Total synthesis of (-)-conophylline and (-)-conophyllidine

Article abstract of DOI:10.1002/anie.201100981

Double take: The total syntheses of the title compounds were accomplished in a highly convergent manner. The approach features the regio- and diastereoselective Polonovski-Potier-type reaction for the coupling of two aspidosperma skeletons and the formation of the dihydrofuran ring. Troc=2,2,2-trichloroethoxycarbonyl. Copyright

Full text of DOI:10.1002/anie.201100981

Communications
DOI: 10.1002/anie.201100981
Natural Products Synthesis
Total Synthesis of (ꢀ)-Conophylline and (ꢀ)-Conophyllidine**
Yuki Han-ya, Hidetoshi Tokuyama, and Tohru Fukuyama*
(ꢀ)-Conophylline (1), isolated from Tabernaemontana divar-
icata in 1992,[1] is a bis(indole) alkaloid consisting of two
pentacyclic aspidosperma skeletons (Figure 1). This com-
Scheme 1. Synthetic strategy for the central part of (ꢀ)-conophylline
(1).
represents our synthetic strategy for the construction of the
dihydrofuran moiety. We envisioned that if we could generate
Figure 1. Structures of (ꢀ)-conophylline (1) and (ꢀ)-conophyllidine
(2).
an iminium ion regioselectively in the lower-half (segment 5),
the electron-rich aromatic ring of the upper-half (segment 4)
would attack the iminium ion to form the C3–C10’ linkage
with a subsequent intramolecular epoxide opening to furnish
the dihydrofuran structure with the requisite stereochemistry.
Husson, Lounasmaa, and co-workers reported their seminal
work on the regiochemical issue of the Polonovski–Potier
reaction[5] in which vincadifformine (6) gave the 5a-cyano
compound 8 via the iminium ion at C5–N4, whereas the
reaction of epipandoline (9) bearing a hydroxy group at C14
provided the 3-cyano compound 11 via an iminium ion at C3–
N4 (Scheme 2).[6] On the basis of these observations as well as
the potential cyclopropane-like role of the epoxide in
stabilizing the adjacent carbocation, we hoped that a substrate
having an epoxide at C14–C15 might also generate the C3–N4
pound acts as a potent inhibitor of the ras function[2] and,
furthermore, has been found to induce beta-cell differentia-
tion in rat pancreatic acinar carcinoma cells and in cultured
fetal rat pancreatic tissue.[3] Thus, this compound has the
potential to be a lead compound in the development of novel
drugs for cancer chemotherapy as well as for regeneration
therapy used for the treatment of diabetes mellitus type 1. As
a result of the important biological activity as well as the
complex structure possessing a unique connectivity between
its two segments, conophylline has attracted considerable
attention from the synthetic community as a challenging
target.[4] Herein, we report the first total synthesis of (ꢀ)-
conophylline (1) and its congener, (ꢀ)-conophyllidine (2),[1b]
in a convergent route utilizing the Polonovski–Potier-type
coupling reaction.
A convergent and potentially biomimetic synthesis of 1
should consist of the construction of the central dihydroben-
zofuran ring at the final stage of the synthesis. Scheme 1
[*] Dr. Y. Han-ya, Prof. Dr. H. Tokuyama, Prof. Dr. T. Fukuyama
Graduate School of Pharmaceutical Sciences
University of Tokyo
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033 (Japan)
Fax: (+81)3-5841-4777
E-mail: fukuyama@mol.f.u-tokyo.ac.jp
Prof. Dr. H. Tokuyama
Graduate School of Pharmaceutical Sciences
Tohoku University, Sendai 985-8578 (Japan)
[**] This work was financially supported by KAKENHI and JSPS
(predoctoral fellowship for Y. H.), and the “Funding Program for
Next Generation World-Leading Researchers”, Cabinet Office,
Government of Japan (for H.T.). We thank Prof. K. Umezawa for
providing the natural sample of conophylline.
Scheme 2. Regiochemistry of Polonovski–Potier reaction reported by
Husson, Lounasmaa, and co-workers.[6] Reagents and conditions:
a) p-O2N-C6H4CO3H, CHCl3, 08C; b) TFAA, CH2Cl2, 08C!RT; c) KCN,
CH2Cl2/H2O (4:1), TFA/CH3CO2K, pH 4, RT, 50% (3 steps); d) aq
H2O2, EtOH/CH2Cl2 (1:1), 608C; e) TFAA, CH2Cl2, 08C!RT; f) KCN,
CH2Cl2/H2O (4:1), TFA/CH3CO2K, pH 4, RT, 30% (3 steps). TFA=tri-
fluoroacetic acid, TFAA=trifluoroacetic anhydride.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201100981.
4884
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4884 –4887

iminium ion 5. For the synthesis of both the upper and lower
aspidosperma indole segments, we planned to modify our
previously established route to (ꢀ)-tabersonine[7] utilizing a
combination of the tin-mediated indole synthesis[8] and the
biomimetic cascade reaction for the formation of an aspido-
sperma skeleton.[7,9]
The synthesis of indole 12, the intermediate to the lower
segment, commenced with the nitration of the commercially
available phenol 13 and the silylation of the sterically less-
hindered phenolic hydroxy group to give 14, which was then
substituent at the 2 position by Stille coupling[11] with the 2-
stannylacrylate derivative 20[12] and removal of the THP
group furnished the indole intermediate 12.
We then constructed the aspidosperma skeleton by the
intramolecular Michael addition/Mannich reaction cascade.
Dinitrobenzenesulfonamide 21, which was prepared by our
previously reported route,[9] was coupled with alcohol 12 by
using
a
Mitsunobu protocol (Scheme 4).[13] Next, both
converted into mesylate 15 by
a three-step sequence
(Scheme 3). The ester was then elongated by reduction to
the aldehyde and subsequent Wittig reaction to give cinna-
mate 16. Next, the nitro group was transformed into an
isocyano group by a conventional three-step sequence. The
isocyanide 17 thus obtained was subjected to the tin-mediated
radical cyclization[8] to produce 2-stannylindole, which was
then converted into 2-iodoindole 18 by in situ treatment with
iodine.[10] Then, reduction of the ester, protection of the
resultant alcohol as THP ether, and Boc protection of the
indole nitrogen atom gave 19. Finally, introduction of the
Scheme 4. Synthesis of the lower-segment 27. Reagents and condi-
tions: a) PPh3, DEAD, benzene, 0C8!RT, 76%; b) TFA, Me2S, CH2Cl2,
RT; c) pyrrolidine, MeOH/CH3CN (5:1), 0!608C, 65% (2 steps);
d) PPh3, CCl4, 2-methyl-2-butene, CH3CN, 608C, 35%; e) tBuOK,
TrocCl, DMAP, THF, 08C; f) mCPBA, aq HClO4, MeOH, 0!508C,
42% (2 steps); g) LDA, THF, ꢀ78!08C, 60%. DEAD=diethyl azodi-
carboxylate, DNs=2,4-dinitrobenzenesulfonyl, LDA=lithium
diisopropylamide, mCPBA=meta-chloroperbenzoic acid, Troc=2,2,2-
trichloroethoxycarbonyl.
Scheme 3. Synthesis of the indole segment 12. Reagents and condi-
tions: a) HNO3, AcOH, RT; b) TBDPSCl, 2,6-lutidine, THF/DMF (4:1),
08C!RT, 42% (2 steps); c) tBuOK, Me2SO4, THF/DMF (3:1), 08C!
RT, 96%; d) TBAF, THF, RT; e) MsCl, Et3N, CH2Cl2, 08C!RT, 60%
(2 steps); f) DIBAL-H, CH2Cl2, ꢀ788C; g) TPAP, NMO, 4ꢀ M.S.,
removal of the Boc group and hydration of the enol ether
were effected by treatment with TFA to give lactol 23. After
removal of the DNs group,[14] the reaction mixture was stirred
at 508C to promote sequential reactions involving cyclic
enamine formation, Michael addition of the enamine to a,b-
unsaturated ester, and Mannich reaction of the indole to the
resultant iminium ion to furnish the desired 25 as the sole
isomer. Finally, regioselective dehydration and stereoselec-
tive epoxidation in the presence of perchloric acid[15]
furnished the desired lower segment 27. The structure of 27
was unambiguously confirmed by transformation into (ꢀ)-
taberhanine (28)[16] in a one-pot removal of both the mesyl
and Troc groups under conditions reported by Carreira and
co-workers[17] in which excess LDA was used.
=
CH2Cl2, RT; h) Ph3P CHCO2Et, toluene, RT, 61% (3 steps); i) Zn,
AcOH, CH2Cl2, 08C!RT; j) HCO2H, Ac2O, CH2Cl2, 08C; k) POCl3, Py,
CH2Cl2, 08C, 64% (3 steps); l) nBu3SnH, AIBN, CH3CN, reflux; I2, RT
81% (2 steps); m) DIBAL-H, CH2Cl2, ꢀ78!08C; n) DHP, CSA,
CH2Cl2, RT; o) Boc2O, DMAP, CH3CN, 08C!RT, 94% (3 steps);
p) [BnPd(PPh3)2Cl], CuI, (2-furyl)3P, methyl 2-(tributylstannyl)acrylate
(20), DMF/HMPA (2:1), 808C, 63%; q) CSA, MeOH, RT, 98%.
AIBN=azobisisobutyronitrile, Boc=tert-butoxycarbonyl, CSA=10-
camphorsulfonic acid, DHP=3,4-dihydro-2H-pyrane, DIBAL-H=diiso-
butylaluminum hydride, DMAP=N,N-dimethyl-4-aminopyridine,
DMF=N,N-dimethylformamide HMPA=hexamethylphosphoric tri-
amide, Ms=methanesulfonyl, M.S.=molecular sieves, NMO=N-
methylmorpholine-N-oxide, TBDPS=tert-butyldiphenylsilyl, TBAF=
tetra-n-butylammonium fluoride, THF=tetrahydrofuran,
TPAP=tetrapropylammonium perruthenate.
Angew. Chem. Int. Ed. 2011, 50, 4884 –4887
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4885

Communications
With the lower segment of (ꢀ)-conophylline (1) in hand,
we then synthesized the upper segment 29 from 7-mesyloxy-
tabersonine 31, which was synthesized from indole derivative
30 and sulfonamide 21 in a similar manner as described in
Scheme 4. After protection of enamine 31 with a Troc group,
stereoselective epoxidation was carried out by treatment with
mCPBA (Scheme 5).[15] Finally, removal of the Troc group
and switching of the mesyl group to an allyl group afforded
the upper segment 29.
Scheme 5. Synthesis of the upper-segment 29. Reagents and condi-
tions: a) NaH, TrocCl, DMAP, THF/DMF (3:1), 08C; b) mCPBA, aq
HClO4, MeOH, 0!508C, 80% (2 steps); c) Zn, aq KH2PO4, THF,
608C; d) 1m KOH, MeOH, 508C; e) AllylBr, K2CO3, DMF, 608C, 82%
(3 steps).
Scheme 6. Completion of the total synthesis of (ꢀ)-conophylline (1).
Reagents and conditions: a) mCPBA, CH2Cl2, 08C; b) TFAA, CH2Cl2, 08C!
RT, 50% (2 steps); c) [Pd(PPh3)4], pyrrolidine, CH2Cl2, RT, 76%; d) LDA,
THF, ꢀ78!08C, 72%.
Having synthesized the requisite upper and lower seg-
ments, we then turned our attention to the coupling reaction
of the two segments (Scheme 6). For the crucial generation of
the iminium ion, we examined the Polonovski–Potier-type
reaction.[5] Thus, mCPBA oxidation of the lower-segment 27
provided the N-oxide 33 as a mixture of diastereomers, which
was then treated with TFAA in the presence of the upper-
segment 29. Fortunately, the desired coupling product 34 was
obtained as a single isomer in 50% yield based on 29. Upon
palladium-mediated removal of the allyl group, a spontaneous
ring closure proceeded to furnish the dihydrofuran ring. This
result unambiguously indicated that regioselective elimina-
tion of the O-acyl N-oxide proceeded to generate the iminium
ion at C3–N4 and the subsequent nucleophilic attack of the
upper-segment 29 occurred from the a face to give 34. Finally,
the Ms group on the phenolic hydroxy group and the Troc
group were successfully removed using excess LDA[17] to
furnish (ꢀ)-conophylline (1). All of the data regarding the
synthetic conophylline were identical to those reported for
the natural compound.[1]
Synthetic utility of the regio- and diastereoselective
Polonovski–Potier-type reaction and the one-pot deallyl-
ation/cyclization cascade established for the synthesis of
(ꢀ)-conophylline (1) proved to be applicable to the synthesis
of conophyllidine (2; Scheme 7). Thus, the crude N-oxide 33
was subjected to the coupling reaction with the upper-
segment 36, which was prepared from 31 by switching the
protective group from an Ms group to an allyl group, giving
the desired coupling product 37 in 55% yield. Finally, the
endgame sequence involving formation of the dihydrofuran
ring initiated by the palladium-mediated deallylation and
removal of both the Ms and Troc groups under Carreiraꢀs
conditions[17] provided (ꢀ)-conophyllidine (2).[1b]
Scheme 7. Completion of the total synthesis of (ꢀ)-conophyllidine (2).
Reagents and conditions: a) 1m KOH, MeOH, 508C; b) AllylBr, K2CO3,
DMF, 608C, 82% (2 steps); c) mCPBA, CH2Cl2, 08C; d) TFAA, CH2Cl2,
08C!RT, 55% (2 steps); e) [Pd(PPh3)4], pyrrolidine, CH2Cl2, RT, 72%;
d) LDA, THF, ꢀ78!08C, 67%.
In conclusion, efficient total syntheses of (ꢀ)-conophyl-
line (1) and (ꢀ)-conophyllidine (2) have been accomplished
by using a newly developed coupling strategy of two
aspidosperma indole segments through the regio- and diaste-
4886 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4884 –4887

[5] A. Ahond, A. Cavꢁ, C. Kan-Fan, H.-P. Husson, J. de Rostolan, P.
Potier, J. Am. Chem. Soc. 1968, 90, 5622.
reoselective Polonovski–Potier reaction and the dihydrofuran
ring formation.
[6] A. Henriques, C. Kan, A. Chiaroni, C. Riche, H.-P. Husson, S.-K.
Kan, M. Lounasmaa, J. Org. Chem. 1982, 47, 803.
[7] a) S. Kobayashi, T. Ueda, T. Fukuyama, Synlett 2000, 883; b) S.
Kobayashi, G. Peng, T. Fukuyama, Tetrahedron Lett. 1999, 40,
1519.
[8] a) T. Fukuyama, X. Chen, G. Peng, J. Am. Chem. Soc. 1994, 116,
3127; b) H. Tokuyama, T. Fukuyama, Chem. Rec. 2002, 2, 37.
[9] S. Yokoshima, T. Ueda, S. Kobayashi, A. Sato, T. Kuboyama, H.
Tokuyama, T. Fukuyama, J. Am. Chem. Soc. 2002, 124, 2137.
[10] H. Tokuyama, Y. Kaburagi, X. Chen, T. Fukuyama, Synthesis
2000, 429.
Received: February 8, 2011
Published online: April 15, 2011
Keywords: alkaloids · cascade reactions · heterocycles ·
.
natural products · total synthesis
[1] a) T.-S. Kam, K.-Y. Loh, L.-H. Lim, W.-L. Loong, C.-H. Chuah,
C. Wei, Tetrahedron Lett. 1992, 33, 969; b) T.-S. Kam, K.-Y. Loh,
C. Wei, J. Nat. Prod. 1993, 56, 1865.
[11] J. K. Stille, B. L. Groh, J. Am. Chem. Soc. 1987, 109, 813.
[12] H. X. Zhang, F. Guibꢁ, G. Balavoine, J. Org. Chem. 1990, 55,
1857.
[2] a) K. Umezawa, T. Ohse, T. Yamamoto, T. Koyano, Y. Takaha-
shi, Anticancer Res. 1994, 14, 2413; b) N. Amino, T. Ohse, T.
Koyano, K. Umezawa, Anticancer Res. 1996, 16, 55.
[13] O. Mitsunobu, Synthesis 1981, 1.
[3] a) K. Umezawa, A. Hiroki, M. Kawakami, H. Naka, I. Takei, T.
Ogata, I. Kojima, T. Koyano, T. Kowithayakorn, H.-S. Pang, T.-S.
Kam, Biomed. Pharmacother. 2003, 57, 341; b) T. Ogata, L. Li, S.
Yamada, Y. Yamamoto, Y. Tanaka, I. Takei, K. Umezawa, I.
Kojima, Diabetes 2004, 53, 2596; M. Fujii, I. Takei, K. Umezawa,
Biomed. Pharmacother. 2009, 63, 710; M. Kawakami, A.
Hirayama, K. Tsuchiya, H. Ohgawara, M. Nakamura, K.
Umezawa, Biomed. Pharmacother. 2010, 64, 226.
[14] T. Fukuyama, M. Cheung, C.-K. Jow, Y. Hidai, T. Kan,
Tetrahedron Lett. 1997, 38, 5831.
[15] J. ꢂles, G. Kalaus, I. Greiner, M. Kajtꢃr-Peredy, P. Szabꢄ, G. M.
Keserꢅ, L. Szabꢄ, C. Szꢃntay, J. Org. Chem. 2002, 67, 7255.
[16] T.-S. Kam, H.-S. Pang, T.-M. Lim, Org. Biomol. Chem. 2003, 1,
1292.
[17] T. Ritter, K. Stanek, I. Larrosa, E. M. Carreira, Org. Lett. 2004, 6,
1513.
[4] S. Ando, Y. Okamoto, K. Umezawa, M. Otsuka, J. Heterocycl.
Chem. 2008, 45, 1803.
Angew. Chem. Int. Ed. 2011, 50, 4884 –4887
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 4887