Biomimetic total synthesis of (+)-gelsemine

Article abstract of DOI:10.1002/anie.201201736

Challenging: (+)-gelsemine was synthesized from (R,R)-aziridine 1 in 25 steps with approximately 1% overall yield. A multistep, one-pot enol-oxonium cyclization cascade was used to construct, simultaneously, the E ring, F ring, C3 stereocenter, and C7 quaternary stereocenter. This synthesis using the enol-oxonium cyclization reaction as a key step to make the cage structure has demonstrated the proposed biosynthetic pathway of the gelsemine family. Copyright

Full text of DOI:10.1002/anie.201201736

Angewandte
Chemie
DOI: 10.1002/anie.201201736
Natural Products
Biomimetic Total Synthesis of (+)-Gelsemine**
Xuan Zhou, Tao Xiao, Yusuke Iwama, and Yong Qin*
(+)-Gelsemine, a major alkaloid component of Gelsemium
sempervirens, was first isolated in 1876.[1] Its intriguing
structure was elucidated in 1959 by NMR spectroscopy[2]
and X-ray crystallographic analysis.[3] To date, seven members
of the gelsemium alkaloid family have been characterized (1–
7).[4] These oxindole alkaloids have a hexacyclic architecture
reaction involves an intramolecular enol–oxonium cyclization
that assembles the cage structure in a highly efficient way.
Besides, a new biosynthetic pathway for the late stage of the
biosynthesis of gelsemine is proposed based on the achieved
enol–oxonium cyclization procedure.[7]
The retrosynthetic analysis of (+)-gelsemine is shown in
Scheme 1. We speculated that the cage structure of (+)-gel-
semine could be assembled from an intermediate such as 8 in
and seven carbon stereocenters, including two quaternary
stereocenters compacted into a small cage. The challenging
structure of gelsemine (1) has been the target of extensive
synthetic efforts around the world;[5] within the last two
decades a number of total syntheses were reported by the
research groups of Johnson, Speckamp, Fukuyama, Hart,
Overman, and Danishefsky.[6] Among these total syntheses,
there was only one asymmetric total synthesis accomplished
by the Fukuyama group.[6g] Furthermore, none of these total
syntheses has sufficiently illustrated the possible biosynthetic
pathway of the gelsemine family. Herein, we report a bio-
mimetic total synthesis of (+)-gelsemine, in which the key
Scheme 1. Retrosynthetic analysis of (+)-gelsemine. TBDMS=tert-
butyldimethylsilyl, MOM=methoxymethyl, Ts=toluene-4-sulfonyl.
a single step by acid-catalyzed enol–oxonium cyclization to
form, simultaneously, the E ring, the F ring, the C7 quaternary
stereocenter of spirooxindole, and the C3-stereocenter. We
reasoned that the D ring could be constructed by intra-
molecular Michael addition of highly functionalized nitrile 9,
which can be readily prepared by intermolecular aldol
addition of oxindole 10 to cis tetra-substituted piperidinoal-
dehyde 11, followed by dehydration. The 15S configuration in
11 could be generated by stereoselective formation of the
C ring by an intramolecular Michael addition of oxepinone
12. Partial hydrogenation of alkynyl acid 13 could yield the cis
double bond in 12. The S configuration at C16 in 13 could be
produced by regio- and stereoselective ring opening of
aziridine (R,R)-15 induced by the hydroxy group. The
stereochemistry in (R,R)-15 could be generated from the
known ester (R,R)-16,[8] which is easily prepared from d-
diethyl tartrate in a four-step procedure.
[*] X. Zhou, T. Xiao, Prof. Dr. Y. Qin
Key Laboratory of Drug Targeting and Drug Delivery Systems of the
Ministry of Education, and State Key Laboratory of Biotherapy
Sichuan University
Chengdu, 610041 (China)
E-mail: yongqin@scu.edu.cn
Y. Iwama
Graduate School of Phamaceutical Sciences, Tohoku University
Sendai, 980-8578 (Japan)
Prof. Dr. Y. Qin
Innovative Drug Research Centre and College of Chemistry and
Chemical Engineering, Chongqing University
Chongqing, 401331 (China)
[**] We are grateful for financial supports from NSFC (20825207,
21021001, and 21132006), the National Basic Research Program of
China (973 program, 2010CB833200), the State Key Laboratory of
Bioorganic and Natural Products Chemistry, Shanghai Institute of
Organic Chemistry, and Tohuku University Global COE Program and
JSPS Predoctoral Fellowship for Y.I.
Our biomimetic total synthesis of (+)-gelsemine began by
converting (R,R)-16 into the chiral cis tetra-substituted
piperidinoaldehyde 11 (Scheme 2). Two steps of reduction
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201201736.
Angew. Chem. Int. Ed. 2012, 51, 4909 –4912
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4909

.
Angewandte
Communications
addition at low temperature yielded two inseparable diaste-
reomers (15S,20R)-21a and (15S,20S)-21b in 66% yield in
a 7:1 ratio. The mixture of diastereomers 21a and 21b became
separable after the lactone group in 21 was reduced to an
aldehyde with DIBAL-H at low temperature, and the
resulting aldehyde group was protected with a dimethoxy
group, affording the major isomer (15S,20R)-22 in 48% yield
in two steps. Oxidation of the hydroxy group in 22 provided
the cis tetra-substituted piperidinoaldehyde 11 in 92% yield.
The cis relationship was confirmed by a strong NOE observed
between the two protons at carbon atoms C15 and C20.
After having synthesized the cis tetra-substituted piper-
idinoaldehyde 11 with the required stereochemistry, we began
to prepare compound 9, which is the precursor for forming the
D ring through a Michael addition (Scheme 3). Condensation
of 11 with N-methoxyoxindole 10 in the presence of LDA at
ꢀ788C, followed by dehydration of the resulting hydroxy
group with SOCl2/pyridine provided two separable geometric
isomers (Z)-9a and (E)-9b in a 1.5:1 ratio. The Z conforma-
tion of the double bond in 9a was confirmed by a strong NOE
observed between the olefin proton at C6 and the aromatic
proton at C9. As expected, the mixture of 9a and 9b easily
underwent Michael addition in the presence of LDA in THF
at ꢀ788C to yield two inseparable diastereomers 23a and 23b
epimerized at C7 in 75% yield. The Michael addition of 9 not
only formed the D ring, but also created the correct C20
quaternary carbon center with an R configuration. However,
the addition led to an S configuration at C6 rather than the
desired R configuration. The S configuration at C6 prevented
that the oxindole moiety gained access to the oxonium cation
that was formed in situ when the MOM group and the two
methoxy groups in 23 were removed under acidic conditions.
Indeed, initial attempts to form the E and F ring by enol–
oxonium cyclization generated both olefin 24 and pyranol 25
when the mixture of 23a and 23b was treated with either TFA
at room temperature or TsOH in CHCl3 at reflux.
Scheme 2. Preparation of the chiral cis tetra-substituted piperidine
ring. Reagents and conditions: a) LiAlH4, EtOH, 08C!258C, 3 h,
85%; b) TsCl, Et3N, CH2Cl2, 08C!258C, 3 h, 90%; c) (3,3-diethoxy-
prop-1-yn-1-yl)lithium 14, THF, 08C!258C, 5 h, 95%; d) MeI, K2CO3,
DMF, 258C, 98%; e) MOMCl, NaH, THF, 08C, 2 h, 95%; f) Mg,
MeOH, ultrasound, 08C, 3 h, 90%; g) acrylonitrile, MeOH, reflux, 10 h,
95%; h) TFA, CH2Cl2, ꢀ108C, 3 h, 85%; i) NaClO2, NaH2PO3, tBuOH/
H2O 1:1, 258C, 3 h, 94%; j) Lindlar cat, H2, EtOH, 258C, 10 h, 95%;
k) HF, pyridine, 08C!258C, 10 h, 90%; l) DCC, DMAP, CH2Cl2, 258C,
10 h, 85%; m) LiHMDS, THF, ꢀ788C, 10 min, 66%; n) DIBAL-H,
CH2Cl2, ꢀ788C, 30 min, 60%; o) HC(OMe)3, p-TsOH, MeOH, 08C!
258C, 2 h, 80%; p) (COCl)2, DMSO, CH2Cl2, ꢀ608C, 1 h, 92%.
TFA=trifluoroacetic acid, DCC=dicyclohexylcarbodiimide, DMAP=4-
(dimethylamino)pyridine, HMDS=1,1,1,3,3,3-hexamethyldisilazanide,
DIBAL-H=diisobutylaluminum hydride, DMSO=dimethylsulfoxide.
To invert the C6 S configuration in 23 to an R configura-
tion and thereby allow construction of the E and F rings by
enol–oxonium cyclization, a double bond between C6 and C7
was generated to yield two separable geometric isomers 26a
and 26b in 72% yield in a 5:1 ratio by using a standard
protocol of olefination with PhSeCl/LDA/NaIO4 (Scheme 3).
Hydrogenation of the double bond in 26 with Lindlar catalyst
at 1 atm of hydrogen pressure in MeOH within three hours
resulted in two inseparable diastereomers 27a and 27b in
90% yield in a 2:1 ratio. The hydrogenation proceeded such
that the nucleophilic attack occurred from the less hindered
bottom face of the double bond, thereby yielding 27a and 27b
with C6 exclusively in the R configuration. In contrast,
hydrogenation of 26 with the more active catalyst Pd/C
under the same conditions yielded two inseparable N-
demethoxy products 28a and 28b in a 3:1 ratio in quantitative
yield.
and tosylation of 16 provided alcohol (R,R)-15 in moderate
yield. The attack by 14 on 15 occurred from the top face of the
aziridine ring to afford 17 as a single stereomer in 95% yield.
The excellent regio- and stereoselectivity is probably due to
coordination of the lithium reagent 14 with the hydroxy group
and efficient blockage of the bottom face of the aziridine ring
by the bulky OTBDMS group in 15.[9] Two steps of
methylation with MeI and protection of the hydroxy group
in 17 provided compound 18 in excellent yield. After removal
of the tosyl group in 18 with Mg dust using ultrasound,[10] the
resulting amine reacted with acrylonitrile in MeOH at reflux
to give 19 in excellent yield. Treatment of 19 with TFA at
ꢀ108C, and subsequent oxidation of the resulting aldehyde
group with sodium chlorite yielded acid 13 in 85% yield in
two steps. Partial hydrogenation of the alkynyl group with
Lindlar catalyst and removal of the TBDMS group with HF/
pyridine in 13 provided cis olefin 20 in excellent yield. To
prepare the piperidine ring with an S configuration at C15 as
in the (+)-gelsemine structure, an oxepinone ring was first
generated by condensation of the acid group with the hydroxy
group in 20 to give 12 in 85% yield. Intramolecular Michael
With 27a and 27b in hand, we carried out the planned
one-pot, multi-step enol–oxonium cyclyzation reaction cas-
cade to efficiently assemble both the E and F rings, and
simultaneously establish the C3- and C7-stereocenters. When
the mixture of 27a and 27b was treated with stoichiometric
TsOH in CH3Cl at reflux for four hours, two separable
4910 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 4909 –4912

Angewandte
Chemie
29a with the inverse 7R configuration. Interestingly,
when the same reaction conditions were applied to
28a and 28b, a complex mixture resulted, rather
than the corresponding cyclization products 30a and
30b. These results indicate that, in the absence of
a methoxy substituent on the oxindole nitrogen
atom, an enol functionality does not efficiently form
during the reaction.[11]
After having successfully synthesized 29b, which
possesses all the ring systems and correct stereo-
chemistry of the carbon stereocenters of the target
(+)-gelsemine, we turned our effort to the final task
of converting the nitrile substituent at C20 in 29b to
an ethylene substituent (Scheme 3). After removal
of the methoxy group in 29b by hydrogenation with
5% Pd/C, the nitrile group in 30b was reduced to an
aldehyde group to provide 31 in 71% yield in two
steps. Initial efforts with olefination of 31 using the
=
Wittig reagent Ph3P CH2 failed under various
conditions, because the aldehyde group in 31 was
severely hindered. Fortunately, olefination of 31
with Tebbe reagent yielded (+)-gelsemine in 60%
yield.[12] The synthetic (+)-gelsemine showed 1H and
13C NMR spectra that were identical to those of
a previous report[6f] and the commercially available
(+)-natural gelsemine.[13]
In summary, the biomimetic total synthesis of
(+)-gelsemine from (R,R)-aziridine 16 has been
accomplished in 25 steps with approximately 1%
overall yield. A multi-step, one-pot enol–oxonium
cyclization cascade is used in a highly efficient way
to construct, simultaneously, the E ring, F ring, C3-
stereocenter, and C7 quaternary stereocenter. The
current synthesis using the enol–oxonium cyclization
reaction as a key step to make the cage structure has
demonstrated a possible biosynthetic pathway. We
also demonstrated that without a methoxy substitu-
tent on the oxindole nitrogen atom, the enol–
oxonium cyclization does not work (failure of
converting compound 28 to compound 30). This
result is meaningful for elucidating the function of
the methoxy group in the proposed biosynthetic
procedure, and to partially explain why most mem-
bers of the gelsemine family have a methoxy sub-
stitutent on the oxindole nitrogen atom.
Scheme 3. Synthesis of (+)-gelsemine. Reagents and conditions: a) 10, LDA, THF,
ꢀ788C, 12 h, 75%; b) SOCl2, pyridine, 08C, 5 min, 90%; c) LDA, THF, ꢀ788C,
30 min, 75%; d) condition a: TfOH, CHCl3, RT. 24 h or condition b: p-TsOH,
CHCl3, reflux, 10 h. e) LDA, PhSeCl, THF, ꢀ788C, 30 min; then NaIO4, NaHCO3,
MeOH/THF/H2O 5:1:1, 258C, 10 h, 70% in 2 steps; f) Lindlar cat., H2, MeOH,
258C, 3 h, 90% (gave 27a and 27b); g) 5% Pd/C, H2, MeOH, 258C, 10 h, 98%
(gave 28a and 28b); h) p-TsOH, CHCl3, reflux, 4 h, 73%, 29a/29b 1:10 ratio;
i) 5% Pd/C H2, MeOH, 258C, 10 h, 98%; j) DIBAL-H, Tol, ꢀ788C, 10 h, 72%;
k) Tebbe reagent, THF, pyridine, reflux, 10 h, 60%. LDA=lithium diisopropyl-
amide, Tf=trifluoromethanesulfonyl.
Received: March 3, 2012
Published online: April 4, 2012
Keywords: asymmetric synthesis · biomimetic synthesis ·
.
indole alkaloids · natural products · total synthesis
diastereomers 29a and 29b were obtained in 73% yield in
a 1:10 ratio. Apparently, the reaction proceeded via two
transition states 8a and 8b. The natural major isomer 29b
with the desired 7S configuration was produced from the
favored transition state 8b, in which an electronic repulsion
between the oxonium cation and the hydroxy group of the
enol functionality is avoided. In contrast, the disfavored
transition state 8a, in which both electron repulsion and
a steric effect are present, led to the unnatural minor isomer
[1] F. L. Sonnenschein, Ber. Dtsch. Chem. Ges. 1876, 9, 1182 – 1186.
[2] a) H. Conroy, J. K. Chakrabarti, Tetrahedron Lett. 1959, 1, 6 – 13;
b) Y. Schun, G. A. Cordell, J. Nat. Prod. 1985, 48, 969 – 971.
[3] F. M. Lovell, R. Pepinsky, A. J. C. Wilson, Tetrahedron Lett.
1959, 1, 1 – 5.
[4] a) Y. Schun, G. A. Cordell, M. Garland, J. Nat. Prod. 1986, 49,
483 – 487; b) Y. Schun, G. A. Cordell, J. Nat. Prod. 1986, 49, 806 –
Angew. Chem. Int. Ed. 2012, 51, 4909 –4912
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4911

.
Angewandte
Communications
808; c) D. Ponglux, S. Wongseripipatana, S. Subhadhirasakul, H.
Takayama, M. Yokota, K. Ogata, C. Phisalaphong, N. Aimi, S.-I.
Sakai, Tetrahedron 1988, 44, 5075 – 5094; d) L.-Z. Lin, Y. Schun,
G. A. Cordell, C.-Z. Ni, J. Clardy, Phytochemistry 1991, 30, 679 –
683.
Overman, M. J. Sharp, J. Am. Chem. Soc. 2005, 127, 18046 –
18053; ac) A. Madin, C. J. OꢀDonnell, T. Oh, D. W. Old, L. E.
Overman, M. J. Sharp, J. Am. Chem. Soc. 2005, 127, 18054 –
18065; ad) S. Grecian, J. Aube, Org. Lett. 2007, 9, 3153 – 3156;
ae) K. Tchabanenko, N. S. Simpkins, L. Male, Org. Lett. 2008, 10,
4747 – 4750.
[5] For model studies and synthesis of a fragment of gelsemine, see:
a) R. L. Autrey, F. C. Tahk, Tetrahedron 1967, 23, 901 – 917;
b) R. L. Autrey, F. C. Tahk, Tetrahedron 1968, 24, 3337 – 3345;
c) R. S. Johnson, T. O. Lovett, T. S. Stevens, J. Chem. Soc. C
1970, 6, 796 – 800; d) I. Fleming, J. P. Michael, J. Chem. Soc.
Perkin Trans. 1 1981, 1549 – 1556; e) I. Fleming, M. A. Loreto,
J. P. Michael, I. H. M. Wallace, Tetrahedron Lett. 1982, 23, 2053 –
2056; f) I. Fleming, M. A. Loreto, I. H. M. Wallace, J. P. Michael,
J. Chem. Soc. Perkin Trans. 1 1986, 349 – 359; g) G. Stork, M. E.
Krafft, S. A. Biller, Tetrahedron Lett. 1987, 28, 1035 – 1038;
h) R. J. Vijn, H. Hiemstra, J. J. Kok, M. Knotter, W. N. Speck-
amp, Tetrahedron 1987, 43, 5019 – 5030; i) M. M. Abelman, T.
Oh, L. E. Overman, J. Org. Chem. 1987, 52, 4130 – 4133; j) C.
Clarke, I. Fleming, J. M. D. Fortunak, P. T. Gallagher, M. C.
Honan, A. Mann, C. O. Nubling, P. R. Raithby, J. J. Wolff,
Tetrahedron 1988, 44, 3931 – 3934; k) H. Hiemstra, R. J. Vijn,
W. N. Speckamp, J. Org. Chem. 1988, 53, 3882 – 3884; l) W. G.
Earley, E. J. Jacobsen, G. P. Meier, T. Oh, L. E. Overman,
Tetrahedron Lett. 1988, 29, 3781 – 3784; m) W. G. Earley, T. Oh,
L. E. Overman, Tetrahedron Lett. 1988, 29, 3785 – 3788; n) J.-K.
Choi, D.-C. Ha, D. J. Hart, C.-S. Lee, S. Ramesh, S. Wu, J. Org.
Chem. 1989, 54, 279 – 290; o) I. Fleming, R. C. Moses, M. Tercel,
J. Ziv, J. Chem. Soc. Perkin Trans. 1 1991, 617 – 626; p) D. J. Hart,
S. C. Wu, Tetrahedron Lett. 1991, 32, 4099 – 4102; q) W.-J. Koot,
H. Hiemstra, W. N. Speckamp, J. Org. Chem. 1992, 57, 1059 –
1061; r) A. Madin, L. E. Overman, Tetrahedron Lett. 1992, 33,
4859 – 4862; s) L. E. Overman, M. J. Sharp, J. Org. Chem. 1992,
57, 1035 – 1038; t) D. J. Hart, S. C. Wu, Heterocycles 1993, 35,
135 – 138; u) H. Takayama, N. Seki, M. Kitajima, N. Aimi, S.-I.
Sakai, Nat. Prod. Lett. 1993, 2, 271 – 276; v) A. P. Johnson,
R. W. A. Luke, R. W. Steele, A. N. Boa, J. Chem. Soc. Perkin
Trans. 1 1996, 883 – 893; w) F. Ng, P. Chiu, S. J. Danishefsky,
Tetrahedron Lett. 1998, 39, 767 – 770; x) M. J. Sung, C.-W. Lee,
J. K. Cha, Synlett 1999, 561 – 562; y) A. G. Avent, P. W. Byrne,
C. S. Penkett, Org. Lett. 1999, 1, 2073 – 2075; z) J. Dijkink, J.-C.
Cintrat, W. N. Speckamp, H. Hiemstra, Tetrahedron Lett. 1999,
40, 5919 – 5922; aa) A. J. Pearson, X. Wang, J. Am. Chem. Soc.
2003, 125, 13326 – 13327; ab) W. G. Earley, J. E. Jacobsen, A.
Madin, G. P. Meier, C. J. OꢀDonnell, T. Oh, D. W. Old, L. E.
[6] For a review on the total synthesis of gelsemine, see: H. Lin, S. J.
Danishefsky, Angew. Chem. 2003, 115, 38 – 53; Angew. Chem.
Int. Ed. 2003, 42, 36 – 51. For individual total syntheses of
racemic gelsemine, see: a) Z. Sheikh, R. Stell, A. S. Tasker, A. P.
Johnson, J. Chem. Soc. Chem. Commun. 1994, 763 – 764; b) J. K.
Dutton, R. W. Steel, A. S. Tasker, V. Popsavin, A. P. Johnson, J.
Chem. Soc. Chem. Commun. 1994, 765 – 766; c) N. J. Newcombe,
F. Ya, R. J. Vijn, H. Hiemstra, W. N. Speckamp, J. Chem. Soc.
Chem. Commun. 1994, 767 – 768; d) S. Atarashi, J. K. Choi, D. C.
Ha, D. J. Hart, D. Kuzmich, C.-S. Lee, S. Ramesh, S. C. Wu, J.
Am. Chem. Soc. 1997, 119, 6226 – 6241; e) T. Fukuyama, G. Liu,
J. Am. Chem. Soc. 1996, 118, 7426 – 7427; f) A. Madin, C. J.
OꢀDonnell, T. Oh, D. W. Old, L. E. Overman, M. J. Sharpe,
Angew. Chem. 1999, 111, 3110 – 3112; Angew. Chem. Int. Ed.
1999, 38, 2934 – 2936; g) S. Yokoshima, H. Tokuyama, T.
Fukuyama, Angew. Chem. 2000, 112, 4239 – 4241; Angew.
Chem. Int. Ed. 2000, 39, 4073 – 4075; h) F. W. Ng, H. Lin, S. J.
Danishefsky, J. Am. Chem. Soc. 2002, 124, 9812 – 9824.
[7] Previously proposed biosynthetic pathway, see: a) H. Conroy,
J. K. Chakarabarti, Tetrahedron Lett. 1959, 1, 6 – 13; b) J. Stçck-
igt, Tetrahedron Lett. 1979, 20, 2615 – 2618; c) P. Dhavadee, W.
Sumphan, S. Sanan, T. Hiromitsu, Y. Masaki, O. Koreharu, P.
Chada, A. Norio, S. Shinichiro, Tetrahedron 1988, 44, 5075 –
5094; d) Z.-J. Liu, R.-R. Liu in The Alkaloids, Vol. 33 (Ed.:
R. H. F. Manske), Academic Press, New York, 1988, pp. 83 – 140.
[8] R. Hili, A. K. Yudin, J. Am. Chem. Soc. 2006, 128, 14772 – 14773.
[9] B. Nyasse, L. Grehnb, U. Ragnarsson, Chem. Commun. 1997,
1017 – 1018.
[10] K. Fuji, T. Kawabata, Y. Kriyu, Y. Sugiura, Heterocycles 1996, 42,
701 – 722.
[11] A similar enol functionality of N-methoxy oxoindole was
reported in the synthesis of racemic 7-epi-20-desethylgelsedine,
see: A. S. Kende, M. J. Luzzio, J. S. Mendoza, J. Org. Chem.
1990, 55, 918 – 924.
[12] F. N. Tebbe, G. W. Parshall, G. S. Reddy, J. Am. Chem. Soc. 1978,
100, 3611 – 3613.
[13] The standard sample of natural (+)-gelsemine was purchased
from Sichuan Weikeqi BioTech Co., Ltd (China).
4912
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 4909 –4912