phosphate buffer, pH
.0. Incubations were carried out at 30 °C for 12 h. The products were
extracted with 100 mL of ethyl acetate and separated by reverse-phase
HPLC as described. The experiment was repeated 10 times to get ca.
A. arborescens PCS F80A/Y82A/M207G triple mutant was
functionally expressed in Escherichia coli at levels comparable
with the wild-type enzyme and purified to homogeneity as in
6
4
the case of the wild-type PCS. When incubated with malonyl-
1
mg of pure compound for the NMR measurement. 2-Acetonyl-8,10-
1
CoA as a substrate, the triple mutant afforded a novel product
dihydroxy-5-methyl-4H-naphtho[1,2-b]pyran-4-one (5): H NMR (800
MHz, CD OH) δ 7.20 (s, 1H), 6.60 (s, 1H), 6.52 (s, 1H), 6.31 (s, 1H),
.94 (s, 2H), 2.77 (s, 3H), 2.34 (s, 3H); C NMR (200 MHz, CD
δ 204.6, 181.6, 161.8, 161.4 158.9, 140.7, 135.8, 126.7, 117.5, 114.7,
14
(
0.2% yield, which was calculated by incubation with [2- C]-
3
1
3
malonyl-CoA as a substrate under the standard assay condition;
3
3
OH)
1
08.3, 104.2, 102.8, 48.1, 30.2, 23.3; UV λmax 240, 272, 360 nm; LRMS
(5) Lund, O.; Nielsen, M.; Lundegaard, C.; Worning, P. Abstract at the CASP5
conferenceA102, 2002.
+
(FAB) m/z 299 [M + H] , 277, 185, 93, 75; HRMS (FAB) found for
(
6) Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.; Thornton, J. M. J.
+
[C
H O
17 15 5
]
299.0893, calcd 299.0920.
Appl. Crystallogr. 1993, 26, 283-291.
5978 J. AM. CHEM. SOC.
9
VOL. 129, NO. 18, 2007

Structure-Based Engineering of Type III Polyketide Synthase
A R T I C L E S
1
0 times less efficient than that of the pentaketide production
of the nonaketide naphthopyrone 5 is apparently different from
that of the pentaketide 5,7-dihydroxy-2-methylchromone (2),
the normal product of PCS.
by the wild-type PCS) in addition to SEK4 (3) and SEK4b (4)
(Figure 1C). The product gave a UV spectrum (λmax 240, 272,
+
and 360 nm) and a parent ion peak [M + H] at m/z 299 on
LC-ESIMS, indicating formation of a nonaketide. The nona-
ketide-forming activity was maximum at pH 6.0, but decreased
to ca. 30% under alkaline pH. In contrast, the SEK4/SEK4b-
forming activity of the PCS mutants showed a broad pH
optimum within a range of 6.0-8.0. Thus, the pH change did
not significantly affect the yield and the ratio of SEK4/SEK4b
It is remarkable that the PCS F80A/Y82A/M207G mutant
not only catalyzed condensation of nine molecules of malonyl-
CoA but also altered the mechanism of the cyclization to
produce the unnatural novel “nonaketide” naphthopyrone 5,
which is the longest polyketide generated by the structurally
simple type III PKS. Significant part of the reactions was,
however, terminated at the octaketide stage to afford SEK4/
SEK4b as shunt products. The low yield of the enzyme reaction
products could be attributed to the possible conformational
changes caused by the triple mutation. In addition, it should be
noted that the entrance to the newly formed polyketide tunnel
is still narrow in the triple mutant (Figure 3). Further optimiza-
tion of the active-site structure would lead to improvement of
the yield of the nonaketide product. On the other hand, despite
the structural similarity with eleutherinol (8), formation of the
“octaketide” naphthopyrone was not detected either with the
M207G point mutant or with the triple mutant, which was
confirmed by the LC-ESIMS analysis. The naphthalene ring-
forming activity was thus only attained by the F80A/Y82A/
M207G triple mutation.
(1:4). Only the nonaketide-forming activity suffered sharp
decline under the alkaline pH.
1
The H NMR spectrum of the nonaketide product obtained
from a large-scale incubation revealed the presence of three
aromatic protons (δ 7.20, 6.60, and 6.52, each 1H, s), one R,â-
unsaturated olefinic proton (δ 6.31, 1H, s), one methylene (δ
3
.94, 2H, s), and two methyl protons (δ 2.77 and 2.34, each
3H, s). A structure with a naphthopyrone skeleton was uniquely
consistent with both biogenetic reasoning and NMR spectro-
scopic data, including heteronuclear correlation spectroscopy
(
HMQC and HMBC). The structure of the novel nonaketide
was thus determined to be 2-acetonyl-8,10-dihydroxy-5-methyl-
H-naphtho[1,2-b]pyran-4-one (5) (Figure 1C), which has been
4
isolated until now neither from the aloe plant, a rich source of
As mentioned above, our homology model predicted that the
replacement of the three residues (Phe80, Tyr82, and Met197)
resulted in dramatic increase in the active-site cavity volume.
The functional conversion appeared to be caused by the simple
steric modulation of the active site accompanied by conservation
of the Cys-His-Asn catalytic triad. Interestingly, a similar active-
site architecture with the downward expanding polyketide tunnel
has been reported for a bacterial “pentaketide” naphthalene-
producing type III PKS, 1,3,6,8-tetrahydroxynaphthalene syn-
aromatic polyketides such as pharmaceutically important aloenin
2
(hexaketide), aloesin (heptaketide), and barbaloin (octaketide),
nor from other natural sources.
1
0
thase (THNS) from S. coelicolor that shares only ca. 20%
amino acid sequence identity with the plant enzyme. On the
basis of the crystal structure of S. coelicolor THNS, the possible
involvement of an additional catalytic Cys residue has been
proposed for the unusual naphthalene ring formation reaction.10
However, such an additional Cys residue is not present at the
catalytic center of A. arborescens PCS; only the conventional
Cys174 (corresponding to CHS’s Cys164) serves as the covalent
attachment site for the growing polyketide intermediates,
suggesting a different mechanism of the enzyme reaction in the
PCS F80A/Y82A/M207G mutant.
Interestingly, the structure of the novel nonaketide 5 showed
The proposed mechanism of the formation of the nonaketide
naphthopyrone 5 with a fused tricyclic ring system involves a
consecutive intramolecular aldol condensation (Figure 4A).
Presumably, the enzyme catalyzes the first aromatic ring
formation reaction after the sequential decarboxylative conden-
sations of nine molecules of malonyl-CoA. One of the most
important points here is the timing of the cyclization reaction
and the thioester bond cleavage of the nonaketide intermediate
bound to the active-site Cys. Further, it is not certain whether
all three ring formations are enzymatic or not. At least, the final
pyranone ring formation, as in the case of that of SEK4/SEK4b
(Figure 1B), is likely to be a nonenzymatic process. The partially
cyclized aromatic intermediates would be released from the
active site and undergo subsequent spontaneous cyclizations,
thereby completing the formation of the fused ring system.
close similarity with heptaketides aloesone (2-acetonyl-7-
7
hydroxy-5-methylchromone) (6) produced by the aloe plant and
6
-hydroxymusizin (2-acetyl-1,6,8-trihydroxy-3-methylnaphtha-
8
lene) (7) from rhubarb (Polygonaceae), and an octaketide
eleutherinol (8,10-dihydroxy-2,5-dimethyl-4H-naphtho[1,2-b]-
pyran-4-one) (8), a constituent of another medicinal plant
9
Eleutherine bulbosa (Iridaceae), suggesting that these aromatic
polyketides are produced by closely related type III PKSs.
Indeed, we have previously reported that the G207A mutant of
A. arborescens octaketide synthase produced aloesone from
seven molecules of malonyl-CoA.3a In contrast, the structure
(
(
(
7) Abe, I.; Utsumi, Y.; Oguro, S.; Noguchi, H. FEBS Lett. 2004, 562, 171-
76.
8) Tsuboi, M.; Minami, M.; Nonaka, G.; Nishioka, I. Chem. Pharm. Bull.
1
1
977, 25, 2708-2712.
9) (a) Birch, A. J.; Donovan, F. W. Aust. J. Chem. 1953, 6, 373-378. (b)
Ebnother, A.; Meijer, Th. M.; Schmid, H. HelV. Chim. Acta 1952, 35, 910-
9
(10) Austin, M. B.; Izumikawa, M.; Bowman, M. E.; Udwary, D. W.; Ferrer,
28.
J.-L.; Moore, B. S.; Noel, J. P. J. Biol. Chem. 2004, 279, 45162-45174.
J. AM. CHEM. SOC.
9
VOL. 129, NO. 18, 2007 5979

A R T I C L E S
Abe et al.
Figure 4. Proposed mechanism for the formation of (A) naphthopyrone 5 from a nonaketide intermediate, and (B) emodin anthrone (9) from an octaketide
intermediate.
Interestingly, Harris and co-workers have developed a meth-
odology for chemical syntheses of polycyclic polyketides;
â-polyketones having the two terminal carbonyl groups protected
as ketals are efficiently and directly cyclized into aromatic
polyketides including 6-hydroxymusizin (7) and eleutherinol
Structure-based engineering of the CHS superfamily type III
PKS enzymes would thus lead to further production of chemi-
cally and structurally disparate unnatural novel polyketides.
Acknowledgment. This work was supported by the PRESTO
program from Japan Science and Technology Agency, Grant-
in-Aid for Scientific Research (Nos. 18510190 and 17310130),
and the COE21 program from the Ministry of Education,
Culture, Sports, Science and Technology, Japan.
1
1
(8). Finally, it should be noted that the formation of the
nonaketide naphthopyrone with the fused ring system by the
A. arborescens PCS mutant strongly suggests further involve-
ment of the CHS superfamily type III PKS in the biosynthesis
of anthrones and anthraquinones including emodin anthrone (9)
and barbaloin in the aloe plant (Figure 4B).3
a
Supporting Information Available: A complete set of NMR
1
13
spectra of nonaketide naphthopyrone (5) ( H and C NMR,
DEPT135, HMQC, and HMBC). This material is available free
of charge via the Internet at http://pubs.acs.org.
In summary, this is the first demonstration of the formation
of a nonaketide by the structurally simple type III PKS.
(
11) (a) Harris, T. M; Wittek, P. J. J. Am. Chem. Soc. 1975, 97, 3270-3271.
(
b) Harris, T. M.; Harris, C. M. Pure Appl. Chem. 1986, 58, 283-294.
JA070375L
5980 J. AM. CHEM. SOC.
9
VOL. 129, NO. 18, 2007