Novel autophagy modulators: Design and synthesis of (+)-epogymnolactam analogues and structure-activity relationship

Article abstract of DOI:10.1016/j.bmc.2018.09.013

(+)-Epogymnolactam (1) was discovered as a novel autophagy inducer from a culture of Gymnopus sp. in our laboratory. To determine structure-activity relationships among (+)-epogymnolactam analogues comparing with cerulenin (2), we synthesized 5 analogues

Full text of DOI:10.1016/j.bmc.2018.09.013

Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Novel autophagy modulators: Design and synthesis of (+)-epogymnolactam
analogues and structure-activity relationship
⁎
Kazuki Ueda, Yuji Okado, Kengo Shigetomi, Makoto Ubukata
Graduate School of Agriculture, Hokkaido University, Kita-9, Nishi-9, Kita-ku, Sapporo 060-8589, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
(+)-Epogymnolactam (1) was discovered as a novel autophagy inducer from a culture of Gymnopus sp. in our
laboratory. To determine structure-activity relationships among (+)-epogymnolactam analogues comparing
with cerulenin (2), we synthesized 5 analogues including (−)-epogymnolactam (3) having each different
functional group, and 3 analogues with different side-chain lengths. Five analogues, 3, 4, 5, 6, and 7 did not
significantly increase the ratio of LC3-II to LC3-I as an autophagy marker in NIH3T3 cells. These results suggest
that presence and stereochemistry of (2R,3S)-epoxy group and cyclic syn-form (1b) of 1 are important for the
activity as autophagy inducer. Hexyl analogue (8) as well as 1 having butyl side-chain dose-dependently in-
creased the ratio of LC3-II to LC3-I, whereas octyl analogue (9) and 2 rather decreased the ratio. Decyl analogue
Epogymnolactam
Autophagy inducer
Autophagy inhibitor
Structure activity relationship
NIH3T3 cells
(
10) did not give a change in the ratio. Although 8 seemed to be an excellent autophagy inducer, it dose-
dependently increased SQSTM1 (p62) as in the case of 2, whereas 1 showed a slight dose-dependent decrease of
p62 as an index of autophagic protein degradation. These observations suggest that 8 is an autophagy modulator
with different molecular target from 1 or 2.
1. Introduction
(+)-epogymnolactam analogues (3–10) to ascertain the structural re-
quirements of 1 as an autophagy-inducing agent. We also propose hexyl
Autophagy inducer and autophagy inhibitor are important for un-
analogue 8 as a new autophagy modulator and discuss several sug-
gestions on the relationship between autophagy and fatty acid synthesis
in mammalian cells.
derstanding precise mechanisms of autophagy and developing ther-
apeutic agents for such diseases as neurodegenerative diseases, cancer,
1
and infectious diseases. (+)-Epogymnolactam (1) was discovered as a
novel autophagy inducer from a culture of Gymnopus sp. in our la-
2. Results and discussion
2
boratory. Total synthesis of 1 was also achieved to confirm its absolute
3
structure and to evaluate further biological properties compared to
To determine the importance of the stereochemistry of 1 as autop-
hagy inducer, we first synthesized (−)-epogymnolactam (3). Epoxy
acetate 12 was obtained from cis-4-benzyloxy-2-buten-1-ol 11 in 98%
ee using Sharpless epoxidation followed by lipase kinetic resolution.
Hydrolysis of 12 afforded enantiopure epoxy alcohol 13, and 3 was
synthesized from 13 according to essentially the same method as the
total synthesis of 13 (Scheme 1).
(+)-Epogymnolactam (1) existed mainly as a cyclic syn-form (1b) in
3 3 3
CD OD and 1b was formed even in CDCl , whereas NMR of 2 in CDCl
showed keto form as shown in Fig. 1 and Supplemental Table S1. To
determine the importance of syn form (1b) for autophagy-inducing
activity, three chain analogues (4–6) and anti-form analogue (7) were
synthesized.
4
cerulenin (2), an inhibitor of fatty acid synthesis. Synthetic 1 showed
the same activity as natural 1 as autophagy inducer. Although the
structure of 1 was similar to that of 2, major tautomer of 1 seemed to be
different from that of 2 as shown in Fig. 1 and Supplemental Table S1.
In addition, a contrasting autophagic response was observed in NIH3T3
cells. Cerulenin (2) dose-dependently decreased LC3-II/LC3-I, a ratio of
lipidate LC3 (LC3-II) to LC3-I as an autophagic marker, and increased
p62/actin, a ratio of SQSTM1 (ubiquitin-binding protein p62) to actin,
5
as an index of autophagic degradation. Thus, cerulenin (2) was
thought to be an autophagy inhibitor, which is discussed in more detail
in the next section.
We, therefore, designed and synthesized 8 analogues of 1, and
evaluated their autophagic activities compared to 1 and 2 in NIH3T3
cells. In this paper, we describe syntheses and biological evaluations of
Ammonolysis of γ-octalactone 14 with liquid ammonia in a pressure
vessel6 gave amide alcohol 15 in 88%. Dess-Martin oxidation of 15
⁎
Corresponding author.
E-mail address: m-ub@for.agr.hokudai.ac.jp (M. Ubukata).
https://doi.org/10.1016/j.bmc.2018.09.013
Received 3 July 2018; Received in revised form 13 September 2018; Accepted 15 September 2018
0968-0896/ © 2018 Elsevier Ltd. All rights reserved.
Please cite this article as: Ueda, K., Bioorganic & Medicinal Chemistry, https://doi.org/10.1016/j.bmc.2018.09.013

K. Ueda et al.
Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
O
Fig. 1. Structures of (+)-epogymnolactam (1)
and cerulenin (2). Both cyclic syn-form 1b and
anti-form 1c as well as keto form 1a were ob-
O
O
H2N
5
7
2
3
6
8
+
1
4
served in CD
keto form of 2 was observed in CDCl
chain tautomerism of 2 was observed in CD
3
OD or even in CDCl
3
, whereas only
. Ring-
OD.
O
N
H
O
O
O
N
H
OH
3
OH
3
1
a
syn-form 1b (major tautomer)
+)-epogymnolactam (1)
anti-form 1c
The tautomeric ratio changed into different
ratio with time.
(
O
O
5
7
9
11
polar solvent
H2N
1
2
3
4
6
8
10
12
O
O
O
N
H
OH
cerulenin (2) (major tautomer)
L-(+)-DIPT,
PPL,
vinyl acetate,
rt
O
O
i
Ti(O Pr)4, TBHP,
o
CH2Cl2 20 C
BnO
OH
BnO
OH
BnO
OAc
68%
81%
11
13 (86% ee)
12
O
O
O
See
Ref. 3
K2CO3, MeOH,
H2N
o
THF, 0 C
BnO
OH
91%
O
O
O
N
OH
H
13 (98% ee)
)-epogymnolactam (3)
Scheme. 1. Synthesis of (−)-epogymnolactam (3).
Dess-Martin
liq. NH3,
periodinane,
5
0oC
o
H2N
H2N
CH Cl , 0 C
2
2
O
O
88%
60%
O
OH
O
O
14
15
deepoxy anlogue (4)
Scheme. 2. Synthesis of deepoxy analogue (4).
afforded deepoxy analogue 4 as shown in Scheme 2. Deepoxy analogue
existed only as a chain form in CDCl (Supplementary Figs. S1 and S2)
considerations suggest that the syn-form (1b) is the active chemical
species in the active site of the unknown primary molecular target.
Attempts to synthesize syn-form analogue corresponding to 7 have not
yet been successful.
4
3
by releasing from the steric constraint by the epoxy ring.
Epoxy amide alcohol (5) was synthesized as the synthetic precursor3
of 1 (Scheme 3). Treatment of lactone 16 with (CH
was subjected to Dess-Martin oxidation to give N,N-dimethyl analogue
6) as indicated in Scheme 3. Chain forms of 5 and 6 were confirmed by
their NMR in CDCl as shown in Supplementary Figs. S3–S6.
Cyclic O-methyl analogue 7 was prepared by treating 1 with cam-
phor sulfonic acid (CSA) in CH OH as shown in Scheme 3. In order to
avoid steric and electronic repulsion, CH OH attacked from the oppo-
site side of the epoxy group in the carbocation intermediate to give only
3
)
2
NH gave 17 which
Next, we synthesized hexyl, octyl, and decyl analogues (8–10) to
investigate the effect of side-chain length on biological responses in
NIH-3T3 cells. Aldehyde 18, a synthetic intermediate of 1, was reacted
3
(
3
with allyltributyltin in the presence of SnCl
cohol 19 and its chlorohydrin.8 Treatment of the by-product chlor-
ohydrin with K CO gave an additional 19, and the overall yield of 19
was 71%. Selective hydrogenolysis of benzyl group of 19 was accom-
4
to afford homoallylic al-
3
2
3
3
9
,10
plished by using lithium 4,4′-di-tert-butylbiphenylide (LiDBB)
to
7
1
anti form (7) as in the case of previously reported observation. H and
give 20 in 61%. Oxidation of 20 with 1,5-dimethyl-9-azanor-
1
3
11
C NMR spectra of 7 indicated its cyclic anti-form as shown in
adamantane N-oxyl (DMN-AZADO)
in combination with [bis
(acetoxy)iodo]benzene (BAIB) gave lactone 21, which was treated
with NH to afford the desired amide 22. Olefin cross-metathesis of 22
with pent-1-ene gave 23 using Grubbs 2nd-generation catalyst.
12
Supplementary Figs. S7 and S8.
All these analogues (3–7) did not enhance LC3-II/LC3-I, the ratios of
LC3-II to LC3-I, and LC3-II/actin, the ratios of LC3-II to actin in NIH-
3
1
3,14
3
T3 cells as shown in Fig. 2. These observations suggest that the un-
Chemoselective hydrogenation of 23 using Pd/C-ethylenediamine (Pd/
C(en)) catalyst15 gave 24 which was converted into desired hexyl
analogue 8 by Dess-Marin oxidation as shown in Scheme 4. Hexyl
known molecular target of 1 recognizes the absolute configuration of
the epoxy group and the cyclic structure of the syn-form (1b).
In the case of ring-chain tautomerism of 1, the major tautomer is
supposed to be syn-form (1b) in polar solution, because amide nitrogen
of 1a attacks C4 carbonyl from the opposite side of the epoxy group to
avoid steric and electronic repulsion (Supplemental Table S1). These
3
analogue 8 mainly existed as a chain form in CDCl as shown in
Supplementary Figs. S9 and S10.
Similarly, decyl analogue 10 was synthesized by using 1-nonene as
an alkene as shown in Scheme 5. NMR analysis indicated that 10 exists
2

K. Ueda et al.
Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
O
H2N
BnO
OH
O
OH
11
epoxy amide alcohol (5)
NH3, CH3OH,
0o
C
Ref. 3
9
8.5%
Dess-Martin
periodinane
O
O
O
(CH3)2NH,
o
N
o
N
CH3OH, 0 C
CH2Cl2, 0 C
O
6
5%
90%
O
O
OH
O
O
O
H
N,N-dimethyl analogue (6)
16
17
Ref. 3
O
CH3OH, CSA
0oC
4
O
O
24%
N
H
N
OCH3
OH
H
O-methyl analogue (7)
(
+)-epogymnolactam (1)
Scheme. 3. Synthesis of epoxy amide alcohol (5), N,N-dimethyl analogue (6), and O-methyl analogue (7).
as a chain form in CDCl
Octyl analogue 9 was prepared by chemoselective hydrogenation of
cerulenin 2 using Pd/C(en) (Scheme 5). Major tautomer of 9 in CDCl
was chain form, while cyclic syn- and anti-forms were also observed as
shown in Supplementary Figs. S11 and S12.
Fig. 3 shows the profiles of LC3-II/LC3-I and LC3-II/actin as au-
tophagic markers after stimulation with 1, 8, 9, 2, and 10 in NIH3T3
cells. Except for 1, only hexyl analogue 8 increased the relative amount
of LC3-II to LC3-I or actin in a dose-dependent manner. In contrast,
octyl analogue 9 and cerulenin 2 were suggested to cause dose-
3
as shown in Supplementary Figs. S13 and S14.
dependent down-regulation, and decyl analogue 10 did not alter the
relative amount of LC3-II.
3
Our preliminary experiments suggested that cerulenin (9) showed
cytotoxicity to NIH3T3 cells at these concentrations. We therefore
evaluated cytotoxicity of 1, 2, 8, 9, and 10 in NIH3T3 cells. Only hexyl
analogue 8 showed lower cytotoxicity than 2, 9, and 10 as in the case of
1 (Fig. 4). Inhibition of fatty acid synthase with 2 induced apoptosis in
multiple myeloma cells16 and human breast cancer cells. We observed
17
1
8
significant cell shrinkage as a signal to apoptosis in NIH 3 T3 cells
treated with 9 at 50 or 100 μM for 4 h, while 1 did not cause significant
Fig. 2. Evaluation of importance of functional groups on autophagy-inducing activity of (+)-epogymnolactam (1). NIH3T3 cells were treated for 4 h with DMSO as
the vehicle (control), 5 μM rapamycin (positive control), or indicated concentrations of 1, 3, 4, 5, 6, or 7. LC3-II/LC3-I: relative ratio of LC3-II to LC3-I. LC3-II/actin:
relative ratio of LC3-II to actin. The data shown represent the mean ± SD of triplicate assay. All significant correlation values (p < 0.05) are given with the
indicated significance (*p < 0.05, **p < 0.01) compared to the DMSO control.
3

K. Ueda et al.
Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
Scheme. 4. Synthesis of hexyl analogue (8).
cell shrinkage at these concentrations but 8 showed cell shrinkage at
3
existed as a chain form in CDCl . Despite the close resemblance of the
1
00 μM (data not shown).
biological properties of 1 and 8, the physicochemical properties of both
are different.
NMR experiments showed that (+)-epogymnolactam (1) existed
primarily as a cyclic syn-form in CD
the syn-form was observed even in CDCl
3
OD and a considerable amount of
In addition to LC3-II, SQSTM1 (p62) can also be used as autophagy
marker to especially determine autophagy flux.5 Thus, p62/actin, the
3
, while hexyl analogue 8
1
-nonene,
O
Grubbs 2nd generation,
Pd/C(en), H2,
THF, rt
o
2
H N
CH2Cl2, 50 C
22
43%
O
OH
69%
25
Dess-Martin
periodinane
CH2Cl2, rt.
O
H2N
62%
O
OH
26
O
Pd/C(en), H2,
THF, rt
H2N
2
O
O
quant.
decyl analogue (10)
O
O
H2N
O
O
O
N
H
OH
octyl analogue (9)
Scheme. 5. Synthesis of octyl analogue (9) and decyl analogue (10).
4

K. Ueda et al.
Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
Fig. 3. Effect of side-chain length of
(+)-epogymnolactam analogue on autop-
hagy-inducing activity. NIH3T3 cells were
treated for 4 h with DMSO as the vehicle
(
control), 5 μM rapamycin (positive con-
trol) or indicated concentrations of 1, 8, 9,
, or 10. LC3-II/LC3-I: relative ratio of
2
LC3-II to LC3-I. LC3-II/actin: relative ratio
of LC3-II to actin. The data shown re-
present the mean ± SD of triplicate assay.
All
significant
correlation
values
(p < 0.05) are given with the indicated
significance (* p < 0.05, **p < 0.01)
compared to the DMSO control.
up-regulation or degradation ratio of p62 to actin was evaluated in
NIH3T3 cells treating with 1 and 8. (+)-Epogymnolactam (1) did not
show significant upregulation or downregulation of p62 after 4 h,
whereas hexyl analogue 8 upregulated p62 in a dose-dependent manner
as indicated in Fig. 5. Octyl analogue 9 also upregulated as in the case
of 2, but dose-dependency was not observed, presumably due to its
cytotoxicity.
In conclusion, the study on the structural requirement of 1 indicated
that presence and stereochemistry of (2R,3S)-epoxy group, the mole-
cular shape of the syn-form (1b), and butyl side chain are critical for
inducing autophagy in NIH3T3 cells. We also found that hexyl analogue
(data not shown), the upregulation of p62 at 4 h exposure with 8 can be
explained by cellular stress caused by this compound. Thus, we con-
cluded that 1, 9 (or 2), and 8 are different types of autophagy mod-
ulators which might be useful to determine the relationships between
fatty acid synthesis and autophagy induction.
Zimmerman et al. reported that very long chain fatty acid (VLCFA)
synthesis has important implications for autophagy and cell home-
ostasis in yeast.19 After 1 was discovered as an autophagy inducer,
Régnacq et al. reported that the effect of 2 was biphasic that moderate
doses (2 or 4 μM) restored autophagy and high doses (20 or 40 μM)
inhibited autophagy in lipid droplets (LDs)-deficient yeast. The biphasic
effect of 2 was specific to nitrogen-starvation autophagy, and only the
inhibitory effect of 2 was observed in rapamycin-induced autophagy in
the LDs-deficient mutant.20 Shpilka et al. observed that suppression of
autophagy by 2 correlated with a reduction in the number of LDs in
yeast cells.21 Although these observations suggest that fatty acid
synthesis correlates with autophagy in yeast cells, precise mechanism
has not been elucidated yet.
1
8
is a novel autophagy modulator that differs from 1 as an autophagy
inducer and 2 as an autophagy inhibitor. SQSTM1 protein (p62) serves
as a link between LC3 and ubiquitinated substrates. SQSTM1 and
SQSTM1-bound polyubiquitinated proteins become incorporated into
the completed autophagosome and are degraded in autolysosomes, thus
serving as an index of autophagic degradation. Inhibition of autophagy
correlates with increased levels of p62 in mammals, suggesting that
steady state levels of the protein reflect the autophagic status. In con5-
trast, decreased p62 levels are associated with autophagy activation.
These considerations suggest that 8 could be an inhibitor of autolyso-
some formation. However, many cellular-stresses can induce the for-
mation of ubiquitinated protein aggregates. The p62-positive ag-
gregates are also formed by proteasome inhibition or puromycin
treatment and can be found in cells exposed to rapamycin for extended
The inhibitory activity of 9 on fatty acid synthesis is slightly weaker
22
than that of 2. However, the inhibitory activities of 3-hexenyl ana-
logue (an unsaturated derivative of 8) and 3-butenyl analogue (an
unsaturated derivative of 1) are much weaker than 2.2
,23,24
These ob-
servations suggest that both 2 and 9 inhibit basal autophagy and cell
proliferation with inhibition of fatty acid synthesis, while 1 and 8 cause
different types of autophagic responses together with moderate sup-
pression of cell proliferation without potent inhibition of fatty acid
synthesis in NIH3T3 cells.
5
periods where the rates of autophagy are elevated. Because 1 upre-
gulated p62 at 6 h exposure where the rate of autophagy is promoted
Fig. 4. Antiproliferation activities of
(+)-epogymnolactam (1), hexyl ana-
logue (8), octyl analogue (9), decyl
analogue (10), and cerulenin (2).
NIH3T3 cells were treated with
DMSO as the vehicle (Control) or in-
dicated concentrations of 1, 8, 9, 2, or
10. Relative viable cells were de-
termined using Cell Counting Kit-8.
The data shown represent the
mean ± SD of triplicate assay.
5

K. Ueda et al.
Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
Fig. 5. Effect of (+)-epogymnolactam
(
1) on autophagy flux compared to that
of hexyl analogue (8) or octyl analogue
9). NIH3T3 cells were treated for 4 h
with DMSO as the vehicle (control),
(
5
μM rapamycin (positive control) or
indicated concentrations of 1, 8, 9, or 2.
p62/actin: relative ratio of SEQSTM1
(p62) to actin. The data shown re-
present the mean ± SD of triplicate
assay. All significant correlation values
(p < 0.05) are given with the indicated
significance (*p < 0.05, **p < 0.01)
compared to the DMSO control.
In hepatocytes, autophagy regulates lipid content because: inhibi-
(40 ml), and stirred for 2 h. The resultant solution was filtered through
tion of autophagy increased triacylglycerol (TG) store and LDs; loss of
autophagy decreased TG breakdown; TGs and LD structural proteins co-
localized with autophagic compartments; and LC3 associated with
celite, Et
2
O was added, and the mixture was washed with brine. The
SO and evaporated in
organic layer was dried with anhydrous Na
2
4
vacuo. The residue was subjected to silica gel column chromatography
(EtOAc: hexane = 1: 1) to give epoxy alcohol 13 (2.98 g, 68.4%). The
enantiomeric excess of 13 was determined as 85.7% ee from the peak-
area ratio by chiral HPLC analysis (DAICEL Chiralpak AD-H (ϕ
0.45 cm × 25 cm), 254 nm, EtOH : hexane = 1: 9).
2
5
LDs.
Mobilization of cholesterol and fatty acids from cholesteryl ester
(
CE) and TG stores of all cells requires lipolytic enzymes.26 Kobayashi
et al. demonstrated that a 1: 1 mixture of dinapinones A1 and A2
DPAmix) enhanced neutral lipid degradation together with induction of
2
5
(
[α]
D
= −18.4 (c 1.0, CHCl
3
).
2
7
1
autophagy in Chinese hamster ovary-K1 cells. These results suggest
interrelationship between autophagy and TG breakdown in mammalian
cells.
H NMR (CDCl , 270 MHz): δ 3.18–3.31 (2H, m, H-2 and H-3),
3
3.62–3.75 (4H, m, H-1 and H-4), 4.50–4.64 (2H, m, benzyl), 7.30–7.36
(5H, m, aromatic).
1
3
Thus, the use of (+)-epogymnolactam (1) and hexyl analogue (8) in
combination with cerulenin (2) or octyl analogue (9) may be useful for
determining relationship between autophagy and TG breakdown or
fatty acid synthesis. These compounds may be also useful to develop
therapeutic agents for dyslipidemia, and provide a new perspective on
understanding autophagy induction.
Further systematic studies are necessary to determine the effect of
lipid metabolism on autophagy induction in mammalian cells and to
determine the molecular targets of 1 and 8.
3
C NMR (CDCl , 67.5 MHz): δ 54.7 (C-3), 55.6 (C-2), 60.7 (C-1),
68.0 (C-4), 73.5 (benzyl), 127.8 (aromatic), 128.0 (aromatic), 128.5
(aromatic), 137.4 (aromatic).
FI-MS: m/z = 194.1 [M]+
.
3.1.2. (2S,3R)-4-Benzyloxy-2,3-epoxybutan-1-yl acetate (12)
To epoxy alcohol 13 (1.0 g, 5.1 mmol) in vinyl acetate (39.3 ml),
porcine pancreatic lipase (PPL) was added, and the reaction mixture
was stirred for 14 h. The reaction mixture was filtered through celite
and evaporated in vacuo. The residue was subjected to silica gel column
chromatography (EtOAc: hexane = 1: 2) to give epoxy acetate 12
3. Experimental procedure
(
0.99 g, 4.17 mmol). The enantiomeric excess of 12 was determined to
be 98.1% ee from the peak-area ratio by chiral HPLC analysis (DAICEL
Chiralpak AD-H (ϕ 0.45 cm × 25 cm), 254 nm, EtOH : hexane = 1: 9).
3
.1. Chemicals
2
5
[
α]
D
= −19.4 (c 1.0, CHCl
H NMR (CDCl , 270 MHz): δ 2.10 (3H, s, Ac), 3.22–3.32 (2H, m, H-
and H-3), 3.60 (1H, dd, J = 11.6, 5.9 Hz, H-4a), 3.72 (1H, dd,
J = 11.2, 4.2 Hz, H-4b), 4.04 (1H, dd, J = 12.1, 7.0 Hz, H-1a), 4.32
1H, dd, J = 12.4 , 3.8 Hz, H-1b), 4.51–4.64 (2H, m, benzyl), 7.28–7.36
(5H, m, aromatic).
3
).
All commercially available chemicals of highest purity were used
1
without further purification. (+)-Epogymnolactam (1) was synthesized
according to the previously reported procedure. Cerulenin (2) was
3
3
2
purchased from Wako Pure Chemical Co. (now FUJIFILM Wako Pure
Chemical Co., Japan). Rapamycin as an autophagy inducer was ob-
tained from Cayman Chemical (MI, USA). Thin layer and column
chromatography were performed with Merck Silica Gel 60 F254 and
Kanto Chemicals Co. (Japan) Silica Gel 60 N (spherical, neutral), re-
spectively. NMR spectra were measured on a JEOL JMN EX-270 FT-
NMR spectrometer. Chemical shifts are reported in ppm (δ-scale) using
TMS as the internal standard and coupling constants (J) are given in Hz.
The carbon number used for NMR assignment was indicated according
to the number shown in 1 and 2 of Fig. 1. MS spectra were acquired
with FI or FD mode on a JEOL JMS-T100GCV equipment. Optical ro-
tations were measured on a JASCO P-2000 polarimeter.
(
13
C NMR (CDCl
3
3
, 67.5 MHz): δ 20.7 (–CH ), 53.0 (C-2), 54.7 (C-3),
62.6 (C-1), 67.8 (C-4), 73.4 (benzyl), 127.8 (aromatic), 127.9 (aro-
matic), 128.4 (aromatic), 137.5 (aromatic), 170.6 (eC(]O)e).
FI-MS: m/z = 236.1 [M]+
.
3.1.3. (2S,3R)- 4-Benzyloxy-2,3-epoxybutan-1-ol (enantiopure 13)
To epoxy acetate 12 (499.5 mg, 2.11 mmol) in THF (6.09 ml),
CH OH (6.09 ml) was added dropwise and then K CO (58.0 mg) was
3
2
3
added at 0 °C. The reaction mixture was stirred for 2 h at the same
temperature. The reaction was quenched with a saturated NH Cl solu-
tion (40 ml), and the mixture was extracted with Et O. The organic
4
2
3
.1.1. (2S,3R)-4-Benzyloxy-2,3-epoxybutan-1-ol (13, 88% ee)
To 4 Å molecular sieves (2.29 g) in CH Cl (32 ml), Ti(O Pr)
7.2 ml, 24.3 mmol) was added, and L-(+)-DIPT (5.03 ml, 23.9 mmol)
was added dropwise at −25 °C under an argon atmosphere. After stir-
ring for 30 min, allyl alcohol 11 (4.00 g, 22.4 mmol) in CH Cl (32 ml)
was added dropwise over 1.5 h. To the mixture, t-BuOOH in nonane
5.5 M, 8.8 ml) was added dropwise, then the reaction temperature was
raised to −20 °C and stirred for 3 days. The reaction mixture was raised
to room temperature, quenched with a saturated Na solution
i
layer was dried with anhydrous Na SO4 and evaporated in vacuo. The
2
2
2
4
residue was subjected to silica gel column chromatography (EtOAc:
(
hexane = 2: 1) to give enantiopure epoxy alcohol 13 (375.9 mg, 91%).
25
[α]
D
3
= −23.1 (c 1.0, CHCl ).
2
2
(
3.1.4. (−)-Epogymnolactam (3)
(2S,3S)-3-(1-Hydroxypentyl)oxiran-2-carboxamide was derived
from enantiopure 13 using essentially the same method as the synthesis
2 2 3
S O
6

K. Ueda et al.
Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
of 1.3 To the amide (19.5 mg, 0.11 mmol) in CH
Martin periodinane (67.1 mg, 1.45 eq.) was added at 0 °C under an
argon atmosphere. After stirring for 2 h, the reaction mixture was
Cl
(1.6 ml), Dess-
[α]
25
= +49.3 (c 1.0, CH
2
2
D
3
OH).
1
H NMR (CDCl , 270 MHz): δ 0.92 (3H, t, J = 7.2 Hz, H-8),
3
1.32–1.75 (6H, m, H-5, H-6 and H-7), 3.13 (1H, dd, J = 7.8, 4.6 Hz, H-
3), 3.42–3.51 (1H, m, H-4), 3.54 (1H, d, J = 4.6 Hz, H-2)
(Supplementary Fig. S3).
quenched with a mixture of saturated Na
NaHCO and extracted with EtOAc. The organic layer was washed with
brine, dried with anhydrous Na SO and evaporated in vacuo. The re-
2 2 3
S O solution and saturated
3
13
2
4
3
C NMR (CDCl , 67.5 MHz): δ 14.0 (C-10), 22.6 (C-7), 27.0 (C-6),
sidue was subjected to silica gel column chromatography (EtOAc:
34.6 (C-5), 54.3 (C-2), 60.1 (C-3), 69.0 (C-4), 170.2 (C-1)
(Supplementary Fig. S4).
hexane = 2: 1) to give (−)-epogymnolactam 3 (11.3 mg, 59%).
2
5
FI-MS: m/z = 174.1 [M+H]+
.
[
α]
D
= −14.4 (c 1.1, CH
= +47.8 (c 1.0, CH OH, incubation time is different)].
OD, 270 MHz): Keto form: δ 0.88–0.97 (3H, m, H-8),
3
OH), [(+)-epogymnolactam 1:
2
0
2
[
α]
D
1
3
H NMR (CD
3
3.1.8. (2R,3R,4R)-2,3-Epoxy-4-hydroxy-N,N-dimethyloctanamide (17)
Lactone 16 (10.0 mg, 64 μmol) was dissolved in a solution of
(CH ) NH in CH OH (2.0 M, 1.5 ml) under argon atmosphere and the
3 2 3
mixture was stirred for 2.5 h at 0 °C. The reaction mixture was con-
centrated in vacuo and purified by silica gel column chromatography
1
.28–1.58 (4H, m, H-6 and H-7), 2.57–2.69 (m, H-5), 3.70 (d,
J = 5.3 Hz, H-3) , 3.89 (d, J = 5.3 Hz, H-2),; syn-form: 0.88–0.97 (3H,
m, H-8), 1.28–1.58 (4H, m, H-6 and H-7), 1.69–1.75 (m, H-5), 3.58 (1H,
d, J = 2.6 Hz, H-3), 3.80 (1H, d, J = 2.6 Hz, H-2); anti-form: 0.88–0.97
(
3
3H, m), 1.28–1.41 (2H, m), 1.49–1.58 (2H, m), 1.69–1.75 (2H, m),
((CHCl
3
: CH
H NMR (CDCl
1.26–1.73 (6H, m, H-5, H-6 and H-7), 3.00 (3H, s, NCH
3
3
OH = 96 : 4) to afford 17 (8.3 mg, 65%) as a colorless solid.
, 270 MHz): δ 0.92 (3H, t, J = 7.0 Hz, H-8),
), 3.09 (1H, dd,
), 3.58 (1H, d, J = 3.8 Hz, H-2),
1
.57 (d, J = 2.6 Hz, H-3), 3.81 (d, J = 2.6 Hz, H-2),.
3
1
3
C NMR (CD
3
OD, 67.5 MHz): Keto form: δ 14.1 (C-8), 23.2 (C-7),
2
6.1 (C-6), 41.0 (C-5), 55.8 (C-2), 59.4 (C-3), 170.5 (C-1), 205.8 (C-4);
J = 3.8, 4.3 Hz, H-3), 3.20 (3H, s, NCH
3
syn-form: 14.3 (C-8), 24.0 (C-7), 27.0 (C-6), 36.3 (C-5), 50.1 (C-3), 53.1
C-2), 87.2 (C-4), 174.4 (C-1); anti-form: 14.3 (C-8), 23.9 (C-7), 25.9 (C-
3.87 (1H, m, H-4).
13
(
3
C NMR (CDCl , 67.5 MHz): δ 14.0 (C-8), 22.7 (C-7), 27.2 (C-6),
6
), 38.9 (C-5), 54.3 (C-2), 58.1 (C-3), 86.8 (C-4), 173.0 (C-1).
34.5 (C-5), 35.1 (-N(CH
71.8 (C-4), 167.2 (C-1).
3 2 3 2
) ), 36.7 (-N(CH ) ), 53.8 (C-2), 59.9 (C-3),
+
FI-MS: m/z = 172.1 [M+H]
.
FI-MS: m/z = 1201.1 [M]+
.
3.1.5. 4-Hydroxyoctanamide (15)
γ-Octalactone 14 (420.0 mg, 3.0 mmol) was mixed with approxi-
mately 20 ml of liquid NH in a pressure-resistance reaction tube, and
the mixture was stirred at 50 °C for 7 days. The tube was opened at
196 °C, and NH was evaporated at ambient temperature. The re-
sultant white solid was purified by silica gel column chromatography
CHCl : CH OH = 9 : 1) to give amide 15 (411.3 mg, 87%) as a white
solid1 .
H NMR (CD
3.1.9. N,N-Dimethyl Analogue (6)
3
2 2
To a stirred solution of 17 (8.2 mg, 41 μmol) in CH Cl (2.4 ml),
Dess-Martin periodinane (29.0 mg, 1.5 eq.) was added at 0 °C under
argon atmosphere. After stirring for 3 h, the reaction was quenched
−
3
with a mixture of saturated aqueous Na
NaHCO . The resultant solution was extracted with EtOAc and the or-
4
ganic layer was washed with brine, dried over anhydrous Na SO ,
2 2 3
S O and saturated aqueous
(
3
3
3
2
3
OD, 270 MHz): δ 0.82 (3H, t, J = 6.8 Hz, H-8),
concentrated in vacuo. The residue was purified by silica gel column
chromatography (EtOAc : hexane = 2 : 1) to give N,N-dimethyl ana-
1
3
.31–1.84 (8H, m, H-3, H-5, H-6 and H-7), 2.21–2.41 (2H, m, H-2),
.47–3.56 (1H, m, H-4).
logue 6 (7.3 mg, 90%) as a colorless solid.
1
3
25
C NMR (CD
3
OD, 67.5 MHz): δ 14.4 (C-8), 23.8 (C-7), 29.0 (C-6),
[α]
D
= −44.2 (c 0.7, CHCl
, 270 MHz): δ 0.89 (3H, t, J = 7.2 Hz, H-8), 1.30
(2H, sext, J = 7.4 Hz, H-7), 1.48–1.62 (2H, m, H-6), 2.55 (2H, dt,
J = 7.4, 7.4, 2.9 Hz, H-5), 2.93 (3H, s, NCH ), 3.13 (3H, s, NCH ), 3.70
3
).
1
3
2.9 (C-2), 34.1 (C-3), 38.1 (C-5), 71.9 (C-4), 179.2 (C-1).
H NMR (CDCl
3
+
FD-MS: m/z = 160.1 [M+H]
.
3
3
3.1.6. Deepoxy analogue (4)
(1H, d, J = 4.9 Hz, H-3), 3.82 (1H, d, J = 5.1 Hz, H-2) (Supplementary
Dess-Martin periodinane (129.7 mg, 2 eq.) was added to a stirred
solution of amide 15 (23.7 mg, 148.8 μmol) in dry CH Cl (8.4 ml) at
°C under argon atmosphere. After stirring for 3 h, the reaction was
Fig. S5).
13C NMR (CDCl
, 67.5 MHz): δ 13.8 (C-8), 22.1 (C-7), 24.8 (C-6),
), 36.3 (eN(CH ), 39.6 (C-5), 54.7 (C-2), 58.2 (C-3),
2
2
3
0
35.3 (eN(CH
3
)
2
3 2
)
quenched with a solution of saturated aqueous Na
aqueous NaHCO . The solution was extracted with Et
layer was washed with brine, dried over Na SO
vacuo. The resultant residue was subjected to silica gel column chro-
matography (CHCl : CH OH = 19: 1) to afford deepoxy analogue 4
14.1 mg, 60%) as a colorless solid.
2
S
2
O
3
and saturated
O and organic
164.3 (C-4), 205.5 (C-1) (Supplementary Fig. S6).
HR-FI-MS: m/z = 200.1290 [M+H]+ calcd. for C10
200.1287.
3
2
3
H18NO , found
2
4
, and concentrated in
3
3
3.1.10. O-Methyl analogue (7)
(
To a solution of (+)-epogymnolactam 1 (8.5 mg, 50 μmol) in
OH (3.0 ml), 10-camphorsulfonic acid (3.0 mg, 0.26 eq.) was added,
and the reaction mixture was stirred at 50 °C for 6 h. The resulting
mixture was diluted with EtOAc and washed H O and brine. The or-
ganic layer was dried over anhydrous Na SO , evaporated in vacuo. The
1
H NMR (CDCl
3
, 270 MHz): δ 0.90 (3H, t, J = 7.3 Hz, H-8), 1.31
CH
3
(
(
2H, sext, J = 7.3 Hz, H-7), 1.57 (2H, quin, J = 7.5, H-6), 2.43–2.51
4H, m, H-2 and H-3), 2.78 (2H, t, J = 6.5 Hz, H-3) (Supplementary Fig.
2
S1).1
2
4
3C NMR (CDCl
9.1 (C-2), 37.4 (C-3), 42.5 (C-5), 174.4 (C-1), 210.2 (C-4)
, 67.5 MHz): δ 13.8 (C-8), 22.3 (C-7), 25.9 (C-6),
resultant oil was purified by silica gel column chromatography (EtOAc:
hexane = 1: 2) to afford O-methyl analogue 7 (2.2 mg, 24%) as a col-
3
2
(
Supplementary Fig. S2).
orless oil.
HR-FD-MS: m/z = 158.1185 [M+H]+ calcd. for C
H16NO
, found
[α]
25
= +95.3 (c 0.29, CHCl
, 270 MHz): δ 0.93 (3H, t, J = 7.1 Hz, H-8),
.26–1.56 (4H, m, H-6 and H-7), 1.79 (2H, t, J = 7.7 Hz, H-5), 3.65
8
2
D
3
).
1
1
58.1181.
H NMR (CDCl
3
1
3
.1.7. (2R,3R,4R)-2,3-Epoxy-4-hydroxyoctanamide (5)
Lactone 16 (low polar diastereomer, 49.6 mg, 0.317 mmol) was
(1H, t, J = 2.3 Hz, H-3), 3.79 (1H, d, J = 2.7 Hz, H-2) (Supplementary
2
Fig. S7).
dissolved in a solution of NH
atmosphere and the mixture was stirred for 2.5 h at 0 °C. The reaction
mixture was evaporated in vacuo and purified by silica gel column
3
in CH
3
OH (2.0 M, 8 ml) under nitrogen
13C NMR (CDCl
34.1 (C-5), 49.4 (C-2), 51.7 (–OCH
3
, 67.5 MHz): δ 13.9 (C-8), 22.7 (C-7), 25.7 (C-6),
3
), 56.5 (C-3), 89.6 (C-4), 171.5 (C-1)
16NO , found
(Supplementary Fig. S8).
chromatography (CHCl
(54.0 mg, 98.5%).
3
: CH
3
OH = 9 : 1) to afford epoxy amide alcohol
HR-FD-MS: m/z = 186.1166 [M+H]+ calcd. for C
9
H
3
5
186.1130.
7

K. Ueda et al.
Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
3
.1.11. (5R,6S)-7-Benzyloxy-5,6-epoxyhept-1-en-4-ol (19)
To a stirred solution of 18 (100 mg, 0.521 mmol) in dry CH
21 (50.0 mg, 75%) as a colorless oil.
3
1
2
Cl
2
3
H NMR (CDCl , 270 MHz): δ 2.43–2.64 (2H, m, allyl), 3.79 (1H, t,
(
12 ml), SnCl
4
(65 μl, 1.1 eq.) was added dropwise at −78 °C under
J = 2.7 Hz, H-5), 4.03 (d, J = 2.4 Hz, H-1 of major diastereomer), 4.11
(dd, J = 2.6, 1.5 Hz, H-1 of minor diastereomer), 4.50 (dt, J = 7.1,
1.4 Hz, H-4 of minor diastereomer), 4.64 (t, J = 6.2 Hz, H-4 of major
diastereomer), 5.17–5.30 (2H, m, olefinic), 5.70–5.92 (1H, m, olefinic).
argon atmosphere. After stirring for 1 h, allyltributyltin (208 μl, 1.3 eq.)
was added dropwise and the reaction mixture was continuously stirred
for another 1.5 h. The reaction mixture was quenched with sat. aqueous
NH
extracted with Et
over Na SO , and concentrated in vacuo. The resultant residue was
1
3
4
Cl at −78 °C and allowed to room temperature. The solution was
3
C NMR (CDCl , 67.5 MHz): δ 34.0 and 36.1 (C-5), 49.8 and 50.6
2
O and the organic layer was washed with brine, dried
(C-3), 56.2 and 57.6 (C-2), 78.1 and 78.4 (C-4), 119.3 and 120.4 (C-7),
129.8 and 131.2 (C-6), 170.1 (C-1).
2
4
purified by silica gel column chromatography (EtOAc : hexane = 1: 3)
to give a mixture of homoallylic alcohol 19 (74.3 mg) and its chlor-
FI-MS: m/z = 140.1 [M]+
.
ohydrin (20 mg). To a solution of chlorohydrin (20 mg) in CH
1 ml), K CO (15. 3 mg, 1.5 eq.) was added at 0 °C, and the reaction
mixture was allowed to room temperature. After 4 days, additional
CO in CH OH (1 ml) was added and the reaction mixture was stirred
for 5 h. The resultant mixture was quenched with saturated aqueous
NH Cl and extracted with Et O. The organic layer was washed with
brine, dried over Na SO , concentrated in vacuo, and the residue was
3
OH
3.1.14. (2R,3R)-2,3-Epoxy-4-hydroxyhept-6-enamide (22)
The diastereomeric mixture of 21 (9.6 mg, 0.067 mmol) was dis-
3 3
solved in a solution of NH in CH OH (2.0 M, 1.3 ml) under argon at-
mosphere and the mixture was stirred at 0 °C for 3 h. The resulting
solution was concentrated in vacuo and purified by silica gel column
(
2
3
K
2
3
3
4
2
chromatography (CHCl
3 3
: CH OH = 93: 7) to afford a diastereomeric
2
4
mixture of amide 22 (10.3 mg, 96%) as a colorless solid.
1
purified by silica gel column chromatography (EtOAc: hexane = 1: 3)
3
H NMR (CDCl , 270 MHz): δ 2.33–2.55 (2H, m, H-5), 3.14–3.22
to afford an additional 19 (12.2 mg). The total yield of 19 was 71%.
(1H, dd, J = 8.1, 4.6 Hz, H-3), 3.46–3.66 (2H, m, H-2 and H-4),
5.14–5.23 (2H, m, H-7), 5.72–5.97 (1H, m, H-6).
1
H NMR (CDCl , 270 MHz): δ 2.35–2.56 (2H, m, H-3), 2.96 (1H, dd,
3
13
J = 8.1 and 4.3 Hz, H-5), 3.23–3.29 (1H, m, H-6), 3.47–3.55 (1H, m, H-
3
C NMR (CDCl , 67.5 MHz): δ 38.2 and 39.3 (C-5), 54.3 and 54.5
4
6
), 3.66 (1H, dd, J = 10.5, 6.7 Hz, H-7a), 3.81–3.87 (1H, dd, J = 10.7,
.1 Hz, H-7b), 4.58 (2H, m, benzyl), 5.14–5.22 (2H, m, H-1), 5.83–5.93
(C-2), 59.6 and 60.8 (C-3), 67.9 and 68.5 (C-4), 118.4 and 118.8 (C-7),
132.7 and 133.1 (C-6), 169.9 and 170.6 (C-1).
(
1H, m, H-2), 7.30–7.37 (5H, m, aromatic).
FD-MS: m/z = 158.1 [M+H]+
.
1
3
C NMR (CDCl , 67.5 MHz): δ 39.4 (C-5), 54.1 (C-2), 57.9 (C-3),
3
6
1
8.5 (C-1), 69.0 (C-4), 73.6 (benzyl), 118.3 (C-7), 127.9 (aromatic),
3.1.15. (2R,3R)-2,3-Epoxy-4-hydroxydec-6-enamide (23)
28.6 (aromatic), 133.5 (C-6), 137.2 (aromatic).
Grubbs 2nd generation catalyst (3.1 mg, 0.1 eq.) was added to a
stirred solution of 22 (5.7 mg, 36.3 μmol) and 1-pentene (70 mg) in dry
and degassed CH Cl (1.0 ml) under argon and the resulting mixture
2 2
+
FI-MS: m/z = 234.1 [M]
.
3
.1.12. (2S,3R)-2,3-Epoxyhept-6-en-1,4-diol (20)
Lithium wire (20.9 mg, 3.02 mmol) was hammered out into foil, cut
was heated to reflux for 24 h. The mixture was concentrated in vacuo
and purified by silica gel column chromatography (EtOAc: hexane = 4:
1) to afford a diastereomeric mixture of amide 23 (4.4 mg, 61%) as a
into several small pieces and washed with hexane. The pieces were then
dipped, one by one, in CH OH until the surface turns shiny, then im-
3
colorless solid.
1
mediately transferred to THF. The lithium chunks were placed in a
Schlenk flask containing DBB (534.8 mg, 2.01 mmol). The flask was
evacuated and purged with argon several times. Dry THF (12 ml) was
added, and the contents of the flask were sonicated between 0 and 20 °C
for 3 h. The resultant deep forest green solution of radical anion
3
H NMR (CDCl , 270 MHz): δ 0.96 (3H, t, J = 7.2 Hz, H-10),
1.23–1.43 (2H, m, H-9), 1.98–2.11 (2H, m, H-8), 2.17–2.51(2H, m, H-
5), 3.14–3.25 (1H, m, H-3), 3.47–3.65 (2H, m, H-2 and H-4), 5.32–5.65
(2H, m, H-6 and H-7), 6.14–6.22 (1H, s).
13
C NMR (CDCl , 67.5 MHz): δ 13.6 (C-10), 22.4 (C-9), 34.7 (C-8),
3
(
4.7 ml) was then added dropwise to a solution of 19 (47.0 mg,
.201 mmol) in dry THF (5.0 ml) at −78 °C under argon atmosphere.
The mixture was stirred for 1.5 h and quenched with sat. aqueous
NH Cl at −78 °C and allowed to reach room temperature. The resultant
solution was extracted with Et O and the organic layer was washed
with brine, dried over Na SO , concentrated in vacuo. The residue was
37.1 and 38.1 (C-5), 54.2 and 54.3 (C-2), 59.5 and 60.8 (C-3), 68.6 and
69.2 (C-4), 123.5 and 123.9 (C-7), 135.3 and 135.9 (C-6), 169.1 and
169.6 (C-1).
0
FD-MS: m/z = 200.1 [M+H]+
.
4
2
2
4
3.1.16. (2R,3R)-2,3-Epoxy-4-hydroxydecanamide (24)
purified by silica gel column chromatography (EtOAc: hexane = 2: 1)
to give a diastereomeric mixture of diol 20 (17.6 mg, 61%) as a col-
orless oil.
To a solution of 23 (4.5 mg, 22.6 μmol) in THF (0.7 ml), Pd/C(en)
(5.0 mg) was added and the mixture was stirred vigorously overnight
under hydrogen atmosphere. The resulting solution was filtered
through celite pad, concentrated and the residue was purified by silica
gel column chromatography (EtOAc: hexane = 2: 1) to afford a dia-
1H NMR (CDCl
3
, 270 MHz): δ 2.28–2.58 (2H, m, H-5), 2.97–3.08
(
1H, dd, J = 7.8, 4.3 Hz, H-3), 3.20–3.76 (3H, m, H-1a, H-2 and H-4),
3
6
.84–4.01 (1H, m, H-1b), 5.12–5.24 (2H, m, H-7), 5.72–5.96 (1H, m, H-
stereomeric mixture of amide 24 (3.5 mg, 77%) as a colorless solid.
1
). 1
3
H NMR (CDCl , 270 MHz): δ 0.88 (3H, t, J = 7.2 Hz, H-10),
3C NMR (CDCl
, 67.5 MHz): δ 39.8 (C-5), 55.8 (C-2), 58.3 (C-3),
3
2
1.28–1.70 (10H, m, CH ), 3.11–3.21 (1H, m, H-3), 3.43–3.60 (2H, m,
6
0.7 (C-1), 68.8 (C-4), 119.0 (C-7), 133.1 (C-6).
H-2 and H-4).
+
13
FD-MS: m/z = 145.1 [M+H]
.
3
C NMR (CDCl , 67.5 MHz): δ 14.0 (C-10), 22.5 and 22.6 (C-9),
2
4.7 and 24.8 (C-8), 29.1 and 29.2 (C-7), 31.6 and 31.7 (C-6), 33.6 and
3
.1.13. (1R,5R)-4-(Prop-2-enyl)–3,6-dioxa-bicyclo[3.1.0]hexan-2-one
34.8 (C-5), 54.2 and 54.8 (C-2), 60.0 and 61.4 (C-3), 69.3 and 70.0 (C-
4), 169.2 and 169.9 (C-1).
(
21)
PhI(OAc)
68.8 mg, 0.478 mmol) and DMN-AZADO (4.0 mg, 0.05 eq.) in CH
4.82 ml) and aqueous phosphate buffer (0.54 ml, 0.2 M, pH 6.8) at
°C. After stirring for 5 h, the mixture was quenched with saturated
aqueous Na . The solution was extracted with EtOAc and the or-
ganic layer was washed with brine, dried over Na SO , concentrated in
vacuo. The residue was purified by silica gel column chromatography
EtOAc: hexane = 1 : 3) to afford a diastereomeric mixture of lactone
(462 mg, 3.0 eq.) was added to a stirred solution of 20
CN
FD-MS: m/z = 202.2 [M+H]+
.
2
(
(
0
3
3.1.17. Hexyl analogue (8)
2 2
To a stirred solution of 24 (3.4 mg, 16.9 μmol) in dry CH Cl
S
2 2
O
3
(1.0 ml), Dess-Martin periodinane (12.2 mg, 1.7 eq.) was added at room
temperature under argon atmosphere. After stirring for 3 h, the reaction
mixture was purified by silica gel column chromatography (EtOAc :
hexane: toluene = 3: 1: 1) to afford hexyl analogue 8 (2.1 mg, 62%) as a
2
4
(
8

K. Ueda et al.
Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
colorless solid.
solution was filtered through celite pad and concentrated in vacuo. The
residue was purified by silica gel column chromatography (EtOAc:
hexane = 1: 1) to afford octyl analogue 9 (4.4 mg, quant.) as a colorless
1
H NMR (CDCl , 270 MHz): δ 0.88 (3H, t, J = 7.0 Hz, H-10),
3
1
3
.28–1.61 (8H, m), 2.48–2.68 (2H, m, H-5), 3.73 (1H, d, J = 5.4 Hz, H-
), 3.88 (1H, d, J = 5.1 Hz, H-2) (Supplementary Fig. S9).
solid.
1
3
25
C NMR (CDCl
3
, 67.5 MHz): δ 14.0 (C-10), 22.4 (C-9), 23.1 (C-8),
[α]
D
= +17.2 (c 0.45, CHCl
, 270 MHz): δ 0.85–0.91 (3H, m, H-12), 1.26 (10H,
2
), 1.55–1.63 (2H, m, H-6), 1.70–1.87 (m, H-5 of cyclic form),
3
).
1
2
2
8.7 (C-7), 31.4 (C-6), 41.1 (C-5), 55.3 (C-2), 58.3 (C-3), 167.3 (C-1),
02.7 (C-4) (Supplementary Fig. S10).
H NMR (CDCl
br s, CH
3
+
HR-FD-MS: m/z = 200.1294 [M+H] calcd. for C10
H18NO
3
, found
2.48–2.68 (m, H-5 of keto-form), 3.61 (t, J = 2.3 Hz, H-3 of major
cyclic form), 3.65 (t, J = 2.4 Hz, H-3 of minor cyclic form), 3.73 (d,
J = 5.1 Hz, H-3 of keto-form), 3.80–3.83 (m, H-2 of cyclic form), 3.87
(d, J = 5.4 Hz, H-2 of keto-form) (Supplementary Fig. S11).
200.1287.
3
.1.18. (2R,3R)-2,3-Epoxy-4-hydroxytetradec-6-enamide (25)
Grubbs 2nd generation catalyst (2.1 mg, 0.1 eq.) was added to a
stirred solution of 22 (3.8 mg, 24.3 μmol) and 1-nonene (97.2 mg) in
dry and degassed CH Cl (2.0 ml) under argon and the mixture was
heated to reflux for 24 h. The resulting mixture was concentrated in
vacuo and purified by silica gel column chromatography (CHCl
CH OH = 19: 1) to afford amide 25 (2.7 mg, 43%) as a colorless solid.
13
C NMR (CDCl , 67.5 MHz): keto form δ 14.1 (C-12), 22.6 (C-11),
3
23.1 (C-6), 29.0 (C-9 and C-8), 29.1 (C-7), 31.8 (C-10), 41.1 (C-5), 55.3
(C-2), 58.3 (C-3), 167.3 (C-1), 202.7 (C-4); syn- and anti-form δ 14.1 (C-
12), 22.6 (C-11), 23.6 (C-6), 29.1 (C-9), 29.3 and 29.4 (C-8), 29.6 (C-7),
31.8 (C-10), 37.1 and 37.5 (C-5), 51.5 and 53.9 (C-2), 57.5 and 57.8 (C-
3), 86.3 (C-4), 171.4 (C-1) (Supplementary Fig. S12).
2
2
3
:
3
1
H NMR (CDCl
.27–1.61 (10H, m, CH
H-5), 3.18–3.23 (1H, m, H-3), 3.55–3.57 (2H, m, H-2 and H-4),
3
, 270 MHz): δ 0.88 (3H, t, J = 6.8 Hz, H-14),
HR-FD-MS: m/z = 228.1595 [M+H]+ calcd. for C12
228.1599.
3
H22NO , found
1
2
), 1.98–2.06 (2H, m, H-8), 2.25–2.32 (2H, m,
5
.34–5.63 (2H, m, H-6 and H-7).
1
3
3.2. Monitoring of autophagy-inducing activity
C NMR (CDCl
3
, 67.5 MHz): δ 14.1 (C-14), 22.7 (C-13), 29.15 (C-
1
1), 29.17 (C-9), 29.3 (C-10), 31.8 (C-12), 32.6 (C-8), 37.1 (C-5), 54.2
The autophagy-inducing activity of each compound was evaluated
(
C-2), 60.8 (C-3), 69.2 (C-4), 123.2 (C-6), 136.3 (C-7), 169.1 (C-1).
HR-FD-MS: m/z = 256.1917 [M+H]+ calcd. for C14
by slightly modifying the method reported previously.2 Briefly, Mouse
H
26NO , found
3
5
embryo fibroblast NIH 3 T3 cells were inoculated at 3.3 × 10 cells/ml
and cultured at 37 °C under 5% CO atmosphere. After 24 h, medium
2
2
3
56.1913.
change was made in each well, and the DMSO solution of each com-
pound was added at final concentrations of 25 μM, 50 μM, and 100 μM.
The cells were incubated at 37 °C for 4 h, and adherent cells were
harvested in 1 ml of PBS (−) and collected by centrifugation. The
collected cells were washed twice with PBS (−) and sonicated in RIPA
Lysis Buffer (Santa Cruz: sc-24948). After centrifugation, the super-
natant was collected and stored at −20 °C until use. Proteins were se-
parated by SDS-PAGE and immunoblot analysis was performed with
anti-LC3 antibody (Cell Signaling Technology), anti-SQSTM1/p62 an-
tibody (Funakoshi, Japan) and anti-actin antibiotic (Cosmo Bio, Sigma-
Aldrich). The specific signals from LC3, SQSTM1 (p62), and actin were
quantified with the enhanced chemiluminescence reagents (NEL105,
Perkin Elmer) by luminescent analyzer LAS-3000 (FUJI FILM, Japan) or
ChemiDoc MP (Bio-Rad). As indicated in reference 2, it was difficult to
visually judge by conventional immunoblotting analysis. Thus, the re-
lative signal ratio (LC3-II/LC3-I, LC3-II/actin, or p62/actin) was de-
termined from the quantified specific signal of each band in LAS-3000
or ChemiDoc MP to evaluate the autophagy inducing activity or au-
tophagy inhibiting activity.
.1.19. (2R,3R)-2,3-Epoxy-4-hydroxytetradecanamide (26)
Pd/C(en) (6.0 mg) was added to a solution of 25 (3.0 mg, 11.7 μmol)
in THF (0.7 ml) and the reaction mixture was stirred vigorously at room
temperature under H atmosphere. The resulting mixture was filtered
through celite pad and concentrated in vacuo. The residue was purified
by silica gel column chromatography (CHCl : CH OH = 19: 1) to give
amide 26 (2.1 mg, 69%) as a colorless solid.
2
3
3
1
H NMR (CDCl
.25–1.58 (20H, m, CH
H-2 and H-4).
3
, 270 MHz): δ 0.88 (3H, t, J = 6.8 Hz, H-14),
1
2
), 3.16–3.21 (1H, m, H-3), 3.49–3.59 (2H, m,
1
3
C NMR (CDCl , 67.5 MHz): δ 14.1 (C-14), 22.7 (C-13), 24.7 (C-6),
3
2
1
9.3 (C-11), 29.4 (C-8), 29.55 (C-9), 29.57 (C-10), 29.7 (C-7), 31.9 (C-
2), 33.6 (C-5), 54.7 (C-2), 61.4 (C-3), 70.0 (C-4), 169.0 (C-1).
HR-FD-MS: m/z = 258.2076 [M+H] calcd. for C14
58.2069.
+
H28NO
3
, found
2
3.1.20. Decyl analogue (10)
2 2
To a stirred solution of 26 (1.9 mg, 7.4 μmol) in dry CH Cl (1.0 ml),
Dess-Martin periodinane (5.3 mg, 1.7 eq.) was added at room tem-
perature under argon atmosphere. After stirring for 3 h, the reaction
mixture was purified by silica gel column chromatography (CHCl
3
:
3.3. Anti-proliferative activity
CH OH = 19: 1) to afford decyl analogue 10 (1.3 mg, 69%) as a col-
3
3
orless solid. In the NMR of 10 in CDCl , only the keto form was ob-
served.
Anti-proliferative activity was performed with slight modification of
the previously reported method. Using the trypan blue exclusion assay,
2
1H NMR (CDCl
, 270 MHz): δ 0.88 (3H, t, J = 6.7 Hz, H-14), 1.26
), 1.56–1.66 (2H, m, H-6), 2.58 (2H, dt, J = 7.6, 7.6 and
.1 Hz, H-5), 3.73 (1H, d, J = 5.1 Hz, H-3), 3.87 (1H, d, J = 5.4 Hz, H-
) (Supplementary Fig. S13).
3
the number of viable cells of NIH3T3 cells was adjusted to
(
5
2
14H, br s, CH
2
4
0
.53 × 10 cells/ml. The cells were treated with each compound in
DMSO to the final concentrations of 25 μM, 50 μM and 100 μM, and
plated on a 96-well plate. The DMSO concentration in all wells was set
1
3
C NMR (CDCl , 67.5 MHz): δ 14.1 (C-14), 22.7 (C-13), 23.1 (C-6),
3
to 0.1%. After 4 h incubation at 37 °C in a CO incubator, 10 μl of Cell
2
2
4
9.0 (C-7), 29.2 (C-8), 29.3 (C-11), 29.4 (C-10), 29.5 (C-9), 31.9 (C-12),
1.1 (C-5), 55.3 (C-2), 58.3 (C-3), 167.4 (C-1), 202.6 (C-4)
Counting Kit-8 (DOJINDO Laboratories, Japan) was added. After in-
cubating for 14 h, the relative number of viable cells was expressed as
the optical density (OD450) value determined by a microplate reader
(
Supplementary Fig. S14).
HR-FD-MS: m/z = 256.1917 [M+H] calcd. for C14
56.1913.
+
3
H26NO , found
(
Sunrise Remote, Tecan Group Ltd., Switzerland).
2
3
1
.1.21. Octyl analogue (9)
Pd/C(en) (1.0 mg) was added to a solution of cerulenin 2 (4.3 mg,
9.3 μmol) in THF (0.5 ml) and the reaction mixture was stirred at room
Acknowledgements
This work was supported by JSPS KAKENHI (Grant-in-Aid for
Scientific Research (C)) Grant Number JP 26450135, Japan.
temperature overnight under hydrogen atmosphere. The resulting
9

K. Ueda et al.
Bioorganic & Medicinal Chemistry xxx (xxxx) xxx–xxx
Appendix A. Supplementary data
(b) Freeman PK, Hutchinson LL. J Org Chem. 1924;1980:45.
1
1
1
0. Donohoe TJ, House D. J Org Chem. 2002;67:5015.
1. Doi R, Shibuya M, Murayama T, Yamamoto Y, Iwabuchi Y. J Org Chem. 2015;80:401.
2. Kumazaki H, Nakajima R, Bessho Y, Yokoshima S, Fukuyama T. Synlett.
2015;26:2131.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmc.2018.09.013.
1
1
1
1
3. Scholl M, Ding S, Lee CW, Grubbs RH. Org Lett. 1999;1:953.
4. Mukherjee JP, Sil S, Pahari K, Chattopadhyay SK. Synthesis. 2016;48:1181.
5. Sajiki H, Hattori K, Hirota K. Chem Eur J. 2000;6:2200.
6. Okawa Y, Hideshima T, Ikeda H, et al. Br J Haematol. 2008;141:659.
References
1
2
.
.
Mizushima N, Levine B, Cuervo AN, Klionsky DJ. Nature. 2008;451:1069.
Mitsuhashi S, Shindo C, Shigetomi K, Miyamoto T, Ubukata M. Phytochemistry.
17. Thupari JN, Pinn ML, Kuhajda FP. Biochem Biophys Res Commun. 2001;285:217.
8. Friis MB, Friborg CR, Schneider L, et al. J Physiol. 2005;567:427.
1
2
015;114:163.
19. Zimmermann C, Santos A, Gable K, et al. Cell Rep. 2013;5:1036.
20. Régnacq M, Voisin P, Sere YY, et al. BBRC. 2016;477:33.
21. Shpilka T, Welter E, Borovsky N, et al. EMBO J. 2015;34:2117.
22. Lawrence DS, Zilfou JT, Smith CD. J Med Chem. 1999;42:4932.
3
4
5
6
7
.
.
.
.
.
Okado Y, Shigetomi K, Mitsuhashi S, Ubukata M. J Antibiot. 2015;68:721.
Kawaguchi A, Tomoda H, Nozoe S, Omura S, Okuda S. J Biochem. 1982;92:7.
Klionsky DJ, et al. Autophagy. 2016;12:1.
Taylor SK, Ide ND, Silver ME, Stephan ML. Synth Commun. 2001;31:2391.
Kuramochi K, Mizushina Y, Nagata S, Sugawara F, Sakaguchi K, Kobayashi S. Bioorg
Med Chem. 1983;2004:12.
2
3. Shimazawa R, Ogawa Y, Morisaki N, Funabashi H, Kawaguchi A, Iwasaki S. Chem
Pharm Bull. 1992;40:2954.
2
4. Morisaki N, Funabashi H, Shimazawa R, et al. Eur J Biochem. 1993;211:111.
8
.
.
Wang S, Howe GP, Mahal RS, Procter G. Tetrahedron Lett. 1992;33:3351.
(a) Freeman PK, Hutchinson LL. Tetrahedron Lett. 1849;1976:17;
25. Singh R, Kaushik S, Wang Y, et al. Nature. 2009;458:1131.
26. Buchebner M, Pfeifer T, Rathke N, et al. J Lipid Res. 2010;51:2896.
9
2
7. Kobayashi K, Ohte S, Ohshiro T, Ugaki N, Tomoda H. Sci. Rep. 2018;8r:12099.
10