Lanosterol biosynthesis: The critical role of the methyl-29 group of 2,3-oxidosqualene for the correct folding of this substrate and for the construction of the five-membered D ring

Article abstract of DOI:10.1002/chem.201201779

Lanosterol synthase catalyzes the polycyclization reaction of (3S)-2,3-oxidosqualene (1) into tetracyclic lanosterol 2 by folding 1 in a chair-boat-chair-chair conformation. 27-Nor- and 29-noroxidosqaulenes (7 and 8, respectively) were incubated with this enzyme to investigate the role of the methyl groups on 1 for the polycyclization cascade. Compound 7 afforded two enzymatic products, namely, 30-norlanosterol (12) and 26-normalabaricatriene (13; 12/13 9:1), which were produced through the normal chair-boat-chair-chair conformation and an atypical chair-chair-boat conformation, respectively. Compound 8 gave two products 14 and 15 (14/15 4:5), which were generated by the normal and the unusual polycyclization pathways through a chair-chair-boat-chair conformation, respectively. It is remarkable that the twist-boat structure for the B-ring formation was changed to an energetically favored chair structure for the generation of 15. Surprisingly, 14 and 15 consisted of a novel 6,6,6,6-fused tetracyclic ring system, thus differing from the 6,6,6,5-fused lanosterol skeleton. Together with previous results, we conclude that the methyl-29 group is critical to the correct folding of 1, with lesser contributions from the other branched methyl groups, such as methyl-26, -27, and -28. Furthermore, we demonstrate that the methyl-29 group has a crucial role in the formation of the five-membered D ring of the lanosterol scaffold. Ringing in the changes: The incubation of 1 with porcine-liver cyclase afforded new nortriterpenes 2 and 3 with 6,6,6,6-fused tetracyclic skeletons, which were produced by chair-boat-chair-chair and chair-chair-boat-chair conformations, respectively (see scheme), thus indicating that the 29-methyl group is critical to the correct folding of oxidosqualene and to the formation of the five-membered D ring for lanosterol biosynthesis. Copyright

Full text of DOI:10.1002/chem.201201779

FULL PAPER
DOI: 10.1002/chem.201201779
Lanosterol Biosynthesis: The Critical Role of the Methyl-29 Group of
2,3-Oxidosqualene for the Correct Folding of this Substrate and
for the Construction of the Five-Membered D Ring
Tsutomu Hoshino,* Akifumi Chiba, and Naomi Abe[a]
Abstract: Lanosterol synthase catalyzes
the polycyclization reaction of (3S)-2,3-
oxidosqualene (1) into tetracyclic lano-
sterol 2 by folding 1 in a chair-boat-
chair-chair conformation. 27-Nor- and
29-noroxidosqaulenes (7 and 8, respec-
tively) were incubated with this
enzyme to investigate the role of the
methyl groups on 1 for the polycycliza-
tion cascade. Compound 7 afforded
two enzymatic products, namely, 30-
norlanosterol (12) and 26-normalabari-
catriene (13; 12/13 9:1), which were
produced through the normal chair-
boat-chair-chair conformation and an
atypical chair-chair-boat conformation,
respectively. Compound 8 gave two
products 14 and 15 (14/15 4:5), which
were generated by the normal and the
generation of 15. Surprisingly, 14 and
15 consisted of a novel 6,6,6,6-fused
tetracyclic ring system, thus differing
from the 6,6,6,5-fused lanosterol skele-
ton. Together with previous results, we
conclude that the methyl-29 group is
critical to the correct folding of 1, with
lesser contributions from the other
branched methyl groups, such as
methyl-26, -27, and -28. Furthermore,
we demonstrate that the methyl-29
group has a crucial role in the forma-
tion of the five-membered D ring of
the lanosterol scaffold.
unusual
polycyclization
pathways
through a chair-chair-boat-chair confor-
mation, respectively. It is remarkable
that the twist-boat structure for the B-
ring formation was changed to an ener-
getically favored chair structure for the
Keywords: alkenes · cyclization ·
enzyme catalysis · polycycles · ter-
penoids
Introduction
Virgil[4] proposed the 17b-oriented side chain 3 as the true
protosteryl cation intermediate based on the analysis of the
stereochemistry of the enzymatic products of 20-oxa-2,3-oxi-
dosqualene[4] and (20E)-20,21-dehydro-2,3-oxidosqualene[5]
by hog-liver lanosterol synthase. Furthermore, through the
subsequent backbone-rearrangement process, 3 can be con-
verted into 2 with a 20R configuration through a least-
motion pathway that involves only a small rotation (<608)
Oxidosqualene cyclases convert (3S)-2,3-oxidosqualene (1)
into nearly 120 different cyclic triterpene skeletons.[1] These
enzymes impose either the chair or boat conformation on
acyclic compound 1, and the ensuing proton attack on the
ꢀ
epoxide ring triggers the ring-forming reactions (C C bond
formation) with the correct regio- and stereochemical specif-
icities, thus usually leading to the production of tetra- and
pentacyclic scaffolds. Triterpene cyclases generate many
chiral centers during the polyolefin cyclization cascade, and
the mechanism for the cyclization/rearrangement reactions
has attracted considerable attention.[1] The mechanism for
the cyclization pathway was first proposed over a half-centu-
ry ago by Eschenmoser et al. in 1955,[2] who with Corn-
forth[3] proposed that the chair-boat-chair-boat conformation
for lanosterol biosynthesis affords protosterol cation 4 with
a 17a-oriented side chain (Scheme 1b). Later, Corey and
ꢀ
about the C-17 C-20 axis, but a large rotation (1208) is re-
quired prior to proton migration from C-17 to C-20 to gen-
erate the R configuration, as observed in 4.[4] Thus, a chair-
boat-chair-chair conformation (Scheme 1a) that leads to the
protosterol cation 3 is now accepted as an intermediate in
lanosterol biosynthesis.[6,7]
Numerous studies on lanosterol synthase-mediated reac-
tions of substrate analogues of 1 have been reported.[1–8]
Corey et al. and van Tamelen et al. reported the enzymatic
products of 26-noroxidosqualene,[9] 28-noroxidosqualene
(5),[10] and 27,28-bisnoroxidosqualenes (6; Figure 1).[11] Ana-
logues of 26-noroxidosqualene and 5 were converted into
19-nor- and 18-norlanosterols 10,[9,10] respectively, thus indi-
cating that the normal chair-boat-chair-chair conformation
was imposed by lanosterol synthase. The tetracyclic product
11, which consists of the ring system of both a 6,6,5-fused
tricycle and an isolated four-membered ring, was created
from 6, thus indicating that 6 was folded in an abnormal
chair-chair-boat conformation.[11] Interestingly, the confor-
mation of the B ring was converted from a twist boat into a
chair structure. Accordingly, Corey et al. proposed that the
[a] Prof. Dr. T. Hoshino, A. Chiba, N. Abe
Department of Applied Biological Chemistry
Faculty of Agriculture
and Graduate School of Science and Technology
Niigata University, Ikarashi 2-8050
Nishi-ku, Niigata, 950-2181 (Japan)
Fax : (+81)25-262-6854
E-mail: hoshitsu@agr.niigata-u.ac.jp
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201201779.
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

methyl-27 group at C10 of 1 was critical for the correct fold-
ing in the polycyclization and rearrangement reaction.[11]
Clayton et al.[12] and Corey et al.[13] examined the cyclization
reaction of 1-nor and 25-nor analogues and reported that
the 1-nor analogue underwent the polycyclization reaction
to yield 29-norlanosterol, whereas the 25-nor analogue un-
derwent no reaction. However,
no report has appeared con-
cerning the enzymatic reactions
of 7 and 8, in which the methyl
groups present in 1 at C-10 and
C-19, respectively, are absent.
Herein, we describe the enzy-
matic products of 7 and 8 and
discuss how the branched
methyl groups present in 1 in-
fluence
the
conformation
during the polycyclization cas-
cade. We report that the
methyl-29 group at C-19 is
more important to the correct
folding of 1, especially with
regard to the boat structure of
the B ring relative to the
methyl-27 group at C-10. We
also describe the unprecedented
cyclization pathway that gener-
ates a novel skeleton that con-
sists of a 6,6,6,6-fused tetracy-
clic ring system, which differs
from the 6,6,6,5-fused lanoster-
ol skeleton.
Scheme 1. Folding conformation of (3S)-2,3-oxidosqualene 1 for lanosterol biosynthesis.
Results
Syntheses of noroxidosqualenes
5, 7, and 8: The synthesis of 7
was performed according to
Scheme 2. (4E,8E)-5,9,13-Tri-
methyltetradeca-4,8,12-trienal
(19), prepared from squalene
by treating with MCPBA and
periodic acid, was subjected to
the Wittig–Horner reaction
with ethyl diethylphosphonoa-
cetate to yield (2E,6E,10E)-
ethyl-7,11,15-trimethylhexade-
ca-2,6,10,14-tetraenoate
(20;
87%). Ethyl ester 20 was con-
verted into the corresponding
alcohol 21 by reduction with
diisobutylaluminium
hydride
(DIBAL-H; 93%), which was
then transformed into bromide
22 by using PBr3. Commercially
available geraniol 23 was treat-
ed with NBS to give bromohy-
drin 24 (53%), followed by bro-
mination with PBr3 to afford
Figure 1. Structures of substrate analogues 5–9 and the enzymatic products 10–16. The fundamental core of 13
is the same as that of malabaricatriene triterpene, but 13 lacks the Me-26 group of the malabaricatriene skele-
ton that is situated at C-8. The black numbering of the carbon atoms is used for the assignments of the NMR
spectroscopic data and is arbitrary. The red numbering for 10, 12, and 13 is based on the numbering of the
parent skeletons (lanosterol and malabaricatriene with C30).
&
2
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Substrate Folding in Lanosterol Biosynthesis
FULL PAPER
HPLC. Compound 12 was sepa-
rated as a pure compound, and
13 was acetylated before purifi-
cation. The 1H and 13C NMR
(600 and 150 MHz, respective-
ly) spectra of 12 measured in
CDCl3 revealed one olefinic
proton (dH =5.09 ppm, t, J=
6.9 Hz), two vinylic methyl
groups (dH =1.68 and 1.60 ppm,
each 3H, s), and one tetrasub-
stituted double bond (C-8: dC =
128.0 ppm,
135.9 ppm, s). Heteronuclear
multiple-bond correlations
s;
C-9:
dC =
(HMBC) were observed for
Me-19/C-9, H-14/C-8, and H-
14/C-9. No NOE interactions
for H-14/Me-18 and clear NOE
interactions for H-14/H-7/H-17
established the structure of the
lanosterol skeleton for 12 (see
Figure S5 in the Supporting In-
formation), which was ascer-
tained to be 30-norlanosterol
with 20R stereochemistry pro-
duced through the normal cycli-
zation pathway (Scheme 1a).
Compound 12 is one of the in-
termediates in cholesterol bio-
synthesis and is known as a tes-
Scheme 2. Synthetic scheme of 10-noroxidosqualene (7). 1) (Et2O)2POCH2CO2Et, NaH, Et2O, THF, ꢀ408C;
2) (i-C4H9)2AlH, Et2O, ꢀ408C; 3) PBr3, THF, 08C; 4) NBS, THF, 08C; 5) PhSO2Na·2H2O, DMF, 08C;
6) K2CO3, MeOH, room temperature; 7) nBuLi, THF/HMPTA (4:1), ꢀ788C; 8) LiBEt3H, Et2O, 0 8C.
HMPTA=hexamethylphosphorotriamide, MCPBA=meta-chloroperbenzoic acid, NBS=N-bromosuccinimide.
ticular
meiosis-activating
sterol.[14] The acetate of product
13 had three olefinic protons
and four vinylic methyl groups,
thus suggesting a tricyclic com-
pound, which was further veri-
25. Bromide 25 was treated with sodium benzenesulphinate
to yield the phenylsulfone derivative 26, which was convert-
ed into epoxide 27 (61%). The coupling reaction of 27 and
22 was carried out in THF containing HMPTA by adding
nBuLi, thus yielding the phenylsulfone derivative 28 of oxi-
dosqualene (13%). The removal of the phenylsulfonyl
group was carried out with the super hydride reagent to
obtain the desired 7 (76%).
The syntheses of noranalogues 5 and 8 were performed by
using essentially identical methods to that for 7 with the
same reagents. Compounds 21 and 18 were used for the
preparation of 5 and 8, respectively (the experimental de-
tails are described in the Supporting Information). The
ethyl-substituted analogue 9 was prepared according to the
method described in the Supporting Information.
fied by detailed HMBC analysis (see Figure 2a and the Sup-
porting Information). Definitive NOE interactions for H-9/
H-13 and Me-25/H-8 demonstrated that the structure of 13
is as shown in Figure 2, thus allowing 7 to adopt a chair-
chair-boat conformation (Scheme 3a). The cationic inter-
mediate 29 thus formed could be quenched by the deproto-
nation of H-15 to give the E geometry, which was confirmed
by the clear NOE interaction of Me-26/H-16 (see the Sup-
porting Information). This conformation is consistent with
that proposed for the polycyclization reaction of 27,28-bis-
noroxidosqualene 6.[11] However, it should be noted that the
ratio of fully cyclized 12/partially cyclized 13 was deter-
mined to be 9:1 by GC analysis, thus suggesting that the in-
fluence of the small hydrogen substitution at C-10 on the
polycyclization cascade is not significant.
Products 14 and 15 were generated by incubating (ꢁ)-8
The structures of products 12–15 from noroxidosqualenes 7
and 8 and the cyclization pathways: Products 12 and 13
were generated by incubating (ꢁ)-7 with hog-liver lanosterol
synthase and were purified by column chromatography and
with hog-liver lanosterol synthase and their acetates were
1
purified with reversed-phase HPLC. H NMR (see the Sup-
porting Information) and DEPT analysis revealed that both
products had one olefinic proton (dH =5.10 ppm, t, J=
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

T. Hoshino et al.
be noted that the Me-19 and Me-29 groups must be placed
in a trans arrangement because both 14 and 15 had no NOE
interaction between the Me-19 and Me-29 groups, as seen in
lanosterol.
The stereochemistry of 14 and 15 was carefully inferred
on the basis of detailed analysis of the NOESY spectra.
Some of the angular methyl proton signals overlapped or
were contiguously located. The NMR spectra of the acetate
salts of 14 and 15 were measured in different solvents such
as CDCl3, C6D6, and (CD3)2CO to separate these signals
(see the Supporting Information). The NOE interactions of
authentic lanosterol acetate were also collected in the differ-
ent solvents (see the Supporting Information) and used as a
reference to determine the stereochemistry of 14 and 15. In
the NOESY spectra of the 14-acetate (see the Supporting
Information), definitive cross peaks of Me-29/H-5a and Me-
19/Me-20b confirmed the configurations of the Me-29a and
Me-19b groups. Furthermore, a strong NOE interaction be-
tween Me-29 and H-18 was observed for 14, but no NOE in-
teraction for Me-19/H-18 indicated that both Me-29a and
H-18a are arranged in an axial disposition in the chair struc-
ture. A clear NOE interaction for Me-19b/H-21 was ob-
served, which is indicative of Me-19b and H-21b (Fig-
ure 2b). Thus, the following stereochemistry was assigned
for 14: 3S, 5R, 10S, 13R, 14R, 18R. On the other hand, clear
NOE interactions for Me-19/H-5a and Me-20b/Me-29 were
found for the 15-acetate, but no NOE interaction for Me-29/
H-5a (see the Supporting Information), unambiguously indi-
cated the stereochemistry of Me-29b and Me-19a. Definitive
NOE interactions for Me-29b/H-21and Me-19a/H-18 indi-
cated that H-21 and H-18 were oriented in axial and equato-
rial dispositions, respectively (Figure 2c). The chair structure
of the D ring was further supported by strong NOE interac-
tions for Me-19/H-17ax, Me-19/H-15ax, and H-15ax/H-17ax.
Thus, the stereochemistry of 15 was determined as follows:
3S, 5R, 10S, 13S, 14S, 18R. It is noticeable that the configura-
tions at C-13 and C-14 of 14 are the same as those of 2, but
those of 15 are opposite to those of 2. Molecular modeling
of the acetate salts of 14 and 15 supported the configura-
tions at C-13 and C-14 (see the Supporting Information). A
comparison of the proton chemical shifts of the angular Me
groups of the acetate salts of 14 and 15 with those of lano-
sterol acetate (see the Supporting Information) further sup-
ported the configurations. Thus, the structures of 14 and 15
are depicted as shown in Figure 1.
Figure 2. Key data of COSY, HOHAHA, NOE, and HMBC spectra for
proposing the structures of products a) 13, b) 14, and c) 15, which were
measured in CDCl3, C6D6, and [D6]acetone (see the Supporting Informa-
tion).
7.0 Hz) and one tetrasubstituted double bond (dC =136 ppm,
s; dC =134 ppm, s), as seen in lanosterol structure 2. In the
HOHAHA spectra of the acetates of 14 and 15 (see the
Supporting Information), the olefinic proton H-23 of both
products had correlations with Me-25, Me-26, H-22, H-21,
and H-18. The Me-19 group had a HMBC cross-peak for C-
18 (dC =40.43 ppm, d for 14-acetate; dC =45.24 ppm, d for
15-acetate). The following COSY and HOHAHA networks
were found from H-18: H-17, H-16, and H-15, and a defini-
tive HMBC correlation of Me-29/C-15 was found. These
data clearly demonstrated that both 14 and 15 consisted of a
six-membered cycle for the D ring, but not a five-membered
ring. The mass spectra (EI) of both the acetate salts of 14
and 15 were identical to each other (see the Supporting In-
formation), thus suggesting that the structure of 14 differed
from that of 15 only with regard to stereochemistry. It is to
It is possible that 29-noroxidosqualene (8) adopts a chair-
boat-chair-chair conformation (path a in Scheme 3b) to
afford cation intermediate 30, which could then undergo
1,2-hydride and methyl shifts in an antiperiplanar fashion
followed by the deprotonation of H-9 to yield 14. A chair-
chair-boat-chair conformation could be adopted for the for-
mation of 15 (path b in Scheme 3b) to give the 6,6,5-fused
ꢀ
tricyclic cation 31 with H-13a. The C-13 C-14 bond axis
could rotate to afford cation 32 with an a-oriented Me
group at C-14, and the subsequent ring expansion could give
the 6,6,6-fused tricyclic cation 33, followed by further cycli-
zation to produce the 6,6,6,6-fused tetracycle 34. The subse-
&
4
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Substrate Folding in Lanosterol Biosynthesis
FULL PAPER
ring expansion occur, would
lead to 59–62, but these struc-
tures do not agree with the
NOE data of 15-acetate. The
reaction schemes of example 5
(see the Supporting Informa-
tion, page S51) and path b in
Scheme 3 are accompanied by a
ꢀ
C-13 C-14 bond rotation on
Markovnikov cation intermedi-
ates prior to the ring expansion
(5-!6-membered ring), then
further cyclization and back-
bone rearrangements lead to
product 15 (63 is identical to
15) that fully coincides with the
observed NOE data. Thus, the
bond rotation prior to the ring
expansion is required to fulfill
the stereochemistry of 15. A
higher production of 15 than 14
(15/14 5:4) indicates that the
Me-29 group is more crucial for
the correct folding of 1 for the
biosynthesis of
2 relative to
Me-27 (12/13 9:1).
In our experiment, the incu-
bation of 5 (see Figure S3 in the
Supporting Information for the
synthetic details) gave 10 as the
sole enzymatic product (see
Figure S10 and the NMR spec-
troscopic data in the Supporting
Information, pages S54–55),
which was in good agreement
with the results reported by van
Tamelen et al.[10]
Time-dependent inactivation by
noroxidosqualenes 5, 7, and 8:
Compounds 12–15 were pro-
Scheme 3. Folding conformations of substrate analogues 7–9, thus leading to products 13–16.
quent backbone rearrangements and the proton elimination
of H-9 could result in the production of 15.
duced in very small amounts. We investigated the possibility
that time-dependent inactivation occurred during the incu-
bations of 7 and 8. Kinetic analysis revealed the following
parameters: IC50 =180 mm, kinact =0.0037 minꢀ1, KI =0.013 mm
for 7; IC50 =76 mm, kinact =0.0068 minꢀ1, KI =0.015 mm for 8
(see Figure S11 in the Supporting Information). It was re-
ported that 6 displayed time-dependent irreversible inhibi-
tion.[11] Time-dependent inhibition was also observed for 5
in our experiment (kinact =0.0065 minꢀ1, KI =0.020 mm), al-
though this finding has not been previously reported.[10]
Other folding conformations of 8 were envisaged and the
corresponding cyclization products were evaluated to vali-
date the proposed structures of 14 and 15 (see the Support-
ing Information, pages S50–S53). To acquire the configura-
tion of Me-29a for 15, which is identical to that of 2, the
conformation for B-ring construction must be a boat form.
Three other possible folding conformations that lead to
products 56–58 are assumed, but the NOE interactions of
the 15-acetate are inconsistent with the proposed structures
of 56–58 (see the Supporting Information, page S50).
Cyclization pathway of the bulkier ethyl-substituted oxidos-
qualene 9: Previously,[15–17] we reported the results of incuba-
tion experiments of analogues of 1 bearing an ethyl substitu-
ent at C-10 and/or C-15 with hog-liver cyclase (see Fig-
ure S12 in the Supporting Information). Analogue 64 substi-
In contrast, the chair-folding in the B-ring formation re-
sults in a b-oriented Me-29 group. All of the possible folding
conformations are listed in the Supporting Information
(page S51). Examples 1–4, in which no bond rotation and no
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

T. Hoshino et al.
tuted with an ethyl group at both C-10 and C-15 afforded
only the monocyclic products 67 and 69.[15] Analogue 65
with an ethyl substituent at C-10 gave monocycles 68 and
70, tricycle 71, and tetracycle 73 (68/70/71/73 1:1:1:1).[16] An-
alogue 66 with an ethyl group at C-15 yielded tricycle 72
and tetracycle 74 (72/74 1:2.6).[17] The polycyclization reac-
tion of 64 afforded the monocyclic compounds 68 and 70 as
the main products, but the reaction of 66 proceeded to the
tri- and tetracyclic stages 72 and 74, respectively. An ethyl
group is slightly bulkier than a methyl group, thus leading us
to conclude that the binding site that accommodates the
methyl group of C-10 of 1 is more discriminating than the
binding site for the methyl group of C-15.[17]
Discussion
X-ray crystallographic analysis of human lanosterol synthase
was reported in 2004.[18] It revealed that the tyrosine (Tyr)
residue Tyr98 is well positioned to enforce the energetically
unfavorable boat folding for the B-ring formation of 3 by
pushing the methyl group at C-10 of 1 below the molecular
plane (Scheme 1a).[18] A less-bulky hydrogen atom in 7 is
substituted at C-10. The increased space between the Tyr
residue and the hydrogen atom at C-10 may have allowed
freer motion of this substrate, thus leading to the adoption
of the energetically favored chair structure in the B ring and
the production of 13. The altered chair folding for the
B ring could further exert influence on the conformation for
the subsequent C-ring formation; a chair form for the C ring
in lanosterol biosynthesis could be changed into a boat
structure (Scheme 3a).
Wu et al. reported that the F699N or F699M variants of
S. cerevisiae lanosterol synthase produced the tricyclic com-
pound (14E,17E)-(13aH)-malabarica-14,17,21-triene,[19] the
skeleton of which is the same as that of 13 (i.e., 26-normala-
baricatriene). This report is the first that describes the suc-
cessful alteration of the conformation of the B ring from
boat to chair by using mutagenesis experiments. The phenyl-
alanine (Phe) residue Phe699 is located near the D-ring-for-
mation site[19] and is chemically equivalent to the Phe601
residue of squalene-hopene cyclase, which stabilizes the in-
termediate cation through cation–p interactions.[1a,22,23] It
was postulated that these mutations might cause inappropri-
ate positioning of the side chain and partially disrupt the
transient dipole interactions between the carbocationic in-
termediate and the hydroxy group of the Tyr99 (equivalent
to Tyr98 of human lanosterol synthase) and/or Tyr707 resi-
dues, thus leading to the atypical chair-chair-boat conforma-
tion (as seen in 29).[19] Furthermore, it was demonstrated
that by varying the size of the side chain through a single
amino acid substitution at the residues of Tyr99-Tyr707-
Ile705-Phe699, which are highly conserved (see Figure S14
in the Supporting Information) and are in proximity to the
B-, C-, and D-ring-formation sites, the substrate conforma-
tion and polycyclization cascade are altered.[19–21] Thus, the
appropriate sizes of the active-site residues to allow for
steric bulk are crucial to determine the substrate folding. In
turn, the steric size at the methyl-substituted positions on
substrate 1 also influenced the conformation and/or the pol-
ycyclization cascade (e.g., truncation of the annulation reac-
tion), as demonstrated by the incubation experiments of 7,
8, and ethyl-substituted analogues 9 and 64–66 with the
native lanosterol cyclase from hog-liver. Therefore, it is ap-
parent that the steric-size match between the substrate and
the catalytic site is critical to the correct folding of 1 and the
cyclization pathway.
Synthesized (ꢁ)-9 (6 mg) was incubated with partially pu-
rified hog-liver cyclase to examine the effect of the bulkier
substituent at C-19 on the polycyclization pathway. The en-
zymatic product was purified by column chromatography on
SiO2 gel and normal-phase HPLC (hexane/THF 100:0.5) to
1
yield pure 16 (1.8 mg) as the sole enzymatic product. The H
and 13C NMR spectra (in CDCl3) indicated the following
functional groups: one olefinic proton (dH =5.11 ppm, t, J=
6.8 Hz), two vinylic methyl protons (Me-26: dH =1.682 ppm,
s, 3H; Me-27: dH =1.605 ppm, s, 3H), and one tetrasubstitut-
ed double bond (dC =134.4 ppm, s; dC =134.3 ppm, s), thus
indicating a lanosterol skeleton for 16. The detailed analyses
of the 2D NMR spectra were in good agreement with homo-
lanosterol structure, in which ethyl group is substituted at C-
20 (see Figure S13 in the Supporting Information for the
NMR spectroscopic analyses). The 20R stereochemistry is
credible in light of the cyclization pathway (Scheme 3c).
High production of 16 indicates that the binding site that ac-
cepts C-19 is somewhat more loosely packed than the bind-
ing sites for C-10 or C-15 and can accept the bulky ethyl
group, albeit with the somewhat lower affinity of 9 for this
cyclase relative to that observed for 1 (Km =231 mm, Vmax
=
0.47 mmminꢀ1 for 9; Km =75 mm, Vmax =0.33 mmminꢀ1 for 1).
Kinetic analysis showed that 9 did not undergo time-de-
pendent inhibition. The polycyclization reaction of the bulk-
ier ethyl-substituted 65 and 66 halted at the premature stage
to afford mono- and tricyclic products (see Figure S12 in the
Supporting Information). In the reaction of 9, full polycycli-
zation occurred without stopping at the intermediate stage.
In contrast, the methyl-deficient 26-noroxidosqualene and
27-noroxidosqualene 7 (substituted with a less-bulky hydro-
gen atom) gave 19-norlanosterol[9] and 30-norlanosterol as
the main products (ca. 90%; see Figure 1), respectively,
which resulted from the complete polycyclization reaction
that proceeded through the normal folding conformation.
Therefore, these results strongly indicate that the binding
site involved in the early steps of the polycyclization is tight-
ly packed and more discriminating, whereas the binding site
for the later step is less compact and can accept larger sub-
stituents.
Analogue 8 could undergo two different cyclization path-
ways. Product 14 could be produced under a normal poly-
cyclization pathway (through a chair-boat-chair-chair confor-
mation). However, it is difficult to discern why 8 was folded
in a chair-chair-boat-chair during the polycyclization cascade
&
6
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Substrate Folding in Lanosterol Biosynthesis
FULL PAPER
to afford 15 (path b in Scheme 3). Notably, the conforma-
tion of the B ring is changed from the boat to chair form.
As described above, the recognition site of the Me-29 group
of 1 is less compact than necessary because the bulkier
ethyl-substituted 9 was also accepted as the substrate and
folded in the normal conformation (Scheme 3c) to afford
homolanosterol 16. Binding between the less-bulky hydro-
gen atom at C-19 and the loosely packed binding site may
have given rise to excess space and free motion at the B-
ring-formation site, thus allowing the thermodynamically fa-
vored chair structure to be formed instead of the constrain-
ed boat conformation. Furthermore, the improper arrange-
ment of 8 inside the reaction cavity could further contribute
to the misfolded boat form at the C-ring-formation site, as
mentioned in the production of 13. The increase in the cleft
volume at the D-ring-formation site, which could be gener-
ated by the lack of the Me-29 group, could further allow the
Conclusion
In conjunction with previously published results,[9–13] we con-
clude that the Me-29 group of 1 is critical to the correct
folding of the substrate in the enzyme cavity, with lesser
contributions from the other branched methyl groups, such
as Me-26, Me-27, and Me-28. Although lanosterol synthase
folds 1 in a normal chair-boat-chair-chair conformation, the
nor-substrate 8 was folded in two different conformations.
The unusual chair-chair-boat-chair conformation afforded
the 6,6,6,6-fused tetracyclic skeleton 15, whereas the normal
chair-boat-chair-chair conformation gave the 6,6,6,6-fused
tetracycle 14. Products 14 and 15 were diastereomers that
have opposite stereochemistry at C-13 and C-14. This report
is the first to describe the importance of the Me-29 group
for the correct folding of 1 and the generation of the new
triterpene scaffold by using hog-liver cyclase. Furthermore,
this study highlights the reason why a five-membered D ring
is constructed during lanosterol biosynthesis instead of a six-
membered ring.
ꢀ
C-13 C-14 bond rotation in the Markovnikov cation 31
prior to the ring expansion, although this rotational move-
ment is usually unlikely when the substrate is tightly con-
stricted by the enzyme active sites.
Why was the six-membered D ring constructed for 14 and
15 instead of the five-membered D ring? The intermediary
cation 33 cannot generate a stable tertiary cation such as 3
in the subsequent annulation reaction, but affords only a
secondary cation irrespective of the formation of five- or
six-membered D rings. Therefore, a six-membered ring with
less steric strain than a five-membered ring could be con-
structed for both products 14 and 15. The H-9b atom of in-
termediate 30 was eliminated for the formation of 14 (path
a in Scheme 3), as seen in the biosynthesis of 2, but H-9a of
intermediate 34 was abstracted for that of 15 (path b in
Scheme 3). At the present time, the reason remains uncer-
tain why the stereochemically different H-9 could be elimi-
nated. One possible explanation is that the basic (polar) res-
idue responsible for the deprotonation reaction of H-9a
may be different from that for H-9b. The X-ray crystal struc-
ture of human lanosterol synthase has revealed that the phe-
nolic Tyr503 residue that is hydrogen bonded to the histi-
dine (His) residue His232 and situated above the molecular
plain of 3 is involved in the deprotonation reaction of H-
9b.[17] The H-9b atom of 30 could be eliminated in the same
mechanism as that of 3 to form 14. Another polar residue
that is located below the molecular plain of 34 may have ab-
stracted H-9a to yield 15. For the production of 13 and 11,
the different other polar residues would have worked for
the deprotonation reaction.
Experimental Section
General analytical methods: NMR spectra of the enzymic products were
recorded in CDCl3 on Bruker DMX 600 and DPX 400 spectrometers, the
chemical shifts are given in ppm relative to the solvent peak dH =7.26
and dC =77.0 ppm as the internal reference for the 1H and 13C NMR
spectra, respectively. The chemical shifts are given in ppm relative to the
solvent peak in C6D6 (dH =7.28 and dC =128.0 ppm). The chemical shifts
of the solvent peaks were assigned to be dH =2.04 and dC =29.8 ppm in
(CD3)2CO. The coupling constants J are given in Hz. GC analyses were
recorded on a Shimadzu GC-8A chromatograph equipped with a flame-
ionization detector (a DB-1 capillary column; 30 mꢁ0.25 mmꢁ0.25 mm;
J&W Scientific Inc.). GC-MS spectra were recorded on a JEOL SX 100
or a JEOL JMS-Q1000 GC K9 instrument equipped with a ZB-5 ms ca-
pillary column (30 mꢁ0.25 mmꢁ0.25 mm; Zebron) by using the EI mode
operated at 70 eV. HRMS (EI) was performed by using a direct-inlet
system. HPLC was carried out with Hitachi L-1700 (pump) and L-7405
(UV detector), and the HPLC peaks were monitored at l=210 or
214 nm. Specific rotation values were measured with a Horiba SEPA-300
polarimeter.
Syntheses of noroxidosqualenes 5, 7 and 8: All the synthetic experiments
were carried out in a N2 atmosphere. The syntheses of 5, 7, and 8 were
performed by using essentially identical methods and with the same re-
agents. The detailed synthetic processes are described in the Supporting
Information (see Figures S1–S4).
Preparation of ethyl-substituted oxidosqualene 9 (see Figure S4 in the
Supporting Information): Triphenylphosphine (17 g, 64.8 mmol) and ethyl
2-bromobutyrate 48 (10 g, 51.3 mmol) were heated to reflux for 12 h in
toluene to give phosphorane 49. The Wittig reaction of (4E,8E,12E)-
4,9,13,17-tetramethyloctadeca-4,8,12,16-tetraenal (18) with 49 gave
(2E,6E,10E,14E)-ethyl-2-ethyl-6,11,15,19-tetramethylicosa-2,6,10,14,18-
pentaenoate (50; 84%), 1.05 g (2.53 mmol) of which was subjected to re-
Further studies, including X-ray crystallography studies of
the complex between 29-nor-2,3-iminosqualene (a potent in-
hibitor) and lanosterol cyclase are required to validate our
assumptions or to understand precisely the polycyclization
mechanisms of 8. Time-dependent inactivation by 7 and 8
would have occurred due to the covalent-bond formation of
this cyclase with intermediary cations 29 and 30–34, respec-
tively, as discussed in the enzymatic reaction of 6 with hog-
liver cyclase.[11]
duction with DIBAL-H (hexane solution, 0.98 molLꢀ1
,
5.93 mL
(5.81 mmol)) to yield alcohol 51 (96%). Next, 51 was transformed into
the bromohydrin derivative 52, which was then converted into the corre-
sponding bromide 54. The phenylsulfone derivative 41, prepared from 3-
methylbut-2-en-1-ol (39), and 54 were coupled with nBuLi to afford C31-
phenylsufone 55, followed by reduction with super hydride to yield 9.
Preparation of hog-liver cyclase (lanosterol synthase) and the incubation
conditions of 2,3-oxidosqualene and its analogues: Hog liver (100 g) was
homogenized at 48C in a Waring blender with tris(hydroxymethyl)amino-
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

T. Hoshino et al.
methane hydrochloride (Tris-HCl) buffer (90 mL, 0.1m, pH 7.4) contain-
ing ethylenediaminetetraacetic acid (EDTA; 1 mm), phenylmethylsulfon-
yl fluoride (40 mgmLꢀ1), and mercaptoethanol (1 mm). The reaction mix-
ture was centrifuged for 10 min at 17000ꢁg at 48C. The supernatant was
further centrifuged at 100000ꢁg for 90 min. The resulting microsomal
pellets were resuspended and homogenized in the original quantity of
Tris-HCl buffer containing EDTA (1 mm), mercaptoethanol (1 mm), and
0.5% Triton X-100 by using a Potter-Elvehjem homogenizer and was cen-
trifuged at 100000ꢁg for 90 min. The supernatant was used as the
enzyme source with a typical protein content of 20–25 mgmLꢀ1. Substrate
1 and its analogues 5, 7, 8, and 9 (200 mg) were emulsified with Triton X-
100 (2.8 mg) in Tris-HCl buffer (0.1m, 0.2 mL, pH 7.4). The solution was
flushed with nitrogen and incubated for 12 h at 378C under anaerobic
conditions. KOH in methanol (15%, 4 mL) was added to the mixture to
terminate the reaction. The incubation tube was heated to 708C for
30 min to saponify the reaction mixture before extraction with hexane
(3ꢁ1 mL). The organic layer was evaporated under reduced pressure to
give a residue, which was dissolved in hexane/EtOAc (10:2) and applied
to a short column of SiO2 gel to remove the excess Triton X-100. Analy-
sis with GC (Shimadzu, GC-8 A chromatograph) equipped with a DB-1
capillary column (0.53 mmꢁ30 m) was used to examine the product pro-
files (injection temperature=2908C, column temperature=2808C, carrier
gas=0.75 kgcmꢀ2). Large-scale incubations were conducted to isolate the
enzymatic products.
6.5 Hz; H-21), 0.984 (s, 3H; Me-19), 0.996 (s, 3H; Me-28), 1.02 (m, 2H;
H-22), 1.13 (bd, J=12.8 Hz; H-5), 1.14 (m; H-17), 1.26 (m; H-1), 1.28
(m; H-15), 1.30 (m; H-16), 1.37 (m, 1H; H-20), 1.38 (m; H-12), 1.50 (m;
H-6), 1.58 (m, 2H; H-2, H-16), 1.600 (s, 3H; Me-27), 1.68 (m; H-6),
1.6801 (s, 3H; Me-26), 1.69 (m; H-2), 1.74 (m; H-1), 1.82 (m; H-23), 1.84
(m; H-15), 1.90 (m; H-7), 1.98 (m, 2H; H-12, H-23)), 2.02 (m, 1H; H-
14), 2.04 (m; H-7), 2.06 (m, 2H; H-11), 3.24 (dd, J=10.9, 3.3 Hz; H-3),
13
5.09 ppm (t, J=6.9 Hz; H-24);
C NMR (150 MHz, CDCl3): d=11.25
(q; C-18), 15.35 (q; C-29), 17.62 (q; C-27), 18.44 (t; C-6), 18.63 (q; C-21),
19.84 (q; C-19), 22.05 (t; C-11), 23.78 (t; C-16), 24.80 (t; C-23), 25.71 (q;
C-26), 27.91 (t; C-2), 27.95 (q; C-28), 28.45 (t; C-7), 28.79 (t; C-15), 35.76
(t; C-1), 36.04 (t; C-22), 36.07 (d; C-20), 36.98 (t; C-12), 38.93 (s, 2C, C-4,
C-10), 42.12 (s; C-13), 50.24 (d; C-5), 51.89 (d; C-14), 54.79 (d; C-17),
78.97 (d; C-3), 125.2 (d; C-24), 128.0 (s; C-8), 130.9 (s; C-25), 135.9 ppm
(s; C-9); MS (EI): m/z (%):69 (84), 135 (50), 241 (12), 245 (16), 259 (22),
397 (35), 412 (100) [M+]; HRMS (EI): calcd for C29H48O: 412.3705;
found: 412.3715.
Product 13-acetate: [a]2D0 =ꢀ9.375 (c=0.04 in CHCl3) 1H NMR
(600 MHz, C6D6): d=0.843 (s, 3H; Me-25), 0.85 (m; H-5), 0.92 (m; H-9),
0.94 (m; H-12), 1.016 (s, 3H; Me-23), 1.049 (s, 3H; Me-24),1.13 (m; H-1),
1.30 (m; H-11), 1.33 (m; H-6), 1.40 (m; H-8), 1.45 (m; H-1), 1.46 (m; H-
11), 1.62 (m; H-7), 1.63 (m; H-6), 1.673 (s, 3H; Me-29), 1.728 (s, 3H;
Me-26), 1.74 (m; H-2), 1.773 (s, 3H; Me-27), 1.801 (s, 3H; Me-28), 1.873
(s, 3H; acetyl Me), 1.88 (m; H-2), 1.91 (m; H-7), 169.8 (s, COO), 2.08
(m; H-13), 2.11 (m; H-12), 2.23 (bt, 2H, J=7.0 Hz; H-19), 2.29 (m, 2H;
H-20), 3.00 (t, J=6.7 Hz; H-16), 4.85 (dd, J=12.5, 4.5 Hz; H-3), 5.36 (bt,
J=6.8 Hz; H-21), 5.53 (bt, J=6.8 Hz; H-17), 5.56 ppm (bt, J=6.8 Hz; H-
15); 13C NMR (150 MHz, C6D6): d=13.35 (q; C-26), 13.59 (q; C-25),
17.72 (q; C-29), 16.16 (q; C-27), 16.79 (q; C-24), 20.84 (q, acetyl Me),
21.83 (t; C-6), 22.92 (t; C-11), 25.82 (q; C-28), 27.14 (t; C-20), 27.44 (t;
C-16), 28.36 (q; C-23), 29.01 (t; C-12), 24.33 (t; C-2), 31.52 (t; C-7), 36.38
(s; C-10), 37.42 (t; C-1), 37.88 (s; C-4), 40.15 (t; C-19)„ 42.44 (d; C-8),
54.99 (d; C-5), 55.22 (d; C-13), 59.14 (d; C-9), 83.00 (d; C-3), 123.6 (d; C-
17), 124.0 (d; C-15), 124.9 (d; C-21), 131.1 (s; C-22), 134.9 (s; C-18),
136.4 (s; C-14), 169.8 ppm (s, acetyl CO); MS (EI): m/z (%): 69 (97), 81
(52), 109 (68), 123 (100), 135 (63), 177 (33), 215 (35), 329 (32), 454 (23)
[M+]; HRMS (EI): calcd for C31H50O2: 454.3811; found: 454.3820.
Isolation of enzymic products 12 and 13 from 7, 14 and 15 from 8, and 16
from 9:
Chemically synthesized (ꢁ)-7 (60 mg; see Figure S2 in the Supporting In-
formation) was anaerobically incubated with the microsomal protein
from pig liver for 12 h at 378C. The enzymatic reactions were terminated
by adding KOH in methanol (15%) and heating to 708C for 30 min. The
enzymatic products of 7 were extracted with hexane and purified by
column chromatography on SiO2 gel, followed by normal-phase HPLC
with hexane/2-PrOH (100:0.05) as the eluent to afford the pure product
12 (4.0 mg). The fraction containing 13 was contaminated with other sub-
stances, therefore the entire fraction was acetylated with Ac2O/pyridine,
and pure 13-acetate (0.4 mg) was separated by HPLC with hexane/2-
PrOH (100:0.05) as the eluent.
Product 14-acetate: [a]D20 =ꢀ29.32 (c=0.054 in CHCl3); 1H NMR
(600 MHz, CDCl3): d=0.687 (s, 3H; Me-19), 0.83 (m; H-21), 0.876 (s,
6H; Me-27, Me-28), 0.983 (s, 3H; Me-29), 0.990 (s, 3H, Me-20), 1.12 (m;
H-12), 1.13 (bd, J=12.8 Hz; H-5), 1.31 (m; H-1), 1.32 (m; H-15), 1.38
(m; H-18), 1.42 (m, 2H; H-12, H-15), 1.46 (m; H-6), 1.50 (m; H-21), 1.53
(m, 2H; H-16), 1.56 (m; H-12), 1.599 (s, 3H; Me-26), 1.62 (m, 2H; H-2,
H-12), 1.68 (m; H-6), 1.686 (s, 3H; Me-25), 1.70 (m; H-2), 1.76 (m; H-1),
1.80 (m; H-22), 1.95 (m; H-7), 1.98 (m; H-11), 2.03 (m; H-11), 2.04 (m;
H-22), 2.050 (s, 3H; acetyl Me), 2.10 (m; H-7), 4.50 (dd, J=11.8, 4.5 Hz;
H-3), 5.10 ppm (t, J=7.0 Hz; H-23); 13C NMR (150 MHz, CDCl3): d=
14.22 (q; C-19), 16.50 (q; C-28), 17.66 (q; C-26), 18.46 (t; C-6), 19.28 (q;
C-20), 19.62 (t; C-11), 21.32 (q, acetyl Me), 21.44 (t; C-16), 22.01 (q; C-
29), 24.23 (t; C-2), 25.71 (q; C-25), 25.71 (t; C-7), 26.67 (t; C-17), 26.99
(t; C-22), 27.87 (q; C-27), 29.15 (t; C-12), 30.47 (t; C-21), 30.52 (t; C-15),
35.37 (t; C-1), 37.15 (s; C-14), 37.27 (s; C-10), 37.76 (s; C-4), 40.34 (s; C-
13), 40.43 (d; C-18), 49.98 (d; C-5), 80.96 (d; C-3), 125.3 (d; C-23), 131.0
(s; C-24), 134.3 (s; C-9), 135.2 (s; C-8), 170.8 ppm (s, acetyl CO); the fol-
lowing 13C NMR signals are indistinguishable from each other because
they have very similar chemical shifts: C-10/C-14, C-15/C-21; MS (EI):
m/z (%): 69 (100), 229 (92), 241 (54), 289 (43), 301 (32), 379 (39), 379
(38), 439 (63), 454 (73) [M+]; HRMS (EI): calcd for C31H50O2: 454.3811;
Anaerobic incubation of the synthesized (ꢁ)-8 (40 mg; see Figure S2 in
the Supporting Information) afforded two enzymatic products 14 and 15.
Lipophilic materials were extracted with hexane after saponification. The
products were partially purified by column chromatography on SiO2 gel,
followed by acetylation with Ac2O/ pyridine. The pure acetates of 14
(0.7 mg) and 15 (0.8 mg) were obtained by reversed-phase HPLC (C18
;
MeOH/H2O 100:4).
Anaerobic incubation of (ꢁ)-9 (6 mg) afforded 16 (1.8 mg) as a single
product, which was purified by column chromatography on SiO2 gel and
HPLC (hexane/THF 100:0.5).
Spectroscopic data for products 10 and 12–16:
Product 10: [a]2D0 = + 19.08 (c=0.23 in CHCl3); 1H NMR (400 MHz,
C6D6): d=0.960 (s, 3H; Me-28), 1.024 (s, 3H; Me-29), 1.093 (s, 3H; Me-
18), 1.094 (d, 3H, J=6.4 Hz; Me-20), 1.159 (s, 3H; Me-27), 1.23 (m, 2H;
H-12), 1.24 (m; H-1), 1.32 (m; H-21), 1.47 (m; H-15), 1.53 (m; H-16),
1.58 (m; H-15), 1.59 (m, 3H; H-2, H-6), 1.60 (m; H-13), 1.65 (m; H-21),
1.66 (m; H-19), 1.70 (m; H-1), 1.719 (s, 3H; Me-25), 1.76 (m; H-17), 1.77
(m; H-6), 1.821 (s, 3H; Me-26), 1.95 (m; H-16), 2.12 (m; H-11), 2.14 (m;
H-22), 2.22 (m; H-11), 2.27 (m, 3H; H-7, H-22), 3.17 (dd, J=9.2, 7.2 Hz;
H-3), 5.36 ppm (t, J=6.8 Hz; H-23); 13C NMR (100 MHz, C6D6): d=
15.73 (q; C-28), 17.72 (q; C-26), 18.60 (q; C-20), 18.80 (t; C-6), 20.35 (q;
C-18, C-29)), 22.78 (t; C-12), 23.87 (t; C-11), 25.87 (q; C-25), 26.11 (t; C-
16), 26.73 (t; C-22), 26.94 (t; C-7), 28.25 (q; C-27), 28.44 (t; C-2), 34.06
(t; C-21), 35.63 (d; C-19), 35.88 (t; C-15), 36.24 (t; C-1), 37.38 (s; C-10),
39.15 (s; C-4), 45.61 (s; C-14), 45.74 (d; C-17), 48.02 (d; C-13), 50.89 (d;
C-5), 78.44 (d; C-3), 125.6 (d; C-23), 130.9 (s; C-25), 136.2 (s; C-9),
136.9 ppm (s; C-8); MS (EI): m/z (%): 69 (42), 95 (18), 109 (18), 135
(16), 175 (16), 379 (68), 397 (100), 412 (50) [M+]; HRMS (EI): calcd for
C29H48O: 412.3705; found: 412.3703.
1
found: 454.3805; assignments of the H and 13C NMR data in [D6]acetone
are shown in the Supporting Information.
Product 15-acetate: [a]2D0 =67.98 (c=0.038 in CHCl3); 1H NMR
(600 MHz, CDCl3): d=0.872 (s, 6H; Me-27, Me-28), 0.908 (s, 3H; Me-
19), 0.95 (m; H-12), 0.978 (s, 3H; Me-29), 0.991 (s, 3H; Me-20), 1.11 (dd,
J=12.6, 2.1 Hz; H-5), 1.15 (m, 1H; H-18), 1.26 (m; H-21), 1.27 (m; H-
15), 1.33 (m; H-1), 1.42 (m; H-17), 1.44 (m; H-16), 1.46 (m; H-6), 1.55
(m; H-15), 1.56 (m; H-16), 1.592 (s, 3H; Me-26), 1.62 (m, 2H; H-2, H-
21), 1.65 (m; H-6), 1.683 (s, 3H; Me-25), 1.71 (m; H-2), 1.75 (m; H-22),
1.76 (m; H-1), 1.81 (m; H-17), 1.95 (m; H-7), 2.00 (m, 2H; H-11, H-22),
Product 12: [a]2D0 = + 22.9 (c=0.2 in CHCl3); 1H NMR (600 MHz,
CHCl3): d=0.590 (s, 3H; Me-18), 0.810 (s, 3H; Me29), 0.937 (d, 3H, J=
&
8
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Substrate Folding in Lanosterol Biosynthesis
FULL PAPER
2.048 (s, 3H; acetyl Me), 2.07 (m; H-7), 2.08 (m, 2H; H-11, H-12), 4.50
(dd, J=11.8, 4.4 Hz; H-3), 5.10 ppm (t, J=7.0 Hz; H-23); 13C NMR
(150 MHz, CDCl3): d=16.49 (q; C-28), 17.67 (q; C-26), 18.19 (t; C-6),
18.31 (t; C-16), 19.29 (q; C-20), 20.14 (t; C-11), 21.32 (q, acetyl Me),
23.15 (q; C-19), 23.79 (q; C-29), 24.24 (t; C-2), 24.40 (t; C-17), 24.65 (t;
C-7), 25.71 (q; C-25), 27.87 (q; C-27), 29.28 (t; C-12), 29.37 (t; C-22),
29.92 (t; C-21), 31.01 (t; C-15), 35.55 (t; C-1), 37.18 (s; C-14), 37.24 (s; C-
10), 37.77 (s; C-4), 39.54 (s; C-13), 45.24 (d; C-18), 49.96 (d; C-5), 80.93
(d; C-3), 125.1 (d; C-23), 131.2 (s; C-24), 133.9 (s; C-9), 136.1 (s; C-8),
171.0 ppm (s, acetyl CO); the following 13C NMR signals are indistin-
guishable from each other because they have very similar chemical shifts:
C-10/C-14, C-12/C-22, C-6/C-16; MS (EI): m/z (%): 69 (100), 229 (84),
241 (48), 289 (28), 301 (23), 379 (39), 379 (25), 439 (36), 454 (68) [M+];
HRMS (EI): calcd for C31H50O2: 454.3811; found: 454.3817; assignments
of the 1H and 13C NMR data in C6D6 are shown in Figure S9 in the Sup-
porting Information.
Product 16: [a]2D0 =18.62 (c=0.188 in CHCl3); 1H NMR (400 MHz,
CDCl3): d=0.687 (s, 3H; Me-18), 0.795 (t, J=7.2 Hz; Me-31), 0.808 (3H;
s, Me-29), 0.882 (s, 3H; Me-30), 0.978 (s, 3H; Me-19), 0.998 (s, 3H; Me-
28), 1.16 (m; H-15), 1.20 (m; H-22), 1.22 (m; H-1), 1.36 (m; H-22, 1.37
(m; H-20), 1.38 (m; H-21), 1.50 (m; H-6), 1.53 (m, 2H; H-16, H-21), 1.56
(m; H-2), 1.58 (m; H-15), 1.581 (brd, J=11.8, 2.0 Hz; H-5), 1.605 (s, 3H;
Me-27), 1.63 (m; H-2), 1.64 (m; H-12), 1.67 (m; H-6), 1.682 (s, 3H; Me-
26), 1.71 (m; H-17), 1.73 (m; H-1), 1.77 (m; H-12), 1.82 (m; H-23), 1.97
(brm, 2H; H-23), 2.01 a(m, 2H; H-11), 2.03 (m, 2H; H-7), 2.07 (m; H-
16), 3.23 (dd, J=11.6, 4.4 Hz; H-3), 5.11 ppm (t, J=6.8 Hz; H-
24);13C NMR (100 MHz, CDCl3): d=9.15 (q; C-31), 15.42 (q; C-29),
15.67 (q; C-18), 17.62 (q; C-27), 18.26 (t; C-6), 19.15 (q; C-19), 21.05 (t;
C-11), 22.53 (t; C-21), 24.35 (t; C-23), 24.43 (q; C-30), 25.74 (q; C-26),
26.49 (t; C-7), 27.85 (t; C-16), 27.93 (t; C-2), 27.96 (q; C-28), 30.64 (t; C-
22), 30.69 (t; C-12), 30.79 (t; C-15), 35.60 (t; C-1), 37.01 (s; C-10), 38.89
(s; C-4), 40.86 (d; C-20), 44.46 (s; C-13), 45.95 (d; C-17), 49.87 (s; C-14),
50.38 (d; C-5), 78.98 (d; C-3), 125.3 (d; C-24), 130.8 (s; C-25), 134.3 (s;
C-9), 134.4 ppm (s; C-8); the following 13C NMR signals are indistin-
guishable from each other because they have very similar chemical shifts:
C-2/C-16, C-15/C-22; MS (EI): m/z (%): 69 (100), 81 (32), 95 (46), 123
(30), 407 (47), 425 (82), 440 (33) [M+]; HRMS (EI): calcd for C31H52O:
440.4018; found: 440.4018.
Acknowledgements
This work was supported in part by Grant-in-Aid for Scientific Research
from the Ministry of Education, Culture, Sports, Science and Technology
(Japan).
[1] Reviews for the polycyclization of triterpene cyclases: a) T. Hoshino,
T. Sato, Chem. Commun. 2002, 291–301; b) K. U. Wendt, G. E.
Schulz, E. J. Corey, D. R. Liu, Angew. Chem. 2000, 112, 2930–2952;
Angew. Chem. Int. Ed. 2000, 39, 2812–2833; c) G. Siedenburg, D.
Jendrossek, Appl. Environ. Microbiol. 2011, 77, 3905–3915; d) R. A.
Yoder, J. N. Johnston, Chem. Rev. 2005, 105, 4730–4756; e) R. Xu,
G. C. Fazio, S. P. T. Matsuda, Phytochemistry 2004, 65, 261–291; f) T-
K. Wu, C-H. Chang, Y-T. Liu, T-T. Wang, Chem. Rec. 2008, 8, 302–
305; g) I. Abe, Nat. Prod. Rep. 2007, 24, 1311–1331.
[2] A. Eschenmoser, L. Ruzicka, O. Jeger, D. Arigoni, Helv. Chim. Acta
1955, 38, 1890–1940.
[3] J. W. Cornforth, Angew. Chem. Int. Ed. Engl. 1968, 7, 903–911.
[4] E. J. Corey, S. C. Virgil, J. Am. Chem. Soc. 1991, 113, 4025–4026.
[5] E. J. Corey, S. C. Virgil, S. Sarshar, J. Am. Chem. Soc. 1991, 113,
8171–8172.
[6] A. Eschenmoser, D. Arigoni, Helv. Chim. Acta 2005, 88, 3011–3050.
[7] J-L. Giner, S. Rocchetti, S. Neunlist, M. Rohmer, D. Arigoni, Chem.
Commun. 2005, 3089–3091.
[8] I. Abe, M. Rohmer, G. D. Prestwich, Chem. Rev. 1993, 93, 2189–
2206.
[9] E. J. Corey, A. Krief, H. Yamamoto, J. Am. Chem. Soc. 1971, 93,
1493–1494.
[10] E. E. van Tamelen, R. P. Hanzlik, K. B. Sharpless, R. B. Clayton,
W. J. Richter, J. Am. Chem. Soc. 1968, 90, 3284–3286.
[11] E. J. Corey, S. C. Virgil, D. R. Liu, S. Sarshar, J. Am. Chem. Soc.
1992, 114, 1524–1525.
[12] R. B. Clayton, E. E. van Tamelen, R. G. Nadeau, J. Am. Chem. Soc.
1968, 90, 820–821.
[13] E. J. Corey, K. Lin, M. Jautelat, J. Am. Chem. Soc. 1968, 90, 2724–
2726.
[14] A. G. Byskov, C. Y. Andersen, L. Leonardsen, M. Baltsen, J. Exp.
Zool. Part B 1999, 285, 237–242.
Kinetic analyses kinact and Ki of time-dependent inhibition by 5, 7, and 8:
The partially purified lanosterol synthase described above (1.8 mL) was
used. Noroxidosqualenes 5, 7, and 8 were added to the enzyme solution,
which was adjusted to a final volume of 2.0 mL with Tris-HCl buffer
(0.1m, pH 7.4): 30 mm, 60 mm, and 121 mm for 5 and 8; 60 mm, 121 mm, and
242 mm for 7. Each of the solutions was incubated at 378C for the inter-
vals of 0, 30, 60, 120, and 300 min. From each of the reaction tubes, an
aliquot of the preincubated mixture (200 mL) was added to a tube con-
taining (ꢁ)-1 (200 mg) and then further incubation was conducted for 8 h.
To quench the reaction, KOH in MeOH (15%, 1 mL) was added and the
solution was heated at 808C for 30 min. The hexane extract was subjected
to GC analysis to estimate the amount of lanosterol produced. The kinet-
ic parameters of kinact and Ki were determined using Kits–Wilson plots.
Ethyl-substituted 9 did not undergo any time-dependent inhibition.
[15] T. Hoshino, E. Ishibashi, K. Kaneko, J. Chem. Soc. Chem. Commun.
1995, 2401–2402.
[16] T. Hoshino, Y. Sakai, Tetrahedron Lett. 2001, 42, 7319–7323.
[17] T. Hoshino, Y. Sakai, Chem. Commun. 1998, 1591–1592.
[18] R. Thoma, T. Schulz-Gasch, B. DꢂArcy, J. Benz, J. Aebi, H. Dehm-
low, M. Hennig, M. Stihle, A. Ruf, Nature 2004, 432, 118–122.
[19] T.-K. Wu, C.-H. Chang, H.-Y. Wen, Y.-T. Liu, W.-H. Li, T.-T. Wang,
W.-S. Shie, Org. Lett. 2010, 12, 500–503.
[20] T.-K. Wu, Y.-C. Chang, Y.-T. Liu, C.-H. Chang, H.-Y. Wen, W.-H.
Li, W.-S. Shi, Org. Biomol. Chem. 2011, 9, 1092–1097.
[21] T.-K. Wu, T.-T. Wang, C.-H. Chang, Y.-T. Liu, W.-S. Shi, Org. Lett.
2008, 10, 4959–4962.
[22] T. Hoshino, M. Kouda, T. Abe, S. Ohashi, Biosci. Biotechnol. Bio-
chem. 1999, 63, 2038–2041.
[23] N. Morikubo, Y. Fukuda, K. Ohtake, N. Shinya, D. Kiga, K. Saka-
moto, M. Asanuma, H. Hirota, S. Yokoyama, T. Hoshino, J. Am.
Chem. Soc. 2006, 128, 13184–13194.
Received: May 21, 2012
Revised: June 26, 2012
Published online: && &&, 0000
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

T. Hoshino et al.
Enzyme Catalysis
T. Hoshino,* A. Chiba,
N. Abe ......................... &&&& — &&&&
Lanosterol Biosynthesis: The Critical
Role of the Methyl-29 Group of 2,3-
Oxidosqualene for the Correct Folding
of this Substrate and for the Construc-
tion of the Five-Membered D Ring
Ringing in the changes: The incubation
of 1 with porcine-liver cyclase afforded
new nortriterpenes 2 and 3 with
6,6,6,6-fused tetracyclic skeletons,
which were produced by chair-boat-
chair-chair and chair-chair-boat-chair
conformations, respectively (see
scheme), thus indicating that the 29-
methyl group is critical to the correct
folding of oxidosqualene and to the
formation of the five-membered
D ring for lanosterol biosynthesis.
&
10
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!