Structure–activity relationship study of the anti-obesity natural product yoshinone A

Article abstract of DOI:10.1002/chir.23292

Yoshinone A was derived from marine algae and shown to inhibit adipogenic differentiation. The natural compound is composed of a γ-pyrone ring and a side chain and that contains two asymmetric carbons. Although their absolute configuration has been determ

Full text of DOI:10.1002/chir.23292

Received: 18 September 2020
DOI: 10.1002/chir.23292
Revised: 24 November 2020
Accepted: 30 November 2020
R E G U L A R A R T I C L E
Structure–activity relationship study of the anti-obesity
natural product yoshinone A
Yoshinori Kawazoe1
| Yuki Itakura2 | Toshiyasu Inuzuka3
|
Sachikazu Omura2 | Daisuke Uemura2
1Center for Education and Research in
Abstract
Agricultural Innovation, Faculty of
Agriculture, Saga University, Saga, Japan
Yoshinone A was derived from marine algae and shown to inhibit adipogenic
differentiation. The natural compound is composed of a γ-pyrone ring and a
side chain and that contains two asymmetric carbons. Although their absolute
configuration has been determined, there is no information available on the
stereoisomers and their bioactivities. To address this question, we synthesized
all four stereoisomers and measured their activities. We also prepared three
more derivatives of yoshinone A and found that the stereo-configuration inside
the side chain, the γ-pyrone ring, and bulkiness of the side chain all played
important roles in its activity. Our findings should help to elucidate the mecha-
nism of action of yoshinone A.
2Department of Chemistry, Kanagawa
University, Hiratsuka, Japan
3Life Science Research Center, Gifu
University, Gifu, Japan
Correspondence
Yoshinori Kawazoe, Center for Education
and Research in Agricultural Innovation,
Faculty of Agriculture, Saga University,
152-1 Shonan-Cho, Karatsu, Saga
847-0021, Japan.
Email: ykawazoe@cc.saga-u.ac.jp
K E Y W O R D S
Daisuke Uemura, Department of
Chemistry, Kanagawa University, 2946
Tsuchiya, Hiratsuka, Kanagawa 259-1293,
Japan.
adipocytes, natural products, obesity, structure–activity relationship, yoshinone A
Email: uemurad@kanagawa-u.ac.jp
Funding information
The Naito Foundation; JSPS A3 Foresight
Program; Grant-in-Aid for Scientific
Research, Grant/Award Number:
19K05720
1 | INTRODUCTION
thalidomide.3,4 Another example is Bay-K-8644, which
acts on a calcium channels. While the S enantiomer func-
The chirality of bioactive molecules frequently influences
their pharmacological activity. One of the most famous
examples is thalidomide, which was first brought to the
market to treat insomnia. However, it was soon revealed
that the drug affected fetal development, in a condition
called thalidomide embryosis. While the R enantiomer is
believed to induce somnolence, the S enantiomer is
thought to be teratogenic.1,2 A recent study identified the
tions as an agonist, the R enantiomer acts as an
antagonist.5–7 Thus, it is important to investigate the dif-
ferences in the activities of asymmetric compounds for
drug development.
Yoshinone A (1, Figure 1), a natural product derived
from marine blue-green algae, Leptolyngbya sp., has been
identified as an inhibitor of adipogenic differentiation in
the mouse fibroblastic cell line 3T3-L1.8 It has also been
shown to reduce the increase in body weight induced by
the administration of a high-fat diet.9 These results indi-
cated that the natural product should be a promising can-
didate to treat obesity.
cellular binding protein of the
S enantiomer of
This Paper was dedicated to the late Professor Koji Nakanishi for his
great contribution to natural products.
226
© 2021 Wiley Periodicals LLC.
wileyonlinelibrary.com/journal/chir
Chirality. 2021;33:226–232.

KAWAZOE ET AL.
227
5 × 104 cells in each well. After 2 days, the cells were
induced to undergo adipogenic differentiation by
insulin in the presence of various concentrations of
yoshinone A or its derivatives. The culture media were
replaced with fresh media without insulin or
yoshinone A derivatives every 2 days. After 7 days of
induction, intracellular triglyceride levels were deter-
mined as an indicator of the degree of differentiation
by a LabAssay Triglyceride Kit (WAKO) according to
the manufacturer's instructions, and the IC50 values
were then determined.
FIGURE 1 Chemical structure of yoshinone A
In contrast to its pharmacological activity in vivo, a
little was known about the structure of the compound
that is responsible for this activity. Yoshinone A con-
sists of a γ-pyrone ring and a side chain. Although our
previous study showed that a conjugation state between
the γ-pyrone ring and an olefin located in the side
chain was important for blocking adipogenesis in
3T3-L1 cells,8 the contributions of other parts of the
compound to its inhibitory effects were not clear. Espe-
cially, there was no information on the relationship
between the stereo-configuration of yoshinone A and its
bioactivity, although the absolute configuration within
the side chain region of the compound was determined
through its total synthesis.10 In this study, we tried to
study the structure–activity relationship of yoshinone A
using all four stereoisomers. We also tested if the
γ-pyrone ring, hydroxy group, or methoxy group in the
side chain was necessary for the activity. Our efforts
contribute to the development of a yoshinone A-based
anti-obesity medicine or research reagent for studying
fat metabolism.
1.3 | Measurement of lactate levels
HeLa cells were placed in 96-well plates at a density of
1 × 104 cells in each well. On the next day, the culture
media were replaced to with fresh media containing vari-
ous concentrations of yoshinone A or its derivatives and
cultured for an additional 2 days. Lactate levels in the
culture media were measured with a Lactate Assay Kit-
WST (Dojindo) according to the manufacturer's instruc-
tions. The EC50 values were then determined.
2 | RESULTS AND DISCUSSION
2.1 | Synthesis of compound 14
(a synthetic yoshinone A) and its
diastereomer 16
After protection of the secondary alcohol of methyl
D-lactate 2 with TBSCl, reduction with DIBAL led to the
formation of TBS-protected aldehyde 3 in 82% yield in
two steps (Scheme 1). Aldehyde 3 was converted into the
ethyl ester 5 in 74% yield by the Wittig reaction with pho-
sphorane 4. DIBAL reduction of ethyl ester 5 was
expected to give the corresponding aldehyde as in the
case of compounds 2 to 3; however, the reaction gave
corresponding allyl alcohol 6. This was explained by con-
sidering that the α,β-unsaturated hemiaminal intermedi-
ate was rapidly decomposed to give an undesired alcohol
even at low temperature. For conversion to an aldehyde,
alcohol 6 was subjected to oxidation with MnO2 under
reflux to give the α,β-unsaturated aldehyde 7 in 64% yield
in two steps. To link the two fragments, aldehyde 7 was
exposed to a Grignard reaction with easily available bro-
mide 8 to form exo-olefin 9 as a mixture of diastereomers
with respect to the hydroxy group (90). We separated each
diastereomer by silica gel column chromatography and
determined its stereochemistry by means of the modified
Mosher's method. The proton NMR spectra of the two
synthesized MTPA esters (Figures S1 and S2) were
1.1 | General
All reagents were ultra-pure grade and purchased from
WAKO Pure Chemical or Nacalai Tesque. NMR spectra
were recorded by a JEOL JNM-ECS 400 FT NMR system.
MS spectra were recorded with a Bruker APCI TOF-MS.
Cell culture media and insulin were purchased from
Gibco. Mouse fibroblastic 3T3-L1 cells were maintained
in DMEM supplemented with 10% calf serum, and HeLa
cells were maintained in DMEM supplemented with 10%
fetal calf serum at 37ꢀC in a 5% CO2 atmosphere and
divided every 2 or 3 days before they reached confluence.
1.2 | Adipogenic differentiation
The induction of adipose cells from 3T3-L1 cells was
performed as described elsewhere.11 Briefly, 3T3-L1
cells were seeded onto 96-well plates at a density of

228
KAWAZOE ET AL.
SCHEME 1 Synthetic pathway to compounds 11 and 110
carefully compared to investigate their differences. As a
result, the left and right signs were divided around the
MTPA plane (Figure 2). Therefore, we concluded that the
stereochemistry of the stereotype applied was S. Conver-
sion of secondary hydroxyl to a methoxy group occurred
smoothly by methylation using MeI and NaH. On the
other hand, three-step synthesis from dimethyl
3-oxopentandioate gave allyl γ-pyrone 12, as reported
previously.12
To determine the full carbon frameworks, we turned
to the cross-metathesis reaction with Hoveyda-Grubbs
second generation catalyst (HG-II). A terminal olefin can
easily homodimerize during the cross-metathesis
reaction.13 Diluted allyl γ-pyrone was slowly added
2.2 | Synthesis of compound 17, an
enantiomer of compound 14, and its
diastereomer 19
Yoshinone A contains two asymmetric carbon atoms,
indicating that there are four stereoisomers including
yoshinone A itself. We prepared two of these four com-
pounds. To complete structure–activity relationship study
of all stereoisomers of yoshinone A, next, we tried to syn-
thesize the remaining two stereoisomers, 17 and 19. We
commenced the synthesis of 17 from methyl L-lactate 18
via the same pathway to 14. Compound 19, an enantio-
mer of 16, was synthesized by oxidation of allyl alcohol
by means of MnO2 followed by Luche reduction
(Scheme 3). Reduction gave a racemate of 14 and 19. We
separated the diastereomers by column chromatography
and determined their identities by comparing their NMR
spectra.
to
4
equivalents of exo-olefin 11 to prevent
homodimerization to give the desired coupling product
13. Following the deprotection of 13 in AcOH/THF/
H2O = 3:1:1, we obtained the desired compound 14
(Scheme 2). Additionally, we synthesized 16 which is a
diastereomer at the C11 methine of 14 from the epimer
110 by the same reaction sequence.
2.3 | Synthesis of compounds 20, 21, and
23
Next, we focused on the structure–bioactivity relationship
around C11 and C14. A cross-metathesis reaction
between 9 and 12 led to the synthesis of C14-protected
allyl alcohol 20 (Scheme 2).
Furthermore, we examined the effect of the pyrone
ring at the terminal moiety of yoshinone
A. Commercially available safrole 21 was exposed to
cross-metathesis followed by deprotection of silyl ether to
give the desired derivative 22 in 24% yield over two steps
(Scheme 4).
FIGURE 2 ΔδRS values from the 1H-NMR spectra of different
MTPA esters

KAWAZOE ET AL.
229
SCHEME 2 Synthetic pathway to the stereoisomers
SCHEME 3 Synthetic pathway to stereoisomers 17 and 19
SCHEME 4 Synthesis of yoshinone A derivative 22
2.4 | Activities of natural and synthetic
yoshinone A (compounds 1 and 14)
adipogenic differentiation of 3T3-L1 cells has been
used for purpose.8 Although this system was sensitive
enough to evaluate yoshinone A and its derivatives, it
took more than 1 week, indicating that it was not suit-
able for processing many samples. Thus, we tried to
Finally, we evaluated the activities of natural and
synthetic yoshinone
A
and its derivatives. The

230
KAWAZOE ET AL.
develop a novel evaluation system to estimate the
activities of the derivatives. In a previous study, we
showed that the lactate level in the culture medium
increased in parallel with the dose of yoshinone A.9
Based on this observation, we speculated that the
lactate level could be helpful for evaluating yoshinone
A and its derivatives. To verify this idea, we tried to
compare the lactate-producing effect to the adipogenic
differentiation-inhibitory effect of yoshinone A. We
measured the lactate levels in the HeLa cell culture
medium after 2 days of incubation or triglyceride levels
in 3T3-L1 cell adipocytes after 7 days of induction
with insulin in the presence of various concentrations
of natural yoshinone A. As a result, the EC50 value of
the lactate-producing activity was comparable with the
IC50 value of the adipogenesis-inhibitory activity of
yoshinone A (130 and 308 nM, respectively; Figure 3A,
B). We also compared the EC50 and IC50 values of
synthetic yoshinone A, compound 14. Again, these
values were comparable (150 and 297 nM, respectively;
Figure 3C,D). These results suggested that lactate-
2.5 | Evaluation of yoshinone A
derivatives
Next, we examined the lactate-producing activities of
yoshinone A stereoisomers, compounds 16, 17, and
19. Compounds 16 and 17 did not exhibit strong
lactate-producing activity and showed higher EC50
values (64.5 mM for compound 16, 1.15 mM for
compound 17; Figure 4A,B). On the other hand,
compound 19 hardly showed lactate-producing
activity. As shown in Figure 4C, the effect was limited,
and only a trace level of lactate production was
observed.
Based on these results, the activity of compound 14
very closely matched that of the natural product
yoshinone A, suggesting that the configuration of
yoshinone A should be the same as that of compound 14.
This conclusion is consistent with the stereochemistry
determined by total synthesis.10
Next, we focused on the structure–activity relation-
ship around C11 and C14. We changed the C11
methoxy group to a hydroxy group (compound 20) or
introduced a TBS group at a C14 hydroxyl group
producing activity should be
a
simple tool for
evaluating yoshinone A derivatives.
FIGURE 3 Activities of yoshinone A and compound 14. Cells were treated with yoshinone A (A and B) or compound 14 (C and D),
their lactate-producing activity (A and C) and adipogenesis-inhibiting activity (B and D) were measured. Each value is the mean SE of
triplicate determinations

KAWAZOE ET AL.
231
FIGURE 4 Activities of yoshinone A stereoisomers. Lactate-
producing activities of compounds 16 (A), 17 (B), and 19 (C) were
examined. Each value is the mean SE of triplicate determinations
FIGURE 5 Activities of yoshinone A derivatives. Lactate-
producing activities of compounds 20 (A), 13 (B), and 22 (C) were
examined. Each value is the mean of triplicate determinations
(compounds 20 and 13). While the introduction of silyl
group to a medicine or its candidate is rare, there
are some examples, including synthetic retinoid TAC-
101 and DNA topoisomerase inhibitor karenitecin.14
Both compounds showed weak activity only at a
higher concentration (EC50 values were 685 nM
and 1270 nM, respectively), indicating that the
introduction of a bulky group at C14 abolished the
activity (Figure 5A,B). Furthermore, we tried to deter-
mine the importance of the pyrone ring at the termi-
nal moiety of yoshinone A and modified it (compound
22). Interestingly, compound 22 did not enhance
lactate production at all in HeLa cells (Figure 5C).
This result indicated that the pyrone ring was indis-
pensable for this activity.
3 | CONCLUSION
We isolated yoshinone A, which is an anti-obesity candi-
date derived from cyanobacteria in Okinawa, Japan. The
compound showed clear inhibition of the adipogenic dif-
ferentiation of 3T3-L1 cells.8 Although its stereochemis-
try was determined by total synthesis,10 there was no
information available about the structure–activity rela-
tionship of the stereoisomers. In this report, we synthe-
sized all four stereoisomers of yoshinone
A and
compared their activities. We demonstrated that the 11S,
14R isomer (same as the natural product) showed much
stronger bioactivity than the others. Our findings support
the absolute configuration of yoshinone A, which was
previously determined through total synthesis, from the

232
KAWAZOE ET AL.
7. Triggle DJ, Rampe D. 1,4-Dihydropyridine activators and
antagonists: structural and functional distinctions. Trends
Pharmacol Sci. 1989;10(12):507-511.
point of view of bioactivities, suggesting that studies on
the structure–activity relationship of chiral compounds
can help to determine the absolute configuration of natu-
ral compounds. We also showed that both the pyrone
ring and the bulkiness of the side chain were important
for this activity of yoshinone A. These results will be
helpful for elucidating the mechanism of action of
yoshinone A.
8. Inuzuka T, Yamamoto K, Iwasaki A, et al. An inhibitor of the
adipogenic differentiation of 3T3-L1 cells, yoshinone A, and its
analogs, isolated from the marine cyanobacterium Lepto-
lyngbya sp. Tetrahedron Lett. 2014;55(49):6711-6714.
9. Koyama T, Kawazoe Y, Iwasaki A, Ohno O, Suenaga K,
Uemura D. Anti-obesity activities of the yoshinone A and the
related marine γ-pyrone compounds. J Antibiot. 2016;69(4):
348-351.
10. Shinomiya S, Iwasaki A, Ohno O, Suenaga K. Total synthesis
and stereochemical determination of yoshinone A. Phytochem-
istry. 2016;132:109-114.
11. Choi Y, Kawazoe Y, Murakami K, Misawa H, Uesugi M. Identi-
fication of bioactive molecules by adipogenesis profiling of
organic compounds. J Biol Chem. 2003;278(9):7320-7324.
12. De Paolis M, Rosso H, Henrot M, Prandi C, d'Herouville F,
Maddaluno J. A concise route to α'-methoxy-γ-pyrones and ver-
ticipyrone based upon the desymmetrization of α,α'-dimethoxy-
γ-pyrone. Chem a Eur J. 2010;16(37):11229-11232.
ACKNOWLEDGMENTS
The authors are extremely grateful for a Grant-in-Aid for
Scientific Research (C) (19K05720) (D. U.), the JSPS A3
Foresight Program (D. U.), and Research Grant from The
Naito Foundation (Y. K.).
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are avail-
able from the corresponding authors.
13. Chatterjee AK, Choi T-L, Sanders DP, Grubbs RH. A general
model for selectivity in olefin cross metathesis. J am Chem Soc.
2003;125(37):11360-11370.
ORCID
Yoshinori Kawazoe https://orcid.org/0000-0001-7925-
7113
14. Ramesh R, Reddy DS. Quest for novel chemical entities
through incorporation of silicon in drug scaffolds. J Med Chem.
2018;61(9):3779-3798.
REFERENCES
1. Franks ME, Macpherson GR, Figg WD. Thalidomide. Lancet.
2004;363(9423):1802-1811.
2. Vargesson N. The teratogenic effects of thalidomide on limbs.
J Hand Surg Eur Vol. 2019;44(1):88-95.
3. Ito T, Ando H, Suzuki T, et al. Identification of a primary
target of thalidomide teratogenicity. Science. 2010;327(5971):
1345-1350.
SUPPORTING INFORMATION
Additional supporting information may be found online
in the Supporting Information section at the end of this
article.
4. Asatsuma-Okumura T, Ando H, De Simone M, et al. p63 is a
cereblon substrate involved in thalidomide teratogenicity. Nat
Chem Biol. 2019;15(11):1077-1084.
5. Bechem M, Schramm M. Calcium-agonists. J Mol Cell Cardiol.
1987;19(Suppl 2):63-75.
6. Triggle DJ, Langs DA, Janis RA. Ca2+ channel ligands:
structure-function relationships of the 1,4-dihydropyridines.
Med Res Rev. 1989;9(2):123-180.
How to cite this article: Kawazoe Y, Itakura Y,
Inuzuka T, Omura S, Uemura D. Structure–activity
relationship study of the anti-obesity natural
product yoshinone A. Chirality. 2021;33:226–232.
https://doi.org/10.1002/chir.23292