Synthetic approach towards nakadomarin A: Efficient synthesis of the central tetracyclic core

Article abstract of DOI:10.1016/S0040-4039(01)01792-0

An efficient synthesis of the tetracyclic core of nakadomarin A was accomplished starting from methyl 4-oxo-3-piperidinecarboxylate. The key steps were intramolecular cyclization of furan to N-acyliminium ions to construct the strained central cyclopenten

Full text of DOI:10.1016/S0040-4039(01)01792-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 8345–8349
Synthetic approach towards nakadomarin A: efficient synthesis
of the central tetracyclic core
Toshiaki Nagata, Atsushi Nishida* and Masako Nakagawa*
Graduate School of Pharmaceutical Sciences, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan
Received 23 August 2001; revised 17 September 2001; accepted 21 September 2001
Abstract—An efficient synthesis of the tetracyclic core of nakadomarin A was accomplished starting from methyl 4-oxo-3-pipe-
ridinecarboxylate. The key steps were intramolecular cyclization of furan to N-acyliminium ions to construct the strained central
cyclopentene ring. © 2001 Elsevier Science Ltd. All rights reserved.
Nakadomarin A was first isolated from the Okinawan
marine sponge Amphimedon sp. by Kobayashi in 1997,1
N
N
CHO
N
F
B
H
H
H
H
and was found to be biogenetically closely related to
manzamine alkaloids, such as manzamine A and ircinal
A (Fig. 1). Its structure was elucidated spectroscopically
and consists of an unprecedented hexacyclic ring system
(8/5/5/5/15/6) that includes a furan ring. An interesting
biogenetic transformation of ircinal A to nakadomarin
A has been proposed.1 Although some biological activ-
ities of nakadomarin A, such as cytotoxicity against
murine lymphoma L1210 cells, inhibitory activity
against Cdk4, and anti-microbial activity, have been
reported,1 its limited availability (1.8×10−3%, wet
weight) has prevented a complete survey of its biologi-
cal activity. In connection with our ongoing project on
the total synthesis of manzamine alkaloids,2 we were
interested in nakadomarin A because of its unique
structure and began a synthetic study. We report here
an efficient synthesis of the ABCD ring system of
nakadomarin A.3,4
C
OH
OH
O
A
N
N
N
D
N
N
H
H
H
E
Nakadomarin A
Figure 1.
Manzamine A
Ircinal A
be synthesized by cyclization of the furan ring to the
iminium ion generated from the spiro-g-lactam. The
spiro-g-lactam could be further simplified to methyl
4-oxo-3-piperidinecarboxylate hydrochloride.6 One
anticipated problem was generation of the iminium ion
because of the poor reactivity of the carbonyl group
due to steric hindrance.7
Methyl 4-oxo-3-piperidinecarboxylate hydrochloride 1
was first converted to allylated 2 in 77% yield in three
steps (Scheme 2). Oxidative cleavage of the double
bond of 2 gave 3. Reductive cyclization of 3 to the
desired spiro-g-lactam 4 proceeded successfully when 3
was treated with 4-methoxybenzylamine followed by
NaBH3CN. Sequential deprotection of both methoxy-
benzyl and ketal groups with CAN and 70% HClO4
gave g-lactam 5, which was acylated by Boc2O to give
6.
Recent advances in ring-closing metathesis (RCM)5
have facilitated access to unsaturated carbocycles or
heterocycles with medium-sized to large rings. Consid-
ering both RCM and a cyclization methodology at the
late stage of synthesis for constructing the 15-mem-
bered ring, we assumed that the ABCD tetracyclic core
was a key synthetic intermediate for nakadomarin A
(Scheme 1). Synthesis of this ABCD ring system could
be retrosynthetically simplified to a spiro-g-lactam with
a substituted furan. We expected that the B ring could
The keto group of 6 was converted to the enol triflate 7.
Suzuki–Miyaura coupling8 of 7 with furan-3-boronic
acid, prepared from 3-bromofuran, was performed
under standard conditions using Pd(PPh3)4 to give cou-
Keywords: nakadomarin A; manzamine alkaloid; furan; Suzuki–
Miyaura coupling; N-acyliminium ion.
* Corresponding authors.
pling product 8 in 80% yield. For the critical
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01792-0

8346
T. Nagata et al. / Tetrahedron Letters 42 (2001) 8345–8349
PO
H
F
B
H
H
H
C
R=H
C
O
A
X
O
O
O
A
N
B
HN
N
or
N
D
P
D
N
N
N
N
H
H
H
P
H
E
R
P=protecting groups
Nakadomarin A
PO
H
PO
H
O
OH
COOMe
O
O
N
N
O
N
N
COOMe
P
P
P
H
HCl
N
N
P
P
Scheme 1.
O
O
O
O
OH
a, b, c
77%
e
d
O
O
O
HCl•HN
N
N
N
N
CO2Me
CO2Me
CO2Me
Bs
Bs
Bs
Bs
N
53%
44%
PMB
O
CHO
1
2
3
7
4
8
Bs=SO2Ph
h
O
O
O
O
OTf
O
f, g
j
i
O
N
N
N
Bs
Bs
Bs
59%
63%
50%
80%
N
N
N
N
H
Boc
Boc
Boc
5
6
O
m
k, l
O
OAc
N
N
Bs
Bs
74%
76%
N
Boc
N
Boc
9
10
Scheme 2. Reagents and conditions: (a) PhSO2Cl, NaHCO3; (b) allyl bromide, K2CO3; (c) ethylene glycol, pTsOH; (d) OsO4,
NaIO4, aq. THF; (e) PMB–NH2, MeOH, AcOH, rt, 1 h, then NaBH3CN, reflux, 2 h; (f) CAN, aq. MeCN, rt; (g) 70% HClO4,
CH2Cl2; (h) Boc2O, Et3N, DMAP; (i) LiN(TMS)2, THF, −65°C, then PhNTf2, 3°C; (j) furan-3-boronic acid, Pd(PPh3)4, LiCl,
DME, aq. Na2CO3, 80°C, 3 h; (k) DIBAL, toluene, −65°C to rt; (l) Ac2O, pyridine; (m) pTsOH, CH2Cl2.
intramolecular cyclization, the g-lactam 8 was con-
verted to acetoxypyrrolidine 9 by DIBAL reduction
followed by acetylation. The presence of the Boc group
is essential for selective and efficient reduction of 8. No
over-reduction was observed even when a large excess
of DIBAL was used, probably due to stabilization of
alkoxyaluminum intermediate of the hemiaminal by
complex formation. Cyclization of 9 occurred regiose-
lectively at the 2-position of the furan ring to give the
tetracyclic key intermediate 109 in 76% yield upon
We first expected that the stereoselective hydrogenation
of 10 would easily occur to give 13 (Scheme 3). Reduc-
tion from the a-side of 10 was expected according to a
previous result with a related unsaturated spiro-lac-
tam.7 When 10 was hydrogenated under standard con-
ditions, however, only 12, an over-reduced product,
was obtained in 80% yield.11 To determine the presence
of the desired 13, the reduction was stopped before all
of the starting material was consumed. A 1H NMR
study of the product mixture revealed the presence of
11 instead of 13. These results showed that the reduc-
tion of 10 occurred first at the tetra-substituted double
bond in the furan ring. This high reactivity of the
treatment with p-toluenesulfonic acid.10
A mole-
cular model suggested that trans cyclization was not
possible.
H2
5% Pd-C
O
O
O
O
N
N
N
N
Bs
Bs
Bs
Bs
AcOEt
N
N
N
N
Boc
Boc
Boc
Boc
12
13
10
11
Scheme 3.

T. Nagata et al. / Tetrahedron Letters 42 (2001) 8345–8349
8347
double bond in the furan ring can be explained by the
strained character of 10. Therefore, the hydrogenation
of 8 was investigated next (Scheme 4). As expected,
hydrogenation of 8 proceeded smoothly without reduc-
tion of the furan ring to give 14 in 68% yield as a single
diastereomer. In a careful NMR study of 14,12 observa-
tion of nuclear Overhauser effects (NOE) between the
axial methine proton and a proton of the methylene in
g-lactam suggested that reduction occurred from the
a-side of 8. Compound 14 was converted to 15 as
above, which was subjected to the cyclization reaction
under acidic conditions to give the desired tetracyclic
framework of nakadomarin A 1313 in 34% yield from
14. The cis relationship of the AB rings was confirmed
by an NOE study.
The remaining task was elongation of the side chains to
construct the 15-membered ring F. The required
boronic ester of substituted furan was prepared as
shown in Scheme 5. Wadsworth–Emmons–Horner
reaction of 4-bromo-2-furaldehyde followed by careful
reduction of the double bond using NaBH4 in the
presence of NiCl2 gave 17 in 89% yield.14 Compound 17
was then converted to 18 by the Pd-catalyzed reaction15
with diborane.
Suzuki–Miyaura coupling of 18 with the enol triflate 7,
which was obtained from 6, gave highly substituted
spirolactam 19 in fairly good yield (Scheme 6). Hydro-
genation of the double bond followed by reduction of
the lactam carbonyl and acetylation gave 21. The
H
O
H
H
O
O
a
d
b, c
O
OAc
O
O
N
N
N
N
H
Bs
Bs
Bs
Bs
68%
N
34%
(3 steps)
N
N
Boc
N
Boc
Boc
H
Boc
Boc
8
14
15
13
6
4"
5"
Bs
O
5
N
O
N
2" O
2
4
H
H
H
H
1'
N
Bs
H
H
4'
H
NOE
N
H
Boc
H
H
H
NOE
3'
Scheme 4. Reagents and conditions: (a) H2, 10% Pd-C; (b) DIBAL, toluene, −65°C to rt; (c) Ac2O, pyridine; (d) pTsOH, CH2Cl2.
O
B
1)
(EtO)2P(O)
COOEt
Br
O
B
O
Br
O
2
tBuONa, THF
CO2Et
CHO
CO2Et
O
O
O
PdCl2(dppf)
AcOK,
2) NaBH4, NiCl2
EtOH, 0 °C
18
16
17
DMSO
89% (2 steps)
74%
Scheme 5.
COOEt
c
COOEt
d,e
O
O
H
O
OTf
O
a
b
O
O
O
N
N
N
N
Bs
Bs
Bs
Bs
N
N
N
N
Boc
Boc
Boc
Boc
87%
6
7
19
20
68% from 6
OAc
f
OAc
H
H
H
H
g - k
O
OAc
O
O
O
RCM
N
N
N
N
Bs
Bs
N
N
N
N
Boc
Boc
22
Boc
Boc
21
23
24
49% from 20
35% from 22
Scheme 6. Reagents and conditions: (a) LiN(TMS)2, THF, −65°C, then PhNTf2, 3°C; (b) 18, Pd(PPh3)4, LiCl, DME, aq. Na2CO3,
80°C, 7 h; (c) H2, 10% Pd-C; (d) DIBAL, toluene, −65°C to rt; (e) Ac2O, pyridine; (f) pTsOH, CH2Cl2; (g) 1N NaOH, rt, MeOH;
(h) Dess–Martin oxd. rt, CH2Cl2; (i) Ph3PꢀCH2, THF–toluene; (j) Na, anthracene, DME, −65°C; (k) TsO(CH2)4CHꢀCH2,
iPr2NEt, THF.

8348
T. Nagata et al. / Tetrahedron Letters 42 (2001) 8345–8349
cyclization of 21 under the above optimal reaction
conditions proceeded smoothly to give 22 in 49% yield
from 20. The acetoxypropyl side chain at the furan ring
was then converted to a 3-butenyl group by conven-
tional transformations. The phenylsulfonyl group on
9. For 10: According to 1H NMR, compound 10 was
1
observed as a 1:1 mixture of amide rotamers. H NMR
(300 MHz, CDCl3): l 1.53 (9H, s), 1.85–2.15 (1H, m),
2.25–2.55 (1H, m), 2.67 (1H, d, J=10.8 Hz), 3.34 (1H,
brd, J=16.8 Hz), 3.58 (2H, brt, J=10.8 Hz), 3.98 (1H, d,
J=10.8 Hz), 4.25 (1H, dd, J=16.8, 3.1 Hz), 4.69 (0.5H,
brs), 4.83 (0.5H, brs), 5.43 (1H, t, J=3.1 Hz), 6.30 (1H,
brs), 7.38 (1H, brs), 7.52–7.62 (3H, m), 7.82 (2H, d,
J=7.0 Hz); 13C NMR (75 MHz, CDCl3): l 28.8, 34.0,
34.8, 44.9, 45.9, 46.5, 48.9, 59.1, 60.3, 61.7, 80.6, 105.3,
109.7, 126.4, 127.8, 129.6, 133.3, 135.1, 136.9, 148.7,
149.0, 154.6, 155.3, 162.5; IR (neat, cm−1) w 1697, 1394,
1360, 1167, 1146, 1101; LRMS (EI) 442 (M+).
the
A
ring was then deprotected by sodium
anthracenide and the resulting secondary amine was
alkylated with hexenyl tosylate to give 23,16 which is a
good substrate for testing the RCM reaction to prepare
the F ring in nakadomarin A. The results of the RCM
reaction of 23 and further transformations for the total
synthesis of nakadomarin A will be reported in due
course.
10. (a) David, M.; Dhimane, H.; Vanucci-Bacque´, C.; Lhom-
met, G. Synlett 1998, 206; (b) Ohta, T.; Shiokawa, S.;
Iwashita, E.; Nozoe, S. Heterocycles 1992, 34, 895.
Acknowledgements
1
11. For 12: H NMR (300 MHz, CDCl3): l 1.45 (9H, brs),
1.40–1.95 (9H, m), 2.2–2.60 (2H, m), 3.02–3.06 (1H, m),
3.41–3.51 (2H, m), 3.64–3.70 (2H, m), 3.95–4.09 (2H, m),
7.52–7.63 (3H, m), 7.62–7.75 (2H, m); IR (neat, cm−1) w
1693, 1402, 1365, 1167, 578; LRMS (EI) 448 (M+);
HRMS (EI) calcd for C23H32N2O5S 448.2032; found
448.2055.
This research was supported by a Grant-in-Aid for
Scientific Research of Priority Areas (A) ‘Exploitation
of Multi-Element Cyclic Molecules’ from the Ministry
of Education, Culture, Sports, Science, and Technol-
ogy. The financial supports by Uehara Memorial Foun-
dation and the Fugaku Trust for Medical Research are
also acknowledged. We also thank Mr. T. Kano for his
technical assistance and Ms. R. Hara at the Analytical
Center, Chiba University, for performing the mass
spectroscopy.
12. For 14: 1H NMR (300 MHz, CDCl3): l 1.50 (9H, s, Boc),
1.57–1.71 (2H, m, Hb-4%, Ha-5), 1.92–2.03 (1H, m, Ha-
4%), 2.45 (1H, dd, J=12.1, 3.9 Hz, Ha-4), 2.51 (1H, ABq,
J=12.4 Hz, Ha-2), 2.55–2.64 (1H, m, Ha-6), 2.70–2.85
(1H, m, Hb-5), 3.17–3.25 (1H, m, Ha-3%), 3.41–3.67 (1H,
m, Hb-3%), 3.74 (1H, dd, J=11.6, 1.5 Hz, Hb-6), 3.96
(1H, ABq, J=12.4, 1.5 Hz, Hb-2), 6.31 (1H, d, J=1.9
Hz, H-4%%), 7.28 (1H, brd, J=1.4 Hz, H-2%%), 7.32 (1H,
brd, J=1.6 Hz, H-5%%), 7.54–7.61 (3H, m, Bs), 7.88 (2H,
dd, J=8.0, 1.7 Hz, Bs); 13C NMR (75 MHz, CDCl3): l
28.2 (C5), 28.5 (tBu), 28.9 (C4%), 40.5 (C4), 42.8 (C3%),
46.3 (C6), 47.9 (C3), 52.1 (C2), 83.2 (tBu), 110.9 (C4%%),
124.2 (C3%%), 128.4 (Bs), 129.4 (Bs), 133.2 (Bs), 137.2 (Bs),
140.7 (C2%%), 143.5 (C5%%), 150.4 (C=O), 173.1 (C1%); IR
(KBr, cm−1) w 1755, 1712, 1323, 1155; HRMS (FAB)
calcd for C23H28N2OSNa 483.1566 (M+Na+); found
483.1596.
References
1. (a) Kobayashi, J.; Watanabe, D.; Kawasaki, N.; Tsuda,
M. J. Org. Chem. 1997, 62, 9236–9239; (b) Kobayashi, J.;
Tsuda, M.; Ishibashi, M. Pure Appl. Chem. 1999, 71,
1123–1126.
2. (a) Uchida, H.; Nishida, A.; Nakagawa, M. Tetrahedron
Lett. 1999, 40 (1), 113–116; (b) Nakagawa, M.; Torisawa,
Y.; Uchida, H.; Nishida, A. J. Synth. Org. Chem., Jpn.
1999, 57, 1004–1015.
3. A part of this study has been reported in 10th Interna-
tional symposium on Marine Natural Products, Oki-
nawa, Japan, June 24–29 2001, abstract p. 45.
13. For 13: 1H NMR (300 MHz, CDCl3): l 1.53 (9H, s, Boc),
1.57–1.64 (1H, m, Hb–5), 1.87–1.92 (2H, m, Ha,b-4%),
2.01–2.08 (1H, m, Ha-5), 2.76 (1H, brt, J=6.3 Hz, Ha-4),
3.06–3.25 (2H, m, Ha,b-6), 3.08 (1H, ABq, J=12.4 Hz,
Ha-2), 3.26–3.75 (2H, m, H-3%), 3.34 (1H, ABq, J=12.4
Hz, Hb-2), 4.65 (1H, s, Hb-1%), 6.14 (1H, d, J=1.8 Hz,
H-4%%), 7.33 (1H, brs, H-5%%), 7.50–7.63 (3H, m. Bs), 7.77
(2H, dd, J=8.4, 1.5 Hz, Bs); 13C NMR (75 MHz,
CDCl3): l 27.6 (C5), 28.8 (tBu), 36.3 (C4%%), 39.1 (C4),
42.8 (C6), 45.3 (C3%), 50.0 (C2), 61.2 (C3), 62.5 (C1%), 80.5
(tBu), 107.3 (C4%%), 127.7 (Bs), 128.9 (C3%%), 129.6 (Bs),
133.1 (Bs), 137.8 (Bs), 147.7 (C5%%), 154.8 (CꢀO), 157.3
(C2%%); IR (neat, cm−1) w 1693, 1398, 1167, 737; HRMS
(FAB) calcd for C23H28N2O5SNa 467.1617 (M+Na+);
found 467.1629.
4. Very recently, Dr. Fu¨rstner, Max-Planck-Institute, Ger-
many, independently reported a synthetic approach to
nakadomarin A. 18th International Congress of Hetero-
cyclic Chemistry, Yokohama, Japan, July 29–August 3
2001, abstract p. 86. Their preliminary report on the
synthesis of nakadomarin A: Fu¨rstner, A.; Guth, O.;
Rumbo, A.; Seidel, G. J. Am. Chem. Soc. 1999, 121,
11108–11113.
5. Fu¨rstner, A. Angew. Chem., Int. Ed. 2000, 39, 3012–3043.
6. (a) The starting material, methyl 4-oxo-3-piperidinecar-
boxylate, 1 is commercially available and can also be
easily prepared in large scale from simple reagents: Baty,
J. D.; Jones, G.; Moore, C. J. Chem. Soc. (C) 1967,
2645–2647; (b) All of the compounds shown in this paper
are racemic. However, an asymmetric allylation of 1 has
been reported: Shilay, V.; Alyev, A.; Grishina, G. Bull.
Soc. Chim. Belg. 1992, 101, 751–752.
14. Satoh, T.; Nanba, K.; Suzuki, S. Chem. Pharm. Bull.
1971, 19, 817–820. Hydrogenation of the unsaturated
ester moiety only gave ethyl 3-(2-furyl)propionate, a
debrominated product. Sodium borohydride reduction in
the presence of CoCl2 or CuCl2 also gave less satisfactory
results.
7. Brands, K. M. J.; DiMichele, L. M. Tetrahedron Lett.
1998, 39, 1677–1680.
8. Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457.

T. Nagata et al. / Tetrahedron Letters 42 (2001) 8345–8349
8349
15. Ishiyama, T.; Murata, M.; Miyaura, N. J. Org. Chem.
1995, 60, 7508–7510.
16. For 23: H NMR (300 MHz, CDCl3): l 1.20–1.60 (5H,
3.53 (2H, m), 4.66 (1H, brs), 4.92–5.08 (4H, m), 5.75–5.89
(2H, m), 5.81 (1H, s); 13C NMR (75 MHz, CDCl3): l
26.3, 26.8, 28.5 (overlapped peaks), 30.6, 32.1, 33.6, 36.1,
38.0, 45.5, 51.9, 58.6, 59.6, 60.4, 62.5, 79.5, 102.7, 114.5,
115.1, 130.9, 137.6, 138.8, 154.9, 155.7, 159.7; IR (neat,
cm−1) w 1697, 1394, 1173; HRMS (FAB) calcd for
C27H41N2O3 441.3117 (MH+); found 441.3121.
1
m), 1.53 (9H, s), 1.69–1.78 (1H, m), 1.99–2.09 (4H, m),
2.14–2.19 (1H, m), 2.17 (1H, ABq, J=11.9 Hz), 2.30–2.42
(4H, m), 2.45–2.55 (1H, m), 2.59 (1H, t, J=7.5 Hz), 2.68
(2H, t, J=7.1 Hz), 2.79 (1H, ABq, J=11.9 Hz), 3.47–