Cation-π interaction in the polyolefin cyclization cascade uncovered by incorporating unnatural amino acids into the catalytic sites of squalene cyclase

Article abstract of DOI:10.1021/ja063358p

It has been assumed that the π-electrons of aromatic residues in the catalytic sites of triterpene cyclases stabilize the cationic intermediates formed during the polycyclization cascade of squalene or oxidosqualene, but no definitive experimental evidence has been given. To validate this cation-π interaction, natural and unnatural aromatic amino acids were site-specifically incorporated into squalene-hopene cyclase (SHC) from Alicyclobacillus acidocaldarius and the kinetic data of the mutants were compared with that of the wild-type SHC. The catalytic sites of Phe365 and Phe605 were substituted with O-methyltyrosine, tyrosine, and tryptophan, which have higher cation-π binding energies than phenylalanine. These replacements actually increased the SHC activity at low temperature, but decreased the activity at high temperature, as compared with the wild-type SHC. This decreased activity is due to the disorganization of the protein architecture caused by the introduction of the amino acids more bulky than phenylalanine. Then, mono-, di-, and trifluorophenylalanines were incorporated at positions 365 and 605; these amino acids reduce cation-π binding energies but have van der Waals radii similar to that of phenylalanine. The activities of the SHC variants with fluorophenylalanines were found to be inversely proportional to the number of the fluorine atoms on the aromatic ring and clearly correlated with the cation-π binding energies of the ring moiety. No serious structural alteration was observed for these variants even at high temperature. These results unambiguously show that the π-electron density of residues 365 and 605 has a crucial role for the efficient polycyclization reaction by SHC. This is the first report to demonstrate experimentally the involvement of cation-π interaction in triterpene biosynthesis.

Full text of DOI:10.1021/ja063358p

Published on Web 09/20/2006
Cation-π Interaction in the Polyolefin Cyclization Cascade
Uncovered by Incorporating Unnatural Amino Acids into the
Catalytic Sites of Squalene Cyclase
Noriko Morikubo,† Yoriyuki Fukuda,† Kazumasa Ohtake,§ Naoko Shinya,‡
Daisuke Kiga,‡,# Kensaku Sakamoto,‡ Miwako Asanuma,‡ Hiroshi Hirota,‡,|
Shigeyuki Yokoyama,‡,§, and Tsutomu Hoshino*,†
Contribution from the Department of Applied Biological Chemistry, Faculty of Agriculture, and
Graduate School of Science and Technology, Niigata UniVersity, Ikarashi 2 no-cho, 8050,
Niigata, Niigata 950-2181, Japan, RIKEN Genomic Sciences Center, 1-7-22 Suehiro-cho,
Tsurumi-ku, Yokohama, 230-0045, Japan, Department of Biophysics and Biochemistry, Graduate
School of Science, The UniVersity of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan,
International Graduate School of Arts and Sciences, Yokohama City UniVersity,
1-7-29 Suehiro-cho, Tsurumi-ku, Yokohama 230-0045, Japan, and RIKEN SPring-8 Center,
Harima Institute, 1-1-1 Kouto, Sayo, Hyogo 679-5148, Japan
Received May 15, 2006; E-mail: hoshitsu@agr.niigata-u.ac.jp
Abstract: It has been assumed that the π-electrons of aromatic residues in the catalytic sites of triterpene
cyclases stabilize the cationic intermediates formed during the polycyclization cascade of squalene or
oxidosqualene, but no definitive experimental evidence has been given. To validate this cation-π interaction,
natural and unnatural aromatic amino acids were site-specifically incorporated into squalene-hopene cyclase
(SHC) from Alicyclobacillus acidocaldarius and the kinetic data of the mutants were compared with that of
the wild-type SHC. The catalytic sites of Phe365 and Phe605 were substituted with O-methyltyrosine,
tyrosine, and tryptophan, which have higher cation-π binding energies than phenylalanine. These
replacements actually increased the SHC activity at low temperature, but decreased the activity at high
temperature, as compared with the wild-type SHC. This decreased activity is due to the disorganization of
the protein architecture caused by the introduction of the amino acids more bulky than phenylalanine.
Then, mono-, di-, and trifluorophenylalanines were incorporated at positions 365 and 605; these amino
acids reduce cation-π binding energies but have van der Waals radii similar to that of phenylalanine. The
activities of the SHC variants with fluorophenylalanines were found to be inversely proportional to the number
of the fluorine atoms on the aromatic ring and clearly correlated with the cation-π binding energies of the
ring moiety. No serious structural alteration was observed for these variants even at high temperature.
These results unambiguously show that the π-electron density of residues 365 and 605 has a crucial role
for the efficient polycyclization reaction by SHC. This is the first report to demonstrate experimentally the
involvement of cation-π interaction in triterpene biosynthesis.
1. Introduction
diversity is remarkable, and ca. 120 kinds of triterpene skeletons
have been found until now; lanosterol from vertebrates and
fungi, cycloartenol and R-, â-amyrin from plants, and hopene
from prokaryote are well-known. The polycyclization reactions
proceed with complete regio- and stereospecificity, leading to
the formation of new C-C bonds and chiral centers (Scheme
1); seven chiral centers and four C-C bonds for lanostane
skeleton, while nine stereocenters and five C-C bonds for
hopanoid. Scheme 1 shows the cyclization mechanism of (3S)-
2,3-oxidosqualene 1 to form lanosterol 2 (A) and that of
squalene 4 to afford pentacyclic hopene 5 and hopanol 6 (B).
The question of how the structural diversity and multistep
reactions are attained by a single cyclase enzyme has attracted
prominent organic and biological chemists for over a half-
century.2-5 Prokaryotic squalene-hopene cyclase (SHC) is of
Triterpenes are abundant in nature and confer important
biological functions. The polycyclic triterpene, including the
steroid scaffold, is biosynthesized via ring-forming reactions
(polycyclization) of the linear molecule of squalene or 2,3-
oxidosqualene having a chain length C30.1 The structural
† Niigata University.
‡ RIKEN Genomic Sciences Center.
§ The University of Tokyo.
| Yokohama City University.
RIKEN SPring-8 Center, Harima Institute.
# Current address: Interdisciplinary Graduate School of Science and
Engineering, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku,
Yokohama 226-0026, Japan.
(1) For recent reviews, see: (a) Yoder, R. A.; Johnston, J. N. Chem. ReV. 2005,
105, 4730-4756. (b) Segura, M. J. R.; Jackson, B. E.; Matsuda, S. P. T.
Nat. Prod. Rep. 2003, 20, 304-317. (c) Hoshino, T.; Sato, T. Chem.
Commun. 2002, 291-301. (d) Wend, K. U.; Schulz, G. E.; Corey, E. J.;
Liu, D. R. Angew. Chem., Int. Ed. 2000, 39, 2812-2833.
(2) Woodward, R. B.; Bloch, K. J. Am. Chem. Soc. 1953, 75, 2023-2024.
9
13184
J. AM. CHEM. SOC. 2006, 128, 13184-13194
10.1021/ja063358p CCC: $33.50 © 2006 American Chemical Society

Cation−π Interaction in Squalene Cyclization
A R T I C L E S
Scheme 1. Polycyclization Mechanism of 2,3-oxidosqualene 1 by
Eukaryotic Lanosterol Synthase (A) and that of Squalene 4 by
Prokaryotic Hopene Synthase (B)a
dride and methyl groups (Scheme 1). Recently, we have shown
that the substrate specificity of prokaryotic cyclase can be
successfully altered into that of eukaryotic type, which are
specific to (3S)-2,3-oxidosqualene.7 Since the three-dimensional
X-ray crystallographic structure of squalene cyclase has been
elucidated by Schulz and co-workers in 1997,8 there have been
remarkable advances on the catalytic mechanism by SHC,1c
which have been disclosed mainly by the site-directed mutagen-
esis experiments.9 Site-directed mutations led to the early
truncation of the polycyclization cascade9 and/or to the aberrant
cyclization products whose stereochemistry is opposite to that
of the normal cyclization intermediates, strongly indicating that
the stereochemical destiny during the polycyclization cascade
is directed by the steric bulk size of the active-site residues.9g
Functional analyses of many site-directed mutants and the
identification of the enzymic products from substrate analogues10
demonstrated that the reaction cavity is compact and all the
folded pre-chair conformation is firmly constricted by the
cyclase; thus, no flexible motion of the folded conformation is
allowed inside the reaction cavity once the polycyclization has
started.10d,f Transient carbocation intermediates, e.g., 3 and 7-13
are presumed to be involved in the polycyclization cascade
(Scheme 1).1c Previously, it was suggested that the carboxylate
anion of acidic amino acid residues likely works as the cation-
stabilizing groups.11 In contrast, Dougherty proposed that
interaction between the intermediate carbocations and π-elec-
trons of aromatic amino acids (cation-π interaction) is respon-
sible for the polycyclization reactions.12 Indeed, the X-ray crystal
structure revealed that the reaction cavity is lined with aromatic
amino acid residues.8 The aromatic residues would be more
preferable than the carboxylate anion as the role for stabilizing
cationic intermediates, because the carboxylate anion in the
a Transient carbocation intermediates are generated during the sequential
ring-forming reactions. In the polycyclization process, anti-Markovnikov
addition occurs as shown by the bold arrows (f); one point is found for
lanosterol construction, while two points for hopene biosynthesis. To tolerate
the violation of Markovnikov rule found in the polycyclization reaction,
two concepts have been proposed. One is that the discrete carbocations are
involved in the cyclization reactions and the stable tertiary captions 9 and
11 are first generated according to Markovnikov rule, and then the ring
expansion process from five- to six-membered ring occurs as shown in the
reactions of 9 f 10 and 11 f 12. In contrast, the alternative is that the
polycyclization reaction proceeds with a concertedness, whereby anti-
Markovnikov addition is attained by proper folding of squalene 4 with aid
of the cyclase enzyme in the reaction cavity.13 It has been proposed that
the residue Phe365 stabilizes the bicyclic cation 8,9d while Phe605 stabilizes
the 6,6,6,5-fused tetracyclic 11 and the 6/6/6/6-fused tetracyclic cation 12.9h
(7) Hoshino, T.; Shimizu, K.; Sato, T. Angew. Chem., Int. Ed. 2004, 43, 6700-
6703.
(8) (a) Wendt, K. U.; Lenhart, A.; Schulz, G. E. Science 1997, 277, 1811-
1815. (b) Wendt, K. U.; Poralla, K.; Schulz, G. E. J. Mol. Biol. 1999, 286,
175-187. (c) Reinert, D. J.; Balliano, G.; Schulz, G. E. Chem. Biol. 2004,
11, 121-126.
(9) Our studies, see: (a) Sato, T.; Kanai, Y.; Hoshino, T. Biosci. Biotechnol.
Biochem. 1998, 62, 407-411. (b) Sato, T.; Abe, T.; Hoshino, T. Chem.
Commun. 1998, 2617-2618. (c) Sato, T.; Hoshino, T. Biosci. Biotechnol.
Biochem. 1999, 63, 1171-1180. (d) Hoshino, T.; Sato, T. Chem. Commun.
1999, 2005-2006. (e) Hoshino, T.; Kouda, M.; Abe, T.; Ohashi, S. Biosci.
Biotechnol. Biochem. 1999, 63, 2038-2041. (f) Sato, T.; Hoshino, T. Biosci.
Biotechnol. Biochem. 1999, 63, 2189-2198. (g) Hoshino, T.; Abe, T.;
Kouda, M. Chem. Commun. 2000, 441-442. (h) Hoshino, T.; Kouda, M.;
Abe, T.; Sato, T. Chem. Commun. 2000, 1485-1486. (i) Sato, T.; Hoshino,
T. Biosci. Biotechnol. Biochem. 2001, 65, 2233-2242. (j) Sato, T.; Sasahara,
S.; Yamakami, T.; Hoshino, T. Biosci. Biotechnol. Biochem. 2002, 66,
1660-1670. (k) Sato, T.; Kouda, M.; Hoshino, T. Biosci. Biotechnol.
Biochem. 2004, 68, 728-738. Poralla and co-worker’s studies, see: (l)
Full, C.; Poralla, K. FEMS Microbiol. Lett. 2000, 183, 221-224. (m) Ochs,
D.; Kaletta, C.; Entian, K. D.; Beck-Sickinger, A.; Poralla, K. J. Bacteriol.
1992, 174, 298-302. (n) Schmitz, S.; Fu¨ll, C.; Glaser, T.; Albert, K.;
Poralla, K. Tetrahedron Lett. 2001, 42, 883-885. (o) Merkofer, T.; Pale-
Grosdemange, C.; Wendt, K. U.; Rohmer, M.; Poralla, K. Tetrahedron Lett.
1999, 40, 2121-2124. (p) Pale-Grosdemange, C.; Merkofer, T.; Rohmer,
M.; Poralla, K. Tetrahedron Lett. 1999, 40, 6009-6012.
(10) (a) Hoshino, T.; Kondo, T. Chem. Commun. 1999, 731-732. (b) Hoshino,
T.; Ohashi, S. Org. Lett. 2002, 25, 2553-6. (c) Hoshino, T.; Nakano, S.;
Kondo, T.; Sato, T.; Miyoshi, A. Org. Biomol. Chem. 2004, 2, 1456-
1470. (d) Nakano, S.; Ohashi, S.; Hoshino, T. Org. Biomol. Chem. 2004,
2, 2012-2022. (e) Hoshino, T.; Kumai, Y.; Kudo, I.; Nakano, S.; Ohashi,
S. Org. Biomol. Chem. 2004, 2, 2650-2657. (f) Abe, T.; Hoshino, T. Org.
Biomol. Chem. 2005, 3, 3127-3139. (g) Abe, I.; Tanaka, H.; Noguchi, H.
J. Am. Chem. Soc. 2002, 124, 14514-14515. (h) Tanaka, H.; Noguchi,
H.; Abe, I. Org. Lett. 2004, 5, 803-806. (i) Tanaka, H.; Noguchi, H.; Abe,
I. Tetrahedron Lett. 2004, 45, 3093-3096. (j) Tanaka, H.; Noguchi, H.;
Abe, I. Org. Lett. 2005, 7, 5873-5876.
particular note from the aspect of molecular evolution, because
it is believed that eukaryotic cyclases evolved from the
prokaryotic SHC.6 The polycyclization mechanism of acyclic
squalene is analogous to that of eukaryotic oxidosqualene
cyclase (OSC), although OSC involves further intricate back-
bone rearrangement reactions accompanying 1,2-shifts of hy-
(3) Conforth, J. W.; Conforth, R. H.; Donninger, C.; Popjack, G.; Shimizu,
Y.; Ichii, S.; Forchielli, E.; Capsi, E. J. Am. Chem. Soc. 1965, 87, 3224-
3228.
(4) (a) Eschenmoser, A.; Ruzicka, Jerger, O.; Arigoni, D. HelV. Chim. Acta
1955, 38, 1890-1904. (b) Eschenmoser, A.; Arigoni, D. HelV. Chim. Acta
2005, 88, 3011-3050.
(5) (a) Corey, E. J.; Ortiz de Montellano, P. R. J. Am. Chem. Soc. 1967, 86,
3362-3363. (b) Corey, E. J.; Virgil, S. C.; Sarshar, S. J. Am. Chem. Soc.
1991, 113, 8171-8172. (c) Corey, E. J.; Virgil, S. C.; Cheng, H.; Baker,
C. H.; Matsuda, S. P. T.; Singh, V.; Sarshar, S. J. Am. Chem. Soc. 1995,
117, 11819-11820. (d) Corey, E. J.; Cheng, H. Tetrahedron Lett. 1996,
37, 2709-2712.
(6) (a) Rohmer, M.; Bouvier, P.; Ourisson, G. Proc. Natl. Acad. Sci. U.S.A.
1979, 76, 847-851. (b) Ourisson, G.; Rohmer, M.; Poralla, K. Annu. ReV.
Microbiol. 1987, 41, 301-333. (c) Ourisson, G. Pure Appl. Chem. 1989,
61, 345-348.
(11) (a) Johnson, W. S.; Telfer, S. J.; Cheng, S.; Schubert, U. J. Am. Chem.
Soc. 1987, 109, 2517-2518. (b) Johnson, W. S.; Lindell, S. D.; Steele, J.
J. Am. Chem. Soc. 1987, 109, 5852-5853.
(12) (a) Dougherty, D. A. Science 1996, 271, 163-168. (b) Zacharias, N.;
Dougherty, D. A. Trends Pharmacol. Sci. 2002, 23, 281-287.
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13185

A R T I C L E S
Morikubo et al.
cyclases is likely to form the covalent bonding with the cationic
intermediates, but this irreversible bond formation is unlikely
in case of hydrophobic aromatic residues.
To evaluate the cation-π interaction, we previously prepared
the SHC mutants with Tyr or Trp in place of Phe365 or
Phe605.9d,h These have been assumed to be essential residue
for catalysis;1c,d,8b,c it has been proposed that Phe365 stabilizes
the 6,6-fused bicyclic cation 8,9d while Phe605 does the 6,6,6,5-
fused tetracyclic 11 and the 6/6/6/6-fused tetracyclic cation 12,9h
based on the product profiles afforded by the Phe365Ala and
the Phe605Ala mutants.9d,h The replacement by Tyr or Trp was
expected to enhance the reaction rate due to an increased
π-electron density as compared with Phe. However, the specific
activities of the mutants against incubation temperatures showed
an unusual profile significantly differing from a typical bell-
shape; accelerated reaction rate was observed at low incubation
temperature region, but decreased velocity at high-temperature
region when compared with those of the wild-type. We could
not get the exact answer of why the abnormal profile was found
for the Tyr and Trp mutants. Thus we have failed to give clear
experimental evidence for the cation-π interaction, although
the enhanced reaction rates observed at lower temperatures may
suggest the augmented cation-π interaction.9d,h Further study
had been required to establish the cation-π interaction in
triterpene biosynthesis. Two contradictory mechanisms have
been proposed for the ring-propagation reaction; one is a
stepwise mechanism via discrete carbocations as shown in
Scheme 1, which involves the ring enlargement process from
the five- to six-membered ring (9 f 10 and 11 f 12), the five-
membered ring being initially formed according to venerable
Markovnikov rule; on the other hand, the alternative is the
concerted mechanism13,14 that the cascade takes place without
the involvement of discrete cationic intermediates and without
the violation of Markovnikov rule, by holding the substrate in
the proper conformation to effect the cascade in the active
groove of the protein.13 This debate for the two inconsistent
ideas is still ongoing.
Figure 1. Western blot analysis of the incorporation of OMeTyr at position
605 (lanes 1, 3, and 5) or at position 365 (lanes 2, 4, and 6) of SHC under
the indicated conditions. The SHC variants with OMeTyr were detected by
an anti-His-tag antibody.
similar van der Waals radius compared to that of a hydrogen
atom. Thus, substitution with FluoroPhe is anticipated to inhibit
the cation-π interaction without causing a structural alteration
of the cyclase enzyme, possibly leading to a slower reaction
rate with a normal bell-shape profile. We constructed three
variants with FluoroPhe, which have one to three fluorine atoms
on Phe benzene ring, based on the idea that the greater number
of fluorine atoms leads to the slower reaction rate. The decreased
specific activities and the formation of the prematurely cyclized
products with five-membered ring, which were caused by the
decreased π-electron density of the FluoroPhe variants, verified
that cation-π interaction plays a crucial role both for the
efficient annulation reaction and for the ring enlargement process
in triterpene biosynthesis.
2. Results and Discussion
2.1. Site-Specific Incorporation of Unnatural Amino Acids.
To prepare several milligrams of SHC variants with OMeTyr
and those with FluoroPhes, we employed cell-based and cell-
free systems, respectively. The Escherichia coli system for the
site-specific incorporation of OMeTyr, developed by Schultz
and co-workers,15a involves this unnatural amino acid supple-
mented in the growth media and a pair of an engineered tyrosyl-
tRNA synthetase (OMeTyrRS) and an amber suppressor tRNA
from Methanococcus jannaschii. OMeTyrRS specifically rec-
ognizes OMeTyr and the suppressor tRNA to produce O-
methyltyrosyl suppressor tRNA, which in turn translates amber
codons into OMeTyr. OMeTyrRS does not recognize the other
amino acids or endogenous tRNA species in E. coli, while
neither of OMeTyr and the suppressor tRNA is recognized by
the other aminoacyl-tRNA synthetases than OMeTyrRS.
Introduction of an SHC gene with an amber codon at position
365 or 605 into this E. coli system led to the production of the
SHC variant with OMeTyr at the amber position. To prevent
misincorporation of any of the 20 common amino acids at the
amber position, we used E. coli strain BL21(DE3), which has
no endogenous activity of amber suppression. Since the ef-
ficiency of the suppression is less than 100%, the SHC
molecules truncated at the amber position were produced. A
His-tag was added at the C-terminus of SHC to be separated
from the truncated molecules. Figure 1 shows the western blot
analyses of the SHC variants (Phe365OMeTyr and Phe605-
OMeTyr) expressed in the E. coli cells. Only in the presence of
all of OMeTyr, OMeTyrRS, and suppressor tRNA, the full-
length SHC was produced, which indicated that OMeTyr was
site-specifically incorporated at position 365 or 605.
To provide a deep insight into the polycyclization mechanism,
we planned to incorporate unnatural aromatic amino acids into
squalene cyclase site-specifically. The introduction of unnatural
amino acid(s) at specific position(s) provides a powerful tool
for analyzing protein function.15 In the present study, O-
methyltyrosine (OMeTyr) and fluorophenylalanines (Fluo-
roPhes) were incorporated into SHC in place of Phe365 and
Phe605, and the resulting mutants were examined for the
catalytic activity. The aromatic π-electron density of OMeTyr
is greater than that of Phe, and therefore an increased reaction
rate is expected for the variant with OMeTyr. On the other hand,
a fluorine atom has an extremely higher electronegativity but a
(13) (a) Hess, B. A., Jr. J. Am. Chem. Soc. 2002, 124, 10286-10287. (b) Hess,
B. A., Jr. Org. Lett. 2003, 5, 165-167. (c) Hess, B. A., Jr.; Smentek, L.
Org. Lett. 2004, 6, 1717-1720. (d) Hess B. A., Jr. Collect. Czech. Chem.
Commun. 2004, 69, 261-266. (e) Hess B. A., Jr. Eur. J. Org. Chem. 2004,
2239-2242.
(14) (a) Rajamani, R.; Gao, J. J. Am. Chem. Soc. 2003, 125, 12768-12781. (b)
Gao, D.; Pan, Y.-K.; Buyn, K.; Gao, J. J. Am. Chem. Soc. 1998, 120,
10286-10287.
(15) (a) Wang, L.; Brock, A.; Herberich, B.; Schultz, P. G. Science 2001, 292,
498-500. (b) Chin, J. W.; Santoro, S. W.; Martin, A. B.; King, D. S.;
Wang, L.; Schultz, P. G. J. Am. Chem. Soc. 2002, 124, 9026-9027. (c)
Mehl, R. A.; Anderson, J. C.; Santoro, S. W.; Wang, L.; Martin, A. B.;
King, D. S.; Horn, D. M.; Schultz, P. G. J. Am. Chem. Soc. 2003, 125,
935-939. (d) Wang, L.; Schultz, P. G. Chem. Commun. 2002, 1-10. (e)
Wang, L.; Xie, J.; Schultz, P. G. Annu. ReV. Biophys. Biomol. Struct. 2006,
35, 225-249.
9
13186 J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006

Cation−π Interaction in Squalene Cyclization
A R T I C L E S
Figure 2. Western blot analysis of the expressed SHC variants in the cell-
free system. The SHC variants were detected by an anti-His-tag antibody.
(Lane 1) Suppressor tRNAPhe added to the translation reaction was not
charged with any amino acid. (Lane 2) SHC variant gene with the
Phe605Amber and Tyr606Ala mutations were expressed under the addition
of FluoroPhe-charged suppressor tRNAPhe. (Lane 3) Expression of the SHC
(Tyr606Ala).
SHC variants with 4-fluoro-L-phenylalanine (F1-Phe), 3,4-
difluoro-L-phenylalanine (F2-Phe), and 3,4,5-trifluoro-L-phen-
ylalanine (F3-Phe) at position 365 or 605 were produced in a
cell-free translation system, because no cell-based system is
available for the site-specific incorporation of the fluoro-amino
acids. We used a cell-free translation system based on an S30
extract from E. coli BL21(DE3) cells. Before SHC synthesis,
an amber suppressor tRNA derived from E. coli tRNAPhe was
charged with FluoroPhes by E. coli phenylalanyl-tRNA syn-
thetase (PheRS) in the presence of elongation factor Tu (EF-
Tu) in a test tube. PheRS, which can recognize both of
FluoroPhe and phenylalanine, attaches only FluoroPhe to the
suppressor tRNA in the absence of phenylalanine. In addition,
the aminoacylation of the suppressor tRNAPhe, which has a low
activity,16 requires the enzyme of such a high concentration that
can be achieved only in vitro. The added EF-Tu protects the
ester bond formed between FluoroPhe and tRNA from hydroly-
sis. The charged suppressor tRNA was then applied to the cell-
free translation of an SHC gene with an amber codon. The
efficiency of amino acid incorporation in response to amber
codons is generally higher with purine base A or G at the 3′
side of the amber codon than pyrimidine base C or U.17
Therefore, as for the SHC variant with FluoroPhes at position
605, the tyrosine codon, UAC, at position 606 were substituted
with alanine codon, GCG; the Tyr606Ala substitution has no
influence on the kinetics of squalene cyclization by SHC.9i The
subjection of uncharged suppressor tRNAPhe led to no production
of the full length of SHC (Figure 2), indicating that the
suppressor tRNA was not charged in the cell-free translation
system. Only FluoroPhes therefore occupied the amber position
of the full-length SHCs. This cell-free system yielded 0.1 mg/
mL of the SHC variants with FluoroPhes.
2.2. MS Analyses of SHCs Substituted with Unnatural
Amino Acids. To confirm the incorporation of OMeTyr or
FluoroPhe in response to UAG codon accurately, the purified
SHCs were digested with trypsin (limited digestion) and the
fragment mixtures were then subjected to LC/MS measurement
(ESI-Q/TOF). Figure 3A shows the LC/MS data of the
Tyr606Ala mutated SHC and 3B shows the proteins into which
F1-Phe was incorporated at position 605 of the Y606A mutated
SHC (the double mutated SHC, Phe605F1-Phe/Tyr606Ala).
The LC/MS showed m/z 1403.03 (M + 3H)3+, triply charged
ion, which corresponds to the fragment covering the peptide
residues Gly577-Arg613 of the Tyr606Ala mutated SHC
(Calcd: 4205.91, Obsd: 4206.07 as the monoisotopic mass of
the peptide residues). In contrast, the FluoroPhe-incorporated
SHC exhibited m/z 1408.97 (M + 3H)3+, i.e., Calcd: 4223.90,
Obsd: 4223.89 as the monoisotopic mass, verifying that one
fluorine atom was incorporated. Introduction of F2-Phe into
Figure 3. ESI-Q/TOF mass spectra. (A) Tyr606Ala mutated SHC. (B)
double mutant of Phe605F1-Phe/Tyr606Ala, in which 4-fluorophenylalanine
(F1-Phe) was further incorporated into the Tyr606Ala. The increased mass
of m/z 18 verified that a fluorine atom was incorporated.
position 605 was also confirmed by observing the corresponding
mass increase (see Supporting Information). The incorporation
of OMeTyr at position 605 was also demonstrated; m/z 1433.69
(M + 3H)3+ was observed for the wild-type SHC and m/z
1443.71 (M + 3H)3+ for the OMeTyr-incorporated SHC, these
MS data showing the molecular weights of the peptide fragments
to be 4298.04 Da (Calcd: 4297.94) and 4328.11 Da (Calcd:
4327.95), respectively (Supporting Information). Incorporation
of the unnatural amino acids at position 365 was also demon-
strated by observing the peptide fragment mass covering the
residues Arg355-Arg389 (Supporting Information). These MS
data coupled with the SDS-PAGE analyses (Figures 1 and 2)
conclusively show that the unnatural amino acids are being
incorporated with high fidelity and high selectivity.
2.3. Kinetic Analyses of Tyr-, OMeTyr-, and Trp-
Incorporated SHCs with Higher π-Electron Density. Figure
4 (solid lines) shows the specific activities of SHCs mutated at
position 605 vs incubation temperatures. The catalytic optimal
temperature of the wild-type SHC was found at 55 °C
accompanying a typical bell-shaped curve. The SHC used in
this experiment is from Alicyclobacillus acidocaldarius, a
thermoacidophilic bacterium, thus showing the high optimum
temperature (55 °C) for the catalysis.9
If cation-π interaction
occurs during the polyene cyclization reaction, the reaction rates
of all the mutants should be increased at the entire temperature
range due to the enhanced π-electron density. However, all the
mutants had lower activity at high temperature region, but higher
activity at low temperatures, compared to the wild-type, thus
leading to lower optimal temperatures (Figure 4, solid lines).
An unusual profile differing from a normal bell-shape of the
wild-type was found, especially in the case of the Phe605Trp
and Phe605OMeTyr mutants; three distinctive phases were
observed: at 30-40 °C, the activities were increased; at 40-
50 °C, a steady state was reached; at 55-70 °C, the activities
decreased (Figure 4, solid lines). The enhanced reaction rates
at low-temperature region (<40 °C) implicate the occurrence
of cation-π interaction, but why did the reaction rate signifi-
cantly decrease at higher temperatures? The kinetic data were
estimated according to Lineweaver-Burk equation (Table 1).
The Km values of all the mutants, which were estimated at both
40 and 50 °C, were significantly larger than that of the wild-
type, indicating that the affinity between the mutants and the
(16) Peterson, E. T.; Uhlenbeck, O. C. Biochemistry 1992, 31, 10380-10389.
(17) Phillips-Jones, M. K.; Watson, F. J.; Martin, R. J. Mol. Biol. 1993, 233,
1-6.
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13187

A R T I C L E S
Morikubo et al.
Figure 5. CD spectra of the mutated SHCs measured at 50 °C. This figure
shows that the protein architectures of Phe605OMeTyr and Phe605Trp
mutants were noticeably altered at the catalytic optimal temperature of 50
°C (see the solid lines of Figure 4). (b) wild-type, (9) Phe605Tyr, (×)
Phe605Trp, (*) Phe605OMeTyr. At ambient temperature, the CD spectra
of the three mutants were superimposable on that of the wild-typ.9d,h
However, significant local change of the enzyme had indeed occurred by
the introduction of the larger steric bulk sizes, despite no visible effect on
the overall CD having been found at ambient temperature.
Figure 4. Specific activities of the wild-type and the mutated SHCs having
OMeTyr, Tyr, Trp, 4-monofluoroPhe (F1-Phe), 3,4-difluoroPhe (F2-Phe),
and 3,4,5-trifluoroPhe (F3-Phe) at position 605. Solid lines: the wild-type
and the Phe606Ala mutant (b); Phe605Tyr mutant (9); Phe605Trp (×);
Phe605OMeTyr mutant (*). Dotted lines: FluoroPhe mutants, in which
Tyr606 are further mutated to Ala. The specific activity of the Tyr606Ala
is the same as the wild type (see the text and ref 9i). F1-Phe mutant (0);
F2-Phe mutant (]); F3-Phe mutant (4). The kinetic data of the double
mutants correspond to those of the single mutant with FluoroPhes at position
605, because the kinetic values of the wild-type are the same as those of
the site-directed Tyr606Ala mutant.
Table 1. Kinetic Parameters of the Site-Directed Mutants
Measured at 40 °C and at 50 °C, Which Were Determined by
Estimating the Amount of Hopene 5 Produced in the Enzymic
Reactions
measured at 40
°
C
measured at 50
Km
M)
°C
Km
kcat
kcat
(min-
)
1
1
(
µM)
(min-
)
(µ
wild-type
Phe605Tyr
Phe605OMeTyr
Phe605Trp
Phe365Tyr
15.4
52.5
105
118
nda
nd
15
16
27
55
nd
nd
nd
24.2
104
119
109
42.0
79.0
10.2
8.95
-
152
191
130
185
Phe365OMeTyr
Phe365Trp
b
nd
-
Figure 6. CD intensities (molecular ellipticities, ∆ mol ellip.) of the mutated
SHCs subtracted from that of the wild-type, which were plotted at 222 nm
against temperatures. The transparent solutions of the Trp and OMeTyr
mutated SHCs became cloudy at 75 °C, indicating the occurrence of thermal
denaturation. (9) Phe605Tyr mutant; (*) Phe605OMeTyr mutant; (×)
Phe605Trp mutant. The slopes of the Phe605Trp and Phe605OMeTyr
mutants were significantly larger than that of the Phe605Tyr mutant,
indicating that the enzyme structures of the Phe605Trp and Phe605OMeTyr
mutants were more markedly disordered than that of the Phe605Tyr, due
to the replacement of the increased steric bulk sizes.
a nd: not measured. b -: not determined due to marginal activity.
substrate markedly decreased in the following order: Phe (the
wild-type) > Tyr > OMeTyr > Trp; the smallest Km for the
wild-type, with the largest one for the Trp mutant. The more
bulky substituents gave the larger Km (looser binding) (Table
1) with elevating incubation temperatures, which would have
resulted in the profile of steady state (at 40-50 °C), despite
π-electron density having been enhanced. Figure 5 shows the
circular dichroism (CD) of the wild-type and the mutants, which
were measured at 50 °C. The intensities of the negative Cotton
effect were in the following order: Phe (the wild-type) > Tyr
> OMeTyr > Trp. The decreased Cotton effect found in the
mutants suggests the disorganization of the protein architecture.
The difference of Cotton effect (∆ molecular ellipticities)
between the wild-type and the mutants was greatly enhanced
by elevating temperatures in the following order: Trp g
OMeTyr > Tyr (Figure 6), which is in a good accordance with
the increasing order of Km values (Table 1). Thus, it is strongly
indicated that the decreased activities at higher temperatures
(Figure 4) would have resulted from thermal denaturation of
the cyclase structure; the Trp mutant with the largest steric bulk
size underwent the highest susceptibility to high temperatures
among all the mutants. Thus, the wild-type with Phe (the
smallest size) at position 605 showed the highest stability against
thermal denaturation. These results indicate that the geometrical
local change of SHC protein architecture had been induced even
by single amino acid substitution. However, it is noticeable that
the reaction velocities of the Trp, OMeTyr and Tyr mutants
were enhanced at lower temperature region (30-45 °C) (Table
1 and Figure 4, solid lines) despite the affinity having been
looser due to the introduction of larger amino acid, thus implying
9
13188 J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006

Cation−π Interaction in Squalene Cyclization
A R T I C L E S
Table 2. Kinetic Values of the FluoroPhe-Incorporated SHCs
Measured at 50 °Ca
relative
activity
(%)
Km
M)
kcat
(min-
k
cat/Km
1
(
µ
)
(min-1 µM)
/
wild-type or Tyr606Ala mutant
Phe605F1-Phe/Tyr606Ala
Phe605F2-Phe/Tyr606Ala
Phe605F3-Phe/Tyr606Ala
Phe365F1-Phe
24.2
27.9
32.4
-
35.2
41.6
47.9
119
4.92
1.76
0.99
100
35.8
20.2
-
37.6
27.0
19.1
49.2
32.2
-
65.1
55.2
45.1
-
1.85
1.33
0.94
Phe365F2-Phe
Phe365F3-Phe
a All the SHCs into which the FluoroPhes are incorporated at position
605 are double mutants, in which Tyr606 is further replaced by Ala in
order to enhance the expression level of the unnatural FluoroPhe-
incorporated SHCs (see section 2.1 in the text). Kinetic values of the wild-
type are the same as those of the site-directed Tyr606Ala mutant, which
was clearly demonstrated in our previous experiments.9i Thus, the kinetic
data of the double mutants correspond to those of the single mutant with
FluoroPhes at position 605. All the FluoroPhe-incorporated SHCs at position
365 are the single amino acid substituted mutants, but not doubly mutated
ones. The Km and kcat values of Phe605F3-Phe were not determined due
to the negligibly small activity.
Figure 7. Specific activities of the mutated SHCs into which natural and
unnatural amino acids were incorporated at position 365. Wild-type (b);
Phe365Tyr (9); Phe365Trp (×); Phe365OMeTyr (*); Phe365F1-Phe (0)
Phe365F2-Phe (]); Phe365F3-Phe (4).
caused by the incorporation of the FluoroPhes on the enzyme
structure was minimum. Indeed, in the case of the FluoroPhe
mutants at position 605, the changes in the Km’s by increasing
the number of fluorine atoms were modest (Table 2): 1.15-
fold and 1.34-fold looser binding for the F1-Phe and for the
F2-Phe mutants, respectively, as compared with that of the wild-
type. On the contrary, the kcat’s were more significantly
decreased; 2.4-fold slower for the F1-Phe variant and 3.7-fold
reduction for the F2-Phe mutant compared to that of the wild-
type. In other words, with increasing the number of substituted
fluorine atoms, the more decreased reaction rates were clearly
observed, but the binding remained almost unchanged, highly
indicating that Figure 4 (dotted lines) displays essentially the
effect of π-electron density on the enzyme activity; thus, the
occurrence of cation-π interaction at position 605 was unam-
biguously demonstrated. Daugherty and co-workers calculated
the cation-π binding energies for the binding of the sodium
cation lying over the center of the aromatic rings.18 The plots
of the specific activities against the binding energies had a good
correlation (Figure 8, correlation R2 ) 0.96 for the FluoroPhe
mutants at position 605), further supporting that the Phe605
residue actually plays the critical role for cation-π interaction.
On the other hand, the Phe365FluoroPhe variants had the more
increased Km values compared to those of the Phe605FluoroPhe
mutants (Table 2): 1.45-fold, 1.72-fold, and 1.98-fold poorer
binding for the F1-Phe, the F2-Phe, and the F3-Phe variants,
respectively, compared to that of the wild-type, but the kcats
were significantly decreased with increasing the number of the
fluorine atoms (1.83-fold, 2.16- and 2.64-fold decreased for the
F1-, F2-, and F3-Phe variants, respectively). The more
increased Km values of the FluoroPhe variants at position 365,
compared to those at position 605, indicate that the active site
region around position 365 is more tightly packed than that of
position 605. This is consistent with the inference from the
comparison of the Tyr, OMeTyr, and Trp variants at positions
365 with those at position 605, which was discussed in the
previous section. No change in Km and a significantly lowered
kcat by the fluorination are ideal for verifying cation-π
that cation-π interaction is likely during squalene cyclization
reaction.
Figure 7 (solid lines) depicts the specific activities of the
SHCs having electron-donating substituents at position 365. The
Phe365 residue has been suggested to stabilize the 6/6-fused
bicyclic tertiary cation 8 through cation-π interaction.8b,9d These
mutants also showed abnormal profiles in a way similar as the
SHC variants mutated at position 605 (see the solid lines of
Figures 4 and 7). As in case of the mutants at position 605, the
more bulky substituents gave the larger Km (Table 1), leading
to the profile of steady state (30-55 °C for the Phe365OMeTyr
and 35-45 °C for the Phe365Tyr). Despite the indole ring
having richest π-electron density among the aromatic rings used,
the Phe365Trp mutant had no enzymatic activity (Figure 7);
however, the Phe605Trp had still somewhat high activity (Figure
4). This suggests that a local change of the protein structure at
position 365 occurred more significantly, compared to that at
position 605 (Table 1 and compare Figure 4 with 7); thus, the
active-site region at position 365 is more tightly compact than
that at position 605.
2.4. Kinetic Analyses of Fluorophenylalanine-Incorporated
SHCs with Lowered π-Electron Density. Figures 4 and 7
(doted lines) show the specific activities of the Phe605FluoroPhes/
Tyr606Ala and Phe365FluoroPhe mutants against incubation
temperatures, respectively. The profiles showed a typical bell-
shape similar to that of the wild-type. As we have reported
previously, no difference of the kinetic data (Km and kcat) and
CD was found between the Tyr606Ala and the wild-type SHCs.9i
Thus, the kinetic data of the double mutants at positions 605
and 606 correspond to those of the single mutant with
FluoroPhes at position 605. The optimal temperatures (55 °C)
of all the FluoroPhe mutants at positions 365 and 605 for the
catalysis were the same as that of the wild-type and no difference
of the CD spectra between all of the fluorine-inserted SHCs
and the wild-type SHC was found at temperature range of 20-
70 °C (see Supporting Information). Thus, the steric effect
(18) Mecozzi, S.; West, A. P., Jr.; Daugherty, D. A. Proc. Natl. Acad. Sci. U.S.A.
1996, 93, 10566-10571.
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13189

A R T I C L E S
Morikubo et al.
dammarene and 20(R)-(17RH)-12(13)-dammarene, respectively,
by comparing the GC/MS with those of the authentic samples,
which were previously isolated by us from some mutants.9g,h
Production of 14 and 15 having a 6/6/6/5-fused tetracycle was
also found in the incubation mixtures of squalene 4 with F1-
and F3-Phe mutants. Formation mechanism of 14 and 15 is
illustrated in Scheme 2. The tetracyclic dammarenyl cations 16
could be produced according to Markovnikov rule. To gain the
20(R)-stereochemistry, the dammarenyl skeleton must have
17R-H and 17â-side chain as shown in 16,10c,20 indicating that
squalene 4 was folded in a chair/chair/chair/boat conformation
18. In contrast, chair/chair/chair/chair conformation 17 leads
to 20(S)-dammarenyl skeleton.10c,20 The hydride shift of H-17
to C20-cation and the deprotonation of H-13 on 16 in anti-
periplanar manner could give 14. The 1,2-hydride shifts of H-17
and H-13 in an anti-parallel fashion, followed by the deproto-
nation of H-12ax, could afford 15. The π-electron density is
significantly decreased by the introduction of fluorine atom on
Phe, but the size of F2-Phe is almost the same as that of Phe,
which was demonstrated indeed by the kinetic and CD results
as described above. Formation of the tetracyclic ring system
by Markovnikov closure definitively demonstrates that the
π-electron density of Phe605 is important for the five- to six-
membered ring enlargement process to yield 12 from 11. In
fact, the production amounts of 14 and 15 relative to 1 was
markedly increased with increasing the number of fluorine
atoms; the ratio of the sum of 14 and 15 to 1 was 0.09, 0.17,
and 0.94 by introducing F1, F2, and F3, respectively. Cation 16
is preferred to cation 11, because unfavorable steric repulsion
(1,3-diaxial interaction) occurs among 30-Me, 21-Me, and 27-
Me for all-chair conformation 17, but the repulsive interaction
is less in chair/chair/chair/boat conformation 18.10c The prefer-
ence of 16 to 11 is further supported by the computational
energy calculation.21a Failure to form the unstable secondary
cation 12, which was caused by the decreased π-electron density,
would have led to the energetically favored 16 during the
cyclization reaction. Thus, the aromatic Phe is crucial not only
for stabilizing the intermediary secondary cation 12, but also
for posing the energetically unfavorable conformation 17. No
truncated cyclization product was found in the incubation
mixtures with the Tyr, OMePhe, and Trp mutants, further
indicating that the energetically favored chair/chair/chair/boat
conformation 18 was generated due to the reduction in
π-electron density.
Figure 8. Relationship between the specific activities of FluoroPhe-inserted
SHCs and cation-π binding energies. Cation-π binding energies (kcal/
mol) of Phe, F1-Phe, F2-Phe and F3-Phe are reported to be 27.1, 21.9,
16.8, and 12.4, respectively.18 The SHC activities were determined by
estimating the amount of hopene produced at 50 °C for 60 min. Closed
circles and solid line: the mutated SHCs into which FluoroPhes and Ala
were incorporated at both positions 605 and 606, respectively, but the
activities correspond to that of the single mutant at position 605 (see text).
Open circles and dotted line: the FluoroPhe variants mutagenized only at
position 365. The correlations for the site-directed mutants with Trp and
Tyr at positions 365 and 605 were added to clearly show that the Trp and
the Tyr mutants at position 365 greatly shifted from the straight line. The
cation-π binding energies for Trp and Tyr were reported to be 32.6 and
26.9 kcal/mol, respectively,18 the value for OMeTyr having not been
calculated.
interaction. Due to occurrences of somewhat larger changes in
both kcat’s and Km’s for the FluoroPhe variants, it may be
difficult to strongly indicate cation-π interaction at position
365. However, it should be noted that the increased Km values
by incorporating the F1- to F3-Phe at position 365 (Km ) 35-
48) were significantly smaller, compared to those by introducing
the Tyr and OMeTyr (Km ) 130-180, and an exceptionally
large Km for the Trp mutant) (see Tables 1 and 2), thus, leading
to the same catalytic optimal temperature of the F1- to F3-
Phe variants as that of the wild-type (Figure 7, dotted lines),
which permitted us to compare the relative activities kcat/Km
between the FluoroPhe mutants and the wild-type at a given
temperature. The relative activities were inversely proportional
to the number of the attached fluorine atoms (Table 2).
Furthermore, the specific activities of the FluoroPhe-incorpo-
rated mutants at position 365 were well correlated with cation-π
binding energies (Figure 8, R2 ) 0.95). This good correlation
also would have been obtained due to the relatively small
changes, though not to be neglected, in the Km’s of the
FluoroPhe variants at position 365. In contrast, the Trp and the
Tyr mutants at position 365 having the exceptionally large Km
values greatly shifted from the straight line of the cation-π
binding energies (Figure 8). Thus, it would be highly probable
that Phe365 also has a role in stabilizing transient carbocation
intermediate as well as Phe605, for which cation-π interaction
was credibly established as described above.
Previously, we reported that the Phe365Ala mutant afforded
the premature cyclization products having the 6/6-fused bicyclic
skeleton (polypodatetraene and neopolypodatetraene),9d,i which
could be produced from intermediary cation 8. However, no
detectable amount of these bicyclic compounds was found in
the incubation mixtures with the FluoroPhe variants (Figure 9C).
A geometrical local change would have occurred around position
365, due to the replacement of a significantly decreased steric
bulk size of Ala, possibly leading to the termination of the
polycyclization reaction at bicyclic intermediate stage 8. No
detectable amount of the bicyclic products by the FluoroPhe
variants further indicates that the structural alteration of the
2.5. Product Profiles by the SHC Variants with Fluo-
rophenylalanines. Figure 9 shows the GC traces of hexane
extract from the incubation mixture of squalene 4 with the
Phe605F2-Phe mutated SHC. In the case of the wild-type, the
GC peaks other than hopene 5 and hopanol 6 was marginal,19
but the mutant with the FluoroPhe afforded two new peaks of
14 and 15 in significantly large amounts besides 5 and 6.
Products 14 and 15 were identified to be 20(R)-13(17)-
(19) Pale-Grosdemange, C.; Feil, C.; Rohmer, M.; Poralla, K. Angew. Chem.,
Int. Ed. 1998, 37, 2237-2240.
(20) Abe, I.; Rohmer, M. J. Chem. Soc., Perkin Trans. 1 1994, 30, 783-791.
(21) (a) Xiong, Q.; Rocco, F.; Wilson, W. K.; Xu, R.; Ceruti, M.; Matsuda, S.
P. T. J. Org. Chem. 2005, 70, 5362-5375. (b) Matsuda, S. P. T.; Wilson,
W. K.; Xiong, Q. Org. Biomol. Chem. 2006, 4, 530-543.
9
13190 J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006

Cation−π Interaction in Squalene Cyclization
A R T I C L E S
Figure 9. GC traces of hexane extracts from the incubation mixtures of squalene 4 with the wild-type SHC (A), with the mutated SHC having F2-Phe at
position 605 (B) and with the mutant having the F2-Phe at position 365 (C). Incubation conditions: squalene substrate 1.0 mg, SHCs 100 µg, pH 6.0 and
55 °C for 3 h. An excess of Triton X-100 used in the incubations was removed by a short SiO2 column, eluting with a mixture of hexane and EtOAc
(100:20). Peaks 4, 5, and 6 correspond to squalene, hopene, and hopanol, respectively. Peaks 14 and 15 were identified as 20(R)-13(17)-dammarene and
20(R)-(17RH)-12(13)-dammarene, respectively.
Scheme 2. Formation Mechanism of 14 and 15 by the
Phe605F2-Phe Mutated SHC
of the cation-π interaction depends on the nature of the
aromatic amino acids involved in protein. Of the natural
aromatic residues, Trp forms the strongest interaction with
cations due to the greater electron density of Trp relative to
Tyr or Phe. However, the Trp- and OMeTyr-incorporated SHCs
showed significantly decreased CD magnitude with elevating
temperatures and led to a lowered catalytic optimal temperature,
compared to the wild-type. In this study, we showed that
introduction of the larger substituents at the active site led to a
looser binding to the substrate and that the reaction rate was
more dramatically decreased at elevated temperatures due to
enhanced susceptibility of the mutants to thermal denaturation,
compared to that of the wild-type. On the other hand, the
difference between the steric bulk size of FluoroPhe and that
of Phe is marginal, but the π-electron density of the FluoroPhe
is significantly lower than that of Phe. In fact, this study showed
that, in the case of the FluoroPhe variants at position 605, the
increase in Km was small (1.15-fold for the F1-Phe mutant),
but the reduction in kcat was more significant (2.42-fold for the
F1-Phe mutant), compared to those of the wild-type, and the
catalytic optimal temperature was identical to that of the wild-
type. In addition, the enzyme activities of the SHC variants with
FluoroPhes were found to be inversely proportional to the
number of the fluorine atoms on the aromatic ring, and clearly
correlated with cation-π binding energies of the aromatic ring
moiety (Figure 8). These findings definitively established that
the residue Phe605 stabilizes carbocation intermediates via
cation-π interaction. On the other hand, the Phe365FluoroPhe
mutants had relatively larger Km’s, as compared with those of
the Phe605FluoroPhe variants, and also had significantly
decreased kcat. The more increased Km’s indicate the severely
tight packing around position 365. No change in Km and a
significant lowering in kcat for the FluoroPhe mutants are ideal
FluoroPhe-incorporated SHCs was a little. Now many definitive
(22) (a) Joubert, B. M.; Hua, L.; Matsuda, S. P. T. Org. Lett. 2000, 2, 339-
evidences are available that alteration of the steric bulk size of
341. (b) Segura, M. J. R.; Meyer, M. M.; Matsuda, S. P. T. Org. Lett.
active sites affords the aberrant cyclization product(s).1c,9g,22
2000, 2, 2257-2259. (c) Meyer, M. M.; Segura, M. J. R.; Wilson, W. K.;
Matsuda, S. P. T. Angew. Chem., Int. Ed. 2000, 39, 4090-4092. (d) Meyer,
M. M.; Xu, R.; Matsuda, S. P. T. Org. Lett. 2002, 4, 1395-1398. (e) Segura,
M. J. R.; Lodeiro, S.; Meyer, M. M.; Patel, A. J.; Matsuda, S. P. T. Org.
Lett. 2002, 4, 4459-4462. (f) Lodeiro S.; Segura M. J.; Stahl, M.; Schulz-
Gasch, T.; Matsuda, S. P. T. ChemBioChem 2004, 5, 1581-1585. (g) Wu,
T. K.; Chang, C. H. ChemBioChem 2004, 5, 1712-1715.
3. Concluding Remarks
We described here the first site-specific incorporation of
unnatural amino acids into a triterpene cyclase. The magnitude
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13191

A R T I C L E S
Morikubo et al.
for proposing cation-π interaction. Thus, the enhanced Km’s
of the FluoroPhe variants at position 365 may not give
convincing evidence for cation-π interaction at position 365.
However, the increased Km values were significantly smaller
compared to those of the Trp, OMeTyr, and Tyr mutants, leading
to little effect on the catalytic optimum temperature (Figure 7,
dotted line) and to a good correlation between the enzyme
activity and cation-π binding energies (Figure 8, dotted line).
Furthermore, the relative activities (kcat/Km) of the FluoroPhe
variants were decreased progressively with increasing the
number of fluorine atoms (Table 2). Therefore, it is fair to
assume that the Phe365 residue does stabilize an intermediary
cation through cation-π interaction. In addition, the crystal
structure of human lanosterol synthase also revealed that the
Phe444 residue, equivalent to Phe365 of SHC, stabilizes the
tertiary cations at C6 and C10 through cation-π interaction.23
The van der Waals radius of the hydrogen atom (120 pm) is
closest to that of fluorine atom (147 pm). The increased radius
(27 pm), the difference between fluorine and hydrogen, for the
fluorine-inserted mutant is significantly small, but squalene
cyclase differentiated fluorine from hydrogen to some extent,
especially with respect to the site 365. The difference of the
van der Waals radius (152 pm, extra oxygen atom attached on
Phe) between the Tyr mutant and the wild-type (Phe), is
relatively larger, which caused thermal denaturation of the
protein (Figure 7). Thus, the present study affords the additional
evidence that the enzyme cavity of the cyclase is exceptionally
compact; thus, no flexible motion of squalene molecule 4 is
allowed inside the reaction cavity. The tightly packed enzyme
cavity would enforce squalene 4 into a meandering conforma-
tion, which was shown by the X-ray crystallographic analysis;8c
the folded conformation allows the close contact of a double
bond to carbocation intermediate to facilitate the subsequent
ring-forming reaction. Cation-π interaction in conjunction with
the meandering conformation of the substrate would effectively
foster the progressive propagation of carbocyclic rings.
be modest, which was also suggested by Matsuda et al.21b (3S,-
22S)-2,3:22,23-Dioxidosqualene was converted to 17,24-epoxy-
25-hydroxybaccharane-3â,25-diol by lupeol synthase,26 further
supporting the involvement of secondary baccharenyl cation.
In addition, secondary cation 12 with 6/6/6/6-fused tetracycle
was trapped also by incubating squalene 4 with Ile261Gly and
Ile261Ala mutated SHCs,9g in which the crucial catalytic site
of Phe605 was still retained. Thus, the ring expansion would
occur with aid of a greater interaction between the secondary
cation intermediate and π-electrons of the catalytic sites.
Previously, Jensen and Jorgensen suggested, on the basis of the
force-field calculation, that the energy barrier between secondary
and tertiary cations might be lowered by selective placement
of nucleophilic group from the protein backbone or side chain
including the indole ring of tryptophan.27 The present study
clearly validated this idea. In conclusion, this study unambigu-
ously demonstrated that cation-π interaction occupies a key
position in the catalytic mechanisms by triterpene cyclases.
Cation-π interactions are now well recognized to play an
important role in biological systems, including protein struc-
ture,28 protein-ligand interactions,29 ion channels,30 etc. Car-
bocationic intermediates are assumed to be involved in many
key enzymes of isoprenoid biosyntheses,30 as exemplified by
31,33
isopentenyldiphosphate isomerase,31,32 prenyltransferase,
squalene synthase31,34 and many terpene cyclases.31 The cat-
ion-π interaction proposed for these isoprenoid biosyntheses
has been inferred from the X-ray crystallography and/or
mutagenesis using conventional natural amino acids. To estab-
lish the cation-π interactions for the terpenoid biosyntheses at
enzymatic level, the site-specific incorporation experiments of
unnatural amino acids using in vivo and/or in vitro translation
systems would be necessary, because variation of steric and
electronic environments using 20 natural amino acids is limited,
that is, a number of insights that would have been impossible
using conventional site-directed mutations would be gained. The
technology of the site-specific incorporation of unnatural amino
acids into protein is fascinating tool for the correct understanding
enzyme mechanisms.
It is of particular note that the aromatic π-electrons of Phe
can facilitate the ring enlargement process from five- to six-
membered D-ring (from tertiary to secondary cations), as
evidenced by the formation of 6/6/6/5-fused tetracycles 14 and
15 by the Phe605FluoroPhe variants. Corey et al. gave the
experimental evidence that the six-membered C-ring in lanos-
terol skeleton is produced by the ring expansion process from
five-membered C-ring, which was initially formed according
to Markovnikov rule.5c,d The ring expansion process has been
supported also by biomimetic studies.24 Incubation of 22,23-
dihydro-2,3-oxidosqualene, lacking a terminal double bond, with
â-amyrin synthase gave bacchar-12-en-3â-ol via 6/6/6/6-
cation.25 This finding indicates that the D-ring expansion did
occur without the benefit of cation stabilization by a terminal
double bond of the side chain; thus, the anchimeric assistance
by an approach of double bond to intermediate cations would
(26) Shan, H.; Segura, M. J. R.; Wilson, W. K.; Lodeiro, S.; Matsuda, S. P. T.
J. Am. Chem. Soc. 2005, 127, 18008-18009.
(27) Jenson, C.; Jorgensen, W. L. J. Am. Chem. Soc. 1997, 119, 10846-10854.
(28) (a) Fernandez-Recio, J.; Romero, A.; Sancho, J. J. Mol. Biol. 1999, 290,
319-330. (b) Ting, A. Y.; Shin, I.; Lucero, C.; Schultz, G. J. Am. Chem.
Soc. 1998, 120, 7135-7136. (c) Loewenthal, R.; Sancho, J.; Fersht, A. R.
J. Mol. Biol. 1992, 224, 759-770. (d) Burley, S. K.; Petsko, G. A. FEBS
Lett. 1986, 203, 139-143. (e) Perutz, M. F.; Fermi, G.; Abraham, D. J.
Poyart, C.; Bursauz, E. J. Am. Chem. Soc. 1986, 108, 1064-1078.
(29) (a) Susman, J. L.; Harel, M.; Folow, F.; Oefner, C.; Goldman, A.; Toker,
L.; Silman, I. Science 1991, 253, 872-879. (b) Zhong, W.; Gallivan, J. P.;
Zhang, Y.; Li, L.; Lester, H. A.; Dougherty, D. A. Proc. Natl. Acad. Sci.
U.S.A. 1998, 95, 12088-12093. (c) Satow, Y.; Cohen, G. H.; Padlan, E.
A.; Davis, D. R. J. Mol. Biol. 1986, 190, 593-604. (d) Beene, D. L.; Brandt,
G. S.; Zhong, W.; Zacharias, N. M.; Lester, H. A.; Dougherty, D. A.
Biochemistry 2002, 41, 10262-10269. (e) Lummis, S. C. L.; Beene, D.;
Harrison, N. J.; Lester, H. A.; Dougherty, D. A. Chem. Biol. 2005, 12,
993-997.
(30) (a) Okada, A.; Miura, T.; Takeuchi, H. Biochemistry 2001, 40, 6053-
6060. (b) Kumpf, R. A.; Dougherty, D. A. Science 1993, 261, 1708-1710.
(31) Cane, D. E., Barton, D., Nakanishi, K., Meth-Cohn, O., Eds. Isoprenoids
Including Carotenoids and Steroids. In ComprehensiVe Natural Product
Chemistry, 2nd ed.; Pergamon/Elsevier: Amsterdam, 1999.
(32) Durbecq, V.; Sainz, G.; Oudjama, Y.; Clantin, B.; Bompard-Gilles, C.;
Tricot, C.; Caillet, J.; Stalon, V.; Droogmans, L.; Villere, V. EMBO J.
2001, 20, 1530-1537.
(33) Ogura, K.; Koyama, T. Chem. ReV. 1998, 98, 1263-1276.
(34) (a) Tansey, T. R.; Shechter, I. Biochim. Biophys. Acta 2000, 1529, 49-62.
(b) Pandit, J.; Danley, D. E.; Schulte, G. K.; Mazzalupo, S.; Pauly, T. A.;
Hayward, C. M.; Hamanaka, E.; Thompson, J. F.; Harwood, H. J., Jr. J.
Biol. Chem. 2000, 275, 30610-30617.
(23) Thoma, R.; Schulz-Gasch, T.: D’Arcy, B.; Benz, J.; Aebi, J.; Dehmlow,
H.; Hennig, M.; Stihle, M.; Ruf, A. Nature 2004, 432, 118-1222.
(24) For examples: (a) van Tamelen, E. E.; Willet, J.; Schwartz, M.; Nadeau,
R. J. Am. Chem. Soc. 1966, 88, 5937-5938. (b) van Tamelen, E. E. J.
Am. Chem. Soc. 1982, 104, 6480-6481. (c) Nishizawa, M.; Takenaka, H.;
Hayashi, Y. J. Am. Chem. Soc. 1985, 107, 522-523. (d) Nishizawa, M.;
Yadav, A.; Imagawa, H.; Sugihara, T. Tetrahedron Lett. 2003, 44, 3867-
3870. (e) Nishizawa, M.; Yadav, A.; Iwamoto, Y.; Imagawa, H. Tetrahedron
2004, 60, 9223-9234.
(25) Abe, I.; Sakano, Y.; Tanaka, H., Lou, W.; Noguchi, H.; Shibuya, M.;
Ebizuka, Y. J. Am. Chem. Soc. 2004, 126, 3426-3427.
9
13192 J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006

Cation−π Interaction in Squalene Cyclization
A R T I C L E S
polymerase.36 Preparation of EF-Tu was according to the published
protocol.37 To convert EF-Tu from the GDP form to the GTP form,
0.15 µM EF-Tu was incubated with 600 µM GTP in a buffer (50 mM
Tris-HCl at pH7.5, 50 mM MgCl2, 150 mM NH4Cl) on ice for 30 min.
The suppressor tRNAPhe (40 µM), 2 mM ATP, 0.5 mM FluoroPhe,
and PheRS (1 µM) were added to this mixture of EF-Tu and GTP, and
aminoacylation reaction was then performed at 37 °C for 30min. The
reaction mixture was finally subjected to phenol extraction, and the
charged tRNA was then precipitated by ethanol and dissolved in water.
The synthesis of SHC variants was performed at 30 °C for 7 h in a
coupled transcription/translation cell-free system with dialysis.38 The
internal reaction mixture (9 mL) consisted of 58 mM Hepes-KOH
buffer (pH 7.5), containing 1.2 mM ATP, 0.8 mM each of GTP, CTP,
and UTP, 1.7 mM DTT, 0.64 mM 3′,5′-cyclic AMP, 200 mM potassium
glutamate, 27.5 mM ammonium acetate, 10.7 mM magnesium acetate,
80 mM creatine phosphate, 250 µg/mL creatine kinase, 500 µM of
each of the 20 amino acids, 4.0% PEG8000, 25 mM phosphoenolpyru-
vate, 35 µg/mL L-(-)-5-formyl-5,6,7,8-tetrahydrofolic acid, 0.05%
sodium azide, a tRNA mixture (170 µg/mL), the pIVEX2.3d-SHC
plasmid (4 µg/mL) as the template, S30 extract (2.7 mL), T7 RNA
polymerase (66.6 µg/mL), and the charged suppressor tRNAPhe
(0.01µM). The external solution (90 mL) contained all of the
components of the internal solution except for creatine kinase, tRNA,
the template plasmid, T7 RNA polymerase, and S30 extract. The
dialysis membrane, MWCO 50,000 Spectra/Por, was from Spectrum.
Purification of the tagged SHC variants was performed as described
above.
Methods and Materials
Incorporation of OMeTyr into SHC Protein in E. coli. The gene
coding for OMeTyrRS was generated from the wild-type TyrRS gene
from M. jannaschii to contain the substitutions of Tyr32fGln32,
Glu107fThr107, Asp158fAla158, and Leu162fPro162. These substitutions
allow OMeTyrRS to recognize OMeTyr specifically.15a In addition, the
substitution of Asp286fTyr86 was introduced to enhance the recognition
of suppressor tRNA.35 The used suppressor tRNA was derived from
M. jannaschii tRNATyr, containing the base substitutions of C17A,
U17aG, U20C, G34C, G37A, and U47G. The G34C substitution makes
the tRNA to recognize amber codon, and the others prevent misacylation
of this tRNA by the endogenous aminoacyl-tRNA synthetases from E.
coli.15a The genes for OMeTyrRS and the suppressor tRNA were cloned
in the pACYC184 vector (NEB) to create plasmid pLMR-o, where the
enzyme and tRNA were expressed from the trpS and lpp promoters,
respectively.
The phenylalanine codons at positions 365 and 605 of the SHC gene
were mutagenized to be amber codon (UAG) by using the QuickChange
site-directed mutagenesis kit (Stratagene, La Jolla, CA, U.S.A.) to create
SHC (Am365) and SHC (Am605). These mutant genes were cloned
in the expression vector pET26b (Novagene), where a His6 tag was
added to the C-terminus of SHC to facilitate protein purification by
Ni2+ affinity column chromatography. Ocher codon (UAA) was used
to stop the translation of SHC. E. coli BL21(DE3) cells were
transformed with both of pET26b carrying the SHC mutant gene and
pLMR-o, and were then cultured in the LB medium containing
kanamycin of 50 µg/mL and chloramphenicol of 25 µg/mL together
with 1 mM OMeTyr. The harvested cells were subjected to ultrasoni-
cation on ice in a buffer containing 50 mM NaH2PO4, 300 mM NaCl,
and 10 mM imidazole. The obtained extracts were subjected to
centrifugation at 10000g for 20 min. The resultant supernatants were
loaded onto a Ni agarose column (Qiagen) equilibrated with a buffer
(50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole, 1% Triton
X-100). The SHC with the His tag was eluted with 250 mM imidazole
including 1% Triton X-100. The protein concentration was determined
by the Bio-Rad protein assay kit by using a standard curve for bovine
serum albumin.
SDS-PAGE and Western Blot Analysis. Proteins, separated on a
10% SDS-polyacrylamide gel, were analyzed by western blot. Mouse
anti-His-tag monoclonal antibody was used as the primary antibody,
and the goat anti-mouse IgG conjugated with HRP were used as the
secondary antibody. Detection was performed by the ECL plus
immunodetection system (Novagene) after treatment with the antibodies.
The band intensity was measured using an image analyzer, LAS-1000
plus (Fuji Photo Film, Tokyo).
MS Analyses of SHCs Substituted with Unnatural Amino Acids.
Tryptic digestions were performed using standard procedures. For
example, 2 µL (0.2 µg) of 0.1 mg/mL trypsin solution was added to a
solution of SHC protein (∼10 µg or 130-140 pmol) in a sodium citrate
buffer, and 100 mM sodium hydrogen carbonate was also added to
reach a final concentration of 20 mM. The mixture was kept at 37 °C
for overnight. When the protein was available as in gel, in-gel tryptic
digestion was applied.
Incorporation of FluoroPhes into SHC in a Cell-Free Protein
Synthesis. To enhance the expression in a cell-free system based on
an S30 extract of E. coli BL21(DE3) cells, silent mutations, suggested
by the ProteoExpert software (Roche), were introduced into the 5′-
terminal sequence of the SHC gene. The sequence of the first 36 bases
of this gene was thus changed to: ATGGCAGAACAACTTGTA-
GAAGCGCCGGCCTACGCG (the silent mutations are underlined).
The substitution of tyrosine codon with alanine codon at position 606
was performed for SHC (Am605) by using the QuickChange site-
directed mutagenesis kit (Stratagene, La Jolla, CA) to create SHC
(Am605 Ala606). For expression in the cell-free protein synthesis
system, the SHC mutant genes, SHC (Am365) and SHC (Am605
Ala606), were cloned in the pIVEX2.3d vector (Roche), where the His
tag was added at the C-terminus of SHC.
A QSTAR mass spectrometer (Applied Biosystems, U.S.A.) equipped
with an Agilent 1100 series LC system (Agilent Technologies, U.S.A.)
was used for LC/MS analyses. LC conditions were as follows: LC
column: Mightysil RP-18 GP 50-1.0 (3 µm) (Kanto Chemical Co.,
Inc., Japan); flow rate: 5 µL/min; mobile phase A: water and B:
acetonitrile with 0.1% formic acid; gradient: 5-60% B in 80 min).
Electrospray was applied for the ionization and the objective peptide
fragments were obtained from the mass chromatography.
Incubation Conditions for the Determination of Kinetic Data.
A mixed solution, which was composed of 0.5 mM squalene, 0.2%
Triton X-100, and 50 µg of the homogeneously purified enzyme in a
sodium citrate buffer (60 mM, pH 6.0), was prepared in a final volume
of 5 mL for the enzyme reactions. Incubation was carried out for 60
min at different temperatures (30, 35, 40, 45, 50, 55, 60, 65, or 70 °C)
to examine the thermal stability of the cyclases. To determine the kinetic
parameters, the incubation was conducted for 60 min at 40 or 50 °C
(Tables 1 and 2). To terminate the reaction, 15% methanolic KOH (6
mL) was added. The lipophilic enzymatic products (5 and 6) and starting
The E. coli pheS and pheT genes encoding phenylalanyl-tRNA
synthetase (PheRS) were cloned between the SmaI and BamHI sites of
the pUC119 vector (Takara). PheRS was overproduced from this
plasmid in E. coli cells, and was purified by successive chromatography
on the columns of DEAE-Toyopearl and Toyopearl phenyl-650M
(Tosoh, Tokyo) using a linear gradient of 50-250 mM KCl and that
of 800-0 mM (NH4)2SO4, respectively.
The gene coding E. coli suppressor tRNAPhe, with the sequence of
GGCCGGATAGCTCAGTCGGTAGAGCAGGGGATTCTAAAT-
CCCCGTGTCCTTGGTTCGATTCCGAGTCCGGCCACCA, was
transcribed by the method of in vitro transcription using T7 RNA
(36) Milligan, J. F.; Uhlenbeck, O. C. Methods Enzymol. 1989, 180, 51-62.
(37) Sakamoto, K.; Ishimaru, S.; Kobayashi, T.; Walker, J. R.; Yokoyama, S.
J. Bacteriol. 2004, 186, 5899-5905.
(38) Kigawa, T.; Yabuki, T.; Yoshida, Y.; Tsutsui, M.; Ito, Y.; Shibata, T.;
Yokoyama, S. FEBS Lett. 1999, 442, 15-19.
(35) Kobayashi T.; Nureki O.; Ishitani R.; Yaremchuk A.; Tukalo M.; Cusack
S.; Sakamoto K.; Yokoyama S. Nat. Struct. Biol. 2003, 10, 425-432.
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13193

A R T I C L E S
Morikubo et al.
material 4, which remained unchanged, were extracted four times with
hexane (5 mL) from the incubation mixtures, and the amount of 5 thus
produced was evaluated by a Shimadzu GC-8A instrument equipped
with a DB-1 capillary column (0.53 mm × 30 m). Kinetic parameters
Km and kcat were determined according to the Lineweaver-Burk
equation. GC/MS spectra were obtained with a JEOL SX 100 mass
spectrometer.
Circular Dichroism of SHCs. SHCs have been usually purified
and assayed in a sodium citrate buffer (pH 6.0) containing 0.2% Triton
X-100 (buffer A). The Triton X-100 detergent has strong absorption
in the UV region, because of the involvement of its phenyl residue.
Therefore, for CD measurements, we employed Brij 35 (C12E23) as an
alternative detergent, instead of Triton X-100. When dissolved in buffer
B [10 mM sodium phosphate (pH 6.0) containing 0.6% Brij 35], the
wild-type SHC was found to have equivalent activity (115%) to that
in buffer A. The SHCs solution purified by a Ni agarose column was
applied to a DEAE-Toyopearl 650 M column (Tohso), which had been
previously equilibrated with buffer B. The wild-type and mutated SHCs
were homogeneously purified through two steps of DEAE column
chromatography; the enriched fraction of the SHCs, which had been
prepared by the first chromatography with a linear gradient of a 10-
50 mM sodium citrate buffer (pH 6.0) containing 0.6% Brij 35, was
applied again to the column by eluting with a linear gradient of 0-0.3
M NaCl in buffer B to obtain the pure SHCs. The purity was checked
by SDS-PAGE. The CD spectrum of the purified enzyme (80-100
µg/mL) in buffer B was measured at a wavelength between 200 and
250 nm across a 1-mm light path at 20-70 °C by using a JASCO
J-725 instrument.
Acknowledgment. This work was made possible by the
financial support (Nos. 15380081, 16208012, and 18380001)
to T.H. provided by the Ministry of Education, Culture, Sports,
Science and Technology of Japan, and also the support of the
RIKEN Structural Genomics/Proteomics Initiative (RSGI), the
National Project on Protein Structural and Functional Analyses,
the Ministry of Education, Culture, Sports, Science and Tech-
nology of Japan.
Supporting Information Available: Amino acid alignment
of the wild-type and the mutated SHCs having unnatural amino
acids, some additional MS data after tryptic digestion, and CD
spectra measured at 20-70 °C. This material is available free
of charge via the Internet at http://pub.acs.org
JA063358P
9
13194 J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006