Enzymatic cyclization reactions of geraniol, farnesol and geranylgeraniol, and those of truncated squalene analogs having C20 and C25 by recombinant squalene cyclase

Article abstract of DOI:10.1039/b407001a

The substrate specificity of squalene-hopene cyclase was investigated using the C₁₀-C₂₅ analogs including naturally occurring substances, e.g. geraniol (C₁₀), farnesol (C₁₅) and geranylgeraniol (C₂₀). No cyclization occurred for geraniol, but a significantly high conversion ratio (64%) was observed for farnesol, yielding the cyclic sesquiterpenes consisting of 6/6-fused bicyclic ring systems. Among them, an attractive compound having C₃₀ was produced, in the structure of which acyclic the farnesol unit is linked to the bicyclic skeleton through ether linkage. Conversion of geranylgeraniol was low (ca. 12%). The squalene analogs having C₂₀ and C₂₅ also were cyclized in yields of ca. 33-36%, but the analogs having the methyl group at C(7) and/or at C(11) underwent no cyclization; the large steric bulk size of C(7)-Me and/or C(11)-Me, which is arranged in α-disposition for all the pre-chair conformation, would have interacted repulsively with the cyclase recognition site near to the C(7) and/or C(11), resulting in no construction of the all-chair conformation inside the reaction cavity. A relatively low yield of geranylgeraniol indicated that a less bulky hydrogen atom must be located at C(14) for the efficient polycyclization reaction. The squalene cyclase shows remarkably broad substrate specificity to accept the truncated analogs having carbon-chain lengths of C₁₅-C₂₅ in addition to C ₃₀.

Full text of DOI:10.1039/b407001a

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Enzymatic cyclization reactions of geraniol, farnesol and
geranylgeraniol, and those of truncated squalene analogs having
C20 and C25 by recombinant squalene cyclase†
Tsutomu Hoshino,* Yuko Kumai, Isao Kudo, Shin-ichi Nakano and Shumi Ohashi
Department of Applied Biological Chemistry, Faculty of Agriculture,
Graduate School of Science and Technology, Niigata University, Ikarashi, Niigata, Japan,
950-2181. E-mail: hoshitsu@agr.niigata-u.ac.jp; Fax: +81 25 262 6854; Tel: +81 25 262 6637
Received 10th May 2004, Accepted 21st July 2004
First published as an Advance Article on the web 23rd August 2004
The substrate specificity of squalene–hopene cyclase was investigated using the C10–C25 analogs including naturally
occurring substances, e.g. geraniol (C10), farnesol (C15) and geranylgeraniol (C20). No cyclization occurred for geraniol,
but a significantly high conversion ratio (64%) was observed for farnesol, yielding the cyclic sesquiterpenes consisting of
6/6-fused bicyclic ring systems. Among them, an attractive compound having C30 was produced, in the structure of which
acyclic the farnesol unit is linked to the bicyclic skeleton through ether linkage. Conversion of geranylgeraniol was low
(ca. 12%). The squalene analogs having C20 and C25 also were cyclized in yields of ca. 33–36%, but the analogs having the
methyl group at C(7) and/or at C(11) underwent no cyclization; the large steric bulk size of C(7)–Me and/or C(11)–Me,
which is arranged in -disposition for all the pre-chair conformation, would have interacted repulsively with the cyclase
recognition site near to the C(7) and/or C(11), resulting in no construction of the all-chair conformation inside the reaction
cavity. A relatively low yield of geranylgeraniol indicated that a less bulky hydrogen atom must be located at C(14) for
the efficient polycyclization reaction. The squalene cyclase shows remarkably broad substrate specificity to accept the
truncated analogs having carbon-chain lengths of C15–C25 in addition to C30.
Introduction
The polycyclization of squalene 1 into pentacyclic hopene 2 and
hopanol 3 is one of the most complicated biochemical reactions
(Scheme 1)1–4 which is mediated by squalene–hopene cyclases
(SHCs) [EC 5.4.99-] from prokaryotic species. The polycyclization
reaction proceeds under precise enzymatic control to form five new
rings and nine new chiral centers.1 This polycyclization mechanism
is analogous to that catalyzed by eukaryotic oxidosqualene cyclases
(OSCs) which lead to lanosterol and numerous plant triterpenes from
(3S)-2,3-oxidosqualene.2–4 Investigations of site-directed mutants
1.2
and substrate analogs
have led to the proposal that this poly-
cyclization consists of eight reaction steps as shown in Scheme 1: (1)
First, cyclization to form A-ring 4 by proton attack on the terminal
double bond,5 (2) second, ring closure to give the B-ring (6/6-fused
A/B ring system 5),6 (3) third, cyclization to yield a five-membered
C-ring (6/6/5-fusedA/B/C-tricyclic ring system 6) by Markovnikov
closure, (4) which then undergoes ring expansion to form the six-
membered C-ring (6/6/6-fused tricyclic ring system 7),7 (5) fifth,
cyclization to give the thermodynamically favored five-membered
D-ring (6/6/6/5-fused A/B/C/D ring system 8, 17-epi-dammarenyl
cation),8 (6) followed by the second ring enlargement process
to form the six-membered D-ring (6/6/6/6-fused A/B/C/D-ring
system, prohopanyl cation 9),9,10 (7) the last ring closure process to
construct the 6/6/6/6/5-fused A/B/C/D/E-ring system (10, hopanyl
cation),10 and (8) the final deprotonation reaction to introduce the
double bond between C-22 and C-29. This deprotonation reaction
occurs exclusively from the (23Z)-methyl group.1,11,12 The polycycl-
ization reaction is featured by the folding of 1 into an all-pre-chair
conformation during the multiple reaction steps.
The substrate specificity of the SHC from A. acidocaldarius is
remarkably broad. (3R)- and (3S)-oxidosqualenes were converted
into 3- and 3-hydroxyhopane skeletons, respectively.12,13 The
norsqualenes lacking a methyl group at the terminal position
afforded a 6/6/6/6/6-fused pentacyclic ring system (tetrahymanol
skeleton) in high yields.12,14 The C(10)–norsqualene was also
Scheme 1 Cyclization pathway of squalene 1 into hopene 2 and hopanol
3 by squalene–hopene cyclase (SHC).
converted in a high yield into the unprecedented carbocyclic
skeleton(s) having 6/5 + 5/5 + (6) ring system(s).15 On the other
hand, the norsqualene lacking the C(15)–Me was converted into
hopane and isohopane skeletons.15 The bisnorsqualene lacking two
methyl groups at C(23) was also efficiently converted to the products
having 6/6/6/5-fused tetracyclic and 6/6/6/6/6-fused pentacyclic ring
† Electronic supplementary information (ESI) available: additional char-
acterization data for products 21–30. See http://www.rsc.org/suppdata/ob/
b4/b407001a/
2 6 5 0
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 6 5 0 – 2 6 5 7
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
systems.12 The terminal methyl groups have a crucial role for the
stereochemical control during the polycyclization cascade lead-
ing to the hopane skeleton.12 The C(6)–norsqualene was cyclized
into the hopane skeleton in significantly high yields.16 The regio-
isomers of 1 with respect to the methyl position were also examined
whether they can be accepted as the substrates.16 The presence of a
methyl group at C(7) or C(11) gave no cyclization, but the analog
possessing a methyl group at C(18) instead of C(19) of 1 underwent
the polycyclization in a quantitative yield.16 The truncated C27- and
C22-analogs having an alcoholic group also were efficiently cyclized
into the compounds with 6/6/6/5 + tetrahydrofuran (THF) ring1,8 and
6/6/5 + THF ring,1,7 respectively, indicating that the carbocationic
intermediates 6 and 8 are involved in the polycyclization reaction
of 1. The dihydroxysqualenes, e.g. 6,7-erythro-dihydroxysqualene,
were converted into the skeleta having 6/6/5 + THF ring systems.1
These substrate analogs were used to trap the transient carbocationic
intermediates generated during the polycyclization reaction. The
cationic intermediates 4 and 5 also have been successfully trapped
by other squalene analogs having the highly nucleophilic hydroxyl
group(s).17,18 The enzymatic conversions of the substrate analogs by
the SHC were also reported by other workers.19–24 The C31 analogs
having a methylidene appendage were examined to show whether
they can be converted to the enzymic product(s) or irreversible inhib-
itors for the SHC activity.19,22,23 The elongated C35-analog also was
accepted as the substrate of the SHC, affording the product having a
6/6/6/6/6/5-fused hexacyclic ring system in a yield of 10%.24
The broad substrate specificity of the prokaryotic SHC has
prompted us to examine whether the SHC can accept the truncated
analogs of 1. Little is known about the cyclization reaction of the
truncated analogs having C10–C25. The following three acyclic
isoprenoid compounds were selected: geraniol 11 (C10), farnesol
12 (C15) and geranylgeraniol 13 (C20), which are physiologically
important substances frequently encountered in nature. In addition,
the chemically synthesized analogs 14–16 having C20 and C25 were
also examined.
Fig. 1 Structures of compounds employed in this study. The methyl posi-
tion of 14 is different from that of 15. Analog 14 has a methyl group at C-11,
while 15 at C-10.
Structures of these compounds are illustrated in Fig. 1. Very
interestingly, farnesol 12 was cyclized in a relatively high conver-
sion yield (64%), but geraniol 11 underwent no cyclization reaction.
Geranylgeraniol 13 had a low conversion ratio (12%). Analog 14
had no conversion, but 15 (a regio-isomer with respect to the methyl
position) underwent the polycyclization reaction in a yield of 33%,
despite 14 and 15 having an identical carbon-length (C20). This find-
ing suggested that the methyl position has a great influence upon the
polycyclization cascade. Analog 16 having a longer chain length of
C25 also could be cyclized (36%). The present investigations give
deeper insight into the polycyclization mechanism of 1, especially
regarding the importance of the methyl position and minimum
carbon-chain length required for the occurrence of the enzymatic
cyclization reaction. Herein we describe the experimental results
and discuss the cyclization mechanisms of the truncated analogs.
Scheme 2 Synthetic scheme of truncated analogs 14–16. Reagents and
conditions: (i) PBr3/ THF, 0 °C, (ii) CuI/pyrrolidine/n-BuLi/ Et2O, −38 °C.
of E. coli clone encoding the native SHC from Alicyclobacillus
acidocaldarius.5,13 Identical incubation conditions were employed to
compare the quantities and distribution pattern of enzymic products
obtained from each of the substrate analogs. Incubation conditions
were as follows: substrate analog 3 mol, the cell-free extract 1.5 ml
as the enzyme source, Triton X-100 20 mg, optimal pH 6.0, optimal
temperature 60 °C, incubation time 16 h, total volume 5 ml. To the
reaction mixture, was added 5% KOH/MeOH and the products were
extracted with hexane. Triton X-100 included in the hexane-extracts
was removed with a short SiO2 column chromatography eluting
with a mixed solvent of hexane/EtOAc (100:20). Fig. 2G shows
the product distribution pattern obtained by incubating 1. Only a
small amount of substrate 1 remained in the incubation mixture,
indicating almost full conversion of 1 into 2 and 3. No reaction
was observed for 11 having C10 (Fig. 2A), the conversion being
negligible if present. In contrast, 12 with C15 was converted in high
yields to four products 21–24, as shown in Fig. 2B. As described
below, a strongly high polar product 23 (a diol) was found during
the isolation process of the enzymic products by a SiO2 column
chromatography. Thus, the incubation mixture was extracted with
hexane/EtOAc (50:50) instead of hexane usually used. The extract
was subjected to the passage through a short SiO2 column by eluting
with hexane/EtOAc (100:35), but Triton X-100 was in part detected
on the gas chromatogram (asterisk symbol in Fig. 2B). Product 24
had a longer retention time. Geranylgeraniol 13 afforded a single
product 25 in a yield of 12% (Fig. 2C). No conversion was found
for 14 having C20-chain length (Fig. 2D), but the conversion ratio of
15 having the same carbon number as 14 was relatively high (28%
for 26 and 5% for 27, total 33%, Fig. 2E). Analog 15 has a methyl
group at position of C-10, while 14 does a methyl group at C-11.
Results
Syntheses of truncated analogs 14–16
The synthetic methods are shown in Scheme 2. Farnesyl bromide
17 and 1-bromo-3-methyl-but-2-ene 19 were prepared from the
corresponding farnesol 12 and 3-methyl-but-2-en-1-ol 18, respec-
tively, by a treatment of PBr3 in THF.16 The bromides 17 and 19,
thus prepared, were subjected to the allyl coupling reaction by using
pyrrolidine/n-BuLi/CuI,25,26 leading to a mixture of the desired 15
and 1 as major products. To obtain the pure 15, AgNO3–SiO2
column chromatography was carried out to remove 1. To prepare 14
and 16, a mixture of 17 and geranyl bromide 20 in Et2O was added
into the metal complex of pyrrolidine/n-BuLi/CuI according to the
same method as described above. Each of the products was isolated
by 5% AgNO3–SiO2 column chromatography in a pure state.
GC analyses of enzymatic products from analogs 11–16
Fig. 2 shows the gas chromatograms of the incubation mixtures
prepared by incubating 1 and 11–16 with the cell-free homogenates
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 6 5 0 – 2 6 5 7
2 6 5 1

View Article Online
These compounds are the regio-isomers with respect to the methyl
position. This finding indicates that the occurrence of the polycycl-
ization reaction depends on the position of the branching methyl
group on the linear backbone. Substrate analog 16 having a longer
carbon chain (C25) was converted into 28–30 in a total yield of 36%
(6% for 28, 4% for 29 and 26% for 30, Fig. 2F)
determined by the detailed NMR analyses including DEPT, COSY
45, HOHAHA, NOESY, HMQC and HMBC pulse sequences.
Product 22 had one allylic methyl group (H 1.97, br s, Me-15)
and three methyl groups at higher fields (H 0.949, 6H, s, Me-12
and Me-14; H 0.966, 3H, s, Me-13). The clear cross peaks of Me-
14 and Me-15 with C-9 (C 57.45, d) were observed in the HMBC
spectrum. Furthermore, strong HMBC cross peaks of H-11 (H 3.75,
m; H 3.63, m) with C-9 and C-10 (C 36.19, s) were also found. An
olefinic proton (H 5.62, br s) had HMBC cross peaks with C-5 (C
50.12, d) and C-9. These NMR data indicated a bicyclic skeleton for
22 (Fig. 3). A strong NOE between H-5 (H 1.25, m) and H-9 (1H,
H 1.86, very broad) suggested an -orientation of H-9. Product 21
had a vinyl proton (H 4.99, s; 4.75, s; C 106.6, t), which had clear
HMBC correlations with C-9 (C 59.32, d) and C-7 (C 38.12, t). A
clear spin–spin coupling between H-9 (H 1.94, m) and H-11 (H
3.86, dd, J 10.8, 3.4; H 3.78, dd, J 10.8, 10.8) was observed in the
COSY 45 spectrum. These NMR data clarified the double bond
position. Product 23 was a very highly polar compound, as described
above, and involved no double bond in the 13C NMR. In addition
to hydroxymethyl group (2H, H 3.97, d, J 6.9, H-11; C 60.9, t), a
tertiary alcoholic carbon (C 74.5, s, C-8) was found, thus indicat-
ing that 23 was a diol product. A clear HMBC cross peak between
Me-15 (H 1.38, s) and C-8 suggested that the hydroxyl group was
attached to C-8. A definitive NOE between Me-15 and Me-14 (H
0.70, s) proved the -arrangement for Me-15. Product 24 had an
attractive structure as shown in Fig. 3. The 13C NMR spectrum
showed the presence of C30, but EIMS did not give the molecular
ion. HRMS (FAB, positive, glycerol) showed that the molecular
composition was C30H53O2 (M + H+, observed, 445.4052 requiring
445.4046). The detailed NMR analyses indicated the involvement
of the bicyclic skeleton of 23 in 24. The 1H NMR spectrum showed
the presence of three olefinic protons (H 5.56, br t, J 6.5; H 5.35, t,
J 6.8; H 5.37, t, J 6.8) and four allyl methyl groups (H 1.70, 6H, s;
H 1.72, 3H, s; H 1.81, 3H, s), suggesting that the acyclic farnesyl
moiety is involved in 24. The linkage of the farnesyl residue and the
bicyclic skeleton of 23 was established by the clear HMBC cross
peak between H-16 (H 3.99, 2H, dd, J 6.1, 3.8) and C-8 (C 80.13,
s). Thus, the structure of 24 was determined as shown in Fig. 3,
which is composed of two farnesol units. A nucleophilic attack of
the hydroxyl group of the linear farnesol molecule occurred toward
the C-8 cation of the bicyclic skeleton that was in advance generated
by the cyclization reaction of the alternative farnesol molecule.
Fig. 2 Gas chromatograms of the reaction mixtures obtained by incubat-
ing 11–16 and 1 with the wild-type SHC. A: geraniol 11 (C10). B: farnesol
12 (C15). C: geranylgeraniol 13 (C20). D: analog 14 (C20). E: analog 15 (C20).
F: analog 16 (C25). G: squalene 1 (C30). Column temperatures: A: 200 °C;
B: 150 °C for 0–40 min, and elevated at a rate 5 °C min−1 to 290 °C for
40–68 min, and 68–100 min at 290 °C; C, D and E: 250 °C; F and G: 270 °C,
N2 carrier pressure being 1.0 kg cm−2 except for B, which was conducted at
the pressure of 0.5 kg cm−2. In case of B, Triton X-100 remained in part due
to a high polarity of the elution solvent used, the symbol * shows Triton
X-100 and impurities.
Enzymatic product 25 from geranylgeraniol 13
Sixty mg of geranylgeraniol 13 was incubated with the cell-free
extracts (200 mg) under the optimum catalytic conditions. A SiO2
column chromatography eluting with hexane/EtOAc (100:5) gave
25 in a pure state (6.3 mg). Product 25 had one double bond (H
5.63, br s, H-12; C 123.4, d, C-12; C 133.5, s, C-13). The clear
HMBC cross peaks of H-15 (H 3.75, dd, J 10.8, 3.2; H 3.62, dd, J
10.8, 5.6) were found for C-14 (C 58.04, d), C-8 (C 36.34, s) and
C-13. The double bond position was further confirmed by the clear
HMBC cross peaks of the allyl methyl group (H 1.99, s, Me-20) for
C-12 and C-13. The detailed 2D NMR analyses showed that 25 had
a 6/6/6-fused tricyclic ring system as shown in Fig. 3.
Isolation and structure determination of enzymatic products
21–24 from farnesol 12
To obtain pure products 21–24, a large-scale incubation was
conducted by using 100 mg of 12 and 200 ml of the cell-free
homogenates from 4L-culture of the cloned E. coli. The lipophilic
extract from the incubation mixture was subjected to a SiO2 column
chromatography with a gradient elution using a mixed solvent of
hexane and EtOAc (100:3 ~ 100:30). Products 21, 22 and 24 were
inseparable, but 23 was isolated in a pure state (6.3 mg). A mixture
of 21, 22 and 24 was again subjected to a column chromatography
over SiO2 using hexane/EtOAc (100:1.0), giving a pure sample of
24 (8.7 mg), but the separation of 21 and 22 failed. A rechromato-
graphy using hexane/EtOAc (100:0.5) successfully afforded pure
21 and 22 in the isolation yields of 4.7 mg and 2.0 mg, respectively.
The Rf value of product 24 on SiO2 TLC was 0.40, while those of
21 and 22 were 0.37–0.38 (nearly identical), by developing with
hexane/EtOAc (100:20). On the other hand, a high polar product 23
had the Rf value of 0.38 for the solvent of hexane/EtOAc (50:50).
The product distribution ratio was estimated by the GC analysis
to be as follows: 3, 7, 45, 9 and 35% for 21–24 and the recovered
12, respectively (Fig. 2B). Structures of these products were
Enzymatic products 26 and 27 from analog 15 having C20
Analog 15 (82 mg) was incubated for 20 h with the cell-free extracts
(250 ml) at the optimum catalytic conditions. After removing the
detergent from the hexane-extract, a careful column chromato-
graphy over SiO2 was carried out by eluting with hexane/EtOAc
(100:3), affording a pure 26 (ca. 3 mg), a mixture of 27 and 26
(ca. 3 mg), and pure 27 (ca. 1 mg) in this elution order, although
complete separation was unsuccessful. The Rf values on SiO2 TLC
by developing with hexane/EtOAc (100:15) were as follows: 0.93,
0.32 and 0.31 for 15, 26 and 27, respectively, suggesting that prod-
ucts 26 and 27 were highly polar. This was further supported by the
finding of the tertiary alcoholic carbon (C 73.88, s, for C-14 of 26;
C 72.55, s, for that of 27) in the 13C NMR spectrum. The methyl
2 6 5 2
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 6 5 0 – 2 6 5 7

View Article Online
s, C-17; C 138.0, s, C-13) was found in the 13C NMR spectrum.
Apparent HMBC cross peaks of H-18/C-13, Me-23 (H 1.30, 3H,
s)/C-13 and Me-24(25)/C-17 clarified the double bond position.
The detailed NMR analyses established that 28 had a tetracyclic
6/6/6/5-fused ring system. Product 29 also consisted of a tetracyclic
6/6/6/5-fused ring system and had the trisubstituted double bond
(H 5.47, m, H-12, C 118.1, d, C-12; C 145.7, s, C-13). The double
bond position was determined to be at C12–C13 by the strong
HMBC correlations of Me-23, H-12 and H-18 for C-13. H-17 was
shown to have -orientation, since no NOE between H-17 and
Me-23 was observed. The 6/6/6/5-fused tetracyclic ring skeleton
of 30 was determined by the detailed 2D NMR analyses. The two
methyl group protons (H 1.21, s and H 1.27, s; Me-24 and Me-25)
had clear HMBC correlations for the alcoholic carbon (C 72.98, s,
C-18) and for C-17 (C 50.07, d), indicating that isopropyl alcohol
residue was attached to C-17. The HMBC cross peak of Me-23 (H
1.17, 3H, s)/C-13 (C 44.2) and the HMQC spectrum made clear
the assignment of H-13 (H 2.09, m). The apparently observed
NOEs of Me-22 (H 1.08, 3H, s)/H-13 and H-13/H-17 established
-orientation of H-17.
Enzymatic reactions of geraniol 11 and analog 14
These compounds were subjected to the enzyme reaction by using
the amount of the cell-free homogenates five times higher than
that of incubation conditions usually used, but no conversion was
observed as shown in Fig. 2A and D.
Discussion
Cyclization reactions of geraniol 11, farnesol 12 and
geranylgeraniol 13
If 11 (C10) is cyclized, the monocyclic skeleton like 4 can be formed
as shown in Scheme 3, but no conversion occurred (Fig. 2). On
the other hand, a significantly high conversion ratio (64%) was
observed for 12 (C15) to give four products 21–24, which could be
formed from the common bicyclic intermediate 31 (like 5). The
deprotonation from the methyl group at C-8 could give 21 (path a).
The deprotonation of H-7 gave 22 (path b). A nucleophilic attack
of a water molecule to the C8-cation of 31 afforded 23 (path c).
Attack of the hydroxyl group of 12 to the cation 31 led to 24 (path
d). The significantly high conversion of 12, but no conversion of
11, suggested that the carbon number of C15 at least is required for
starting the cyclization reaction. We have previously reported that
the methyl group at C-10 of 1 has a pivotal role in adopting the
normal folding of all chair conformation. The C(10)–norsqualene
underwent the polycyclization reaction through the unusual folding
conformation, affording the novel carbocyclic skeleton(s) having
the 6/5 + 5/5/ + (6) ring system.15 We have proposed that the strong
binding of the C(10)–Me of 1 is involved in the cyclase enzyme
and suggested that the unusual folding conformation would have
occurred due to the trapping of C(15)–Me by the recognition site
inherent to C(10)–Me involved in the cyclase. 12 has a methyl
group at C-10, therefore 12 could have undergone the cyclization
reaction. However, the methyl group is absent in 11, thus leading
to no cyclization. To ascertain this idea, further studies are neces-
sary. 13 (C20) also underwent the cyclization reaction, but the yield
was low (12%) despite the important C(10)–Me being involved.
Intermediate 32 was produced according to Markovnikov closure.
The proton elimination at C-12 of 32 gave product 25. Why was the
conversion ratio of 13 low, despite the favorable Markovnikov cat-
ion 32 being formed?Abulky methyl group at C-14 of 13 may have
repulsively interacted with the enzyme site located near to C-13 of
32, leading to a decrease in the amount of all the chair conformation
required for the formation of 32 (Scheme 3). In addition, the strong
1,3-diaxial interaction between C(8)–Me and C(13)–Me found in
32 may have decreased the formation of 32. We have previously
demonstrated that squalene analog (C30), which has the methyl
group at C-14 instead of the C(15)–Me of 1, was cyclized into the
6/6/6-fused tricycle in a similar way as 13, but the yield (7%) was
Fig. 3 Enzymic products obtained by incubating truncated analogs 11–16
with the wild-type SHC.
protons at C-19 and C-20 of 26 (H 1.22, 3H, s, and H 1.24, 3H, s)
and those of 27 (H 1.17, 3H, s; H 1.34, 3H, s) had the clear HMBC
cross peaks for C-14, indicating that the hydroxyl groups of 26 and
1
27 were attached to C-14. No olefinic proton resonance in the H
NMR and no sp2 carbon in the 13C NMR spectra suggested that 15
underwent the complete cyclization reaction. The detailed NMR
analyses established that 26 and 27 had a tricyclic 6/6/5-fused ring
system as shown in Fig. 3. The clear NOEs of Me-18 (H 1.16, s)/Me-
19 and Me-18/Me-20 were observed for 27, but no NOE between
them for 26, indicating that H-13 of 27 had -configuration, while
that of 26 possessed -orientation. The finding of a strong NOE
between H-9 (H 1.07, dd, J 13.2, 7.2) and H-13 (H 1.29, t, J 10.1)
for 27 further supported the configuration of 13-H for 27.
Enzymatic products 28–30 from analog 16
Fifty mg of substrate analog 16 was incubated for 20 h with the
cell-free homogenates (200 ml) from 4L-culture of the wild-type
SHC at the optimum catalytic conditions, as described above, and
then the reaction mixture was lyophilized. Triton X-100 included in
the hexane extract was removed by a passage through a short SiO2
column (Hexane/EtOAc = 100:20). A column chromatography
over SiO2 with a gradient elution using hexane to hexane/EtOAc
(100:10) afforded 30 in a pure state. A mixture of 28 and 29 was
subjected to a 5% AgNO3–SiO2 column chromatography eluting
with hexane/EtOAc (100:3), leading to the successful separation
of 28 and 29. Product 28 possessed two doublet methyl groups (H
1.13, 3H, d, J 6.9 and H 1.17, 3H, d, J 6.9 for Me-24 and Me-25),
which had the correlations with H-18 (H 2.85, septet, J 6.9) in
the COSY 45 spectrum. A tetrasubstituted double bond (C 136.4,
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 6 5 0 – 2 6 5 7
2 6 5 3

View Article Online
tions occur in anti-parallel fashion. The 1,2-hydride shift of H-17
to C-18 and the deprotonation of H-13 gave 28 (path d). A series of
1,2-hydride shifts of H-17 and H-13 and the deprotonation reactions
of axial-oriented H-12 afforded 29 (path e). The stereochemistry at
C-17 of intermediate 16e coincides with that of 8 formed during
the polycyclization of 1 (see Scheme 1). The structure of 30 having
17-H unambiguously demonstrated that 30 was produced via 16e,
but not via 16f. The product ratio of 30 via 16e to that of (28 + 29)
via 16f was ca. 72:28; a large portion of 16 was converted into in-
termediate 16e like 8, but a relatively high production of 16f (39%
of 16e) indicated that the stereochemical control is not perfect.
The results from incubation experiments of 15 and 16 may sug-
gest that longer carbon-chain length (>C25) of the substrate, like 1
(C30), is essential to acquiring the complete stereochemical control
during the polycyclization reaction. Previously, Corey and Cheng
reported the polycyclization cascade of the 2,3-epoxide of 15 by
lanosterol synthase; the truncated C20-2,3-oxidosqualene analog
was transformed into a 6/6/5-fused tricyclic A/B/C-ring system
having a chair/boat structure.27 All the products possessed the stereo-
chemistry of 13-H, but no product having 13-H was found in the
incubation mixture. It is well known that the 6/6/5-fused tricyclic
ring system having 13-H, but not 13-H, is formed during the
biosynthetic process of lanosterol through a folding conformation
of chair/boat/chair structure.2,3,28–30 Thus, it could be inferred that
the stereochemical control by lanosterol synthase is more accurate
compared to that by hopene synthase. It should be also noted that
the Markovnikov product having 6/6/5-fused tricyclic ring skeleton
like 15e was not trapped during the polycyclization reaction of the
longer substrate 16. Hess suggested by computational studies that
the C-ring expansion and D-ring formation occur concomitantly
in lanosterol biosynthesis and that the primary role of lanosterol
cyclase in the C-ring expansion and D-ring formation is to hold the
substrate in the proper conformation inside the reaction cavity for
the completion of the cascade.31 It is anticipated that an increase
in the carbon-chain length would give rise to a more precise inter-
action between the cyclase enzyme (SHC) and substrate analogs.
The more correct conformation given by the longer substrate 16
inside the cavity may have led to feasibility of the ring enlargement
process from the five- to the six-membered C ring (6→7) without
violation of Markovnikov’s rule.
In the present study, we further demonstrated that truncated
analogs (C15–C25) also can be accepted as the substrate of the
squalene cyclase. The broad substrate specificity of the squalene
cyclase is remarkable. A surprisingly high conversion of farnesol
12 (64%), but no cyclization of geraniol 11 (C10), unequivocally
demonstrated that at least C15 is necessary for initiating the cycli-
zation reaction. An interesting compound 24 was constructed, in
which two farnesol molecules are linked. The diol 23, enzymati-
cally prepared with a high yield (45%) and in a single step, may
be practically useful as a chiral synthon for the chemical syntheses
of natural terpenoid compounds. Acyclic geranylgeraniol 13 (C20)
was also cyclized to form 6/6/6-fused tricyclic ring system, but the
yield was low. Analog 15 having the same carbon number (C20) as
13 had a higher conversion yield than 13, indicating that the methyl
position of the acyclic isoprenoids is important for the efficient
cyclization reaction. The C(14)–Me of 13 may have repulsively
interacted with the recognition site located near to C-13 of 32.
The importance of the methyl position was further supported by
the findings that conformations 14ab, 15b and 16b underwent no
cyclization, that is, a bulky methyl group at C-7 and/or C-11 can-
not be tolerated by the cyclase enzyme, due to the repulsive inter-
action between -oriented C(7)–Me or C(11)–Me and the cyclase
site(s) located near to their methyl groups. These findings suggest
that the -face domain of the cyclase, which is responsible for the
recognition of the C-7 and C-11 positions, is strictly compact, thus
cannot accept the large -oriented methyl groups. This inference
is in good agreement with our previous findings that the cycliza-
tion reactions of C30-analogs having C(7)–Me or C(11)–Me do
not occur.16 The C25-truncated analog 16 could also undergo the
polycyclization to afford a 6/6/6/5-fused tetracycle via 16a. The
formation of all the enzymic products found in this study accorded
Scheme 3 Cyclization mechanisms of geraniol 11, farnesol 12 and
geranylgeraniol 13 by the SHC.
as low as that of 13.16 Thus, it is inferred that the introduction of the
methyl group at C-14 of the linear backbone of 1 and the analogs
gives a low yield of the cyclization product having the tricyclic
ring system
Cyclization reactions of truncated analogs 14–16
In the previous papers,12,14 we have demonstrated that an isopropyl-
idene moiety is required for initiating the polycyclization reaction.
Analogs 14–16 have this moiety at the both terminal positions,
thus the cyclization reaction can start from the both termini (path
a and b of Scheme 4). Analog 14 is a symmetrical molecule, thus
the identical folding conformation of chair/chair/chair structure
can be adopted during the cyclization reaction as shown in 14a,b
(Scheme 4A). Fig.1D shows that no cyclization reaction occurred
for 14. However, analog 15, the regio-isomer of 14 with respec-
tive to the methyl position, could be cyclized into Markovnikov
cations 15e and 15f via the folding conformations 15c and 15b,
respectively. A water attack to the C-14 cations of 15e and 15f gave
products 26 and 27 having the 6/6/5-fused tricycle in a yield of 28%
and 5%, respectively. The product structures clearly indicated that
the polycyclization occurred exclusively via path a (15a), but not
via path b (15b). No conversion of 14a,b and 15b indicated that
the introduction of a methyl group at C-7 and/or C-11 prevented
the formation of all chair conformation in the reaction cavity. The
C(7)–Me or C(11)–Me of 14a,b and 15b would have repulsively
interacted with the binding pocket(s) near to the C-7 or C-11 in
the folded conformation, suggesting that the -face domain of the
cyclase accepting the region of C-7 and C-11 is compact. Thus,
the steric bulk at C-7 and C-11 of the substrate(s) must be small;
a preferred size is hydrogen atom, and a methyl group is too large.
Products 26 and 27 differ only in the stereochemistry at C-13; 26
has 13-H, whereas 27 has the opposite stereochemistry of 13-H.
As shown in Scheme 1, the true intermediate of the 6/6/5-fused
tricyclic skeleton 6 has the configuration of 13-H during the
polycyclization cascade of 1. The product ratio of 26 to 27 was
ca 85:15; a higher production of 26 than that of 27 coincides with
the cyclization pathway of 1, but a relatively high production of
27 (ca, 18% of 26) having the opposite stereochemistry cannot be
neglected, suggesting that a further long carbon-chain length may
be required for the complete stereochemical control during the
polycyclization reaction. Analog 16 having a longer chain (C25)
underwent the polycyclization reaction via path a (16a), but not via
path b (16b). This finding also indicated that the methyl position is
important for the occurrence of polycyclization reaction; introduc-
tion of methyl group at C-11 prevented the polycyclization reac-
tion, possibly due to the repulsive interaction between C(11)–Me
and the cyclase site located near to the C(11)–Me in the folded
all chair conformation. Conformation 16a led to a Markovnikov
cation 16e having 17-H via the folding of conformation 16c. This
corresponds to 8 formed during the polycyclization of 1. A water
attack to the C-18 cation of 16e led to the production of 30 (path c).
Products 28 and 29 could be produced from intermediate 16f hav-
ing 17-H through the folding conformation 16d. Regarding the
formation of 28 and 29, 16f may be more preferable than 16e from
the point of view that the deprotonation and 1,2-hydride shift reac-
2 6 5 4
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 6 5 0 – 2 6 5 7

View Article Online
solvent peak H 7.280 and C128.0 ppm as the internal reference
1
for H- and 13C NMR spectra, respectively. Some synthetic inter-
mediates were measured in CDCl3. The chemical shifts in CDCl3
solution were given according to the internal solvent peaks of
H 7.26 and C 77.0 ppm. The coupling constants J are given in
Hz. GC analyses were done on a Shimadzu GC-8A chromato-
graph equipped with a flame ionization detector (DB-1 capillary
column (0.53 mm × 30 m). GC-MS spectra were on a JEOL SX
100 spectrometer under electronic impact at 70 eV with a DB-1
capillary column (0.32 mm × 30 m), the oven temperature being
elevated according to the methods described in the legend to Fig. 2.
HR-EIMS was performed by direct inlet system. Specific rotation
values were measured at 25 °C with a Horiba SEPA-300 polari-
meter. FABMS (positive) was measured using glycerol matrix.
Incubation conditions
Standard incubation conditions were performed according to
the published protocols.5,13 The cell-free extracts were prepared
as follows. 1 L culture of E. coli encoding the native SHC was
harvested by centrifugation and to the collected pellets was added
50 cm3 of citrate buffer solution (pH 6.0), and then subjected to
ultrasonication to disrupt the cells. The supernatant was used for the
incubations after removing the cell debris by centrifugation. 1 cm3
of the supernatant contains ca. 200 g of the pure SHC.
Preparation of isoprenyl bromides 17, 19 and 20
Farnesyl bromide 17, 1-bromo-3-methylbut-2-ene 19 and geranyl
bromide 20 were prepared by treatment of the corresponding
alcohols farnesol 12, 3-methylbut-2-en-1-ol and geraniol 11 with
PBr3 in THF at 0 °C for 30 min.16 The reaction mixture was poured
into ice water. The product was extracted with n-hexane and dried
over anhydrous Na2SO4 and used without further purification.
Syntheses of 2,6,11,15-tetramethyl-hexadeca-
(6E,10E)-2,6,10,14-tetraene 14, 2,6,10,15-tetramethyl-
hexadeca-(6E,10E)-2,6,10,14-tetraene 15 and
2,6,10,15,19-pentamethyl-eicosa-(6E,10E,14E)-2,6,10,14,18-
pentaene 16
Scheme 4 Cyclization mechanisms of the truncated squalene analogs (C20
and C25) by the SHC. A: analog 14 (C20) having the methyl group at C-11.
B: analog 15 (C20) having the methyl group at C-10. C: analog 16 (C25).
The stereochemistry at C-13 of 15e (13-H) is identical to that of 6 formed
during the polycyclization reaction of 1, whereas 15f has the opposite con-
figuration of 13-H. 16e has the same configuration of 17-H as 8, but 16f
has 17-H opposed to 16e. The symbol of the closed circle indicates that the
axial-oriented C(7)–Me or C(11)–Me repulsively interacted with the recog-
nition site(s) involved in the cyclase, due to the introduction of large steric
bulk size of the methyl group at this position(s). Possibly, the -face domain
of the cyclase is strictly compact. In contrast, the -face domain would be
less compact, thus can accept the axial-oriented methyl groups at positions
of C-2, C-6 and C-10.
These compounds were synthesized by a coupling reaction of
each bromide according to the literature.25,26 As a typical example,
the syntheses of 14 and 16 are described below. To the ethereal
solution of pyrrolidine (250 mg in 10 ml), 2.1 ml of n-BuLi (1.58 M
hexane solution) was added at 0 °C under N2 atmosphere and stood
for 20 min. To the reaction mixture which was further cooled
at −38 °C, 250 mg of CuI was added, the color quickly turning
into reddish brown, and stirred for 20 min. A mixed solution of
17 (100 mg) and 20 (50 mg) in Et2O (5 ml) was slowly added and
stirred for 60 min. The reaction mixture was poured into brine and
the product was extracted with n-hexane and purified by the follow-
ing three steps: SiO2 column chromatography (n-hexane), an open
column of reverse-phase (C18) with an eluent of 85% CH3CN/H2O
and finally with 5% AgNO3–SiO2 column chromatography by
eluting with a mixture of hexane and EtOAc (100:5). Pure 14 and
16 was isolated in yields of 20 mg and 28 mg, respectively. The
total yield of 14 and 16 was 26%. NMR data of 14 in C6D6: H, 5.44
(2H,br s), 5.36 (br t, J 6.8), around 2.2 (12H, m), 1.80 (6H, s), 1.73
(6H, s), 1.68 (6H, s); C, 135.1(s), 131.1(s), 124.9(d), 124.8(d), 40.2
(t), 28.7 (t), 27.2 (t), 25.8 (q), 17.7 (q), 16.1 (q). EIMS; m/z (%) 69
(100), 81 (48), 137 (15), 150 (15), 274 (M+, 8). NMR data of 16 in
C6D6: H, around 5.4 (5H, m), around 2.4 (16H, m), 1.84 (6H, s),
1.73 (9H, s), 1.68 (6H, s);C, 135.1 (2C, s), 134.9 (s), 131.0 (2C,
s), 124.9 (2C, d), 124.85 (2C, d), 124.8 (d), 40.2 (3C, t), 28.7 (2C,
t), 27.2 (t), 27.19 (t), 27.1(t), 25.8 (2C, q), 17.7 (2C, q), 16.1 (2C,
q), 16.09(1C, q). EIMS; m/z (%), 69 (100), 81 (59), 137 (22), 342
(M+, 5). For the preparation of 15, 19 and 17 were subjected to the
allyl coupling reaction. NMR data of 15 in CDCl3: H, 5.12 (4H,
m), around 2.0 (12H, m), 1.69 (6H, s), 1.60(12H, s); C, 135.1 (s),
134.9 (s), 131.4 (s), 131.2 (s), 124.5 (d), 124.4 (d), 124.3 (2C, d),
39.7 (2C, t), 28.4 (t), 28.3 (t), 26.8 (t), 26.6 (t), 25.70 (q), 25.67 (q),
with the Markovnikov rule. It should be noted that the carbocationic
intermediates 5, 6 and 8 formed during the polycyclization reaction
of 1 were trapped through the present studies using the truncated
analogs and that all the truncated analogs (C15–C25) had higher
conversion ratios than the elongated C35 analog (only 10%).24 It
is unresolved why the conversion ratios of 15 (C20) and 16 (C25)
are relatively lower than that of 12 (C15) despite 15 and 16 having
longer carbon-chain lengths closer to that of 1 (C30) and why the
cyclization of geraniol 11 (C10) was not allowed despite 11 having
the minimum carbon-chain length required for the formation of the
monocyclic skeleton like 4? The stereochemical control by the SHC
was not perfect for the truncated analogs (15 and 16), leading to the
false intermediates having the configurations of 13-H and 17-H.
This finding may suggest that the complete stereochemical control
would be attained only by substrate 1 (C30). Further studies on some
truncated squalene analogs will provide better understanding for the
polycyclization mechanism of the true substrate 1.
Experimental
Analytical methods
NMR spectra were recorded in C6D6 on a Bruker DMX 600 or
DPX 400 spectrometer, the chemical shifts being relative to the
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 6 5 0 – 2 6 5 7
2 6 5 5

View Article Online
17.68 (q), 17.66 (q), 15.99 (2C, q). EIMS m/z (%); 69 (100), 81 (36),
13.6, 3.2, H-2), 1.46 (m, H-6), 1.48 (m, H-7), 1.60 (m, H-2), 1.67
(2H, m, H-1 and H-6), 1.98 (2H, m, H-11), 1.99 (3H, s, Me-20), 1.99
(very broad, s, H-14), 2.05 (ddd, J 12.8, 3.2, 3.2, H-3), 3.62 (dd, J
10.8, 5.6, H-15), 3.75 (dd, J 10.8, 3.2, H-15), 5.63 (br s, H-12). C
15.87 (C-19), 15.92 (C-18), 18.85 (C-6), 19.07 (C-2), 21.85 (C-17),
22.14 (C-20), 22.92 (C-11), 33.21 (C-4), 33.35 (C-16), 36.34 (C-8),
37.39 (C-10), 40.07 (C-1), 41.55 (C-3), 42.17 (C-7), 55.04 (C-9),
56.28 (C-5), 58.04 (C-14), 60.74 (C-15), 123.4 (C-12), 133.5 (C-
13). The assignments of C-18 and C-19 and those of C-2 and C-6
are interchangeable. EIMS (%): 59 (100), 82(57), 95(48), 123(38),
137(31), 177(21), 191 (71), 219 (25), 234 (24), 259 (14), 274
(M+–H2O, 10). HRMS (EI) m/z (M+): calcd. for C20H34O, 290.2610;
found 290.2615. []D25 = 9.23 (c 0.051, EtOH); mp. 110–113 °C.
137 (10), 274 (M+, 3).
Spectroscopic data of enzymic products 21–30
Product 21. 1H-NMR (C6D6),  0.77 (3H, s, Me-14), 0.86 (3H, s,
Me-13), 0.91 (3H, s, Me-12), 1.02 (dd, J 12.7, 2.4, H-5), 1.10 (m,
H-1), 1.17 (m, H-3), 1.30 (m, H-6), 1.42 (m, H-3), 1.47 (m, H-2),
1.58 (m, H-2), 1.61 (m, H-1), 1.66 (br d, J 15.0, H-6), 1.94 (m, H-
9), 1.95 (br s, H-16), 2.01 (m, H-7), 2.40 (br d, J 12.8, H-7), 3.78
(dd, J 10.8, 10.8, H-11), 3.86 (dd, J 10.8, 3.4, H-11), 4.75 (s, H-15),
4.99 (s, H-15). 13C-NMR (C6D6),  15.36 (C-14), 19.50 (C-2), 21.84
(C-13), 21.84 (C-14), 24.39 (C-6), 33.45 (C-4), 33.68 (C-12), 38.12
(C-7), 39.05 (2 × C, C-1 and C-10), 42.19 (C-3), 55.14 (C-5), 58.73
(C-11), 59.32 (C-9), 106.6 (C-15), 147.9 (C-8). EIMS (%): 69 (25),
81 (23), 109 (100), 124 (42),191 (10), 222 (M+, 8). HRMS (EI):
calcd. for C15H26O, 222.1984; found 222.1964. []D25 = +20.4 (c
0.015, EtOH). oil.
Product 26. NMR data (C6D6). H 0.98 (6H, s, Me-16 and Me-
17), 0.99 (m, H-5), 1.01(3H, s, Me-15), 1.03 (3H,s, Me-18), 1.05
(m, H-1), 1.22 (3H, s, Me-20), 1.24 (3H, s, Me-19), 1.29(m, H-3),
1.44 (m, H-11), 1.46 (m, H-2), 1.49 (m, H-9), 1.50 (m, H-3), 1.52
(m, H-6), 1.58 (m, H-11), 1.60 (m, H-1), 1.62 (m, H-13), 1.70(m,
H-2), 1.71 (m, H-6), 1.82 (2H, m, H-12), 1.90 (m, H-7), 2.12 (br t, J
10.4, H-7). The assignments of Me-19 and Me-20 are interchange-
able. C 16.52 (C-17), 18.89 (C-2), 20.11 (C-6), 21.54 (C-16), 21.71
(C-11), 25.52 (C-12), 26.82 (C-18), 28.54 (C-20), 30.17 (C-19),
33.18 (C-4), 33.68 (C-15), 37.43 (C-10), 38.09 (C-7), 41.17 (C-
1), 42.72 (C-3), 45.20 (C-8), 57.20 (C-5), 59.92 (C-9), 60.74 (C-
13), 73.88 (C-14). EIMS (%): 59(100), 95(39), 123(32), 137(27),
191(68), 219(22), 234(24), 274(M+–H2O, 15) []D25 = +6.11(c 0.01,
EtOH). Mp. 115–117 °C.
Product 22. NMR data (C6D6). H 0.95 (6H, s, Me-12 and Me-
14), 0.97 (3H, s, Me-13), 1.06 (m, H-1), 1.24 (m, H-3), 1.25 (m,
H-5), 1.49 (2H, m, H-2 and H-3), 1.62(m, H-2), 1.86 (very broad,
H-9), 1.92 (m, H-6), 1.95 (m, H-1), 1.97 (3H, s, Me-15), 2.04 (m,
H-6), 3.63 (m, H-11), 3.75 (m, H-11), 5.62 (br s, H-7). C 14.93
(C-14), 19.14 (C-2), 22.15 (C-13), 22.28 (C-15), 23.96 (C-6), 33.01
(C-4), 33.50 (C-12), 36.19 (C-10), 39.95 (C-1), 42.45 (C-3), 50.12
(C-5), 57.45 (C-9), 60.82 (C-11), 123.6 (C-7), 133.7 (C-8). The
assignments of C-13 and C-15 may be exchangeable. EIMS (%):
69 (61), 81 (70), 95 (58), 137 (100), 189 (11), 222 (M+, 10). HRMS
(EI): calcd. for C15H26O, 222.1984; found 222.1964. []D25 = −28.2
(c 0.045, EtOH). Mp. 87–89 °C.
Product 27. NMR data (C6D6). H 0.87 (dd, J 12.4, 2.4, H-5),
0.95 (m, H-1), 0.98 (3H, s, Me-16), 0.97 (3H, s, Me-17), 1.03 (3H,
s, Me-15), 1.07 (dd, J 13.2, 7.2, H-9), 1.16 (3H,s, Me-18), 1.17 (3H,
s, Me-20), 1.19 (m, H-7), 1.29 (t, J 10.1, H-13), 1.31 (m, H-3), 1.34
(3H, s, Me-19), 1.43 (m, H-6), 1.49 (m, H-11), 1.52 (3H, m, H-1,
H-2 and H-3), 1.59 (m, H-6), 1.72 (m, H-12), 1.73 (2H, m, H-2 and
H-11), 1.88 (m, H-12), 2.10 (m, H-7). The assignments of Me-19
and Me-20 are interchangeable. C 15.73 (C-18), 15.92 (C-17),
18.72 (C-11), 19.56 (C-2), 19.61 (C-6), 21.43 (C-16), 22.84 (C-
12), 30.01 (C-19), 32.06 (C-20), 33.10 (C-4), 33.70 (C-15), 37.24
(C-10), 40.58 (C-1), 42.78 (C-3), 42.88 (C-7), 44.38 (C-8), 57.46
(C-5), 60.92 (C-13), 63.26 (C-9), 72.55 (C-14). The assignments of
C-19 and C-20 are interchangeable. EIMS (%): 59 (100), 95 (39),
123 (32), 137 (27), 191 (68), 219 (22), 234 (24), 274 (M+–H2O,
15). HRMS (EI) m/z (M+–H2O): calcd. for C20H34, 274.2661; found
274.2626. []D25 −7.16 (c 0.034, EtOH). Mp. 115–117 °C.
Product 23. NMR data (C6D6). H 0.70 (3H, s, Me-14), 0.81 (3H,
s, Me-13), 0.91 (3H, s, Me-12), 0.936 (dd, J 12.3, 1.8, H-5), 0.94
(m, H-6), 1.03 (m, H-2), 1.07 (ddd, J 12.9, 12.9, 3.5, H-1), 1.16 (m,
H-6), 1.18 (m, H-3), 1.38 (3H,s, Me-15), 1.40 (m, H-3), 1.42 (m,
H-2), 1.60 (m, H-7), 1.63 (m, H-9), 1.66 (m, H-1), 1.92 (ddd, J 2.9,
6.0,12.6, H-7), 3.97 (d, J 6.9, H-11). C 15.99 (C-14), 18.91 (C-2),
21.74 (C-13), 21.74 (C-6), 24.44 (C-15), 33.26 (C-4), 33.61 (C-12),
37.53 (C-10), 40.09 (C-1), 41.92 (C-3), 44.44 (C-7), 55.93 (C-5),
60.59 (C-9), 60.9 (C-11), 74.53 (C-8). EIMS (%): 69 (70), 81 (72),
95 (100), 109 (95), 123 (61), 137 (46), 164 (28), 189 (22), 207 (16),
222 (M+–H2O, 13). HRMS (EI) m/z (M+–H2O): calcd. for C15H26O,
222.1984; found 222.1964. []D25 = +14.3 (c 0.015, EtOH). Mp.
110–114 °C.
Product 24. NMR data. H 0.74 (3H, s, Me-14), 0.83 (3H, s, Me-
13), 0.88 (br d, J 11.0, H-5), 0.93 (3H, s, Me-12), 1.08 (m, H-1),
1.10 (m, H-6), 1.14 (m, H-3), 1.34 (3H, s, Me-15), 1.37 (m, H-3),
1.39 (m, H-2), 1.55 (m, H-2), 1.55 (m, H-7), 1.56 (m, H-6), 1.70
(6H, s, Me-29 and Me-30), 1.72 (3H, s, Me-28), 1.78 (m, H-1), 1.81
(3H,s, Me-27), 1.82 (m, H-9), 1.88 (m, H-7), 2.12 (2H, br t, J 7.5,
H-19), 2.22 (dt, J 7.2, 7.2, H-23), 2.25 (2H, dt, J 7.2, 7.2, H-20),
2.30 (dt, J 7.3, 7.3, H-24), 3.99 (2H, dd, J 6.1, 3.8, H-16), 4.13 (2H,
dd, J 16.7, 8.6, H-11), 5.35 (t, J 6.8, H-21), 5.37 (t, J 6.8, H-25),
5.56 (br t, J 6.5, H-17). The assignments of Me-27 and Me-30 are
interchangeable. C 16.08 (C-29), 16.37 (C-14), 16.41 (C-28), 17.73
(C-30), 18.86 (C-2), 19.66 (C-15), 20.01 (C-6), 21.68 (C-13), 25.84
(C-27), 26.65 (C-20), 27.21 (C-24), 33.25 (C-4), 33.57 (C-12),
37.79 (C-10), 39.14 (C-7), 39.78 (C-19), 40.11 (C-23), 40.15 (C-1),
41.90 (C-3), 55.76 (C-5), 56.72 (C-16), 60.11 (C-9), 60.47 (C-11),
80.13 (C-8), 122.21 (C-17), 124.37 (C-21), 124.99 (C-25), 131.04
(C-26), 135.27 (C-22), 139.24 (C-18). The assignments of C-1 and
C-23 and those of C-14 and C-28 may be interchangeable. EIMS
(%): 80 (82), 91 (100), 135 (48), 159 (21), 186 (24). HRMS (FAB)
m/z (M+ + H): calcd. for C30H53O2, 445.4046; found 445.4052.
[]D25 −0.31 (c 0.04, EtOH); oil.
Product 28. NMR data (C6D6). H 0.90 (m, H-1), 0.91(m, H-5),
0.96 (3H, s, Me-20), 0.997(3H, s, Me-21), 1.03 (3H, s, Me-19), 1.05
(3H,s, Me-22), 1.13 (d, J 6.9, Me-24), 1.17 (d, J 6.9, Me-25), 1.28
(m, H-3), 1.30 (3H, s, Me-23), 1.36 (m, H-11), 1.47 (m, H-2), 1.48
(m, H-15), 1.50 (m, H-3), 1.52 (2H, m, H-2 and H-7), 1.58 (m, H-9),
1.65 (m, H-7), 1.66 (m, H-11), 1.67 (H-6), 1.74 (m, H-6), 1.78 (m,
H-1), 2.03 (m, H-12), 2.11 (ddd, J 10.2, 10.2, 4.9, H-15), 2.34 (m,
H-16), 2.47 (m, H-16), 2.56 (m, H-12), 2.85 (septet, J 6.9, H-18).
The assignments of H-2 and H-6 and those of Me-24 and Me-25
are interchangeable. C 16.67 (C-21), 17.06 (C-22), 18.93 (C-2),
19.07 (C-6), 21.53 (C-24), 21.89 (C-20), 21.89 (C-25), 22.23 (C-
11), 23.22 (C-23), 23.26 (C-12), 27.02 (C-18), 29.63 (C-16), 31.13
(C-15), 33.50 (C-4), 33.66 (C-19), 35.63 (C-7), 37.99 (C-10), 40.85
(C-1), 41.98 (C-8), 42.33 (C-3), 52.19 (C-9), 56.73 (C-14), 57.30
(C-5), 136.37 (C-17), 137.99 (C-13). The assignments of C-2 and
C-6 and those of C-24 and C-25 are interchangeable. EIMS (%):
69 (58), 67 (81), 95 (89), 121 (91), 161(89), 191 (84), 299 (42),
342 (M+, 100). HRMS (EI): calcd. for C25H42, 342.3287; found
342.3278. []D25 = −11.7 (c 0.01, EtOH).
Product 29. NMR data (C6D6). H 0.96 (3H, s, Me-21), 0.97 (3H,
s, Me-19), 1.02 (3H, s, Me-22), 1.03 (3H, s, Me-20), 1.03 (m, H-1),
1.04 (3H, d, J 6.0, Me-24), 1.08 (3H, d, J 6.0, Me-25), 1.17 (3H,s,
Me-23), 1.29 (m, H-3), 1.38 (m, H-16), 1.45 (m, H-2), 1.50 (2H,
Product 25. NMR data (C6D6). H 0.80 (m, H-1), 0.84 (dd, J 12.0,
2.0, H-5), 0.953 (9H, s, Me-17, Me-18 and Me-19), 1.01 (3H, s, Me-
16), 1.17 (m, H-9), 1.20(m, H-3), 1.22 (m, H-7), 1.38 (ddd, J 13.6,
2 6 5 6
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 6 5 0 – 2 6 5 7

View Article Online
3 R. Xu, G. C. Fazio and S. P. T. Matsuda, Phytochemistry, 2004, 65,
261–290.
4 M. J. R. Segura, B. E. Jackson and S. P. T. Matsuda, Nat. Prod. Rep.,
2003, 20, 304–317.
5 T. Sato and T. Hoshino, Biosci. Biotechnol. Biochem., 1999, 63,
2189–2198.
6 T. Hoshino and T. Sato, Chem. Commun., 1999, 2005–2006.
7 T. Hoshino, M. Kouda, T. Abe and S. Ohashi, Biosci. Biotechnol. Bio-
chem., 1999, 63, 2038–2041.
8 T. Sato, T. Abe and T. Hoshino, Chem. Commun., 1998, 2617–2618.
9 T. Hoshino, T. Abe and M. Kouda, Chem. Commun., 2000,
441–442.
10 T. Hoshino, M. Kouda, T. Abe and T. Sato, Chem. Commun., 2000,
1485–1486.
11 T. Sato, M. Kouda and T. Hoshino, Biosci. Biotechnol. Biochem., 2004,
68, 728–738.
12 T. Hoshino, S. Nakano, T. Kondo, T. Sato and A. Miyoshi, Org. Biomol.
Chem., 2004, 2, 1456–1470.
m, H-3 and H-17), 1.52 (m, H-5), 1.60 (m, H-18), 1.63 (m, H-6),
1.66 (m, H-6), 1.67 (m, H-2), 1.68 (m, H-15), 1.75 (m, H-1), 1.77
(m, H-15), 1.78 (m, H-7), 1.89 (m, H-7), 2.01 (m, H-11), 2.06 (m,
H-16), 2.27 (m, H-11), 2.44 (m, H-9), 5.47 (m, H-12). The assign-
ments of Me-24 and Me-25 are interchangeable. C 12.98 (C-21),
18.04 (C-6), 19.06 (C-2), 21.13 (C-20), 21.94 (C-22), 22.11 (C-25),
22.95 (C-24), 24.33 (C-11), 27.20 (C-23), 28.30 (C-16), 31.26
(C-18), 32.79 (C-19), 32.92 (C-4), 33.66 (C-7), 34.09 (C-15), 35.02
(C-10), 38.88 (C-1), 42.30 (C-3), 43.41 (C-14), 49.08 (C-9), 51.17
(C-8), 51.30 (C-5), 54.93 (C-17), 118.08 (C-12), 145.70 (C-13).
The assignments of C-20, C-24 and C-25 are interchangeable.
EIMS (%): 69 (26), 95 (23), 175 (13), 203 (15), 327 (100), 342
(M+, 11). HRMS (EI): calcd. for C25H42, 342.3287; found 342.3278.
[]D25 = −21.4 (c 0.036, EtOH).
Product 30. NMR data (C6D6). H 0.92(2H, m, H-1 and H-5),
0.99 (3H,s, Me-20), 1.00 (3H,s, Me-21), 1.02 (3H,s, Me-19), 1.08
(3H,s, Me-22), 1.17 (3H,s, Me-23), 1.21 (3H,s, Me-24), 1.23 (m,
H-11), 1.24(m, H-15), 1.27 (3H,s, Me-25), 1.39(2H, m, H-3 and
H-7), 1.41 (m, H-6), 1.50 (2H, m, H-2 and H-3), 1.51 (m, H-9),
1.52 (m, H-15), 1.58 (m, H-12), 1.67 (3H, m, H-6, H-11 and H-16),
1.73 (3H, m, H-2, H-7 and H-16), 1.75 (m, H-1), 1.98 (ddd, J 9.9,
9.9, 9.9, H-17), 2.09 (m, H-13), 2.10 (m, H-12). The assignments
of Me-24 and Me-25 are interchangeable. C 16.16 (C-22), 16.48
(C-21), 17.10 (C-23), 19.12 (C-2), 19.12 (C-6), 21.77 (C-20), 22.86
(C-11), 26.05 (C-12), 27.56 (C-16), 29.74 (C-24), 31.55 (C-25),
32.79 (C-15), 33.54 (C-4), 33.61 (C-19), 35.73 (C-7), 37.73 (C-10),
40.90 (C-1), 41.19 (C-8), 42.49 (C-3), 44.19 (C-13), 49.39 (C-14),
50.07 (C-17), 51.23 (C-9), 57.37 (C-5), 72.98 (C-18). EIMS (%):
95 (42), 191 (100), 302 (33), 342 (M+–H2O,13). The assignments of
C-24 and C-25 are interchangeable. HRMS (EI): calcd. for C25H42,
342.3287; found 342.3274. []D25 = −3.14 (c 0.113, EtOH). Mp.
98–100 °C.
13 T. Sato and T. Hoshino, Biosci. Biotechnol. Biochem., 1999, 63,
1171–1180.
14 T. Hoshino and T. Kondo, Chem. Commun., 1999, 731–732.
15 T. Hoshino and S. Ohashi, Org. Lett., 2002, 4, 2553–2556.
16 S. Nakano, S. Ohashi and T. Hoshino, Org. Biomol. Chem., 2004, 2,
2012–2022.
17 T. Abe, T. Kondo, S. Ohashi and T. Hoshino, to be published, presented
at 44th Symposium on the Chemistry of Essential oils, Terpenes and
Aromatics, Sapporo, 2000, pp. 129–131.
18 Y. Sugino, T. Abe and T. Hoshino, to be published, presented at 47th
Symposium on the Chemistry of Essential oils, Terpenes and Aromatics,
Tokyo, 2003, pp. 330–332.
19 I. Abe, T. Dang, Y. F. Zheng, B. A. Madden, C. Feil, K. Poralla and
G. D. Prestwich, J. Am. Chem. Soc., 1997, 119, 11333–11334.
20 B. J. Robustell, I. Abe and G. D. Prestwich, Tetrahedron Lett, 1998, 39,
957–960.
21 B. J. Robustell, I. Abe and G. D. Prestwich, Tetrahedron Lett, 1998, 39,
9385–9388.
22 H. Tanaka, H. Noguchi and I. Abe, Org. Lett., 2004, 6, 803–806.
23 H. Tanaka, H. Noguchi and I. Abe, Tetrahedron Lett, 2004, 45,
3093–3096.
24 I. Abe, H. Tanaka and H. Noguchi, J. Am. Chem. Soc., 2002, 124,
14514–14515.
Acknowledgements
25 Y. Kitagawa, K. Oshima, H. Yamamoto and H. Nozaki, Tetrahedron
Lett., 1975, 16, 1859–1862.
26 T. Hoshino, H. J. Williams, K. Shishido and A. I. Scott, J. Labelled
Compd. Radiopharm., 1990, 28, 1285–1292.
27 E. J. Corey and H. Cheng, Tetrahedron Lett., 1996, 37, 2709–2712.
28 E. J. Corey, S. C. Virgil, H. Cheng, C. H. Baker, S. P. T. Matsuda,
V. Singh and S. Sarshar, J. Am. Chem. Soc., 1995, 117, 11819–
11820.
This work was done by the financial supports to T. H. (No.
15380081), provided by the Ministry of Education, Science and
Culture of Japan. Farnesol and geranylgeraniol were kindly gifted
from Kuraray Co. LTD (Nakajoh, Japan).
References
1 T. Hoshino and T. Sato, Chem. Commun., 2002, 291–301.
2 K. U. Wendt, G. E. Schulz, E. J. Corey and D. R. Liu, Angew. Chem.,
Int. Ed. Engl., 2000, 39, 2812–2833.
29 T. Hoshino and Y. Sakai, Chem. Commun., 1998, 1591–1592.
30 T. Hoshino and Y. Sakai, Tetrahedron Lett., 2001, 42, 7319–7323.
31 B. A. Hess Jr., J. Am. Chem., Soc., 2002, 124, 10286–10287.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 6 5 0 – 2 6 5 7
2 6 5 7