β-Amyrin Biosynthesis: Effect of Steric Bulk at the 6-, 10- and 15-Positions in the 2,3-Oxidosqualene Backbone on Polycyclisation Cascades

Article abstract of DOI:10.1002/ejoc.201601306

β-Amyrin synthase incubation experiments have been conducted to determine the influence of steric effects at the 6-, 10- and 15-positions of 2,3-oxidosqualene on the polycyclisation pathway. Nor- and ethyl-substituted oxidosqualene analogues were synthesised. Cyclisation of the ethyl-substituted analogues did not occur, but the nor analogues underwent a polycyclisation cascade to yield fully cyclised products with 6/6/6/6/6-fused pentacyclic scaffolds generated via a final oleanyl cation. Previously, we reported that 19- and 23-ethyl-substituted analogues underwent polycyclisation reactions. Therefore, the catalytic domain involved in earlier cyclisation steps is notably compact. In contrast, the catalytic domain in the later cyclisation steps is more loosely packed (less compact) to accommodate the bulky ethyl group. The reaction cavities for recognising branched methyl groups are discussed by comparing β-amyrin synthase with other triterpene cyclases such as lanosterol and hopene synthases.

Full text of DOI:10.1002/ejoc.201601306

A Journal of
Accepted Article
Title: β-Amyrin biosynthesis: Effect of steric bulk at the 6-, 10-, and 15-positions in
the 2,3-oxidosqualene backbone on polycyclisation cascades
Authors: Yuri Terasawa; Yusuke Sasaki; Yuki Yamaguchi; Kazunari Takahashi; Tsu-
tomu Hoshino
This manuscript has been accepted after peer review and the authors have elected
to post their Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by using the Digital
Object Identifier (DOI) given below. The VoR will be published online in Early View as
soon as possible and may be different to this Accepted Article as a result of editing.
Readers should obtain the VoR from the journal website shown below when it is pu-
blished to ensure accuracy of information. The authors are responsible for the content
of this Accepted Article.
To be cited as: Eur. J. Org. Chem. 10.1002/ejoc.201601306
Link to VoR: https://doi.org/10.1002/ejoc.201601306
Supported by

European Journal of Organic Chemistry
10.1002/ejoc.201601306
-Amyrin biosynthesis: Effect of steric bulk at the 6-, 10-, and
15-positions in the 2,3-oxidosqualene backbone on polycyclisation
cascades
Yuri Terasawa, Yusuke Sasaki, Yuki Yamaguchi, Kazunari Takahashi and Tsutomu
Hoshino*
Department of Applied Biological Chemistry, Faculty of Agriculture and Graduate
School of Science and Technology, Niigata University, Ikarashi 2-8050, Nishi-ku,
Niigata, Japan 950-2181
E-mail: hoshitsu@agr.niigata-u.ac.jp
Supporting information for this article is available. on the WWW under …..
Table of Content: Polycyclization cascade
On the methyl binding sites in the reaction cavity of -amyrin synthase: catalytic
domain responsible for the earlier stage of the polycyclization reaction is notably
compact, while that for the latter stage is somewhat loosely packed.
1
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Abstract
-Amyrin synthase incubation experiments were conducted to determine the
influence of sterics at the C-6, C-10, and C-15 positions of 2,3-oxidosqualene on the
polycyclisation pathway. Nor- and ethyl-substituted oxidosqualene analogues were
synthesised. Cyclisation of the ethyl-substituted analogues did not occur, but the
nor-analogues underwent a polycyclisation cascade, yielding fully cyclised products
with 6/6/6/6/6-fused pentacyclic scaffolds, generated via a final oleanyl cation.
Previously, we reported that C-19 and C-23 ethyl-substituted analogues underwent
polycyclisation reactions. Therefore, the catalytic site domain involved in earlier
cyclisation steps was notably compact. In contrast, the catalytic domain in later
cyclisation steps was more loosely packed (less compact) to accommodate the bulky
ethyl group. The reaction spaces for recognising branched methyl groups are discussed
by comparing -amyrin synthase with other triterpene cyclases, such as lanosterol and
hopene synthases.
Keywords: Alkene; cyclization, enzyme catalysis; polycycle; terpenoids
Introduction
The polycyclisation cascades of squalene and (3S)-2,3-oxidosqualene (1) have attracted
organic chemists and biochemists ever since Ruzicka and coworkers proposed their
“biogenetic isoprene rule” more than 70 years ago.[1] The reactions proceed with
complete regio- and stereospecificity, yielding sterols and triterpenes with remarkable
structural diversity,[2] of which there are more than 100 different carbon frameworks
exhibiting important biological activities.[3] -Amyrin (2) is widely found in plants and
contains a pentacyclic scaffold with eight chiral centres. Recently, we succeeded in
isolating pure -amyrin synthase from Euphorbia tirucalli (EtAS) and characterised its
enzymatic properties.[4] -Amyrin synthase folds oxidosqualene, 1, into a
chair–chair–chair–boat–boat conformation (Scheme 1).[1,5] Proton attack at the epoxide
ring initiates a polycyclisation reaction, yielding dammarenyl cation 3 containing a
6/6/6/5-fused A/B/C/D-tetracyclic ring system, which then undergoes ring expansion to
afford 6/6/6/6-fused tetracyclic baccharenyl cation 4. Further cyclisation affords
6/6/6/6/5-fused pentacyclic lupanyl cation 5.
＜Figure 1 & Scheme 1>
2
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
A second ring expansion of the five-membered E-ring generates 6/6/6/6/6-fused
pentacyclic oleanyl cation 6. Subsequent 1,2-hydride shifts (H-18 to C-19 cation and
H-13 to C-18) and deprotonation of H-12, introduces a double bond between C-12
and C-13, finally generating the -amyrin skeleton (Scheme 1). The enzymatic reactions
of substrate analogues with -amyrin synthase have been reported,[6–9] but little is
known about the effect of steric bulk at the methyl-group position in 1 on the
polycyclisation cascade. We have reported the polycyclisation reactions of substrate
analogues 13–18 and 24,30-bisnoroxidosqualene using EtAS enzyme.[10–12] The
23Z-methyl group is critical for the correct folding of 1 to yield 2,[11] with large amounts
of tetracyclic products formed in its absence, resulting from the polycyclisation cascade
terminating at the intermediate stage. In contrast, the absence of the 23E-Me group had
little influence on the polycyclisation cascade, with the -amyrin skeleton afforded in
approx. 96% yield. We previously proposed that a CH– interaction operates between
23Z-Me and the binding site.[12] A larger ethyl group is also tolerated at the C-23
position, furnishing the -amyrin skeleton irrespective of whether the E- (15) or
Z-substituted (16) analogue is used.[12] Ethyl group substitution at the 19-position in 14
is also tolerated, with polycyclisation affording the -amyrin skeleton and 6/6/6/5-fused
tetracyclic skeleton in a 1:3 ratio.[10] Furthermore, 29-noroxidosqualene (13) afforded a
novel scaffold consisting of a 6/6/6/6-fused tetracycle. However, the effects of changing
sterics at the 6-, 10-, and 15-positions of 1 on the polycyclisation have not been reported.
Herein, we describe the polycyclisation cascades of 7–12, showing that
ethyl-substituted oxidosqualenes 7–9 underwent no reaction, while noroxidosqualenes
10–12 afforded oleanyl cation-derived products. The reaction spaces for recognising
branched methyl groups are discussed by comparing -amyrin synthase with other
triterpene cyclases, such as lanosterol and hopene synthases.
Results
Syntheses of substrate analogues 7–14.
Synthesis of substrate analogues: Ethyl-substituted oxidosqualenes 8 and 9
The starting material, C17-aldehyde 19, was prepared according to our previous
report.[13] Squalene was treated with mCPBA (m-chloroperbenzoic acid) in CH2Cl2 to
produce an epoxide mixture, which was then treated with H5IO6, yielding a mixture of
C17-19, C22-20, and C27 aldehydes. Scheme 2A shows the synthetic design of 8 and 9.
Aldehyde 19 was subjected to a Wittig–Horner reaction using triethyl
3
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
2-phosphonobutyrate to yield an E/Z-mixture of alkenes, which were separated by SiO2
column chromatography to isolate E-isomer 21. Compound 21 was reduced with
DIBAL-H to give alcohol 22, which was then converted to bromide 23 using PBr3.
Geraniol (26) was also transformed into the corresponding bromide using PBr3,
followed by treatment with sodium benzenesulfinate, to afford phenylsulfone derivative
27. Compound 27 was then converted to a bromohydrin using N-bromosuccinimide
(NBS), and subsequent epoxide 28 using K2CO3. Compounds 23 and 28 were subjected
to a coupling reaction using n-BuLi, yielding the phenylsulfone derivative of
ethyl-substituted analogue 29. To remove the phenylsulfonyl group, 29 was treated with
Super-Hydride reagent, yielding desired compound 8. Alcohol 22 was also treated with
NBS to give bromohydrin 24, which was then transformed into bromide 25. The
coupling reaction of 25 and 27 gave 30, which was treated with Super-Hydride reagent
to afford desired compound 9.
<Scheme 2>
Synthesis of substrate analogues: Ａ mixture of ethyl-derivatives 7 and 14 and that of
nor-derivatives 10 and 13.
Ethyl- and nor-derivatives were prepared using similar synthetic procedures to
those for 8 and 9, with the synthetic plans shown in Scheme 2B. To prepare
ethyl-substituted 7 and 14, aldehyde 20 was first converted into alkene 31 using triethyl
2-phosphonobutyrate. Next, 31 was reduced to alcohol 33, which was then treated with
PBr3 to afford bromide 35. Commercially available 3-methylbut-2-en-1-ol (37) was
transformed into phenylsulfone 38, as detailed in our previous report.13 The coupling
reaction of 35 and 38 yielded 39. Desulfonation was carried out using Super-Hydride
reagent to afford 41. Subsequent reactions of 41 with NBS and K2CO3 furnished a
mixture of 7 and 14 in an approx. 1:1 ratio (Supporting Information, Figs. S2–S5). To
prepare noroxidosqualenes 10 and 13, triethyl 2-phosphonoacetate was employed in the
Wittig–Horner reaction to produce 32. Subsequent reactions to prepare the mixture of
10 and 13 were the same as those to prepare 7 and 14. The final ratio of 10 and 13 was
also approx. 1:1 (Figs. S6–S8)
Details of the preparation of synthetic intermediates and their characterisation by
NMR and MS are given in Supporting Information. In particular, detailed NMR
analyses are given for the mixtures of 7 and 14, and 10 and 13 (Supporting Information,
pp. S11–S16).
4
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Enzymatic reaction of 7–14 with -amyrin synthase
A buffered reaction mixture (2.5 mL, pH 7.4, 100 mM potassium phosphate buffer),
consisting of Triton X-100 (0.1%, w/v), 1 [200 g, (3R/S)-racemic mixture], BSA (1
mg/mL), DTT (dithiothreitol, 1 mM), and purified histidine-tagged EtAS (5 g), was
incubated at 30 °C for 9 h.[4,12] Next, 15% KOH/MeOH (7.5 mL) was added, and the
resulting mixture was heated to 70–80 °C for 30 min. The lipophilic materials were
extracted with hexane, and 20 g of squalene was added to the hexane extract as the
internal standard. Triton X-100 present in the extract was removed by elution through
an SiO2 column (hexane/EtOAc, 100:10), and the eluent was evaporated to dryness.
Next, hexane (150 L) was added to the residue and 1.0 L of the resultant solution was
subjected to GC analysis in order to quantify the lipophilic materials. Several peaks
corresponding to oxidosqualene (1) and its analogues (7–14) indicated that these
substrates were thermally degraded during GC analyses (see Figs. 2 and S9).
Figure S9 depicts GC analyses obtained from incubation of the ethyl-substituted
analogues of 8, 9, 14, and the mixture of 7 and 14. No conversion of 8 and 9 was
observed. In contrast, analogue 14 was transformed into 47 and 48 (1:3 ratio), as we
reported previously.[10] Incubation results for a mixture of 7 and 14 indicated that only
products 47 and 48 were produced. Furthermore, the product peak areas of only 14 and
the mixture of 7 and 14 were indistinguishable in the GC profile. Therefore, substrate 7
underwent no or little transformation.
Figure 2 shows the GC profiles of the hexane extract obtained from incubation
experiments of 1, 11–13, and the mixture of 10 and 13. Unlike the ethyl-substituted
analogues, these nor-analogues underwent enzymatic reactions, forming cyclic products.
Figure 3 shows the relative conversion yields of substrate analogues 7–14 relative to
that of 1. As these synthesised analogues were (3R/S)-racemic mixtures, and only the
cyclisation reaction of (3S)-oxidosqualene proceeds, the conversion yields were
calculated as twice the value of the GC peak areas. Compound 11 was converted into
49–51 in a substantial yield (34%). Analogue 12 was also transformed into 52–54 in a
fair yield (22%). However, the reaction yield of 55 from 13 was poor (3.4%), due to the
occurrence of time-dependent inhibition of this reaction, as described in our previous
report.10 Incubation of a mixture of 10 and 13 generated 56 in a negligible yield (0.8%),
in addition to 55, suggesting that time-dependent inhibition caused by 10 was stronger
than that of 13.
<Figures 2 & 3>
Product isolation and chemical structure elucidation.
5
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Analogue 11 (50 mg) was incubated with partially purified -amyrin synthase (30
mg) under optimal conditions. The lipophilic materials, including the enzymatic
products and unreacted 8, were extracted with hexane, and Triton X-100 present in the
hexane extract was removed by elution through a short SiO2 column (hexane/EtOAc,
100:20). In subsequent SiO2 column chromatography (n-hexane/EtOAc, 100:0–100:5),
substrate 11 was eluted first, with a yield of 4.0 mg, with a fraction containing a mixture
of enzymatic products collected subsequently. The latter fraction (12.6 mg) was
converted into acetates with Ac2O/Py, and subjected to separation by normal phase
HPLC (n-hexane/THF, 100:2), affording pure 50 (6.5 mg). Separation of the mixed
fraction containing 49 and 51 was also successful using HPLC (n-hexane/THF, 100:2),
giving isolated yields of 0.8 and 0.1 mg for 51 and 49, respectively. Analogue 12 was
also incubated under similar conditions, and acetate products were isolated by normal
phase HPLC, affording 53 (2.83 mg), 52 (0.66 mg), and 54 (0.52 mg).
We reported the structures of 47 and 48 previously (Fig. 4).[10] The structures of
products 49–54 were determined by detailed NMR analyses. The NMR spectra,
including 1H, 13C, 1H–1H COSY, HOHAHA, NOESY, HSQC, and HMBC experiments,
are included in Figs. S10–S15. DEPT 45, 90, and 135 spectra were also measured to
differentiate methyl, methylene, methine, and quaternary carbons. In the 13C NMR
spectrum of product 49 acetate (150 MHz, C6D6), one double bond was found at C =
137.7 (s) and 131.6 (s), but no olefinic protons were observed in the 1H NMR spectrum
(600 MHz, C6D6). Me-26 (H 1.20, 3H, s) and Me-27 (H 1.23, s, 3H) had clear HMBC
cross-peaks with C 137.7 (s) and 131.6 (s), respectively, indicating that the double bond
was located at C-13/C-18. A clear nuclear Overhauser effect (NOE) was observed
between Me-27 and Me-29 (H 1.12, s, 3H), and between Me-26 and H-9 (H 1.06, m,
1H). These findings confirmed the assignments of Me-26, Me-27, and Me-29. Me-27
was -oriented, because unambiguous NOEs were found for Me-29 and H-19 (1.89, m,
1H), H-19 and Me-27, and H-19 and H-12 (1.83, m, 1H). The presence of a
1,1-dimethylcyclohexane ring (E-ring) was confirmed as follows: Me-28 (H 0.978, s,
3H; C 24.1, q) and Me-29 (C 32.6, q) showed strong NOE and HMBC correlations
with each other. Furthermore, both Me-28 and Me-29 had definitive HMBC cross-peaks
for C-20 (C 33.4, s), C-19 (C 38.9, t), and C-21 (C 35.8, t). Thus, the polycyclisation
reaction proceeded toward the final stage without truncation of the polycyclisation
cascade. Me-26 had a strong HMBC correlation with C-8 (C 48.7, d), indicating a
hydrogen atom substituent at C-8, which was further supported by the definitive NOE
between H-8 (H 1.18, m, 1H) and Me-25 (H 0.810, s, 3H). Detailed NMR analyses of
acetate 49 (Fig. S10.9) allowed us to determine 49 to be 26-nor--amyrin, as shown in
6
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Figs. 4 and S10.9.
<Figure 4>
One double bond was also found in the 1H (400 MHz, C6D6) and 13C NMR spectra
(100 MHz, C6D6) of product 50 acetate: H 5.47 (t, J = 3.6 Hz, 1H, H-12); C 123.1 (d,
C-12), and C 146.2 (s, C-13). The double-bond position was confirmed by HMBC
correlations between Me-26 (H 1.17, s, 3H) and C-13, and H-18 (H 2.23, m, 1H) and
C-12/C-13. Strong NOEs were observed for Me-29 (H 1.17, s, 3H) and H-18, and
Me-27 (H 1.07, s, 3H) and H-18, supporting the assignment of the -amyrin skeleton in
50 acetate. However, the C-8 Me group in -amyrin had been replaced with a hydrogen
atom, determined by observation of a strong NOE for Me-25 (H 0.840, s, 3H) and H-8
(H 1.34, m, 1H), and a clear HMBC cross-peak for Me-26 and C-8 (C 44.8, d). All
detailed NMR analyses (Fig. S11.9) verified that 50 was 26-nor--amyrin, as depicted
in Figs. 4 and S11.9.
One double bond was found in the 1H (600 MHz, C6D6) and 13C NMR spectra (150
MHz, C6D6) of 51 acetate: H 5.19 (s, 1H); C 129.9 (d, C-19), and 142.4 (s, C-18).
Definitive HMBC correlations between Me-27 (H 1.23, s, 3H) and C-18, Me-29 (H
1.20, s, 3H) and C-19, and Me-28 (H 1.20, s, 3H) and C-19 demonstrated that the
double bond was positioned at C-18/C-19. The following definitive NOEs were
observed: Me-27/H-13 (H 1.94, bd, J=11.8 Hz, 1H), H-13/H-8 (H 1.17, m, 1H),
H-8/Me-25 (H 0.882, s, 3H), and Me-26 (H 0.817, s, 3H)/H-9 (H 0.92, m, 1H),
indicating the structure and stereochemistry shown in Fig. 4, with compound 51
determined to be 26-norgermanicol. Detailed analyses of the NMR data are shown in
Fig. S12.9.
1H (600 MHz) and 13C NMR spectra (150 MHz) of 52 acetate were measured in
C6D6. Acetylated 52 also contained one double bond at C 130.0 (s, C-13) and 135.5 (s,
C-14), but no olefinic protons were found. The double-bond position was determined by
detailed HMBC analyses, showing that C-14 had definitive cross-peaks with the
following protons via 3-bond C-H couplings: H-9 (H 1.03, m, 1H), Me-26 (H 1.10, s,
3H), H-16 (H 1.05, m, 1H; H 2.10, m, 1H). Me-26 was determined to be -oriented,
because of the strong NOE observed for Me-26 and Me-25 (H 0.896, s, 3H). HMBC
cross-peak between Me-27 (H 1.08, s, 3H) and C-16 (C 27.2, t) and HSQC correlation
for C-16 and H-16 allowed the assignment of H-16. The position of another sp2 carbon
(C-13) was determined by clear HMBC correlations of C-13 with H-19 (H 1.30, m, 1H)
and H-11 (H 1.59, m, 1H). Unambiguous NOEs were observed for -oriented Me-26
7
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
and H-12 (H 1.86, m, 1H), H-12 and H-18 (H 1.72, m, 1H), and H-18 and Me-27.
Further detailed analyses of NMR data (Fig. S13.9) indicated structure of 52 as depicted
in Figs. 4 and S13.9. This nortriterpene skeleton has not previously been isolated from
nature. Compound 52 was previously prepared artificially from -peltoboykinolic acid
(-amyrin homologue with a carboxyl group at the C-14 position) via heat treatment of
the carboxylic acid at 230 °C, resulting in the production of 52 by CO2 gas evolution.[14]
Detailed NMR analyses (600 MHz, C6D6) indicated that, like product 50, 53 acetate
had a -amyrin skeleton, but with the C-14 Me group missing. The HMBC spectrum
showed the following correlations: Me-26 (H 0.974, s, 3H)/C-14 (C 48.8, d), H-14 (H
1.97, m, 1H)/C-13 (C 141.1, s), H-14/C-12 (C 121.1, d), H-18 (H 2.16, dd, 12.9, 4.0
Hz)/C-13, H-18/C-12, and H-19 (H 1.84, m, 1H)/C-13, verifying that the double-bond
position was C-12/C-13. A clear NOE between H-9 (H 1.00, m, 1H) and H-14 indicated
that H-14 was -oriented. Further detailed NMR analyses (Fig. S14.9) allowed us to
propose that 53 was 27-nor--amyrin, as shown in Fig. 4 and S14.9.
A double bond was also involved in 54 acetate. Both Me-28 (H 1.21, s, 3H) and
Me-29 (H 1.21, s, 3H) had HMBC cross-peaks for C-19 (C 127.8, d), while Me-27 (H
1.19, s, 3H) also exhibited HMBC correlation with C-18 (C 144.6, s), supporting that
the double-bond position was C-18/C-19. This finding suggested a germanicol skeleton
for product 54, as well as in the carbon framework of 51. Unambiguous NOEs were
found for Me-26 (H 0.948, s, 3H)/H-13 (H 2.07, br t, J=11.4 Hz, 1H)/Me-27 (H 1.19,
s, 3H), and for H-9 (H 0.605, br d, J=11.4 Hz, 1H)/H-14 (H 0.79, m, 1H), confirming
the -orientation of H-13 and -disposition of H-14. These detailed analyses indicated
that product 54 was 27-norgermanicol (see also Fig. S15.9).
We have previously reported the incubation experiment of 13 by EtAS and
elucidated the structure of product 55.[10,13] Isolation of product 56 was very difficult,
because the conversion yield from 10 was significantly low (approx. 0.80%, see Fig. 3).
Figure S16 shows the electron ionisation mass spectrometry (EIMS) results for product
56 (A) and authentic -amyrin 2 (B). The mass spectrum of 56 was superimposable on
that of 2, except for the molecular weight (M+), suggesting that product 56 had a
-amyrin scaffold. The fragment ions at m/z 203 and 218 present in the EIMS spectrum
of 2, were assigned as shown in Fig. S16 (C).[5d,15] These characteristic ions were also
found in the spectrum of 56, strongly indicating that substrate 10 underwent the full
polycyclization reaction to yield 56. Thus, product 56 can be proposed to be
26-nor--amyrin.
Discussion
8
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
All products, 49–54 and 56, were produced from nor-oleanyl cations (6’, the final
intermediates). Formation mechanisms are shown in Scheme 3. Products 50, 53, and 56,
possessing -amyrin scaffolds, were generated as shown in Scheme 1, via 1,2-shifts of
H-18 to the C-19 cation and H-13 to the C-18 cation in an antiperiplanar fashion,
followed by H-12 proton elimination, affording the double bond at the C-12/C-13
position.
<Scheme 3>
Non-natural product 52 was produced by deprotonation at H-14, furnishing the double
bond at C-13/C-14. Germanicol skeletons 51 and 54 were provided by deprotonation at
H-18 in cation 6’, affording the double bond at C-18/C-19. Notably, the major products
of analogues 11 and 12 possess -amyrin scaffolds, indicating that the less-bulky
hydrogen atom substituents at C-10 and C-15 did not provide significant perturbation of
the normal folding conformation of the substrates, although aberrant polycyclisation
products, in which the double-bond positions were different, were also generated. The
substantial amounts of products (22–34%, see Fig. 3) generated from 11 and 12
suggested that time-dependent inhibition by 11 and 12 was small or negligible. However,
the conversion yield of analogue 10 was only 0.80%, clearly indicating that substrate
analogue 10 underwent a time-dependent inhibition, as observed in the incubation
experiment of 13.10 In contrast, analogues 7–9, with the bulkier ethyl group
(C1-appendage) substituents at C-6, C-10, and C-15, underwent no cyclisation reaction,
demonstrating that the EtAS enzyme was unable to accept the larger ethyl substituent.
However, analogues 14–16, with ethyl-group substituents at C-19 and C-23, did
undergo polycyclisation, as we reported before.[10,12] The catalytic site domain
responsible for the earlier cyclisation stages (mono-, bi-, and tricyclic stages) would be
tightly packed (compact). Thus, the larger ethyl moieties could not be accepted due to
steric repulsion, preventing cyclisation. This indicated that the correct size (Me group)
at C-6, C-10, and C-15 positions in 1 is required for a normal polycyclisation cascade.
Conversely, the catalytic domain involved in later cyclisation steps is somewhat loosely
packed (less compact), and thus able to accommodate the bulkier ethyl group.
Next, we discuss the comparison of the substrate specificities of -amyrin
synthase (EtAS) and lanosterol synthases (LaS). Figures S17 and S18 depict the
enzymatic products of analogues 7–18 by EtAS and LaS, respectively. The enzymatic
cyclisation reactions of substrate analogues by -amyrin synthase have only recently
been disclosed.[7–12] In contrast, the many investigations into lanosterol synthase go
9
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
back to the 1960s. Corey et al. and van Tamelen et al. examined the cyclisation
reactions of 1[16] and 12 by LaS,[17] which were converted into 19-norlanosterol and
18-norlanosterol, respectively. Very recently, we reported the incubation experiment of
11 with LaS, resulting in the production of 30-norlanosterol and
(14E,17E)-(13H)-26-normalabarica-14,17,21-triene consisting of a 6/6/5-fused
tricyclic ring system, with a product ratio of 9:1.[13] A normal folding conformation
(chair–boat–chair–chair) and the correct backbone rearrangements resulted in
production of the 30-norlanosterol scaffold. However, the minor tricyclic product was
generated via an unusual chair–chair–boat folding conformation.[13] This suggested that
replacement of the less bulky hydrogen atom at C-10 afforded the aberrant cyclisation
product. Furthermore, bulky ethyl-substituted analogue 8 afforded some mono- and
tricyclic products (see Fig. S18.1).[18] Thus, the correct steric bulk of Me-27 was
required for the correct folding of 1 to yield lanosterol. Analogue 9 also underwent
polycyclisation to afford the tricyclic product and homolanosterol with the Et group at
C-13 (see Fig. S18.1), indicating that a precise steric bulk at C-15 was necessary for the
normal folding conformation. LaS tolerated the bulky ethyl-substituted analogues (8 and
9), as described above, but EtAS never did accept the bulky analogues (7–9) as shown
in Fig. 3, indicating that the reaction cavity of EtAS was more compact than that of LaS.
Squalene is converted into hopene, which is produced by Alicyclobacillus
acidocaldarius squalene-hopene cyclase (SHC).[2a, 2b] The hopane skeleton, consisting
of the pentacyclic 6/6/6/6/5-fused ring system, is biosynthesised via the all-chair folding
conformation of squalene. Figure S19 illustrates the enzymatic products of norsqualenes.
26-Norsqualene is transformed into 26-norhopene.[19] The incubation of 27-norsqualene
by SHC affords unprecedented cyclisation products consisting of 6/5+5/5+(6) ring
systems,[20] with a cyclisation pathway quite different to those of other triterpene
cyclases. [20]28-Norsqualene afforded hopene and isohopene scaffolds as the enzymatic
products, indicating that the effect of Me-28 on the cyclisation pathway was less
significant than that of Me-27. The enzymatic reaction of 10(15)-ethyl-substituted
squalene with SHC gave mono-, bi-, and tricyclic products, but did not afford tetra- or
pentacyclic products (our unpublished results), indicating that the precise steric bulk at
C-10 or C-15 was required for the normal squalene cyclisation pathway to occur.
Detailed results will be reported in due course.
Conclusion
In summary, we have successfully investigated the influence of sterics on the
polycyclisation pathway of oxidosqualene by-amyrin synthase. Specifically, steric
10
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
bulk at the C-6, C-10, and C-15 positions was examined by substitution of the Me group
with a less bulky hydrogen atom and a larger ethyl group. -Amyrin synthase did not
tolerate the bulkier ethyl-substituted analogues 7–9, indicating that the catalytic domain
responsible for the earlier cyclisation steps was tightly packed. Other triterpene cyclases,
such as LaS and SHC, were able to accept 7–9 and their squalene analogues
(deoxy-derivatives of 7–9), respectively, to yield the cyclisation products. Thus, LaS
and SHC had larger reaction spaces and more flexible substrate specificities than those
of EtAS. The superiority of substrate flexibility by SHC is further supported by the fact
that SHC can accept truncated analogues, such as farnesol and geranylgeraniol, as
substrates,[22] but no or little reaction of the truncated analogues (epoxy-derivatives)
occurred with EtAS (our unpublished results). However, abortive cyclisation products
were accumulated by LaS and SHC without the polycyclisation going to completion.
Therefore, the catalytic site domain involved in the earlier cyclisation steps was
revealed to be significantly compact, irrespective of the type of triterpene cyclase. This
tight binding of Me-26, Me-27, and Me-28 with the catalytic domains possibly leads to
the highly efficient polycyclisation cascade of (oxido)squalene substrates.
Experimental
Analytical Methods
NMR spectra of the enzymatic products were recorded in C6D6 on a Bruker DMX 600
and DPX 400 spectrometers, the chemical shifts being given in ppm relative to the
solvent peak δH =7.28 and δC=128.0 ppm as the internal reference for 1H- and 13C NMR
spectra, respectively. The coupling constants J are given in Hz. GC analyses were done
on a Shimadzu GC2014 chromatograph equipped with a flame ionization detector (a
DB-1 capillary column, 30m x 0.25 mm x 0.25 m, J&W Scientific. Inc.). GC-MS
spectra were on a JEOL SX 100 or a JEOL JMS-Q1000 GC K9 instrument equipped
with a ZB-5ms capillary column (30m x 0.25 mm x 0.25 m; Zebron) by using the EI
mode operated at 70 eV. High resolution-mass spectrometry (HR-MS) was performed
on a JMS-T100GCV using electron ionization (EI) mode. HPLC was carried out with
Hitachi L-1700 (pump) and L-7405 (UV detector), the HPLC peaks having been
monitored at 210 or 214 nm. Specific rotation values were measured with a Horiba
SEPA-300 polarimeter.
Incubation experiments of analogues 7~12.
The purification of EtAS and the incubation conditions were carried out according to
11
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
the published protocol. [4] The detailed incubation conditions are described in the text.
Synthetic experiments for substrate analogues 7~12
The detailed synthetic procedures and the spectral data of the synthetic compounds are
described in Supporting Information.
Spectroscopic data of enzymatic products 49─54.
Product 49 acetate. 1H NMR (600 MHz, C6D6): =0.810 (3H, s, Me-25), 0.874 (1H, dd,
J=12.0, 2.0 Hz, H-5), 0.978 (3H, s, Me-28), 1.01 (1H, m, H-1), 1.05 (1H, m, H-11), 1.05
(6H, s, Me-23 & Me-24), 1.06 (1H, m, H-9), 1.12 (3H, s, Me-29), 1.17 (1H, m, H-7),
1.18 (1H, m, H-8), 1.20 (3H, s, Me-26), 1.23 (3H, s, Me-27), 1.30 (1H, m, H-6), 1.53
(1H, m, H-21), 1.54 (1H, m, H-16), 1.56 (1H, m, H-22), 1.57 (1H, m, H-15), 1.63 (1H,
m, H-1), 1.66 (2H, m, H-16 & H-21), 1.67 (1H, m, H-22), 1.69 (1H, m, H-11), 1.70 (1H,
m, H-2), 1.76 (1H, m, H-15), 1.83 (1H, m, ax.), 1.88 (3H, s, CH3CO), 1.88 (2H, m, H-2
& H-6), 1.89 (1H, m, H-19), 1.91 (1H, m, H-7), 2.47 (1H, d, J=12.2 Hz, H-19), 2.78
(1H, bd, J=14.2 Hz, eq.), 4.85 (1H, dd, J=11.8, 4.2 Hz, H-3). 13C NMR (150 MHz,
C6D6): =14.3 (q, C-25), 17.1 (q, C-24), 20.7 (q, CH3CO), 20.8 (q, C-26), 21.5 (t, C-6),
24.1 (q, C-28), 24.4 (t, C-2), 24.5 (q, C-27), 25.5 (t, C-12), 26.7 (t, C-11), 27.6 (t, C-7),
28.3 (q, C-23), 32.6 (q, C-29), 33.4 (s, C-20), 34.8 (t, C-15), 35.0 (s, C-17), 35.8 (t,
C-21), 36.8 (s, C-10 & t, C-22), 37.4 (t, C-1), 37.9 (s, C-4), 38.9 (t, C-19), 39.1 (s,
C-14), 39.7 (t, C-16), 48.7 (d, C-8), 50.1 (d, C-9), 54.3 (d, C-5), 80.6 (d, C-3), 131.6 (s,
C-18), 137.7 (s, C-13), 169.9 (s, CH3CO). MS (EI) of 49 (not acetate): see Fig. S10.1.
HRMS (EI) of the acetate: m/z: calcd. for C31H50O2: 454.38108; found: 454.38048;
[]D20 = +22.4 (c = 0.12, CHCl3)
Product 50 acetate. 1H NMR (400 MHz, C6D6): =0.840 (3H, s, Me-25), 0.902 (1H,
J=dd, 12.4, 2.8 Hz, H-5), 0.94 (1H, m, H-16), 0.98 (1H, m, H-1), 1.05 (3H, s, Me-23),
1.069 (6H, s, Me-24 & Me-27), 1.074 (3H, s, Me-29), 1.10 (3H, s, Me-28), 1.14 (1H, m,
H-7), 1.17 (3H, s, Me-26), 1.26 (1H, m, H-9), 1.28 (1H, m, H-22), 1.34 (2H, m, H-6 &
H-8), 1.36 (1H, m, H-19), 1.42 (1H, m, H-21), 1.49 (1H, m, H-1), 1.51 (1H, m, H-22),
1.56 (2H, m, H-15), 1.63 (1H, m, H-21), 1.64 (1H, m, H-6), 1.66 (1H, m, H-2), 1.83
(1H, m, H-7), 1.84 (1H, m, H-2), 1.88 (3H, s, CH3CO), 1.93 (2H, m, H-11), 1.95 (1H, m,
H-19), 2.18 (1H, m, H-16), 2.23 (1H, m, H-18), 4.83(1H, dd, J=11.6, 4.4 Hz, H-3), 5.47
(t, J=3.6 Hz, H-12). 13C NMR (100 MHz, C6D6): =14.1 (q, C-25), 17.1 (q, C-24), 20.9
(q, CH3CO), 21.1 (t, C-6), 22.8 (q, C-26), 23.8 (q, C-29), 24.2 (t, C-2), 24.9 (t, C-11),
26.1 (t, C-7), 27.4 (t, C-16), 28.3 (q, C-23), 28.8 (q, C-27), 31.3 (s, C-20), 32.9 (s, C-17),
33.6 (q, C-28), 34.7 (t, C-15), 35.0 (t, C-22), 35.9 (s, C-14), 36.9 (t, C-1), 37.1 (s, C-10),
12
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
37.2 (t, C-21), 37.8 (s, C-4), 44.8 (d, C-8), 45.9 (d, C-9), 46.2 (t, C-19), 48.7 (d, C-18),
54.3 (d, C-5), 80.6 (d, C-3), 123.1 (d, C-12), 146.2 (s, C-13), 169.9 (s, CH3CO). MS
(EI) of 50 (not acetate): see Fig. S11.1. HRMS (EI) of the acetate: m/z: calcd. for
C31H50O2: 454.38108; found: 454.38107; []D20 = +99.6 (c = 1.02, CHCl3). m.p.
154─158C.
Product 51 acetate. 1H NMR (600 MHz, C6D6): =0.817 (3H, s, Me-26), 0.835 (1H, dd,
J=12.4, 2.1 Hz, H-5), 0.882 (3H, s, Me-25), 0.92 (1H, m, H-9), 0.97 (1H, m, H-1), 1.03
(3H, s, Me-23), 1.04 (2H, m, H-7 & H-11), 1.07 (3H, s, Me-24), 1.17 (1H, m, H-8), 1.20
(6H, s, Me-28 & Me-29), 1.23 (3H, s, Me-27), 1.32 (1H, m, H-12), 1.33 (1H, m, H-6),
1.42 (1H, m, H-15), 1.51 (1H, m, H-16), 1.53 (1H, m, H-21), 1.58 (1H, m, H-16), 1.62
(1H, m, H-1), 1.63 (1H, m, H-22), 1.64 (1H, m, H-6), 1.65 (1H, m, H-12), 1.66 (1H, m,
H-21), 1.68 (1H, m, H-23), 1.71 (1H, m, H-15), 1.72 (1H, m, H-11), 1.73 (1H, m, H-2),
1.88 (1H, m, H-2), 1.89 (3H, s, CH3CO), 1.93 (1H, m, H-7), 1.94 (1H, bd, J=11.8 Hz,
H-13), 4.82 (1H, dd, J=11.8, 4.5 Hz, H-3), 5.19 (1H, s, H-19). 13C NMR (150 MHz,
C6D6): =13.1 (q, C-26), 14.3 (q, C-25), 17.1 (q, C-24), 20.9 (q, CH3CO), 21.6 (t, C-6),
24.4 (t, C-2), 24.8 (t, C-12), 25.4 (q, C-27), 25.5 (t, C-11), 27.2 (t, C-7), 28.3 (q, C-23),
29.4 (q, C-29), 31.5 (q, C-28), 32.6 (s, C-20), 33.8 (t, C-21), 34.6 (s, C-17), 35.6 (t,
C-15), 36.8 (s, C-10), 37.2 (t, C-1), 37.3 (s, C-4), 37.9 (t, C-22), 38.1 (t, C-16), 39.4 (s,
C-14), 46.3 (d, C-13), 47.2 (d, C-8), 49.3 (d, C-9), 54.4 (d, C-5), 80.6 (d, C-3), 129.9 (d,
C-19), 142.4 (s, C-18), 169.9 (s, CH3CO). The assignment of C-28 and C-29 may be
exchangeable, because of very similar chemical shifts. MS (EI) of 51 (not acetate): see
Fig. S12.1. HRMS (EI) of the acetate: m/z: calcd. for C31H50O2: 454.38020; found:
454.38107; []D20 = +56.9 (c = 0.31, CHCl3).
Product 52 acetate. 1H NMR (600 MHz, C6D6): =0.77 (1H, dd, J=12.2, 2.0, H-5),
0.88(1H, m, H-1), 0.896 (3H, s, Me-25), 1.03 (1H, m, H-9), 1.035 (6H, s, Me-23 &
Me-24), 1.05 (1H, m, H-16), 1.076 (3H, s, Me-27), 1.081(3H, s, Me-29), 1.09 (3H, s,
Me-28), 1.10 (3H, s, Me-26), 1.18 (1H, ddd, J=12.6, 12.6, 3.6, H-7), 1.29 (1H, m, H-21),
1.30 (1H, m, H-19), 1.38 (1H, m, H-6), 1.42 (1H, m, H-22), 1.48 (1H, m, H-11), 1.54
(1H, m, H-6), 1.57 (1H, m, H-21), 1.59 (1H, m, H-11), 1.64 (1H, m, H-1), 1.67 (1H, m,
H-22), 1.72 (1H, m, H-18), 1.73 (1H, m, H-19), 1.74 (1H, m, H-2), 1.86 (1H, m, H-12),
1.87 (1H, m, H-2), 1.88 (1H, m, H-7), 1.89 (3H, s, CH3CO), 2.09 (1H, m, H-15), 2.10
(1H, m, H-16), 2.19 (1H, m, H-15), 2.40 (1H, m, H-12), 4.80 (1H, dd, J=11.8, 4.5 Hz,
H-3). 13C NMR (150 MHz, C6D6): =16.6 (2C, q, C-24 & C-25), 18.1 (d, C-11), 18.7 (t,
C-6), 20.3 (t, C-15), 20.6 (q, C-26), 20.8 (q, CH3CO), 24.1 (t, C-2), 24.7 (q, C-29), 27.2
(2C; q, C-27; t, C-16), 28.0 (q, C-23), 31.06 (s, C-20), 31.10 (s, C-17), 32.4 (t, C-12),
33.2 (q, C-28), 34.9 (t, C-21), 36.5 (t, C-22), 37.3 (s, C-4), 37.8 (t, C-1), 37.9 (s, C-10),
13
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
38.0 (s, C-8), 39.5 (t, C-7), 43.2 (t, C-19), 45.2 (d, C-18), 55.1 (d, C-5), 56.2 (d, C-9),
80.5 (d, C-3), 130.0 (s, C-13), 135.5 (s, C-14), 169.9 (s, CH3CO). The assignments of
C-8 and C-10 and those of C-17 and C-20 may be exchangeable, because of the very
close chemical shifts. MS (EI) of 52 (not acetate): see Fig. S13.1. HRMS (EI) of the
acetate: m/z: calcd. for C31H50O2: 454.38042; found: 454.38107; []D20 = +168.6 (c =
0.06, CHCl3).
Product 53 acetate. 1H NMR (600 MHz, C6D6): =0.793 (1H, m, H-5), 0.862 (1H, ddd,
J=13.0, 13.0, 3.6Hz, H-1), 0.934 (3H, s, Me-25), 0.974 (3H, s, Me-26), 1.00 (1H, m,
H-9), 1.03 (1H, m, H-16), 1.04 (3H, s, Me-24), 1.06 (3H, s, Me-23), 1.086 (3H, s,
Me-29), 1.09 (1H, m, H-7), 1.10 (3H, s, Me-27), 1.11 (3H, s, Me-28), 1.28 (1H, m,
H-19), 1.29 (1H, m, H-21), 1.38 (2H, m, H-6 & H-22), 1.54 (1H, m, H-1), 1.55 (1H, m,
H-6), 1.58 (2H, m, H-15 & H-21), 1.59 (1H, m, H-22), 1.70 (1H, m, H-2), 1.82 (1H, m,
H-2), 1.84 (2H, m, H-11 & H-19), 1.88 (3H, s, CH3CO), 1.89 (1H, m, H-15), 1.93 (1H,
m, H-7), 1.94 (1H, m, H-11), 1.97 (1H, m, H-14), 2.07 (1H, m, H-16), 2.19 (1H, dd,
J=12.9, 4.0Hz, H-18), 4.81 (1H, dd, J=11.8, 4.5 Hz, H-3), 5.54 (1H, bs, H-12). 13C
NMR (150 MHz, C6D6): =15.3 (q, C-26), 16.3 (q, C-25), 16.9 (q, C-24), 18.7 (t, C-6),
20.7 (t, C-15), 20.83 (q, CH3CO), 22.82 (t, C-11), 24.0 (t, C-2), 24.5 (q, C-29), 28.2 (q,
C-23), 28.4 (q, C-27), 29.4 (t, C-16), 31.2 (s, C-20), 32.8 (s, C-17), 33.5 (q, C-28), 34.8
(t, C-21), 35.7 (s, C-8), 36.89 (t, C-22), 36.94 (s, C-10), 37.9 (s, C-4), 38.1 (t, C-1), 41.8
(t, C-7), 44.7 (t, C-19), 48.2 (d, C-18), 48.8 (d, C-14), 55.39 (d, C-5), 55.43 (d, C-9),
80.4 (d, C-3), 121.1 (d, C-12), 141.1 (s, C-13), 169.9 (s, CH3CO). MS (EI) of 53 (not
acetate): see Fig. S14.1. HRMS (EI) of the acetate: m/z: calcd. for C31H50O2: 454.38057;
found: 454.38107; []D20 = +76.7 (c = 0.28, CHCl3). m.p. 152─155C.
Product 54 acetate. 1H NMR (600 MHz, C6D6): =0.60 (1H, bd, J=11.4 Hz, H-9), 0.77
(1H, bd, J=11.3 Hz, H-5), 0.79 (1H, m, H-14)), 0.87 (1H, m, H-7), 0.88 (3H, s, Me-25),
0.90 (1H, m, H-1), 0.95 (3H, s, Me-26), 1.04 (3H, s, Me-24), 1.05 (3H, s, Me-23), 1.19
(3H, s, Me-27), 1.208 (3H, s, Me-29), 1.214 (3H, s, Me-28), 1.28 (1H, m, H-11), 1.35
(1H, m, H-6), 1.36 (1H, m, H-15), 1.40 (1H, m, H-16), 1.51 (1H, m, H-6), 1.54 (1H, m,
H-21), 1.55 (2H, m, H-11 & H-12), 1.62 (1H, m, H-22), 1.63 (1H, m, H-1), 1.68 (2H, m,
H-21 & H-22), 1.69 (1H, m, H-16), 1.75 (1H, m, H-2), 1.79 (1H, m, H-7), 1.889 (3H, s,
CH3CO), 1.89 (1H, m, H-12), 1.92 (1H, m, H-2), 1.93 (1H, m, H-15), 2.07 (1H, bdd,
J=11.4, 11.4 Hz, H-13), 4.84 (1H, dd, J=11.9, 4.5 Hz, H-3), 5.28 (1H, bs, H-19). The
assignments of H-16 and H-22 may be exchangeable. 13C NMR (150 MHz, C6D6):
=15.1 (q, C-26), 16.7 (2C, q, C-24 & C-25), 18.5 (t, C-6), 20.76 (t, C-12), 20.79 (2C; q,
CH3CO; t, C-11), 24.1 (t, C-2), 24.9 (q, C-27), 28.1 (q, C-23), 29.5 (q, C-29), 31.3 (t.
C-15), 31.6 (q, C-28), 32.6 (s, C-20), 33.7 (t, C-21), 34.8 (s, C-17), 36.0 (d, C-13), 37.2
14
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
(s, C-10), 37.6 (s, C-8), 37.8 (t, C-22), 38.0 (s, C-4), 38.1 (t, C-1), 41.2 (t, C-7), 42.3 (t,
C-16), 55.4 (d, C-5), 57.5 (d, C-14), 59.4 (d, C-9), 80.6 (d, C-3), 127.8 (d, C-19), 144.6
(s, C-18), 169.9 (s, CH3CO). The assignments of C-11 and C-12, those of C-16 and
C-22 and those of C-28 and C-29 may be interchangeable, because of the very similar
chemical shifts. MS (EI) of 54 (not acetate): see Fig. S15.1. HRMS (EI) of the acetate:
m/z: calcd. for C31H50O2: 454.38014; found: 454.38107; []D20 = +43.8 (c = 0.05,
CHCl3).
Acknowledgements
This work was supported in part by Grant-in-Aid for Scientific Research from Japan
Society for the Promotion of Science, Nos. 25450150 (C) and 18380001(B).
References
[1] A. Eschenmoser, L. Ruzicka, O. Jeger and D. Arigoni, Helv. Chim. Acta 1955, 38,
1890–1904.
[2] Review. (a) K. U. Wendt, G. E. Schulz, E. J. Corey and D. R Liu, Angew. Chem., Int.
Ed. 2000, 39, 2812–2833. (b) T. Hoshino and T. Sato, Chem. Commun. 2002,
291–301. (c) R. A. Yoder and J. N. Johnston, Chem. Rev. 2005, 105, 4730–4756. (d)
I. Abe, Nat. Prod. Rep. 2007, 24, 1311–1331. (e) T.-K. Wu, C.-H. Chang, Y.-T. Liu
and T.-T. Wang, Chem. Rec. 2008, 8, 302–325. (f) W. D. Nes, Chem. Rev. 2011, 111,
6423–6451.
[3] R. Xu, G. C. Fazio and S. P. T. Matsuda, Phytochemistry 2004, 65, 261–291.
[4] R. Ito, Y. Masukawa and T. Hoshino, FEBS J. 2013, 280, 1267–1280.
[5](a) Rees, H. H.; Mercer, E. I.; Goodwin, T. W., Biochem. J. 1966, 99, 726-734. (b) H.
H. Rees, G. T. Britton and T. W. Goodwin, Biochem. J. 1968, 106, 659-665. (c) D.
H. R. Barton, G. Mellows, D. A. Widdowson, and J. J. Wright, J. Chem. Soc., C
1971, 1142-1148. (d) T. Suga and T. Shishibori, Phytochemistry, 1975, 14,
2411-2417. (e) S. Seo, Y. Tomita and K. Tori, J. Am. Chem. Soc., 1981, 103,
2075-2780. (f) S. Seo, Y. Yoshimura, A. Uomori, K. Takeda, H. Seto, Y. Ebizuka
and U. Sankawa, J. Am. Chem. Soc., 1988, 110, 1740-1745.
[6] E. J. Corey and S. K. Gross, J. Am. Chem. Soc., 1968, 90, 5045-5046.
[7] I. Abe, Sakano, H. Tanaka, W. Lou, H. Noguchi, M. Shibuya and Y. Ebizuka, J.
Am. Chem. Soc., 2004, 126, 3426-3427.
[8] I. Abe, Y. Sakano, M. Sodeyama, H. Tanaka, H. Noguchi, M. Shibuya and Y.
15
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Ebizuka, J. Am. Chem. Soc., 2004, 126, 6880–6881.
[9] H. Noma, H. Tanaka, H. Noguchi, M. Shibuya, Y. Ebizuka and I. Abe, Tetrahedron
Lett., 2004, 45, 8299-8301.
[10] T. Hoshino, Y. Yamaguchi, K. Takahashi, R. Ito, Organic Lett., 2014, 16,
3548-3551.
[12] T. Hoshino, Y. Miyahara, M. Hanaoka, K. Takahashi, I. Kaneko, Chem. Eur. J.
2015, 21, 15769-15784.
[13] I. Kaneko and T. Hoshino, J. Org. Chem., 2016, 81, 6657-6671.
[14] T. Hoshino, A. Chiba and N. Abe, Chem. Eur. J. 2012, 18, 13108-13116.
[15] M. Nagai, K. Izawa, T. Inoue, Chem. Pharm. Bull., 1969, 17, 1438-1443.
[16] (a) H. Budzikiewicz, J. M. Wilson and C. Djerassi, J. Am. Chem. Soc., 1963, 85,
3688-3699. (b) J. Karlinen and C. Djerassi, J. Org. Chem., 1966, 31, 1945-1946.
[17] E. J. Corey, A. Krief, H. Yamamoto, J. Am Chem. Soc., 1971, 93, 1493-1494.
[18] E. E. van Tamlen, R. P. Hanzlik, K. B. Sharpless, R. B. Clayton, W. J. Richter, J.
Am Chem. Soc., 1968, 90, 3284-3286.
[19] T. Hoshino and Y. Sakai, Tetrahedron Lett., 2001, 42, 7319-7623.
[20] S. Nakano, S. Ohashi, and T. Hoshino, Org. Biomol. Chem., 2004, 2, 2012-2022.
[21] T. Hoshino and S. Ohashi, Org. Lett., 2002, 4, 2553-2556.
[22] T. Hoshino, Y. Kumai, I. Kudo, S. Nakano and S. Ohashi, Org. Biomol. Chem.,
2004, 2, 2650-2657.
16
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Figures and Schemes
Scheme 1. Polycyclisation pathway of (3S)-2,3-oxidosqualene (1) into -amyrin (2).
17
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Figure 1. Structures of substrate analogues discussed in the text.
18
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Scheme 2. Synthetic schemes for substrate analogues 8 and 9, and mixtures of 7 and 14,
and 10 and 13. (i) triethyl-2-phosphonobutyrate ((EtO)2P(O)CH(Et)CO2Et), n-BuLi, rt;
(ii) iBu2AlH/Et2O, –40 °C; (iii) PBr3, Et2O, 0 °C; (iv) NBS, THF-H2O, 0 °C; (v)
PhSO2Na, DMF, rt; (vi) K2CO3, MeOH, rt; (vii) n-BuLi, THF–HMPA (9:1), –78 °C;
(viii) LiBEt3H, Et2O, PdCl3(dppp), 0 °C; (ix) triethyl-2-phosphonoacetate
((EtO)2P(O)CH2CO2Et).
19
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
2: β-amyrin
Squalene
(SQ)
Substrate 1
50
51
SQ
×2
49
Analog 11
53
52
SQ
SQ
54
×4
Analog 12
Analog 13
55
×4
Analog
10 &13
SQ
55
×8
56
18
21
Retention time (min)
15
36
12
24
27
30
33
Figure 2. GC traces of hexane extracts from the incubation mixtures 1 and 11–13, and a
mixture of 10 and 13. SQ denotes squalene (internal standard).
GC conditions: DB-1 capillary column (30 m × 0.25 mm); injection temp., 300 °C;
column temp., 190–250 °C (10 °C/min) and 250–260C (0.35 °C/min); flow rate (He
gas), 1.50 mL/min.
20
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Figure 3. Conversion yields of substrate analogues relative to that of 1. Relative
activities were determined by estimating the total amounts of each product.
Substrate analogues 11 and 12 produced 49–51 and 52–54, respectively. The production
ratios were as follows: 49/50/51 = 1:4:2; 52/53/54 = 2:8:1. Conversion yields represent
the total amounts of enzymatic products. Substrate 14 was converted into 47 and 48 in a
1:3 ratio, thus the total amount of 47 and 48 was calculated using the conversion yield
of substrate 14. A mixture of 10 and 13 (100 g each) was incubated under the same
conditions as the other substrates (5 g EtAS at 30 °C for 9 h). Conversion yields for 10
and 13 were determined as 0.8% and 3.4%, respectively, from the peak areas of 56 and
55.
21
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Figure 4. Products obtained from substrate analogues 7–14. Carbon numbering systems
are shown to aid interpretation of NMR data.
22
This article is protected by copyright. All rights reserved

European Journal of Organic Chemistry
10.1002/ejoc.201601306
Scheme 3. Enzymatic product formation mechanisms
23
This article is protected by copyright. All rights reserved