Abe. 2007. Enzymatic Synthesis of Cyclic Triterpenes PDF

View Article Online / Journal Homepage / Table of Contents for this issue
REVIEW www.rsc.org/npr | Natural Product Reports
Enzymatic synthesis of cyclic triterpenes

Ikuro Abe*
Received (in Cambridge, UK) 29th June 2007
First published as an Advance Article on the web 7th August 2007
Published on 07 August 2007. Downloaded by University of Wisconsin - Madison on 12/05/2017 02:57:24.
DOI: 10.1039/b616857b
Covering: 2002 to 2006
This review covers recent advances in the chemistry and enzymology of squalene cyclase and
oxidosqualene cyclase. The enzymatic cyclizations of squalene and oxidosqualene are the most
remarkable steps in the biosynthesis of sterols and triterpenes. The polyenes are converted to various
polycyclic triterpenes by different enzyme systems employing only small modification of the active-site.
Recent crystallographic and structure-based mutagenesis studies as well as utilization of chemically
synthesized active-site probes have begun to reveal intimate structural details of the enzyme-templated
cyclization reactions. 126 References are cited.
1 Introduction 3 Site directed mutagenesis

2 Structural biology 3.1 Squalene : hopene cyclase
2.1 Squalene : hopene cyclase 3.2 Oxidosqualene : lanosterol cyclase
2.2 Oxidosqualene : lanosterol cyclase 3.3 Oxidosqualene : cycloartenol cyclase
4 Substrate specificity and catalytic potential
School of Pharmaceutical Sciences, University of Shizuoka, and PRESTO, 4.1 Squalene : hopene cyclase
Japan Science and Technology Agency, 52-1 Yada, Shizuoka, 422-8526, 4.2 Squalene : tetrahymanol cyclase
Japan. E-mail: [email protected]; Fax: +81-54-264-5662; 4.3 Oxidosqualene : b-amyrin cyclase
Tel: +81-54-264-5662
4.4 Oxidosqualene : lupeol cyclase
5 Cyclases with novel catalytic functions
Ikuro Abe graduated from the University of Tokyo in 1984, and 5.1 Arabidopsis oxidosqualene cyclases
obtained his PhD degree in 1989 from the same university under 5.2 Other oxidosqualene cyclases
the direction of Professor Yutaka Ebizuka, where he studied the 6 Conclusions
chemistry and biochemistry of natural products biosynthesis. After 7 Acknowledgements
two years postdoctoral research with Professor Guy Ourisson at the 8 References
CNRS Institut de Chimie des Substances Naturelles, and mostly
with Professor Michel Rohmer at the Ecole Nationale Supérieure 1 Introduction
de Chimie de Mulhouse (1989–1991), he moved to the USA to
work with Professor Glenn D. Prestwich at the State University The remarkable cyclization of squalene and oxidosqualene to
of New York at Stony Brook (1991–1996) and the University of sterols and triterpenes has fascinated organic chemists and bio-
Utah (1996–1998) as a Research Assistant Professor. In 1998, chemists for over a half century.1–18 Bacterial squalene : hopene
he moved back to Japan to the University of Shizuoka, School cyclase (SHC) binds squalene (1) in the all-chair conformation,
of Pharmaceutical Sciences. Now he is also an investigator at and initiates the sequential ring forming reaction by protonating
PRESTO, Japan Science and Technology Agency (2005–2009). His the terminal double bond. The cyclization reaction produces the
research interests involve exploring and engineering the biosynthesis 6.6.6.6.5-fused pentacyclic hopanyl C-22 carbocation (2), which
of natural products. undergoes either proton elimination or addition of water to
produce a 5 : 1 mixture of hop-22(29)-ene (3) and hopan-22-ol
(4) (Scheme 1A).19 No methyl or hydride migration takes place
during the enzyme reaction. On the other hand, in animals and
fungi, oxidosqualene : lanosterol cyclase (OSLC) catalyzes the
formation of lanosterol (7) initiated by an oxirane ring opening
of (3S)-2,3-oxidosqualene (5) folded in the chair–boat–chair
conformation.20,21 The cyclization reaction produces the 6.6.6.5-
fused tetracyclic protosteryl C-20 cation (6) with a pseudoaxial
17b-side chain,12,22 which then undergoes transformation into the
final product via a sequence of two 1,2-hydride shifts and two
1,2-methyl migrations (H-17a→20, H-13a→17a, CH3 -14b→13b,
CH3 -8a→14a) followed by removal of H-9 to generate the D8
Ikuro Abe
double bond of 7 (Scheme 1B).12,22
This journal is © The Royal Society of Chemistry 2007 Nat. Prod. Rep., 2007, 24, 1311–1331 | 1311
Published on 07 August 2007. Downloaded by University of Wisconsin - Madison on 12/05/2017 02:57:24. View Article Online
Scheme 1 Proposed mechanism for formation of (A) hopene by SHC and (B) lanosterol by OSLC.
The enzyme-templated sequential carbon–carbon bond for-

mation reactions are thought to proceed through a series of
discrete, conformationally rigid, partially cyclized carbocationic
intermediates.10 On the basis of experimental evidence, it is now
largely accepted that, after formation of a 6.6 bicyclic A/B-ring
system, the six-membered C-ring is formed by initial ring closure
to a tricyclic Markovnikov tertiary cation with a five-membered
C-ring (8), then followed by ring expansion to a six-membered
C-ring to generate a 6.6.6-fused tricyclic anti-Markovnikov sec-
ondary cation (9).14,23 However, it still remains controversial how
the C-ring process overcomes the energy barrier required to
expand the tertiary carbocation to the thermodynamically less
stable secondary carbocation.24 Jenson and Jorgensen suggested,
based on force-field calculations on a model system, that this
barrier may be lowered by selective replacement of nucleophilic
groups from the protein backbone or side chains including
aromatic amino acid residues through cation–p interactions.25 On
the other hand, Johnson proposed a mechanism in which axial
delivery of negative point charges by the enzyme could stabilize Scheme 2 Proposed mechanism for concerted C-ring expansion/D-ring
the developing cationic centers as ion pairs, and that such charge formation by Hess.
delivery is important in enhancing the rate and efficiency of the
overall cyclization process.26,27 Recently, Hess proposed, on the
basis of ab initio calculations, that the 6.6.5 tricyclic Markovnikov sively summarized recently.16 The relationships between enzyme
tertiary cation (8) is the first intermediate in the cascade of the structure, the cyclization mechanism, and the products formed
cyclization reactions, and that conversion of 8 to 9 might instead are extremely interesting. In principle, only small modifications of
involve expansion of the C-ring in concert with the formation the active-site structure of the enzyme are sufficient to produce
of the five-membered D-ring in 11, via transition state structure dramatically different products.13 It is now accepted that the
10 (Scheme 2).24,28 In contrast, for the enzymatic formation of cyclase enzymes provide (i) a catalyst, presenting a general acid of
hopenes from squalene folded in the all-chair conformation, sufficient strength to initiate the cyclization cascade by protonating
Rajamani and Gao suggested, on the basis of computational the terminal double bond/oxirane ring, (ii) a template that
model studies, that the 6.6 bicyclic tertiary cation (13) is the chaperones the flexible substrate and intermediate carbocations
only reaction intermediate with a significant life time, and then through a series of precise conformations leading to one unique
formation of rings C, D, and E takes place via a single transition cyclization product, (iii) a shielding of carbocationic intermedi-
state to produce a 6.6.6.6.5 pentacyclic intermediate cation (14) ates from premature addition of water or elimination, and (iv)
(99%) in addition to a 6.6.6.5 tetracyclic tertiary cation by-product accelerates the sequential carbon–carbon bond forming reactions
(15) (1%) (Scheme 3).29 by stabilizing intermediate carbocations with an electron-rich
Squalene and oxidosqualene are thus converted to various environment.14 Thus, the active-site of the enzyme would have
skeletal types of sterols and triterpenes by different enzyme a molecular geometry which allows cyclization to proceed with
systems. Nearly 200 different triterpene skeletons are known from a minimum of conformational changes in the substrate once the
natural sources or enzymatic reactions, whose molecular diversity reaction is initiated.14 Stereoelectrostatic factors, such as cation–p
and the proposed mechanisms of formation have been comprehen- interactions involving the incipient cation and the electron-rich
1312 | Nat. Prod. Rep., 2007, 24, 1311–1331 This journal is © The Royal Society of Chemistry 2007
View Article Online
in 2002,15 and on enzymatic cyclization reactions of squalene and

oxidosqualene by Abe et al. in 1993.13
2 Structural biology
2.1 Squalene : hopene cyclase
As described above, the X-ray crystal structure of A. acidocaldarius

SHC was first reported at 2.9 Å resolution in 1997,34 and later
refined to 2.0 Å resolution in 1999.35 Recently, the Schulz group
reported another new X-ray crystal structure of A. acidocaldarius
SHC complexed with the inhibitor 2-azasqualene at 2.1 Å
resolution.36 It is remarkable that, in the crystal, the substrate
analogue binds with a conformation very close to that required for
the chair conformations of hopene rings A through D by forming
a salt bridge to Asp376, suggesting that the bound squalene
indeed undergoes only small conformational changes during the
formation of rings A through D (Fig. 1).
This finding led Schluz and co-workers to propose an elegant
Scheme 3 Proposed mechanism for formation of hopene by Rajamani and comprehensive illustration of the enzyme-templated cycliza-
and Gao.
tion reaction. First, as previously proposed, the enzyme reaction
is initiated by protonation of the terminal double bond by the
faces of aromatic side-chains of Phe, Tyr, or Trp, could shelter catalytic acid Asp376 which is hydrogen bonded to His451.34 The
transient cationic cyclization species, as in the case of known newly obtained crystal structure further suggested that, after the
receptor–ligand and enzyme–substrate complexes.30,31 protonation, Asp376 is most likely to be stabilized by the positively
Following the pioneering work by Poralla and co-workers on charged His451 that in turn is electrostatically stabilized by the
the purification and molecular cloning of SHC from a ther- solvent accessible Glu454.36 Moreover, the cyclization cascade
moacidophilic bacteria Alycyclobacillus acidocaldarius,32,33 Schulz up to the 6.6.6.5 tetracyclic intermediate cation is thought to
and co-workers first reported the X-ray crystal structure of A. proceed almost concomitantly without significant conformational
acidocaldarius SHC at 2.9 Å resolution in 1997,34 and later changes.36 Thus, a Markovnikov-type cation–olefin addition first
refined to 2.0 Å resolution in 1999.35 The crystal structures of produces a 6.6 bicyclic tertiary cation, which is stabilized by the
the homodimeric monotopic membrane-bound 72 kDa protein p-electrons and dipoles of aromatic residues including Phe365,
revealed dumbbell-shaped (a/a) barrel domains connected by long Tyr420, Trp489, and Tyr609. Then, a six-membered C-ring is
loops which together enclose a large central cavity of 1200 Å3 . The formed directly in concert with formation of a Markovnikov-
active-site of the enzyme is located in the central cavity, consisting type five-membered D-ring, generating a 6.6.6.5-fused tetracyclic
of an extended central hydrophobic section lined with conserved intermediate tertiary cation, which is stabilized by Phe601, Phe605,
aromatic residues, and the top and bottom sections of the cavity are and Trp169 through cation–p interactions. This proposal appears
formed by polar hydrogen networks around Asp376 and Glu45, to be in contrast to that of Rajamani and Gao mentioned above,
which are thought to participate in the initial protonation and in who suggested that the 6.6 bicyclic tertiary cation is the only
the final deprotonation reaction, respectively.14 Further, substrate reaction intermediate with a significant life time (Scheme 3).29
uptake and product release are likely to occur through a nonpolar Nonetheless, the proposed relatively long life time of the 6.6.6.5
channel connecting the active-site cavity to the membrane- tetracyclic intermediate cation seems to be consistent with the
immersed region of the enzyme and thus to the membrane experimental observation that 6.6.6.5 tetracyclic products have
interior.36 Interestingly, the prokaryotic SHC and the eukaryotic been obtained as minor products of wild-type SHC,39 and also as
OSLC share a highly-conserved unique sequence repeat, b-strand- products of various site-directed mutants of SHC.15
turn motif, rich in aromatic amino acids (the QW-sequence motif; The cyclization reaction is subsequently followed by D-ring
[K/R][G/A]XX[F/Y/W][L/I/V]XXXQXXXGXW),37 which is expansion and Markovnikov E-ring closure, yielding the tertiary
now thought to stabilize the enzyme structure by connecting hopanyl C-22 cation with the 6.6.6.6.5-fused pentacyclic ring
surface a-helices during the highly exergonic polyene cyclization system, which is likely to be stabilized by the conserved Phe605 p-
reaction. electrons.35 Finally, the enzyme reaction is terminated by either
Recent advances in crystallographic and structure-based mu- regiospecific elimination of the Z-methyl group to the major
tagenesis studies as well as utilization of synthetic active-site product hop-22(29)-ene (80%),40,41 or by addition of water to
targeted probes have begun to reveal intimate structural details of hopan-22-ol (20%) (Scheme 1A). It has been proposed that the
the catalytic mechanism of the enzyme-templated sequential ring water molecules are polarized by residues in the hydrogen-bonding
forming reactions. This review covers the most recent progress network around Glu45 at the bottom of the active-site cavity;
in these areas from the last 5 years, and continues coverage of a polarized water molecule abstracts the proton or attacks the
the literature from previous reviews on enzyme mechanisms by E-ring cation, to terminate the reaction.35 In fact, the product
Wendt et al. in 2001,14 on mutagenesis approaches by Matsuda ratio of hopene and hopanol has been reported to be significantly
and coworkers in 2003,38 on bacterial SHC by Hoshino and Sato altered in E45A and E45D mutants of A. acidocaldarius SHC.42
Fig. 1 Stereoview of the active-site of A. acidocaldarius SHC. (A) Squalene model (green) indicating the first four carbocation additions to double bonds
(green dots). The observed 2-azasqualene structure (thin dark lines) is given for reference. (B) Models of the 6.6.6.5 tetracycle (green) and the hopanyl
cation (blue). Reprinted from ref 36;Chem. Biol., 2004, 11, 121. Copyright Elsevier 2004.
Moreover, it should also be noted that most of the partially cyclized additional isoprene unit, which involves regiospecific final proton
products obtained from SHC enzyme reactions with substrate elimination from the terminal cationic intermediate (Scheme 6),
analogues are derived from final proton elimination rather than suggesting enough space is left for the putative deprotonating
cation hydration.14 Interestingly, as described later, this is also the water.43
case for the enzymatic formation of an unnatural 6.6.6.6.6.5-fused After the enzyme reaction, the cyclized products eventually have
hexacyclic polyprenoid from a C35 substrate analogue with an to pass through the nonpolar channel connecting the active-site
View Article Online
cavity to the membrane interior. In the crystal structures, however, 2.2 Oxidosqualene : lanosterol cyclase
this channel appears to be too narrow for the bulky products.35 On
the basis of the observation that the wall between the channel The long-awaited X-ray crystal structure of human OSLC was
and dimer surface is rather mobile, the channel cross-section is finally solved by Ruf and co-workers at Hoffmann-La Roche in
proposed to be appreciably enlarged by melting and displacing 2004 (Fig. 2).44,45 The crystal structure revealed two (a/a) barrel
the mobile peptide. This seems likely since the thermoacidophilic domains connected by loops and smaller b-sheet structures with
enzyme reaction is optimal at 60 ◦ C, and the cyclization reaction identical topology to that observed in A. acidocaldarius SHC.34
is highly exergonic (about 200 kJ mol−1 ).36 The active-site of the 83 kDa membrane-bound enzyme, sharing
Fig. 2 Stereoview of the active-site of human OSLC. (a) The omit electron density map showing the binding of product lanosterol in the active-site.
(b) A- and B-rings region with the active-site residues. (c) C- and D-rings region with the active-site residues. Reprinted from ref. 45; Nature, 2004, 432,
118. Copyright 2004, with permission from Nature Publishing Group.
View Article Online
25% amino acid sequence identity with A. acidocaldarius SHC, aromatic residues.45 This knowledge of the inhibitor binding mode
is located in a large central cavity lined with aromatic residues, in the enzymes may help develop more potent and efficient enzyme
and a hydrophobic plateau serves as the monotopic membrane inhibitors of both OSLC and SHC.
insertion motif.45 As in the case of SHC, the QW-sequence motif37
appears to stabilize the enzyme structure by connecting surface
a-helices during the cyclization reaction. The structure of human
OSLC co-crystallized with lanosterol at 2.1 Å resolution indicated
a hydrogen bond formed between the lanosterol 3-hydroxy group

and the catalytic Asp455 that constitutes the polar cap of the
mainly hydrophobic cavity, which is consistent with the proposed
role of Asp455 (corresponding to Asp376 in A. acidocaldarius
SHC)34 as the general acid that donates a proton to the epoxide 3 Site directed mutagenesis
oxygen of the substrate to initiate the cyclization cascade.45 The 3.1 Squalene : hopene cyclase
side chain of Asp455 also forms hydrogen bonds with Cys456 and
Cys533, which may enhance the acidity of Asp455.45 The p-electrons of the aromatic residues of Phe, Tyr, or Trp, in
The conserved aromatic side-chains of Phe444, Tyr503 and the active-sites of the cyclases have been proposed to stabilize
Trp581 are appropriately positioned to stabilize the intermedi- the partially cyclized cationic intermediates during the cyclization
ate tertiary cations at C-6 and C-10, after A-ring and B-ring cascade through cation–p interactions.30,31 However, as pointed out
formation, through cation–p interactions. On the other hand, previously, the repeated observation of partially cyclized products
Tyr98 is well positioned to enforce the energetically unfavourable arising from many of the site-directed mutants reflects the general
boat conformation of oxidosqualene required for the formation problem of separating specific cation–p interactions from simple
of lanosterol B-ring, while for the C-ring formation, His232 steric effects that may arise from displacement of the forming
and Phe696 are positioned to stabilize the anti-Markovnikov product in the active-site cavity or from subtle changes in the
secondary cation through cation–p interactions.45 The cyclization structure of the active-site.14 To solve the problems, Hoshino and
cascade is interrupted at the 6.6.6.5-fused tetracyclic protosteryl co-workers recently constructed mutants in which the active-site
C-20 cation without formation of E-ring. This would be caused Phe365 and Phe605 of A. acidocaldarius SHC are replaced with
by the fact that the aromatic residues Trp169 and Phe605 in mono-, di-, and tri-fluorophenylalanines; these unnatural amino
A. acidocaldarius SHC, which stabilizes the long-lived 6.6.6.5 acids reduce cation–p binding energies but have van der Waals
tetracyclic secondary cation required for hopene E-ring, are radii similar to that of Phe.50 Kinetic analyses revealed that the
replaced with His and Cys, respectively, in OSLC.36 In addition, enzyme activities of the fluorophenylalanine mutants are inversely
the active-site cavity of OSLC appears to be smaller than that of proportional to the number of fluoro atoms on the aromatic ring
SHC which has to accommodate the additional E-ring.45 and clearly correlates with the cation–p binding energies of the
The tetracyclic protosteryl cation subsequently undergoes the ring moiety, confirming that the p-electron density of the residues
backbone rearrangement followed by removal of H-9 to generate 365 and 605 is indeed critical for the cyclization reactions.50 In
the final product (Scheme 1B). During the rearrangement reaction, addition, the fluorophenylalanine mutants of Phe605 produced
premature termination of the reaction by the nucleophilic addition tetracyclic minor products with the Markovnikov 6.6.6.5-fused
of a water molecule is avoided by the predominantly hydrophobic ring system; (20R)-13(17)-dammarene (25) and (20R)-17-epi-
nature of the active-site cavity. Further, the highly conserved dammar-12(13)-ene (26) (Fig. 3), along with hopene and hopanol,
His232 is suggested to be a possible candidate for the base that which clearly indicated that the p-electron density of Phe605 is
abstracts the H-9 proton to terminate the enzyme reaction.45 On crucial for the anti-Markovnikov D-ring expansion and E-ring
the other hand, the hydroxyl group of Tyr503 that is hydrogen- formation.50
bonded to His232 is also proposed to be in a proper position for Earlier site-directed mutagenesis studies by Poralla and co-
the final proton abstraction.45 workers have established that the active-site residues Asp376 and
Interestingly, crystal structures of human OSLC45 and A. acido- Asp377 of the conserved DDTAVV motif of A. acidocaldarius
caldarius SHC46 complexed with the potential anticholesteremic SHC are essential for the catalytic activity.51 As described, the
drug Ro48-8071 have been also reported. The benzophenone- conserved Asp376 in SHC and the corresponding Asp455 in
containing Ro48-8071 (17) is known to be an extremely potent OSLC are the general acid that initiates the cyclization cascade
inhibitor of both the vertebrate OSLC (IC50 = 6.5 nM)47 and the by protonating the terminal double bond. Later, Dang and
bacterial SHC (IC50 = 9.0 nM).48 The crystal structure revealed Prestwich constructed a A. acidocaldarius SHC mutant in which
that Ro48-8071 is bound in the central cavity of the active-site of the DDTAVV motif is changed to DCTAEA, as in the case of the
SHC and extends into the channel that connects the cavity with the corresponding conserved catalytic region of eukaryotic OSLC,
membrane.46 The binding site is largely identical with the expected by a triple mutation of D377C/V380E/V381A.42 Interestingly,
site of squalene, however, it differs from a previously reported the triple mutant no longer accepted squalene as a substrate and
model based on a photoaffinity labeling experiment.49 On the other completely lost the hopene-forming activity, however, the mutant
hand, in the crystal structure of human OSLC, Ro48-8071 is bound still accepted both enantiomers of 2,3-oxidosqualene to produce
in a different conformation and the basic nitrogen atom forms a a mixture of 3-hydroxyhopene as in the case of wild-type SHC.11
charged hydrogen bond with the catalytic Asp 455 that directs both A similar altered substrate specificity of A. acidocaldarius SHC
oxygen atoms towards the active-site cavity, and the charged amino has also been observed for another mutant in which Gly600, the
group is also stabilized through cation–p interactions with several conserved active-site residue located in a close proximity to the
Fig. 3 Structures of cyclic triterpenes produced by mutants of A. acidocaldarius SHC.
D-ring formation site, is deleted as in the case of eukaryotic amino acid sequence identity with the yeast OSLC) are essential
OSLC.52 Interestingly, the deletion mutant converted (3S)-2,3- for the catalytic activity.65 Recently, Wu and co-workers repeated
oxidosqualene into a mixture of monocyclic achilleol A (50) the mutagenesis experiments on the His234 residue of yeast
and 6.6.5-fused tricyclic products (36, 38, 39, and 40), but OSLC, and thoroughly analyzed the enzyme reaction products,
did not accept both squalene and (3R)-2,3-oxidosqualene. Since leading to identification of novel products; a 6.6.5-fused trans–
other site-directed mutagenesis studies on A. acidocaldarius SHC syn–trans tricyclic (13aH)-isomalabarica-14(26),17E,21-trien-3b-
were extensively reviewed by Hoshino and Sato in 2002,15 this ol (47), 6.6.6.5-fused tetracyclic protosta-12,24-dien-3b-ol (48)
review just briefly summarizes the structures of the cyclization and protosta-20,24-dien-3b-ol (49), in addition to the monocyclic
products (18–46) obtained by site-directed mutagenesis of A. achilleol A (50) and parkeol (53) (Scheme 4).66,67 Interestingly,
acidocaldarius SHC, which includes point mutants of Trp169,53 almost all of the mutants exhibited altered product specificities,
Ile261,54 Gln262,55 Pro263,55 Trp312,53 Phe365,56,57 Asp377,57 and most remarkably, by H234S and H234T substitution, OSLC
Tyr420,58–60 Trp489,53 Gly600,52 Phe601,58,61 Phe605,50,62 Leu607,60 was completely functionally converted to protosta-12,24-dien-3b-
Tyr609,57,63,64 and Tyr61257 (Fig. 3). ol synthase and parkeol synthase, respectively.66 Notably, this is the
long-sought mechanistic evidence for the formation of the 6.6.5-
3.2 Oxidosqualene : lanosterol cyclase fused tricyclic Markovnikov cation and the 6.6.6.5-fused tetra-
cyclic protosteryl cation during the cyclization of oxidosqualene
In 1997, Corey and co-workers first demonstrated that the folded in the chair–boat–chair conformation. Moreover, the results
conserved active-site residues His146, His234, and Asp456 in Sac- also demonstrate the catalytic plasticity of OSLC; how subtle
charomyces cerevisiae OSLC (ERG7) (corresponding to His144, changes in the catalytic environment may have stereoelectronic
His232, and Asp455, respectively, in human OSLC that shares 36% effects on the intrinsic His234 : Tyr510 hydrogen-bonding network
Scheme 4 Cyclization of oxidosqualene by mutants of S. cerevisiae OSLC.
(corresponding to His232 : Tyr503 in human OSLC), leading to a crucial role in guiding the course of the final rearrangement
formation of the structurally diverse product profiles.66 and proton elimination, but does not influence the construction
On the other hand, site-directed mutagenesis of the active-site of the tetracyclic ring system itself.70 On the other hand, the
Tyr510 of S. cerevisiae OSLC causes perturbation in the substrate Phe445 mutants (replaced with Cys, Met, Asn, Thr, or Asp)
folding, which results in formation of monocyclic, tricyclic, and produced the tricyclic (13aH)-isomalabarica-14(26),17E,21-trien-
tetracyclic by-products.68,69 Thus, the Y510F mutant produces the 3b-ol (47), in addition to 9b-lanosta-7,24-dien-3b-ol (52), parkeol,
tricyclic (13aH)-isomalabarica-14(26),17E,21-trien-3b-ol (47) in and lanosterol, suggesting that cation–p interactions between
addition to lanosterol, while Y510H and Y510A mutants afford a carbocationic intermediate and an enzyme can be replaced
a mixture of the monocyclic achilleol A (50) and the tetracyclic by an electrostatic or polar side chain to stabilize the cationic
parkeol (53) (Scheme 4). Considering the proposed catalytic role intermediates, but with product differentiation.71
of the Tyr510 hydroxyl group in the final H-9 proton abstraction
during lanosterol formation,45 it is interesting that the Y510F 3.3 Oxidosqualene : cycloartenol cyclase
mutant still predominantly retains the lanosterol-forming activity
despite the loss of the deprotonating hydroxyl group.69 Matsuda The proposed mechanism of cyclization of (3S)-2,3-oxidosqualene
and co-workers explained that the point mutant may have a similar into cycloartenol (55), the sterol precursor in plants, is essentially
active-site geometry; the native phenolic OH is replaced by an the same as that for lanosterol; only the difference is the final 9b,19-
ordered water that could abstract the H-9 proton and transfer a cyclopropane ring formation and elimination of H-19 instead of
proton to its hydrogen-bonding partner His234.69 Alternatively, it H-9 proton (Scheme 5).13 Here, the cyclopropane ring closure pro-
would also be possible that the new hydrogen-bond network may ceeds with retention of configuration, which was earlier confirmed
locate His234 close enough to H-9 for direct deprotonation.69 by incubation of 2,3-oxidosqualene bearing a chiral methyl group
The functional role of the conserved aromatic residues Trp232 (H, D, T) at C-6 with the phytoflagelate Ochromonas malhamensis
and Phe445 in S. cerevisiae OSLC have been also investigated.70,71 cell-free system.72 Cycloartenol synthase (CAS), sharing 30–46%
First, almost all of the Trp232 mutants produced protosta-12,24- amino acid sequence identity with OSLCs, was first cloned from
dien-3b-ol (48), parkeol (53), and lanosterol, indicating that the model plant Arabidopsis thaliana (CAS1, At2g07050),73 and
Trp232, located in proximity of the lanosterol C/D-ring, also plays have been extensively investigated by the Matsuda group. Since
Scheme 5 Cyclization of oxidosqualene folded in chair–boat–chair conformation by plant OSCs.
the previous Nat. Prod. Rep. article in 2003 comprehensively base. In contrast, when Tyr410 is mutated to Thr, H477N cannot
summarizes structure-function analysis of the CAS enzyme,38 this interact with the smaller and more distant Thr side chain and,
review just briefly complements recent research progresses. therefore, cannot influence catalysis.79
Previous studies have demonstrated that the strictly conserved
active-site residues Tyr410, His477, and Ile481 in A. thaliana CAS1
synergize to promote cycloartenol formation.74–77 Thus, Ile481, 4 Substrate specificity and catalytic potential
replaced with Val in OSLC, is at the top of the active-site cavity, 4.1 Squalene : hopene cyclase
and, by interacting with the A-ring, it orients the substrate in the
binding site to avoid early reaction termination.74 On the other The bacterial SHC, which catalyzes a stereochemically and
hand, Tyr410, replaced with Thr in OSLC, is hydrogen-bonded mechanistically simpler process than eukaryotic OSCs, shows
to His257, forming part of the ceiling of the active-site cavity broad substrate tolerance and remarkable catalytic potential; the
and positioned close to the C-19 angular methyl group. Finally, enzyme accepts a wide variety of squalene analogues (C25 –C31 ) and
His477, replaced with Gln or Cyc in OSLC, is not in the active-site, efficiently performs sequential ring-forming reactions to produce
but affects the product profile through interactions with Tyr410.78 a series of unnatural cyclic polyprenoids.11,13,15 On the basis of
Most interestingly, point mutations at these positions lead to the crystal structure of the enzyme,34 the hydrophobic active-site
production of lanosterol; I481V mutant produces 25% lanosterol cavity (ca. 1200 Å3 ) lined with aromatic residues appears to have
in addition to cycloartenol and parkeol (53),74 Y410T mutant enough space to accept bulky substrate analogues. By utilizing
forms 65% lanosterol along with 9b-lanosta-7,24-dien-3b-ol (52) such properties of the enzyme, it becomes possible to generate
and parkeol,75 while H477N mutant yields 88% lanosterol and unnatural novel polycyclic polyprenoids by enzymatic conversion
12% parkeol (Scheme 5).76 On the basis of homology modeling, of chemically synthesized substrate analogues.
H477N is thought to produce lanosterol by positioning the base One of the most impressive examples is the cyclization of a
near C-9/C-8, but close enough to C-11 to form some parkeol, C35 analogue (58) into a “supra-natural” hexacyclic polyprenoid
while I481V allows some lanosterol synthesis by introducing a (59); A. acidocaldarius SHC efficiently accepts a C35 analogue in
smaller side chain, which enlarges the active-site cavity, permitting which a farnesyl C15 unit is connected in a head-to-head fashion
rotation of the intermediate cation.78 to a geranylgeranyl C20 unit, and efficiently catalyzes sequential
Recently, Matsuda and co-workers further succeeded in engi- ring-forming reactions to produce a novel hexacyclic polyprenoid
neering CAS to more accurately produce lanosterol by combi- with a 6.6.6.6.6.5-fused ring system in 10% yield (Scheme 6A).43
nation of the above mentioned point mutations.79 Thus, CAS1 Interestingly, the cyclization of the C35 analogue (C15 unit + C20
H477N/I481V double mutant was found to be the most accurate unit) is directional; it is initiated from the C15 end, and not by
example of an enzyme mutated to produce lanosterol with 99% a proton attack on the terminal double bond of the C20 unit
accuracy, while Y410T/H477N/I481V triple mutant still yielded (Scheme 6B). This suggests that a-orientation of the pro-C14
22% 9b-lanosta-7,24-dien-3b-ol. The synergetic effects of the methyl group is crucial for the substrate binding by the bacterial
H477N and I481V mutations have been explained as follows; SHC, just as in the case of the cyclization of squalene (C15 unit × 2)
the H477N/I481V double mutant relocates polarity to a position into hopene. Moreover, the analogue should be folded in
more favorable for lanosterol formation, and the decreased sterics chair–chair–chair–chair–boat–boat conformation to achieve the
allow the intermediate to rotate, moving C-9/C-8 toward the stereochemistry of the cyclization product. The cyclization
View Article Online
(61) is efficiently converted to a 2 : 1 mixture (total yield 8%)

of a tricyclic (62) and a bicyclic (63) unnatural novel product
(Scheme 7B). In this case, the cyclization reaction, initiated
in a chair–chair–chair conformation, is interrupted at the bi-
cyclic or tricyclic stage due to the stereoelectronic effect of the
bulky, electron rich, indole moiety. As a result, from a 6.6.5-
fused tricyclic Markovnikov cation, a backbone rearrangement
(H-13b→14, CH3 -8b→13b, H-9a→8a) with elimination of H-

11b yields the major product, while from a 6.6-fused bicyclic
Markovnikov cation, a rearrangement (H-9a→8a, CH3 -10b→9b,
H-5a→10a) with elimination of H-6b produces the minor product.
Similar results have been obtained with enzyme reactions with
pyrrole-containing analogues; 2-(geranylgeranyl)pyrrole and 2-
(farnesyldimethylallyl)pyrrole. When incubated with A. acido-
caldarius SHC, 2-(farnesyldimethylallyl)pyrrole was enzymati-
cally converted to a 10 : 1 mixture (total yield 1%) of a
tricyclic and a bicyclic unnatural novel polyprenoids, whereas 2-
(geranylgeranyl)pyrrole was found to be not a substrate for the
enzyme.84
Scheme 6 Cyclization of a C35 analogue by A. acidocaldarius SHC.
reaction proceeds without rearrangement of carbon and hydrogen,

and is terminated specifically by elimination of H-27 proton.
It is remarkable that even for the unnatural substrate with an
additional isoprene unit, the stereochemistry of the cyclization
reaction is tightly controlled by the enzyme.
In contrast, enzyme reactions of a C40 analogue (C20 unit +
C20 unit) and an unnatural C30 analogue (C20 unit + C10 unit)
just afforded bicyclic products, while the C25 analogue (C15 unit +
C10 unit) yielded a tetracyclic sesterterpene with a 6.6.6.5-fused
ring system.80 It is thus confirmed again that the presence of the
farnesyl C15 unit and the a-orientation of the pro-C14 methyl group
are crucial to the correct folding and binding of the substrate by
A. acidocaldarius SHC.43 Interestingly, the bacterial enzyme also
efficiently accepts the truncated analogues having carbon-chain
lengths of C15 and C20 , while geraniol (C10 ) is not a substrate for
the enzyme.81
Indole-containing substrate analogues, 3-(geranylgeranyl)-
indole (60) and 3-(farnesyldimethylallyl)indole (61), in which
a C20 isoprene unit is connected to indole, were synthesized
and tested for the enzymatic cyclization.82 Interestingly, 3-
(geranylgeranyl)indole (60) was not a substrate for A. acidocal-
darius SHC (Scheme 7A), which is in sharp contrast with the Scheme 7 Cyclization of indole-containing substrate analogues by A.
enzymatic cyclization of 3-(x-oxidogeranylgeranyl)indole (117) acidocaldarius SHC.
into a hexacyclic petromindole (119) by a plant OSC (Scheme 13)
as described later.83 The bacterial SHC is in fact particularly To investigate the effect of the presence of a methylidene group
sensitive to the b-orientation of the pro-C14 methyl group, and fails on the sequential ring forming reactions, a series of methylidene-
to bind the substrate.43 In contrast, 3-(farnesyldimethylallyl)indole extended squalenes were newly synthesized and tested for the
View Article Online
enzyme reaction. First, both 1-methylidenesqualene (64) and

25-methylidenesqualene (65) were efficiently converted to 30-
methylidene-hop-22(29)-ene (66) (35% and 38% yield, respec-
tively) by A. acidocaldarius SHC (Scheme 8A).40 In both cases, the
proton initiated cyclization first generates a pentacyclic hopenyl
C-22 cation, and the subsequent proton elimination from the
terminal methyl group produces the same conjugated diene system.
Notably, a significant life time of the intermediate allylic cation,

stabilized by the methylidene residue, enables a rotation of the
isobutenyl group around the C-21–C-22 bond prior to the final
proton abstraction. In contrast, in the biosynthesis of hop-22(29)-
ene, the final proton elimination takes place regiospecifically from
the Z-methyl (Me-30) group, which we have confirmed by cycliza-
tion of (1,1,1,24,24,24-2 H6 )squalene (71) into (23,23,23,30,30,30-
2
H6 )hop-22(29)-ene (72) (Scheme 8D).40 On the other hand, 26-
methylidenesqualene (67) is converted to a novel unusual C31
17-epi-dammarene derivative (68) with a 6.6.6.5+6 ring system
(10% yield) (Scheme 8B);85 the same C31 6.6.6.5+6 skeleton
has been previously obtained from enzymatic conversion of 29-
methylidene-2,3-oxidosqualene (29-MOS) by A. acidocaldarius
SHC.86 It is also known that 29-MOS functions as a time-
dependent irreversible inhibitor of the bacterial SHC as in the
case of vertebrate OSLC.86–88 Finally, 27-methylidenesqualene (69)
is cyclized into 26-methylidene-hop-22(29)-ene (70) (3% yield)
(Scheme 8C).85 Interestingly, the methylidene-extended squalenes
were found to be poor inhibitors of A. acidocaldarius SHC (IC50 =
ca. 100 lM), except 27-methylidenesqualene (IC50 = 5 lM).85
In a previous review in 2002,15 Hoshino and Sato summarized
their enzymatic conversion experiments of 6-, 10-, 15-, 19, and 23-
desmethylsqualenes by recombinant A. acidocaldarius SHC, which
provided important insights into the substrate recognition and the
role of the methyl group on the cyclization reaction.41,89–91 Recently,
the same group further investigated enzymatic cyclizations of
squalene analogs with threo- and erythro-diols at the 6,7- and
10,11-positions (73–80) by A. acidocaldarius SHC (Scheme 9).92
Remarkably, the enzyme reactions afforded ether-ring containing
polycyclic products derived from 6.6.5-fused tricyclic, 6.6-fused
bicyclic, and monocyclic Markovnikov tertiary cations. The ob-
served stereochemical course of the cyclization reactions suggests
that the folding conformation of squalenediols is tightly con-
stricted by the enzyme, and the substrate and product specificities
are dominantly directed by the least motion of the nucleophilic
hydroxyl group toward the intermediate carbocations.92
4.2 Squalene : tetrahymanol cyclase
A squalene cyclase from the ciliate protozoan Tetrahymena pyri-

formis produces a pentacyclic, gammacerane triterpene, tetrahy-
manol (95). The pioneering work by Caspi and co-workers
has established that the cyclization is initiated by protonation
of a terminal double bond of squalene folded in the all-chair
conformation, and followed by addition of water at the result-
ing gammaceranyl C-21 cationic center without carbon skeletal
rearrangement (Scheme 10A).93
It has been previously reported that T. pyriformis cyclase accepts
2,3-dihydrosqualene (96), an analogue lacking one of the terminal
Scheme 8 Cyclization of methylidene-extended substrate analogues by
double bonds of squalene, and efficiently catalyzes cyclization A. acidocaldarius SHC.
into a 6.6.6.5-fused tetracyclic, euph-7-ene (97), with the 20R
configuration (Scheme 10B).94,95 In addition, two new bicyclic
Scheme 9 Cyclization of squalenediols by A. acidocaldarius SHC.
compounds (98 and 99) were later identified as by-products.96 the exact mechanism and the stereochemistry of the cyclization
This unexpected result requires the formation of a 6.6.6.5 reaction still remains to be elucidated.
tetracyclic intermediate cation having either a 17a- or 17-b
side chain (cyclization reaction folded in all-chair conformation
4.3 Oxidosqualene : b-amyrin cyclase
produces a dammaranyl cation (100a) with a 17a-side chain,
while in chair–chair–chair–boat conformation yields a 17-epi- Since the historical proposal of the biogenetic isoprene rule
dammaranyl cation (100b) with a 17-b side chain; as described by Eschenmoser, Ruzicka, Arigoni, and Jeger in 1955,2,3 the
later, similar mechanisms are discussed for the biosynthesis of remarkable cyclization of (3S)-2,3-oxidosqualene into b-amyrin
tirucalla-7,24-dien-3b-ol, see Scheme 17), which in the original (105) has fascinated organic chemists for over a half century. The
work was considered to subsequently undergo transformation formation of b-amyrin is initiated in the chair–chair–chair–boat
into the final product via a sequence of two 1,2-hydride shifts conformation of (3S)-2,3-oxidosqualene, and the proton-initiated
and two 1,2-methyl migrations (H-17→20, H-13b→17b, CH3 - cyclization first produces the 6.6.6.5-fused tetracyclic dammarenyl
14a→13a, CH3 -8b→14b), followed by removal of the axial C-20 cation (101) (Scheme 11A). After D-ring expansion, an
H-7a to generate the D7 double bond of 97.94,95 Arigoni and electrophilic addition of the tetracyclic baccharenyl secondary
co-workers recently repeated the experiments using [7-2 H]- cation (102) on to the terminal double bond generates the lupanyl
and [11-2 H]-labeled isotopomers of 96, and confirmed that the tertiary cation (103) with a five-membered E-ring, which is
formation of 97 involves two consecutive 1,2-hydride shifts, followed by ring expansion to yield the oleanyl secondary cation
instead of a putative 1,3-hydride shift from C-13 to C-20, to (104). Finally, the oleanyl C-19 cation undergoes sequential 1,2-
achieve the 20R configuration of 97, just as in the case of the hydride shifts (H-18a→19, H-13b→18b) followed by elimination
formation of lanosterol from the protosteryl cation.96 However, of H-12a to yield the 6.6.6.6.6-fused pentacyclic ring system with
View Article Online
29,30-bisnorgermanicol (108), and 29,30-bisnor-d-amyrin (109)

(Scheme 11A).102 Moreover, the enzyme reactions with [23-13 C]-
and [23,23-2 H]-labeled isotopomers of 106 indicated that the
cyclization reaction does not proceed through formation of a
lupanyl primary cation with a five-membered E-ring (route A),
but instead, an electrophilic addition of the tetracyclic C-18
cation on to the terminal double bond directly generated a
thermodynamically favored oleanyl secondary cation (106) (route

B).102 The E-ring formation of the bisnor products is thus
apparently under thermodynamic control, and the formation of
three regioisomers suggests lack of control on the stereochemistry
of the final rearrangement reactions. Presumably, the absence
of the terminal methyl groups causes structural perturbation in
the folding conformation of the E-ring of the intermediate C-20
cation, leading to the interruption of the sequential rearrangement
reactions.
22,23-Dihydro-2,3-oxidosqualene (110) is an analogue lacking
the terminal double bond of 2,3-oxidosqualene, therefore it is
no longer possible to form the pentacyclic ring system. Enzyme
reaction with P. sativum bAS afforded a 4 : 1 mixture (total
5% yield) of 6.6.6.5-fused tetracyclic products; euph-7-en-3b-ol
(112) and bacchar-12-en-3b-ol (114) (Scheme 11B).103 Thus, in
the absence of the terminal double bond, the enzyme reaction
was interrupted at the tetracyclic dammaranyl C-20 cation (111)
with the 17b-side chain. Subsequently, a backbone rearrangement
(H-17a→20, H-13b→17b, CH3 -14a→13a, CH3 -8b→14b) with
elimination of H-7a yields euph-7-en-3b-ol (112), as in the case
of above described enzymatic formation of euph-7-ene by T.
pyriformis cyclase,94,95 while D-ring expansion to the baccharenyl
C-18 cation (113) with the 17R configuration, and subsequent
hydride shift (H-13b→17b) with loss of H-12a as in the case of
b-amyrin synthesis, produces bacchar-12-en-3b-ol (114). This is
the first demonstration of the enzymatic formation of baccharene
skeleton with the 6-membered D-ring. Most remarkably, the D-
Scheme 10 Cyclization into (A) tetrahymanol and (B) euph-7-ene by T.
pyriformis cyclase. ring expansion takes place even in the absence of the terminal dou-
ble bond. Thus, the enzymatic formation of the anti-Markovnikov
the D12 double bond.13,97 b-Amyrin synthases (bAS) from several six-membered D-ring does not depend on the participation of
plants including Pisum sativum have been cloned and functionally the terminal p-electrons. Instead, properly positioned active-site
expressed in Saccharomyces cerevisiae.98,99 bAS shows only ca. 20% aromatic residues may play a crucial role for the ring expansion
overall amino acid sequence identity to bacterial A. acidocaldarius reaction. Further, it should be noted that the stereochemistry of
SHC. Ebizuka and co-workers have demonstrated, on the basis of the cyclization reaction is strictly controlled by the enzyme. The
site-directed mutagenesis experiments of bAS from Panax ginseng, formation of the C-20R configuration of 112 from the dammarenyl
that the active-site aromatic residues Tyr261 and Trp259 play C-20 cation is likely to involve the pathway with least motion; i.e.
a critical role for the D/E-ring formation of b-amyrin; Tyr259 only 60◦ rotation around the C-17–C-20 bond prior to the proton
stabilizes one of the cationic intermediates generated after the migration from C-17 to C-20. Finally, P. sativum bAS afforded a
dammarenyl cation, while Trp259 controls b-amyrin formation tetracyclic sesterterpene alcohol with the same 6.6.6.5-fused ring
presumably through stabilization of oleanyl cation.100,101 system as 112 from a C25 analogue in which a farnesyl C15 unit is
Recent advances in enzymolgy and molecular biology of bAS connected in a head-to-head fashion to a geranyl C10 unit.104
offer us the opportunity to reinvestigate the still cryptic questions
regarding the mechanism and stereochemistry of the complex 4.4 Oxidosqualene : lupeol cyclase
ring-forming reaction, which includes the enzymatic conversion
of 24,30-bisnor-2,3-oxidosqualene (106).102 As earlier described The proposed mechanism of the biosynthesis of lupeol (115)
by Corey and Gross in 1968, the removal of the terminal methyl and b-amyrin from (3S)-2,3-oxidosqualene is identical up to the
groups could significantly affect the stereoelectronic course of the formation of the 6.6.6.6.5-fused pentacyclic lupanyl intermediate
D/E-ring formation reaction, which is the most complex and cation (103).13,105 Subsequent proton elimination from one of the
interesting process of the b-amyrin biosynthesis.126 Our enzyme terminal methyl groups produces lupeol, while E-ring expansion
reaction of 106 with recombinant P. sativum bAS afforded a to oleanyl cation followed by 1,2-hydride shifts lead to formation
3 : 1 : 0.2 mixture (total 27% yield) of three regioisomers of of b-amyrin. Remarkably, Ebizuka and co-workers have previ-
6.6.6.6.6-fused pentacyclic products; 29,30-bisnor-b-amyrin (107), ously demonstrated that lupeol synthase from Olea europaea is
Scheme 11 Cyclization of (A) 24,30-bisnoroxidosqualene and (B) 22,23-dihydrooxidosqualene by P. sativum bAS.
functionally converted into b-amyrin synthase by replacement and lupan-3b,20-diol (116) (Scheme 15).106–108 The Ebizuka group
of a single amino acid residue L256W.101 On the other hand, has established that, during the biosynthesis of lupan-3b,20-diol,
recombinant lupeol synthase from Arabidopsis thaliana (LUP1, the water addition to the lupanyl C-20 cation by LUP1 takes place
At1g78970) has been shown to produce a mixture of lupeol, stereospecifically,108 while nonstereospecific deprotonation from
b-amyrin, germanicole, taraxasterol (139), w-taraxasterol (140), the terminal methyl groups produces lupeol (Scheme 12).100
Scheme 13 Cyclization of 3-(x-oxidogeranylgeranyl)indole by A. thaliana

LUP1.
including OSLC that efficiently catalyze cyclization into (24S)-

24,25-epoxylanosterol (124).109,110 These cyclases all exclude the
Scheme 12 Proposed mechanism for formation of lupeol and lupandiol
terminal epoxide from cyclization and convert DOS to triterpene
by A. thaliana LUP1.
epoxides rather than heterocycles. In contrast, A. thaliana LUP1
has been shown to accept DOS and catalyze cyclization reactions
Matsuda and co-workers have demonstrated that A. thaliana to produce a 3 : 4 : 2 mixtures (total 56% yield) of oxacyclic
LUP1 efficiently converted 3-(x-oxidogeranylgeranyl)indole (117) triterpene diols; 127 with an unusual 17,24-epoxybaccharane
(racemic) into a 1 : 5 : 1 mixture of a hexacyclic indole diterpene, skeleton, and (20R,24S)- and (20S,24S)-20,24-epoxydammaranes,
petromindole (119), along with monocyclized products 120 and 128 and 129 (Scheme 14).111 As described in the biosynthesis
121 (Scheme 13).83 The product 119 is a single enantiomer having of lupeol, after D-ring expansion from the dammarenyl C-20
the 3S configuration as in the case of the authentic sample cation, an electrophilic addition of the baccharenyl C-18 cation
from the soil fungus Petromyces muricatus, which excludes the on to the terminal double bond generates the pentacyclic lupanyl
possibility of the formation by nonenzymatic cyclization. It is tertiary cation. It has been reasoned that, in the analogous way,
remarkable that the enzyme accepts such a bulky substrate and the distal epoxide in 126 could attack the cationic center to
catalyzes the sequential ring forming reaction to generate the generate a 6.6.6.6.6-fused oxacyclic ring system.111 Notably, a
indole containing 6.6.6.6.6.5-fused heptacyclic ring system. This is ginseng and other medicinal plants accumulate biologically active
in sharp contrast with the above described A. acidocaldarius SHC epoxydammarene and dammarenediol saponins generally having
that does not accept 3-(geranylgeranyl)indole (60) as a substrate a 20S configuration.111 The results suggest new roles for DOS
(Scheme 7A).82 The cyclization of 117 has been proposed to pro- in triterpene biosynthesis and illuminate substrate folding during
ceed in chair–chair–chair conformation to produce a 6.6.6 tricyclic cyclization.
tertiary Markovnikov cationic intermediate 118. Petromindole is
apparently formed by cyclization of cation 118 to indole at C-
4 . This regioselectivity corresponds to anticipated constraints of 5 Cyclases with novel catalytic functions
the LUP1 active-site cavity, as judged by the shape of lupeol. In 5.1 Arabidopsis oxidosqualene cyclases
contrast, enzymatic formation of indole diterpene 122 arising from
C-2 annulation was not detected.83 Although no indole diterpene So far more than 30 OSCs have been cloned and sequenced,
synthases have yet been characterized, the result suggests that including the above described enzymes that produce lanosterol,
petromindole formation may be catalyzed by a fungal enzyme cycloartenol, b-amyrin, and lupeol (Schemes 5 and 15).16,112
closely related to plant triterpene synthases.83 The completion of the genome sequencing project revealed the
(3S,22S)-2,3:22,23-Dioxidosqualene (DOS) (123) is pro- presence of 13 OSC homologues in the model plant A. thaliana,
duced in limited amount by further epoxidation of (3S)-2,3- and 10 out of 13 OSCs have been identified for their catalytic
oxidosqualene, and is known to be substrate of several OSCs, function, demonstrating that A. thaliana has the ability to produce
View Article Online
pressed in a lanosterol synthase-deficient yeast strain, the yeast

transformant cell accumulated C-ring-seco-b-amyrin (142) and C-
ring-seco-a-amyrin (147) (Scheme 16B), in addition to lupeol,
a-amyrin (145), and bauerenol (146) (Scheme 15). The seco-
triterpenes are thought to be generated from a 6.6.6.6.6-fused
pentacyclic C-13 tertiary cation 137 and 144, respectively, through
bond cleavage between C-8 and C-14.115 The mechanism of the
formation has been explained as follows; if the C-13 cation 137

had a significant lifetime and the D/E-rings were forced to bend
slightly upward as a possible consequence of the steric hindrance
exerted by the enzyme protein, the empty orbital at C-13 would
no longer be anti to H-12a or 26-Me. Instead, it would become
nearly anti to the C-8–C-14 bond, and under this stereoelectronic
condition the bond cleavage would take place, which is followed by
removal of the axial H-7a, leading to formation of the D7 , D13 diene
system of 142 (Scheme 16B).115 A similar sequence of reactions
from the cation 144 could provide 147. It is remarkable that OSC
enzymes not only catalyze sequential bond forming reactions but
also mediate fragmentation of the once formed ring systems of
cationic intermediates by carbon–carbon bond cleavage. On the
basis of the experimental results, direct production of other natural
seco-triterpenes including camelliol A, sasanquol, and helianol by
OSCs has been proposed.115
Further genome mining uncovered two more novel OSCs
that catalyze formation of tricyclic triterpenes with a 6.6.5-
fused ring system, which have not been isolated from any plants
including A. thaliana. Thus, genes At5g48010 and At4g15340
encode thalianol synthase116 and arabidiol synthase,117,118
which produce 6.6.5-fused trans–anti–trans tricyclic triterpenes;
(3S,13S,14R)-malabarica-8,17E,21-trien-3b-ol (thalianol) (131)
Scheme 14 Cyclization of 2,3:22,23-dioxidosqualene by A. thaliana and (3S,13R,14R)-malabarica-17E,21-dien-3b,14-diol (arabidiol)
LUP1. (132), respectively (Scheme 15). The proposed mechanism of the
cyclization involves initial formation of a 6.6.5-fused tricyclic
more than 15 triterpenes, although triterpene constituents of the malabaricanyl C-14 cation (130) from 2,3-oxidosqualene folded
plant have not been well identified yet.112 in the all-chair conformation. Subsequent hydride and methyl
Matsuda and co-workers have identified that the Arabidopsis 1,2-shift (H-13a→14, CH3 -8b→13b) with elimination of H-9a
gene At5g42600 encodes a novel OSC (marneral synthase) that produces thalianol with the D8 double bond, while stereospecific
catalyzes formation of an iridal (marneral) (56), the precursor addition of a water molecule to the C-14 cationic center yields
of iridals, unusual triterpenoids known only in the distant Iris arabidiol with the 14R configuration118 (Scheme 15). In thalianol
family (Scheme 16A).113 The proposed mechanism of the enzyme synthase, Tyr410 and Ile481, the conserved cycloartenol synthase
reaction involves initial cyclization of (3S)-2,3-oxidosqualene into active-site residues guiding catalytic events in the B-ring,74,75 are
a 6.6 bicyclic C-8 tertiary cation (148) in chair–boat conformation, uniquely replaced with Phe and Val, respectively, as in the case
which subsequently undergoes 1,2-shifts to a C-5 tertiary cation of pentacyclic triterpene synthases. This similarity may reflect the
(149). Ring A is then cleaved to form the 3,4-seco aldehyde 56 by fact that common catalytic forces are used to promote formation
a Grob fragmentation. The 8a-methyl configuration of marneral of a chair conformation in the B-ring,116 which is in contrast
supports the original hypothesis of Marner that cyclization of oxi- with the above described formation of the trans–syn–trans (13aH)-
dosqualene into iridals proceeds via a B-ring boat conformation.114 isomalabarica-14(26),17E,21-trien-3b-ol (47) folded in the chair–
Sequence analysis revealed that, unlike other OSCs, marneral boat–chair conformation (Scheme 4).66,67
synthase contains Gly in place of the active-site Cys456 which The gene At1g66960 encodes another novel OSC that produces
is hydrogen bonded to the catalytic protonating Asp455, which tirucalla-7,24-dien-3b-ol (135), the 20S epimer of eupha-7,24-
may facilitate the deprotonation of the axial hydroxy group in the dien-3b-ol (butyrospermol) (134), with additional uncharacterized
C-5 cationic intermediate by increasing the mobility and basicity products.119 As in the case of the above discussed enzymatic
of the Asp residue (Scheme 16A).113 In addition, the enzyme also formation of euph-7-ene by tetrahymanol synthase,94,95 there are
differs from other OSCs by a deletion between positions 730 and two possible pathways for the biosynthesis of tirucalla-7,24-
731; residues in this region are thought to be involved in substrate dien-3b-ol. Thus, cyclization of (3S)-2,3-oxidosqualene in all
folding to form rings C and D).113 chair conformation produces 17-epi-dammarenyl cation (150),
Very recently, Ebizuka and co-workers have identified that while in chair–chair–chair–boat conformation yields dammarenyl
another gene At1g78500 encodes a novel OSC that catalyzes cation (101) with the 17b-side chain (Scheme 17).13 Subsequently,
formation of other seco-triterpenes.115 When the gene was ex- from 150, a backbone rearrangement (H-17b→20, H-13b→17b,
Scheme 15 Cyclization of oxidosqualene folded in chair–chair–chair conformation by plant OSCs.
Scheme 17 Proposed mechanism for formation of tirucalla-7,24-

dien-3b-ol by Arabidopsis OSC.
Finally, the gene At3g45130 was found to encode an accurate

lanosterol synthase (OSLC), which is the first example of OSLC
cloned from a plant.120,121 Phylogenetic reconstructions reveal that
OSLCs are broadly distributed in eudicots but evolved indepen-
dently from those in animals and fungi. Notably, A. thaliana OSLC
has the above described H477N/I481V mutations79 relative to
cycloartenol synthases.120 The coexistence of cycloartenol synthase
and lanosterol synthase suggests specific roles for both cyclopropyl
Scheme 16 Proposed mechanism for formation of seco-triterpenes by and conventional sterols in plants. Physiological functions of
Arabidopsis OSCs. lanosterol and lanosterol metabolites in plants are not yet clear,
but expression profiles of A. thaliana OSLC suggest that lanosterol
CH3 -14a→13a, CH3 -8b→14b) with elimination of H-7a produces metabolites may participate in defense responses.120
135 (Scheme 17A). Alternatively, the C-20S configuration of 135
may be also derived from dammarenyl C-20 cation with the 17b- 5.2 Other oxidosqualene cyclases
side chain; in this case 120◦ rotation instead of 60◦ rotation, around
the C-17–C-20 bond is required prior to the proton migration Cucurbitacins, a group of highly oxygenated tetracyclic triter-
from C-17 to C-20 (Scheme 17B). The exact mechanism and penes, occur in the restricted plant families, such as Cucurbitaceae,
the stereochemistry of the cyclization reaction still remains to be as bitter tasting and purgative constituents. A novel OSC, catalyz-
elucidated. ing the first committed step of cucurbitacin biosynthesis, has been
View Article Online
cloned and sequenced from seedlings of Cucurbita pepo.122 When example, in lanosterol formation, it will be of interest to know
expressed in a OSLC-deficient yeast strain, the transformant cell why the back-bone rearrangement takes place after the initial
accumulated 10a-cucurbita-5,24-diene-3b-ol (57). The postulated cyclization to the protosteryl cation. The relationship between
mechanism of cyclization of (3S)2,3-oxidosqualene proceeds cyclization mechanism and active-site structure of the various SC
through initial formation of protosteryl C-17 cation followed by and OSC enzymes, including the newly discovered seco-tritepene
a series of hydride and methyl shifts to generate the C-9 cation synthases, presents a formidable challenge. Finally, further analy-
as in the case of cyclization into lanosterol and cycloartenol. The ses of the catalytic potential and plasticity of functionally divergent
cationic intermediate then undergoes transformation into the final cyclase enzymes promise to reveal intimate structural details of
product via a further methyl and hydride 1,2-shift (CH3 -10b→9b, the enzyme catalyzed processes, and suggest strategies for the
H-5a→10a) followed by elimination of H-6b to form the D5(6) dou- structure-based rational engineering of the enzyme proteins and
ble bond (Scheme 5). Cucurbitadienol synthase is closely related manipulating substrate and product specificities of the polyene
(65–70% identity at the amino acid level) to both cycloaretnol and cyclization reactions.
lanosterol synthases. Site-directed mutagenesis L488P and I489V
did not change the product specificity, suggesting that the role
of the active-site residue Ile489 in CAS is different from that of
7 Acknowledgements
cucurbitadienol synthase.122 The author expresses his appreciation to an excellent group of
Panax ginseng is one of the most famous oriental medicinal co-workers at Shizuoka whose contributions are cited in the
plants used as crude drugs in Asian countries. Dammarenediol- text, in particular Dr Hideya Tanaka, Dr Yuichi Sakano, and
II synthase, the first committed enzyme for the biosynthesis of Professor Hiroshi Noguchi. Financial support at Shizuoka has
sapogenin of dammarane type ginsenosides, has been successfully been provided by the PRESTO program from the Japan Science
cloned from hairy root cultures of P. ginseng by homology and Technology Agency, and Grant-in-Aid for Scientific Research
based PCR method.123 When expressed in the yeast system, the from the Ministry of Education, Culture, Sports, Science and
transformant cell accumulated (20S)-dammar-24-ene-3b,20-diol Technology, Japan. This contribution is dedicated to the memory
(133) as the sole cyclization product of (3S)-2,3-oxidosqualene, of Professor Guy Ourisson, a mentor who introduced the author
demonstrating that water addition to dammrenyl C-20 cation is to the mystique of the molecular evolution of the cyclase enzymes.
stereospecific (Scheme 15).
Finally, recently disclosed genomic information on the fungus
Aspergillus fumigatus led to identification of a gene cluster involved 8 References
in the biosynthesis of an antibiotic, helvolic acid (152), with 1 R. B. Woodward and K. Bloch, J. Am. Chem. Soc., 1953, 75, 2023.
the unique protosterol skeleton as in the case of an antibiotic 2 A. Eschenmoser, L. Ruzicka, O. Jeger and D. Arigoni, Helv. Chim.
fusidic acid.13,124 The Ebizuka group successfully demonstrated Acta, 1955, 38, 1890.
3 A. Eschenmoser and D. Arigoni, Helv. Chim. Acta, 2005, 88, 3011.
that an OSC homologue cDNA heterologously expressed in the 4 G. Stork and A. W. Burgstahler, J. Am. Chem. Soc., 1955, 77, 5068.
yeast converted 2,3-oxidosqualene into protosta-17(20)Z,24-dien- 5 J. W. Cornforth, R. H. Cornforth, C. Donninger, G. Popják, Y.
3b-ol (151).125 In this case, cyclization of (3S)-2,3-oxidosqualene Shimizu, S. Ichii, E. Forchielli and E. Caspi, J. Am. Chem. Soc., 1965,
proceeds without the backbone rearrangement, the initially 87, 3224.
6 E. J. Corey, W. E. Russey and P. R. Ortiz de Montellano, J. Am. Chem.
formed tetracyclic protosteryl C-20 cation is stabilized by proton Soc., 1966, 88, 4750.
elimination of H-17 to form a double bond between C-17 and 7 E. E. van Tamelen, J. D. Willett, R. B. Clayton and K. E. Lord, J. Am.
C-20. Chem. Soc., 1966, 88, 4752.
8 J. D. Willett, K. B. Sharpless, K. E. Lord, E. E. van Tamelen and R. B.
Clayton, J. Biol. Chem., 1967, 242, 4182.
9 D. H. R. Barton, T. R. Jarman, K. C. Watson, D. A. Widdowson,
R. B. Boar and K. Damps, J. Chem. Soc., Perkin Trans. 1, 1975, 1134.
10 E. E. van Tamelen, J. Am. Chem. Soc., 1982, 104, 6480.
11 G. Ourisson, M. Rohmer and K. Poralla, Annu. Rev. Microbiol., 1987,
41, 301.
12 E. J. Corey and S. C. Virgil, J. Am. Chem. Soc., 1991, 113, 4025.
13 I. Abe, M. Rohmer and G. D. Prestwich, Chem. Rev., 1993, 93, 2189.
14 K. U. Wendt, G. E. Schulz, E. J. Corey and D. R. Liu, Angew. Chem.,
Int. Ed., 2000, 39, 2812.
15 T. Hoshino and T. Sato, Chem. Commun., 2002, 291.
6 Conclusions 16 R. Xu, G. C. Fazio and S. P. Matsuda, Phytochemistry, 2004, 65, 261.
17 R. A. Yoder and J. N. Johnston, Chem. Rev., 2005, 105, 4730.
18 K. U. Wendt, Angew. Chem., Int. Ed., 2005, 44, 3966.
In the last 5 years, there have been significant advances in under- 19 K. Poralla, in Cycloartenol and other Triterpene cyclases, ed.
standing the structures and mechanisms of SC and OSC enzymes. D. H. R. Barton and K. Nakanishi, Elsevier, Oxford, UK, 1999.
However, many questions still remain unanswered, particularly in 20 I. Abe, J. C. Tomesch, S. Wattanasin and G. D. Prestwich, Nat. Prod.
the areas of structure and mechanism. First of all, elucidation of Rep., 1994, 11, 279.
21 I. Abe and G. D. Prestwich, in Squalene epoxidase and oxidosqua-
the three-dimensional structures of the enzymes are prerequisite lene:lanosterol cyclase. Key enzymes in cholesterol biosynthesis, ed.
for further understanding of the molecular interactions involved D. H. R. Barton and K. Nakanishi, Elsevier, Oxford, UK, 1999.
in substrate binding and catalysis. So far crystal structures of 22 E. J. Corey, S. C. Virgil and S. Sarshar, J. Am. Chem. Soc., 1991, 113,
8171.
only two enzymes; bacterial A. acidocaldarius SHC and human
23 E. J. Corey, S. C. Virgil, H. Cheng, C. H. Baker, S. P. T. Matsuda, V.
OSLC, have been solved. Further, there are still many fundamental Singh and S. Sarshar, J. Am. Chem. Soc., 1995, 117, 11819.
questions of enzyme mechanism that remain to be elucidated. For 24 B. A. Hess, Jr., J. Am. Chem. Soc., 2002, 124, 10286.
View Article Online
25 C. Jenson and W. L. Jorgensen, J. Am. Chem. Soc., 1997, 119, 10846. 70 T. K. Wu, M. T. Yu, Y. T. Liu, C. H. Chang, H. J. Wang and E. W.
26 W. S. Johnson, S. D. Lindell and J. Steele, J. Am. Chem. Soc., 1987, Diau, Org. Lett., 2006, 8, 1319.
109, 5852. 71 T. K. Wu, Y. T. Liu, F. H. Chiu and C. H. Chang, Org. Lett., 2006, 8,
27 W. S. Johnson, S. J. Telfer, S. Cheng and U. Schubert, J. Am. Chem. 4691.
Soc., 1987, 109, 2517. 72 L. J. Altman, C. Y. Han, A. Bertorino, G. Handy, D. Laugani, W.
28 B. A. Hess, Jr., Org. Lett., 2003, 5, 165. Muller, S. Schwartz, D. Shanker, W. H. de Wolf and F. J. Yang, J. Am.
29 R. Rajamani and J. Gao, J. Am. Chem. Soc., 2003, 125, 12768. Chem. Soc., 1978, 100, 3235.
30 D. A. Dougherty, Science, 1996, 271, 163. 73 E. J. Corey, S. P. Matsuda and B. Bartel, Proc. Natl. Acad. Sci. U. S. A.,
31 J. C. Ma and D. A. Dougherty, Chem. Rev., 1997, 97, 1303. 1993, 90, 11628.
32 D. Ochs, C. Kaletta, K.-D. Entian, A. Beck-Sickinger and K. Poralla, 74 E. A. Hart, L. Hua, L. B. Darr, W. K. Wilson, J. Pang and S. P. T.
J. Bacteriol., 1992, 174, 298. Matsuda, J. Am. Chem. Soc., 1999, 121, 9887.
33 D. Ochs, C. H. Tappe, P. Gärtner, R. Kellner and K. Poralla, 75 J. B. Herrera, W. K. Wilson and S. P. T. Matsuda, J. Am. Chem. Soc.,
Eur. J. Biochem., 1990, 194, 75. 2000, 122, 6765.
34 K. U. Wendt, K. Poralla and G. E. Schulz, Science, 1997, 277, 1811. 76 M. J. Segura, M. M. Meyer and S. P. Matsuda, Org. Lett., 2002, 8,
35 K. U. Wendt, A. Lenhart and G. E. Schulz, J. Mol. Biol., 1999, 286, 1395.
175. 77 T. K. Wu and J. H. Griffin, Biochemistry, 2002, 41, 8238.
36 D. J. Reinert, G. Balliano and G. E. Schulz, Chem. Biol., 2004, 11, 78 S. Lodeiro, M. J. Segura, M. Stahl, T. Schulz-Gasch and S. P. Matsuda,
121. ChemBioChem, 2004, 5, 1581.
37 K. Poralla, A. Hewelt, G. D. Prestwich, I. Abe, I. Reipen and G. 79 S. Lodeiro, T. Schulz-Gasch and S. P. Matsuda, J. Am. Chem. Soc.,
Sprenger, Trends Biochem. Sci., 1994, 19, 157. 2005, 127, 14132.
38 M. J. R. Segura, B. E. Jackson and S. P. T. Matsuda, Nat. Prod. Rep., 80 I. Abe, H. Tanaka, Y. Takahashi, W. Lou and H. Noguchi, Proceedings
2003, 20, 304. of 45th Symposium on the Chemistry of Natural Products, Kyoto, 2003.
39 C. Pale-Grosdemange, C. Feil, M. Rohmer and K. Poralla, Angew. 81 T. Hoshino, Y. Kumai, I. Kudo, S. Nakano and S. Ohashi, Org. Biomol.
Chem., Int. Ed., 1998, 37, 2237. Chem., 2004, 2, 2650.
40 H. Tanaka, H. Noguchi and I. Abe, Org. Lett., 2004, 6, 803. 82 H. Tanaka, H. Noguchi and I. Abe, Org. Lett., 2005, 7, 5873.
41 T. Hoshino, S. Nakano, T. Kondo, T. Sato and A. Miyoshi, Org. 83 Q. Xiong, X. Zhu, W. K. Wilson, A. Ganesan and S. P. Matsuda,
Biomol. Chem., 2004, 2, 1456. J. Am. Chem. Soc., 2003, 125, 9002.
42 T. Dang and G. D. Prestwich, Chem. Biol., 2000, 7, 643. 84 H. Tanaka, H. Noma, H. Noguchi and I. Abe, Tetrahedron Lett.,
43 I. Abe, H. Tanaka and H. Noguchi, J. Am. Chem. Soc., 2002, 124, 2007, 47, 3085.
14514. 85 H. Tanaka, H. Noguchi and I. Abe, Tetrahedron Lett., 2004, 45, 3093.
44 A. Ruf, F. Muller, B. D’Arcy, M. Stihle, E. Kusznir, C. Handschin, 86 I. Abe, T. Dang, Y.-F. Zheng, B. A. Madden, C. Feil, K. Poralla and
O. H. Morand and R. Thoma, Biochem. Biophys. Res. Commun., 2004, G. D. Prestwich, J. Am. Chem. Soc., 1997, 119, 11333.
315, 247. 87 X.-y. Xiao and G. D. Prestwich, J. Am. Chem. Soc., 1991, 113, 9673.
45 R. Thoma, T. Schulz-Gasch, B. D’Arcy, J. Benz, J. Aebi, H. Dehmlow, 88 I. Abe and G. D. Prestwich, J. Biol. Chem., 1994, 269, 802.
M. Hennig, M. Stihle and A. Ruf, Nature, 2004, 432, 118. 89 T. Hoshino and T. Kondo, Chem. Commun., 1999, 731.
46 A. Lenhart, W. A. Weihofen, A. E. Pleschke and G. E. Schulz, Chem. 90 T. Hoshino and S. Ohashi, Org. Lett., 2002, 4, 2553.
Biol., 2002, 9, 639. 91 S. Nakano, S. Ohashi and T. Hoshino, Org. Biomol. Chem., 2004, 2,
47 O. H. Morand, J. D. Aebi, H. Dehmlow, Y.-H. Ji, N. Gains, H. 2012.
Lengsfeld and J. Himber, J. Lipid Res., 1997, 38, 373. 92 T. Abe and T. Hoshino, Org. Biomol. Chem., 2005, 3, 3127.
48 I. Abe, Y. F. Zheng and G. D. Prestwich, Biochemistry, 1998, 37, 5779. 93 E. Caspi, Acc. Chem. Res., 1980, 13, 97.
49 T. Dang, I. Abe, Y. F. Zheng and G. D. Prestwich, Chem. Biol., 1999, 94 I. Abe and M. Rohmer, J. Chem. Soc., Chem. Commun., 1991, 902.
6, 333. 95 I. Abe and M. Rohmer, J. Chem. Soc., Perkin Trans. 1, 1994, 783.
50 N. Morikubo, Y. Fukuda, K. Ohtake, N. Shinya, D. Kiga, K. 96 J. L. Giner, S. Rocchetti, S. Neunlist, M. Rohmer and D. Arigoni,
Sakamoto, M. Asanuma, H. Hirota, S. Yokoyama and T. Hoshino, Chem. Commun., 2005, 3089.
J. Am. Chem. Soc., 2006, 128, 13184. 97 S. P. Matsuda, W. K. Wilson and Q. Xiong, Org. Biomol. Chem., 2006,
51 C. Feil, R. Süssmuth, G. Jung and K. Poralla, Eur. J. Biochem., 1996, 4, 530.
242, 51. 98 T. Kushiro, M. Shibuya and Y. Ebizuka, Eur. J. Biochem., 1998, 256,
52 T. Hoshino, K. Shimizu and T. Sato, Angew. Chem., Int. Ed., 2004, 238.
43, 6700. 99 M. Morita, M. Shibuya, T. Kushiro, K. Masuda and Y. Ebizuka,
53 T. Sato and T. Hoshino, Biosci., Biotechnol., Biochem., 1999, 63, 1171. Eur. J. Biochem., 2000, 267, 3453.
54 T. Hoshino, T. Abe and M. Kouda, Chem. Commun., 2000, 441. 100 T. Kushiro, M. Shibuya and Y. Ebizuka, J. Am. Chem. Soc., 1999, 121,
55 T. Sato, M. Kouda and T. Hoshino, Biosci., Biotechnol., Biochem., 1208.
2004, 68, 728. 101 T. Kushiro, M. Shibuya, K. Masuda and Y. Ebizuka, J. Am. Chem.
56 T. Hoshino and T. Sato, Chem. Commun., 1999, 2005. Soc., 2000, 122, 6816.
57 T. Sato and T. Hoshino, Biosci., Biotechnol., Biochem., 2001, 65, 2233. 102 I. Abe, Y. Sakano, M. Sodeyama, H. Tanaka, H. Noguchi, M. Shibuya
58 T. Merkofer, C. Pale-Grosdemange, K. U. Wendt, M. Rohmer and K. and Y. Ebizuka, J. Am. Chem. Soc., 2004, 126, 6880.
Poralla, Tetrahedron Lett., 1999, 40, 2121. 103 I. Abe, Y. Sakano, H. Tanaka, W. Lou, H. Noguchi, M. Shibuya and
59 C. Pale-Grosdemange, T. Merkofer, M. Rohmer and K. Poralla, Y. Ebizuka, J. Am. Chem. Soc., 2004, 126, 3426.
Tetrahedron Lett., 1999, 40, 6009. 104 H. Noma, H. Tanaka, H. Noguchi, M. Shibuya, Y. Ebizuka and I.
60 T. Sato, S. Sasahara, T. Yamakami and T. Hoshino, Biosci., Biotech- Abe, Tetrahedron Lett., 2004, 45, 8299.
nol., Biochem., 2002, 66, 1660. 105 Q. Xiong, F. Rocco, W. K. Wilson, R. Xu, M. Ceruti and S. P. Matsuda,
61 T. Hoshino, M. Kouda, T. Abe and S. Ohashi, Biosci., Biotechnol., J. Org. Chem., 2005, 70, 5362.
Biochem., 1999, 63, 2038. 106 J. B. R. Herrera, B. Bartel, W. K. Wilson and S. P. T. Matsuda,
62 T. Hoshino, M. Kouda, T. Abe and T. Sato, Chem. Commun., 2000, Phytochemistry, 1998, 49, 1905.
1485. 107 M. J. Segura, M. M. Meyer and S. P. Matsuda, Org. Lett., 2000, 2,
63 C. Full and K. Poralla, FEMS Microbiol. Lett., 2000, 183, 221. 2257.
64 C. Full, FEBS Lett., 2001, 509, 361. 108 T. Kushiro, M. Hoshino, T. Tsutsumi, K. I. Kawai, M. Shiro, M.
65 E. J. Corey, H. Cheng, C. H. Baker, S. P. T. Matsuda, D. Li and X. Shibuya and Y. Ebizuka, Org. Lett., 2006, 8, 5589.
Song, J. Am. Chem. Soc., 1997, 119, 1289. 109 E. J. Corey and S. K. Gross, J. Am. Chem. Soc., 1967, 89, 4561.
66 T. K. Wu, Y. T. Liu, C. H. Chang, M. T. Yu and H. J. Wang, J. Am. 110 O. Boutaud, D. Dolis and F. Schuber, Biochem. Biophys. Res.
Chem. Soc., 2006, 128, 6414. Commun., 1992, 188, 898.
67 T. K. Wu, Y. T. Liu and C. H. Chang, ChemBioChem, 2005, 6, 1177. 111 H. Shan, M. J. Segura, W. K. Wilson, S. Lodeiro and S. P. Matsuda,
68 T. K. Wu and C. H. Chang, ChemBioChem, 2004, 5, 1712. J. Am. Chem. Soc., 2005, 127, 18008.
69 S. Lodeiro, W. K. Wilson, H. Shan and S. P. Matsuda, Org. Lett., 2006, 112 D. R. Phillips, J. M. Rasbery, B. Bartel and S. P. Matsuda, Curr. Opin.
8, 439. Plant Biol., 2006, 9, 305.
View Article Online
113 Q. Xiong, W. K. Wilson and S. P. Matsuda, Angew. Chem., Int. Ed., 120 M. D. Kolesnikova, Q. Xiong, S. Lodeiro, L. Hua and S. P. Matsuda,
2006, 45, 1. Arch. Biochem. Biophys., 2006, 447, 87.
114 L. Jaenicke and F.-J. Marner, Pure Appl. Chem., 1990, 62, 1365. 121 M. Suzuki, T. Xiang, K. Ohyama, H. Seki, K. Saito, T. Muranaka, H.
115 M. Shibuya, T. Xiang, Y. Katsube, M. Otsuka, H. Zhang and Y. Hayashi, Y. Katsube, T. Kushiro, M. Shibuya and Y. Ebizuka, Plant
Ebizuka, J. Am. Chem. Soc., 2007, 129, 1450. Cell Physiol., 2006, 47, 565.
116 G. C. Fazio, R. Xu and S. P. Matsuda, J. Am. Chem. Soc., 2004, 126, 122 M. Shibuya, S. Adachi and Y. Ebizuka, Tetrahedron, 2004, 60, 6995.
5678. 123 P. Tansakul, M. Shibuya, T. Kushiro and Y. Ebizuka, FEBS Lett.,
117 T. Xiang, M. Shibuya, Y. Katsube, T. Tsutsumi, M. Otsuka, 2006, 580, 5143.
H. Zhang, K. Masuda and Y. Ebizuka, Org. Lett., 2006, 8, 124 A. Kawaguchi, H. Kobayashi and S. Okuda, Chem. Pharm. Bull.,
2835. 1973, 21, 577.

118 M. D. Kolesnikova, A. C. Obermeyer, W. K. Wilson, D. A. Lynch, Q. 125 H. Mitsuguchi, Y. Seshime, T. Kushiro, M. Shibuya, I. Fujii and Y.
Xiong and S. P. Matsuda, Org. Lett., 2007, 9, 2183. Ebizuka, The 127th Annual Meeting of the Pharmaceutical Sciences of
119 Y. Ebizuka, Y. Katsube, T. Tsutsumi, T. Kushiro and M. Shibuya, Pure Japan, Toyama, Japan, 2007, p. 30P1pm026.
Appl. Chem., 2003, 75, 369. 126 E. J. Corey and S. K. Gross, J. Am. Chem. Soc., 1968, 90, 5045.

Abe. 2007. Enzymatic Synthesis of Cyclic Triterpenes PDF

Uploaded by

Copyright:

Available Formats

Abe. 2007. Enzymatic Synthesis of Cyclic Triterpenes PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Abe. 2007. Enzymatic Synthesis of Cyclic Triterpenes PDF

Uploaded by

Copyright:

Available Formats

View Article Online / Journal Homepage / Table of Contents for this issue

REVIEW www.rsc.org/npr | Natural Product Reports

Enzymatic synthesis of cyclic triterpenes

Covering: 2002 to 2006

1 Introduction 3 Site directed mutagenesis

The enzyme-templated sequential carbon–carbon bond for-

in 2002,15 and on enzymatic cyclization reactions of squalene and

As described above, the X-ray crystal structure of A. acidocaldarius

a hydrogen bond formed between the lanosterol 3-hydroxy group

Fig. 3 Structures of cyclic triterpenes produced by mutants of A. acidocaldarius SHC.

Scheme 4 Cyclization of oxidosqualene by mutants of S. cerevisiae OSLC.

Scheme 5 Cyclization of oxidosqualene folded in chair–boat–chair conformation by plant OSCs.

(61) is efﬁciently converted to a 2 : 1 mixture (total yield 8%)

(H-13b→14, CH3 -8b→13b, H-9a→8a) with elimination of H-

Scheme 6 Cyclization of a C35 analogue by A. acidocaldarius SHC.

reaction proceeds without rearrangement of carbon and hydrogen,

enzyme reaction. First, both 1-methylidenesqualene (64) and

Notably, a signiﬁcant life time of the intermediate allylic cation,

4.2 Squalene : tetrahymanol cyclase

A squalene cyclase from the ciliate protozoan Tetrahymena pyri-

Scheme 9 Cyclization of squalenediols by A. acidocaldarius SHC.

29,30-bisnorgermanicol (108), and 29,30-bisnor-d-amyrin (109)

thermodynamically favored oleanyl secondary cation (106) (route

Scheme 11 Cyclization of (A) 24,30-bisnoroxidosqualene and (B) 22,23-dihydrooxidosqualene by P. sativum bAS.

Scheme 13 Cyclization of 3-(x-oxidogeranylgeranyl)indole by A. thaliana

including OSLC that efﬁciently catalyze cyclization into (24S)-

pressed in a lanosterol synthase-deﬁcient yeast strain, the yeast

formation has been explained as follows; if the C-13 cation 137

Scheme 15 Cyclization of oxidosqualene folded in chair–chair–chair conformation by plant OSCs.

Scheme 17 Proposed mechanism for formation of tirucalla-7,24-

Finally, the gene At3g45130 was found to encode an accurate

2835. 1973, 21, 577.

You might also like