Synthesis of riccardin C and its seven analogues. Part 1: The role of their phenolic hydroxy groups as LXRα agonists

Article abstract of DOI:10.1016/j.bmcl.2008.12.022

Riccardin C, a nuclear receptor LXRα selective agonist, is an 18-membered macrocyclic bisbibenzyl isolated from several liverworts. Synthesis of riccardin C and its seven O-methylated derivatives was accomplished. The synthetic sequence highlights an intramolecular Suzuki-Miyaura coupling in the formation of the 18-membered biaryl linkage present in riccardin C. The structure-activity relationship of these compounds suggests that all of the phenolic hydroxy groups present in riccardin C are essential for the activation of LXRα.

Full text of DOI:10.1016/j.bmcl.2008.12.022

Bioorganic & Medicinal Chemistry Letters 19 (2009) 738–741
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis of riccardin C and its seven analogues. Part 1: The role of their
phenolic hydroxy groups as LXRa agonists
Hideaki Hioki a, Naoki Shima a, Kota Kawaguchi a, Kenich Harada a, Miwa Kubo a, Tomoyuki Esumi a,
Tomoko Nishimaki-Mogami b, Jun-ichi Sawada b, Toshihiro Hashimoto a, Yoshinori Asakawa a,
Yoshiyasu Fukuyama a,
*
a Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho, Tokushima 770-8514, Japan
b National Institute of Health Sciences, Kamiyoga 1-18-1, Setagaya-ku, Tokyo 158-8501, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Riccardin C, a nuclear receptor LXRa selective agonist, is an 18-membered macrocyclic bisbibenzyl iso-
Received 31 October 2008
Revised 20 November 2008
Accepted 5 December 2008
Available online 10 December 2008
lated from several liverworts. Synthesis of riccardin C and its seven O-methylated derivatives was accom-
plished. The synthetic sequence highlights an intramolecular Suzuki–Miyaura coupling in the formation
of the 18-membered biaryl linkage present in riccardin C. The structure–activity relationship of these
compounds suggests that all of the phenolic hydroxy groups present in riccardin C are essential for the
activation of LXRa.
Keywords:
Ó 2008 Elsevier Ltd. All rights reserved.
Bisbibenzyl
Riccardin C
LXRs
Agonist
Synthesis
Structure–activity relationship
Suzuki–Miyaura coupling
Macrocyclic bisbibenzyls such as marchantins, riccardins and
plagiochins are phenolic constituents that are characteristic of liv-
erworts and exhibit a variety of biological activities.1 Riccardin C
(1) was originally isolated from Reboulia hemisphaerica (L.) Raddi2a
and was later found in several liverworts.2 Recently riccardin C was
Four groups have reported the synthesis of riccardin C (1). The
key step in the synthesis of 1 is the construction of the 18-mem-
bered macrocyclic ring. All of their formations of the corresponding
18-membered ring were employed at the benzilic position (paths a
and b in Fig. 1) by Wurtz,5a Wittig-type5b,d, and McMurry reac-
tion.5c Previously we reported the synthesis of plagiochin A and
D,6a,b and isoplagiochin D,6c which are variant members of the
16-membered macrocyclic bisbibenzyls with a biaryl linkage, by
applying intramolecular Stille–Kelly or Suzuki–Miyaura reaction
as the key step to form the biaryl linkage. Our approach aiming
discovered to bind directly to liver X receptor
of the nuclear receptor superfamily, leading to the activation of
LXR
/RXR-dependent reporter gene transcription.3 Interestingly,
a (LXRa), a member
a
riccardin C functions as an antagonist but not an agonist of LXRb.
In addition, it has no ability to activate other nuclear receptors,
such as PPARc, RARa, RARb, RARc, FXR, and RXRa, and increases
LXR-target gene expression in macrophages. From these results
riccardin C was expected to be a lead compound for the treatment
of cardiovascular-related diseases such as arteriosclerosis because
LXRs are thought to regulate cholesterol metabolites.4 In this
paper, we report our independent synthesis of riccardin C and its
seven O-methylated analogues and their selective manner of bind-
OH
O
path a
path b
ing to LXRs,
a and b, and discuss the role of three hydroxy groups
HO
related to the activation of LXRs through structure–activity rela-
tionship of the synthesized compounds.
OH
path c
riccardin C (1)
Figure 1. Riccardin C (1), a LXRa receptor agonist. Paths a and b: macrocyclization
* Corresponding author. Tel.: +81 88 602 8435; fax: +81 88 655 3051.
E-mail address: fukuyama@ph.bunri-u.ac.jp (Y. Fukuyama).
points to synthesize 1 by other groups. Path c: a macrocyclization point in this
study.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.12.022

H. Hioki et al. / Bioorg. Med. Chem. Lett. 19 (2009) 738–741
739
respectively, according to the literature precedents shown in
Br
Scheme 1. Segment A was prepared by nucleophilic ipso substitu-
tion at the C-4 position of 4-bromobenzaldehyde (2) with methyl
3-hydroxy-4-methoxybenzoate (3).9 Michaelis–Arbuzov reaction
of 2-bromo-5-methoxybenzyl bromide (5) afforded segment (6)6a
in good yield. Protection of the phenolic hydroxy group in vanilin
(7) as a methoxymethyl (MOM) ether provided segment C (8).10
With all segments in hand, a precursor for macrocyclization was
prepared by successive Horner–Wadsworth–Emmons (HWE) reac-
tions shown in Scheme 2. Segment A was coupled with segment B
by the first HWE reaction to give 9 in 98% yield. The ester group in
9 was reduced by LiAlH4 and the resulting hydroxy group in 10 was
converted to phosphonate 11 by consecutive bromination and
phosphorylation. The second HWE reaction of 11 with segment C
also proceeded smoothly to give 12 which have all carbon atoms
required for the synthesis of riccardin C. Two double bonds in 12
have to be reduced with keeping a bromine atom on an aromatic
group intact. After several unsuccessful attempts of the selective
hydrogenation under H2 atmosphere by using Pd or Pt catalyst,
the olefins in 12 were selectively reduced by using triethylsilane
in trifluoroacetic acid11 at 60 °C. A MOM group was also removed
during the reaction. The resulting phenolic hydroxy group in 13
was converted to trifluoromethanesulfonate 14 in 96% yield. Selec-
tive halogen-boronate exchange is required in order to prepare the
cyclization precursor 15 even in the presence of aryl-
trifluoromethanesulfonyl group. The reaction selectively pro-
ceeded by treatment of 14 with bis(pinacolato)diboron and
10 mol % of Pd(PPh3)4 affording the desired product 15 in 95%
yield.
OHC
HO
OMe
2
O
a
+
OHC
OMe
CO2Me
segment A (4)
CO2Me
3
O
Br
P(OMe)2
b
MeO
MeO
Br
Br
5
segment B (6)
CHO
CHO
c
OMe
OH
OMe
OMOM
7
8
segment C ( )
Scheme 1. Preparation of all building blocks 4, 6 and 8 for the synthesis of riccardin
C. Reagents and conditions: (a) K2CO3, DMSO, 140 °C, 4 h, 68%; (b) P(OMe)3, 90 °C,
2 h, 98%; (c) MOMCl, NaH, THF, 1.5 h, 84%.
at synthesis of the biaryl-type macrocyclic bibenzyls is most likely
to follow the biosynthetic pathway involved in the series of macro-
cyclic bisbibenzyls.7 Thus we have applied the same strategy (path
c in Fig. 1) for the synthesis of riccardin C as part of our studies on
natural product synthesis utilizing transition metal catalysts.8
The synthesis commenced with preparation of segments A (4), B
(6), and C (8) from commercially available chemicals in one step,
The crucial macrocyclization Suzuki–Miyaura coupling to
form the 18-membered macrocyclic ring was examined. The em-
ployed reaction conditions are shown in Table 1. Pd(PPh3)4
(10 mol %)–K3PO4–DMF system, which was effective for the syn-
thesis of isoplagiochin D,6c was employed for this cyclization at
first. The desired product 16, however, was obtained in only
OMe
OH
OMe
O
O
OMe
O
a
b
c, d
MeO
MeO
CO2Me
OHC
CO2Me
Br
Br
segment A (4)
10
9
OMe
OMe
O
O
OMe
O
e
f
MeO
MeO
MeO
P(OMe)2
O
Br
Br
13
Br
OMe
OMOM
OMe
OH
11
12
OMe
OMe
O
O
h
g
MeO
MeO
O
B
O
Br
OMe
OMe
OTf
OTf
14
15
Scheme 2. Preparation of cyclization precursor 15. Reagents and conditions: (a) NaH, THF,10 min, 0 °C then segment B (6), 20 h, 99%; (b) LiAlH4, THF, 0.5 h, 90%; (c) NBS,
Me2S, CH2Cl2; 2 h –10 °C to rt; (d) P(OMe)3, 100 °C, 1 h, 98% (2 steps); (e) NaH, THF, 10 min, 0 °C, then segment B, 3 h 87%; (f) Et3SiH, TFA, 60 °C, 25 min, 74%; (g) N-phenyl-
bis(trifluoromethanesulfonimide), Cs2CO3, MeCN/DMF (9/1), 8.5 h, 96%; (h) 10 mol % Pd(PPh3)4, bis(pinacolato)diboron, K3PO4, dioxane, 100 °C, 4.5 h, 95%.

740
H. Hioki et al. / Bioorg. Med. Chem. Lett. 19 (2009) 738–741
Table 1
Macrocyclization by Suzuki–Miyaura coupling
OMe
OMe
O
OMe
MeO
O
OMe
Pd cat (10 mol %)
additive (20 mol %)
base (3 eqiv.)
O
MeO
MeO
solvent (5 × 10-3 M)
100˚C, 12h
+
O
B
O
OMe
MeO
OTf
OMe
OMe
O
OMe
15
16
17
Entry
Pd cat.
Additive
Base
Solvent
Isolated yield (%)
16
17
1
2
3
4
Pd(PPh3)4
Pd2(dba)3
Pd2(dba)3
Pd2(dba)3
Pd2(dba)3
Pd2(dba)3
Pd2(dba)3
K3PO4
K3PO4
DMF
DMF
DMF
THF
DMSO
DMF
DMF
9
SPhos (18)a
SPhos (18)a
SPhos (18)a
SPhos (18)a
SPhos (18)a
SPhos (18)a
16
37
34
3
39
15
b
Na2CO3
Na2CO3
Na2CO3
Na2CO3
Na2CO3
27
24
b
b
b
b
5
6c
7d
33
4
a
PCy2
OMe
MeO
18
SPhos (
)
b
c
3 equiv of 2 M aqueous Na2CO3 solution were added.
20 mol % of Pd2(dba)3 and 40 mol % of SPhos were used.
d
Higher diluted conditions were applied (2.5 Â 10À3 M) and the reaction period was 24 h.
OR1
OMe
OH
b
c
O
O
O
16
1
a
R3O
MeO
OR2
HO
OMe
OH
16
1
1
2
3
19a 19f
-
; R , R , R = H or Me
6 possible analogues
Scheme 3. Synthesis of riccardin C (1) and its six analogues. Reagents and conditions: (a) 12 equiv of BBr3, CH2Cl2, 0 °C to rt, 7 h, 97%; (b) 5 equiv of BBr3, CH2Cl2, 0 °C to rt,
20 min, yields are shown in Table 2; (c) 12 equiv of trimethylsilyldiazomethane, Et2O, 20 h, yields are shown in Table 2.
9% yield (entry 1 in Table 1). The yield was improved when
SPhos (18), a highly effective phosphine ligand for the prepara-
20
LXRα
tion of extremely hindered biaryls in Suzuki–Miyaura coupling
15
10
5
RC
S1
developed by Buchwald et al.,12 was applied to our reaction (en-
try 2). This macrocyclization was highly dependent on the base.
Na2CO3 enhanced reactivity to give desired product 16 in 37%
LXRβ
RC
S1
Table 2
Isolated yields of 19a–f
0
0
1
10
100
Concentration (μM)
Compound
R1
R2
R3
Isolated yield (%)
From 16 From 1
Figure 2. Activation of LXR
cells were transfected with a reporter plasmid (pLXREX4-tk-Luc) and expression
plasmids for LXR (or LXRb) and RXR together with a plasmid for b-galactosidase
a and LXRb by natural and synthetic riccardin C. COS-1
19a (riccardin A)
19b
19c (riccardin F)
19d
19e
19f
H
H
Me
H
Me
Me
H
Me
H
Me
H
Me
Me
H
H
Me
Me
H
14
21
—
22
—
8
7
10
—
10
5
a
a
as an internal control, and were treated for 24 h with various concentrations of
natural riccardin C (RC) or synthetic riccardin C (S1) in the presence of compactin
which depletes endogenous LXR ligands. The luciferase activity in the cell extracts
was normalized using b-galactosidase and expressed as the fold induction relative
to the vehicle-treated cells.
—

H. Hioki et al. / Bioorg. Med. Chem. Lett. 19 (2009) 738–741
741
25
20
15
10
5
LXRα
10 μM
30 μM
TO
LXRβ
10 μM
30 μM
TO
0
Control RC
S1
16
19a
19b
19c 19d
19e
19f TO901317
(100 nM)
Figure 3. Agonistic activities on LXRa and LXRb by natural (RC) and synthetic riccardin C (S1) and seven analogues (16 and 19a–f). TO901317 was used as a positive control.
Experimental procedure is same as that described in the captions for Figure 1.
yield along with 27% of cyclic dimer 17 (entry 3). The yield was
reduced in THF or DMSO (entries 4 and 5). Larger amounts of Pd
catalyst (20 mol %) and SPhos (40 mol %) gave slightly improved
yields (entry 6). Higher diluted conditions were not effective for
suppressing the dimerization (entry 7).
Acknowledgments
This study was financially supported by a Grant-in-Aid from the
Ministry of Education, Culture, Sports, Science, and Technology of
the Japanese Government (No. 18590113).
Finally, treatment of 16 with BBr3 removed three O-methyl
groups giving rise to riccardin C in 97% yield (Scheme 3). The spec-
troscopic data of synthetic compound 1 were in agreement with
those of the natural riccardin C.
References and notes
1. Asakawa, Y.. In Progress in the Chemistry of Organic Natural Products; Herz, W.,
Kirby, G. W., Moore, R. E., Steglich, W., Tamm, C., Eds.; Springer: Vienna, 1995;
Vol. 65, pp 618–621.
2. (a) Asakawa, Y.; Matsuda, R. Phytochemistry 1982, 21, 2143; (b) Lu, Z. Q.; Fan, P.
H.; Ji, M.; Lou, H. X. J. Asian Nat. Prod. Res. 2006, 8, 187; (c) Harinantenaina, L.;
Asakawa, Y. Chem. Pharm. Bull. 2004, 52, 1382; (d) Bardón, A.; Kamiya, N.;
Toyota, M.; Takaoka, S.; Asakawa, Y. Phytochemistry 1999, 52, 1323; (e) Anton,
H.; Schoeneborn, R.; Mues, R. Phytochemistry 1999, 52, 1639. and references
cited therein.
In our previous study, riccardin F (19c), a 14-methoxy derivative
of riccardin C, activated neither LXRa
nor LXRb.3 This drastic differ-
enceintheactivitybetweenriccardinCandFpromptedusto synthe-
size partially O-methylated analogues to explore the structure–
activity relationship of three hydroxy groups. Treatment of perme-
thylated riccardin C (16) with 5 equiv of BBr3 led to three partially
O-methylated compounds 19a, 19b, and 19d. On the other hand,
reaction of riccardin C with 12 equiv of trimethylsilyldiazomethane
provided five compounds 19a, 19b, 19c, 19e, and 19f. The isolated
yields of these analogues are shown in Table 2. Their structures were
carefully determined by NOESY spectroscopic analyses and compar-
ison with the spectroscopic data of natural riccardins C and F. Prep-
aration of all possible partially O-methylated analogues were
completed by these reactions.
3. Tamehiro, N.; Sato, Y.; Suzuki, T.; Hashimoto, T.; Asakawa, Y.; Yokoyama, S.;
Kawanishi, T.; Ohno, Y.; Inoue, K.; Nagao, T.; Nishimaki-Mogami, T. FEBS Lett.
2005, 579, 5299.
4. Li, A. C.; Glass, C. K. J. Lipid Res. 2004, 45, 2161.
5. (a) Gottsegen, Á.; Nógrádi, M.; Vermes, B.; Kajtár-peredy, M.; Bihátsi-karsai, É.
Tetrahedron Lett. 1988, 29, 5039; (b) Eicher, T.; Fey, S.; Puhl, W.; Büchel, E.;
Speicher, A. Eur. J. Org. Chem. 1998, 1998, 877; (c) Harrowven, D. C.; Woodcock,
T.; Howes, P. D. Angew. Chem., Int. Ed. 2005, 44, 3899; (d) Dodo, K.; Aoyama, A.;
Noguchi-Yachide, T.; Makishima, M.; Miyachi, H.; Hashimoto, Y. Bioorg. Med.
Chem. 2008, 16, 4272.
6. (a) Fukuyama, Y.; Yaso, H.; Nakamura, K.; Kodama, M. Tetrahedron Lett. 1999, 40, 105;
(b) Fukuyama, Y.; Yaso, H.; Mori, T.; Takahashi, H.; Minami, H.; Kodama, M.
Heterocycles 2001, 54, 259; (c) Esumi, T.; Wada, M.; Mizushima, E.; Sato, N.; Kodama,
M.; Asakawa, Y.; Fukuyama, Y. Tetrahedron Lett. 2004, 45, 6941.
7. (a) Friederich, S.; Maier, U. H.; Deus-Neumann, B.; Yoshinori Asakawa, H.; Zenk,
M. Phytochemistry 1999, 50, 589; (b) Keserü, G. M.; Nógrádi, M. Phytochemistry
1992, 31, 1573.
8. Esumi, T.; Zhao, M.; Kawakami, T.; Fukumoto, M.; Toyota, M.; Fukuyama, Y.
Tetrahedron Lett. 2008, 49, 2692. and references cited therein.
9. Ha, T. T. N.; Nogradi, M.; Brlik, J.; Kajtar-Peredy, M.; Wolfner, A. J. Chem. Res.
Synop. 1991, 137.
We first confirmed the LXRs agonistic activity of synthetic ric-
cardin C by comparing it with that of natural riccardin C. Upon
cotransfection with LXRa
/RXR, 30 mmol LÀ1 of synthetic riccardin
C raised the transactivation of an LXR-response element-driven re-
porter gene by approximately 15-fold. The dose–response curve
was in good accordance with that of natural riccardin C shown in
Figure 2. In addition, neither of them activated LXRb. Synthetic ana-
logues of riccardin C were also screened by the transient transfec-
tion assay shown inFigure 3. Unexpectedly, none of them activated
10. Mabic, S.; Vaysse, L.; Benezra, C.; Lepoittevin, J. Synthesis 1999, 1127.
11. Kursanov, D. N.; Parnes, Z. N.; Bassova, G. I.; Loim, N. M.; Zdanovich, V. I.
Tetrahedron 1967, 23, 2235.
12. (a) Walker, S. D.; Barder, T. E.; Martinelli, J. R.; Buchwald, S. L. Angew. Chem., Int.
Ed. 2004, 43, 1871; (b) Barder, T. E.; Walker, S. D.; Martinelli, J. R.; Buchwald, S.
L. J. Am. Chem. Soc. 2005, 127, 4685.
LXR
a or LXRb. These results clearly indicate that all hydroxy
groups in riccardin C are essential for its binding to LXRa 13
.
In conclusion, riccardin C and seven analogues were synthe-
sized, via the intramolecular Suzuki–Miyaura coupling to construct
the necessary 18-membered biaryl linkage. Three phenolic hydro-
xy groups in riccardin C were clarified to be indispensable for bind-
13. Recently, Hashimoto et al. reported that riccardin C had no agonistic activity
toward LXRs by using
a chimeric receptor system in which the ligand
binding domain of LXR was fused to the yeast GAL4 DNA binding domain
(Ref. 5d). In contrast, in the present and previous studies, we have shown
the agonist activity of riccardin C in a physiological assay system using a
heterodimer of full-length LXR and RXR and an LXR-responsive element-
driven reporter gene.
ing to LXRa according to structure–activity relationship studies.
The present structure–activity relationship studies on riccardin C
provide beneficial information for the design and synthesis of more
potent and selective LXRa agonists.