Total synthesis of the putative structure of stemonidine: The definitive proof of misassignment

Article abstract of DOI:10.1021/ol070486p

The total synthesis of the putative structure of the Stemona alkaloid stemonidine has been completed. The key transformations include a 1,3-dipolar cycloaddition of a chiral nitrone derived from (S)-prolinol and a spirolactonization process involving the

Full text of DOI:10.1021/ol070486p

ORGANIC
LETTERS
2007
Vol. 9, No. 9
1769-1772
Total Synthesis of the Putative Structure
of Stemonidine: The Definitive Proof of
Misassignment
Francisco Sa´nchez-Izquierdo, Pilar Blanco, Fe´lix Busque´, Ramo´n Alibe´s,
Pedro de March, Marta Figueredo,* Josep Font, and Teodor Parella†
UniVersitat Auto`noma de Barcelona, Departament de Qu´ımica,
08193 Bellaterra, Spain
marta.figueredo@uab.es
Received February 26, 2007
ABSTRACT
The total synthesis of the putative structure of the Stemona alkaloid stemonidine has been completed. The key transformations include a
1,3-dipolar cycloaddition of a chiral nitrone derived from (S)-prolinol and a spirolactonization process involving the generation of the critical
stereocenter. The NMR data of the synthetic material do not match those reported for the natural product. It is concluded that the structure
assigned to stemonidine is incorrect, and it must be reassigned as stemospironine.
The extracts of several plants of the Stemonaceae family have
long been used in Chinese and Japanese traditional medicine
for the treatment of respiratory disorders, as antihelmintics
and also as insecticides.1 Significant constituents of these
extracts are a series of structurally related alkaloids, which
may be responsible for their medicinal properties. Nowadays,
around 90 Stemona alkaloids are known, whose structures
were elucidated by X-ray analysis, spectroscopic techniques,
and/or chemical correlation, but there is a continuous flow
in the literature of new reports describing the isolation of
previously unknown members of the family. All the Stemona
alkaloids are polycyclic, and most of them present a central
pyrrolo[1,2-a]azepine system as a common characteristic
structural feature. The majority also incorporates at least one
substructure of R-methyl-γ-butyrolactone, which can be
linked to the azabicyclic core in a spiro or fused manner or
as a substituent. Considering their structural diversity, Pilli
and co-workers have recently classified the Stemona alkaloids
in eight groups,1a whereas Greger has suggested a different
classification in only three groups taking into account their
biosynthetic connections.1b One of these groups, the
tuberostemospironine1a or croomine1b type, concurs in both
classifications and is characterized by the inclusion of a spiro-
γ-lactone at C9 of the basic azabicyclic nucleus (Figure 1).
† Servei de Ressona`ncia Magne`tica Nuclear, Universitat Auto`noma de
Barcelona. 08193-Bellaterra, Spain.
(1) For recent reviews, see: (a) Pilli, R. A.; Rosso, G. B.; de Oliveira,
M. C. F. In The Alkaloids; Cordell, G. A., Ed.; Elsevier: New York, 2005;
Vol. 62, pp 77-173. (b) Greger, H. Planta Med. 2006, 72, 99.
Figure 1. Some alkaloids of the tuberostemospironine group.
10.1021/ol070486p CCC: $37.00
© 2007 American Chemical Society
Published on Web 03/31/2007

Scheme 1. Retrosynthetic Analysis for Some Stemona Alkaloids
In 1982, Xu and co-workers assigned to stemonidine, an
flexible approach, with some intermediates being common
precursors of various alkaloids (Scheme 1).13A main advan-
tage of this methodology is the high antifacial selectivity
accomplished in the 1,3-dipolar cycloadditions of nitrones
such as 6 to electron-deficient olefins of type 5, delivering
isoxazolidine adducts 4 with relative trans configuration of
the stereogenic centers at C3 and C9a, as required for the
target alkaloids.13b,14
Previously, we have reported the one-step preparation of
(S)-5-hydroxymethyl-1-pyrroline N-oxide, 7, by treatment of
L-prolinol, 8, with dimethyldioxirane in acetone at low
temperature and its isolation in 32% yield.14a Although this
straightforward methodology competed favorably with other
preparations of related nitrones in terms of brevity and yield,
difficulties associated with the purification of 7 and scaling-
up of the procedure led us to temporarily abandon its use
and explore different alternatives. However, recent reports
disclosing a new protocol for in situ dioxirane reactions15
led us to reinvestigate the possibility of preparing the required
nitrone by direct oxidation of a prolinol derivative. Hence,
the TBDPS derivative of L-prolinol 9 was prepared16 and
treated with Oxone under different reaction conditions
(Scheme 2). As the best result, it was found that 2.1 equiv
of the oxidant in THF-CH3CN (1:4) in the presence of
EDTA and NaHCO3 at 0 °C furnished a chromatographically
separable mixture of nitrones 6 and 10 in 1.3:1 ratio and
90% total yield.
alkaloid isolated from the roots of Stemona tuberosa, the
1
structure depicted as 1, on the basis of its H NMR data.2
Later on, the same group revised the former stereochemical
assignment and proposed the new structure that is depicted
as 2.3 More recently, Williams and co-workers completed
the first synthesis of (-)-stemospironine and found that its
spectral and physical data matched those reported for the
natural material, whose structure had been unequivocally
established by X-ray analysis.4 They also found that the 13
C
NMR spectra of synthetic stemospironine and natural ste-
monidine were virtually identical; the authors did not exclude
the possibility of the two compounds being spirocyclic
diastereomers. Herein, we describe the total synthesis and
NMR data of the putative structure of stemonidine 2, which
are definitive proof of the incorrect original assignments of
the natural product.
The challenging molecular architecture of the Stemona
alkaloids has attracted considerable interest among synthetic
organic chemists, and several total syntheses have been
published, although they are limited to a quite small number
of targets.4-12 We designed a strategy in which the azabi-
cyclic core was generated at an early stage of the sequence
and the R-methyl-γ-butyrolactone and other specific frag-
ments were then incorporated, with the aim of developing a
(2) Xu, R.; Lu, Y.; Chu, J.; Iwashita, T.; Naoki, H.; Naya, Y.; Nakanishi,
K. Tetrahedron 1982, 38, 2667.
(3) He, X.; Lin, W.; Xu, R. Huaxue Xuebao 1990, 48, 694.
(4) (-)-Stemospironine: Williams, D. R.; Fromhold, M. G.; Early, J.
D. Org. Lett. 2001, 3, 2721.
According to the plan, the first step in the synthesis of
stemonidine was the 1,3-dipolar cycloaddition between
(5) (+)-Croomine: (a) Williams, D. R.; Brown, D. L.; Benbow, J. W.
J. Am. Chem. Soc. 1989, 111, 1923. (b) Martin, S. F.; Barr, K. J. J. Am.
Chem. Soc. 1996, 118, 3299.
(6) (()-Stenine: (a) Chen, C.-Y.; Hart, D. J. J. Org. Chem. 1990, 55,
6236. (b) Ginn, J. D.; Padwa, A. Org. Lett. 2002, 4, 1515. (c) Golden, J.
E.; Aube´, J. Angew. Chem., Int. Ed. 2002, 41, 4316. (-)-Stenine: (d) Wipf,
P.; Kim, Y.; Goldstein, D. M. J. Am. Chem. Soc. 1995, 117, 11106. (e)
Morimoto, Y.; Iwahashi, M.; Nishida, K.; Hayashi, Y.; Shirahama, H.
Angew. Chem., Int. Ed. Engl. 1996, 35, 904.
(9) (()-Stemonamide and (()-isostemonamide: Kende, A. S.; Hernando,
J. I. M.; Milbank, J. B. J. Org. Lett. 2001, 3, 2505.
(10) (-)-Tuberostemonine: Wipf, P.; Rector, S. R.; Takahashi, H. J.
Am. Chem. Soc. 2002, 124, 14848.
(11) (-)-Stemonine: Williams, D. R.; Shamim, K.; Reddy, J. P.; Amato,
G. S.; Shaw, S. M. Org. Lett. 2003, 5, 3361.
(12) (()-Didehydrostemofoline and (()-isodidehydrostemofoline: Bru¨gge-
mann, M.; McDonald, A. I.; Overman, L. E.; Rosen, M. D.; Schwink, L.;
Scott, J. P. J. Am. Chem. Soc. 2003, 125, 15284.
(7) (()-Stemoamide: (a) Kohno, Y.; Narasaka, K. Bull. Chem. Soc. Jpn.
1996, 69, 2063. (b) Jacobi, P. A.; Lee, K. J. Am. Chem. Soc. 1997, 119,
3409. (-)-Stemoamide: (c) Williams, D. R.; Reddy, J. P.; Amato, G. S.
Tetrahedron Lett. 1994, 35, 6417. (d) Kinoshita, A.; Mori, M. J. Org. Chem.
1996, 61, 8356. (e) Jacobi, P. A.; Lee, K. J. Am. Chem. Soc. 2000, 122,
4295. (f) Gurjar, M. K.; Reddy, D. S. Tetrahedron Lett. 2002, 43, 295. (g)
Sibi, M. P.; Subramanian, T. Synlett 2004, 1211. (h) Olivo, H. F.; Tovar-
Miranda, R.; Barraga´n, E. J. Org. Chem. 2006, 71, 3287. (i) Bogliotti, N.;
Dalko, P. I.; Cossy, J. J. Org. Chem. 2006, 71, 9528.
(13) (a) Cid, P.; Closa, M.; de March, P.; Figueredo, M.; Font, J.;
Sanfeliu, E.; Soria, A. Eur. J. Org. Chem. 2004, 4215. (b) Alibe´s, R.; Blanco,
P.; Casas, E.; Closa, M.; de March, P.; Figueredo, M.; Font, J.; Sanfeliu,
E.; AÄ varez-Larena, A. J. Org. Chem. 2005, 70, 3157.
(14) (a) Closa, M.; de March, P.; Figueredo, M.; Font, J. Tetrahedron:
Asymmetry 1997, 8, 1031. (b) Busque´, F.; de March, P.; Figueredo, M.;
Font, J.; Gallagher, T.; Mila´n, S. Tetrahedron: Asymmetry 2002, 13, 437.
(15) (a) Evarts, J. B.; Fuchs, P. L. Tetrahedron Lett. 2001, 42, 3673. (b)
Li, W.; Fuchs, P. L. Org. Lett. 2003, 5, 2853.
(8) (()-Isostemofoline: Kende, A. S.; Smalley, T. L., Jr.; Huang, H. J.
Am. Chem. Soc. 1999, 121, 7431.
(16) Gerasyuto, A. I.; Hsung, R. P.; Sydorenko, N.; Slafer, B. J. Org.
Chem. 2005, 70, 4248.
1770
Org. Lett., Vol. 9, No. 9, 2007

Scheme 2. Preparation of Nitrone 6 from L-Prolinol
Scheme 4. Synthetic Plan from 3 to the Target Alkaloids
nitrone 6 and diester 5. This reaction was performed in
toluene at reflux for 10 h and delivered the endo isoxazo-
lidine 4 as the major product as expected,17 along with a
minor quantity of the exo isomer 11 (Scheme 3). Hydro-
the spirolactone with the appropriate configuration at C9.
Conversely, spirolactonization on ketone 14, available from
13, should presumably proceed to give the opposite config-
uration at the spiro stereocenter, as required for stemonidine.
Dehydration of 3 was accomplished in 88% yield under
Mitsunobu conditions and, as anticipated, the dihydroxylation
of the R,â-unsaturated ester 12 occurred with complete facial
selectivity in 92% yield (Scheme 5). From the crucial
intermediate 13, we pursued the synthesis of stemonidine to
shed light on its uncertain stereochemical assignment.
Regioselective methylation of diol 13 delivered the corre-
sponding methyl ether 15, which was converted to the ketone
14 by consecutive treatment with LiBH4 and then lead
tetraacetate. Treatment of 14 with ethyl bromomethylacrylate,
17, and zinc in THF18gave the spiro-R-methylen-γ-lactone
18 with complete facial selectivity in 86% yield. After
removal of the protecting silyl group, Dess-Martin oxidation
furnished aldehyde 20, where the configuration of the spiro
stereocenter was unambiguously ascertained by the remark-
able NOE between H9a and one of the protons H10. Further
treatment of 20 with 17 and zinc produced a roughly 1:1
mixture of bislactones 21 and 22 in 73% overall yield. Their
relative erythro/threo configuration was established by NMR
and comparison with literature data.5b,19 Thus, the proton H3
in the erythro isomer 21 (δ 4.24) is upfield shifted in relation
to the threo isomer 22 (δ 4.51), whereas the opposite occurs
for the carbon atom C3 (δ 62.0 and 60.4, respectively). This
assignment is consistent with NOE experiments performed
in subsequent derivatives.
All that remained to convert the threo bislactone 22 into
stemonidine was to reduce the lactam and the C-C double
bonds. Preparation of the corresponding thiolactam followed
by treatment with Ra-Ni was devised as a straightforward
manner to achieve this goal, and this protocol was tested
starting from the erythro isomer 21. Thus, treatment of 21
with Lawesson’s reagent furnished the expected thiolactam,
which by reaction with Ra-Ni in EtOH delivered a mixture
of amines 23 and 24, epimers at C11, in 54% total yield for
the two steps (Scheme 6). Unfortunately, when the same
procedure was applied to the threo bislactone 22, the
corresponding thiolactam underwent rapid decomposition.
We found that hydrogenation of 22 under 6 bar pressure in
Scheme 3. Construction of the Azabicyclic Core
genolysis of 4, performed by treatment with Zn in glacial
acetic acid, followed by basic treatment, and then heating
furnished lactam 3 in 84% overall yield.
Once the azabicyclic core had been elaborated, the forma-
tion of the surrounding γ-lactones was our next concern.
Obviously, the silyloxymethyl substituent at C3 in 3 should
work as the pivotal feature to install the east-side lactone,
which is common to all the target alkaloids shown in Scheme
1, as well as to many other Stemona alkaloids. Conversely,
the lactone at the west region differs from one alkaloid to
another and its formation must be specifically devised for
each target. However, for both the tuberostemospironine and
stemoamide groups, all the alkaloids bearing an oxygen atom
at C8 present the same configuration at this center, opposite
to that in compound 3. On the other hand, preliminary studies
with model compounds had shown that attempted manipula-
tion of the hydroxyl group at C8 for further synthetic elabor-
ation usually led to elimination products,13a and a simple con-
formational analysis of the perhydropyrroloazepinone skel-
eton of these compounds shows that the concave face is
relatively inaccessible. Taking these facts into account, it
seemed to us that dehydration of 3 followed by diastereo-
selective dihydroxylation could be a good strategy for further
studies (Scheme 4). Diol 13 was considered a suitable pre-
cursor for stemospironine because it would present the correct
configuration at C8 and a convenient functionalization to form
(18) Rauter, A. P.; Figueiredo, J.; Ismael, M.; Canda, T.; Font, J.;
Figueredo, M. Tetrahedron: Asymmetry 2001, 12, 1131.
(19) Blanco, P.; Busque´, F.; de March, P.; Figueredo, M.; Font, J.;
Sanfeliu, E. Eur. J. Org. Chem. 2004, 48.
(17) de March, P.; Figueredo, M.; Font, J. Electron-deficient 1,2-
disubstituted olefins of E configuration lead mainly to the formation of
endo cycloadducts. Heterocycles 1999, 50, 1213 and references therein.
Org. Lett., Vol. 9, No. 9, 2007
1771

Scheme 5
the presence of Pd/C in EtOH/2 M HCl produced a mixture
-5.4 (c 0.9, acetone)2 (Table 1), definitively shows that the
structure assigned to the alkaloid isolated from natural
sources is incorrect and that it must be reassigned as
stemospironine.
of C11-epimeric azepinones 25 and 26 in 68% yield.
Formation of the derived thiolactams and then treatment with
Ra-Ni furnished the corresponding azepines in 45% yield
for the two steps (Scheme 7). Analytical samples of each
isomer could be chromatographically isolated.
Table 1. Comparison of 13C NMR Data (δ) in CDCl3
stemonidine4
179.3 (2)
stemospironine4
179.5 (2)
synthetic 2
epi-2
Scheme 6
180.8
179.5
90.8
87.2
79.1
69.3
68.5
57.5
52.1
35.69
35.67
35.60
34.8
28.3
26.9
26.9
25.3
16.8
14.8
180.6
179.5
91.3
85.9
79.6
69.3
68.1
57.4
51.8
35.5
35.0
34.9
33.9
27.4
26.83
26.79
25.4
16.4
14.8
90.3
85.2
80.0
67.7
63.0
58.0
48.9
35.7
35.1
35.0
34.6
27.0
26.5
25.6
22.3
17.6
14.8
90.5
85.2
80.0
67.7
63.1
58.0
48.9
35.7
35.0
35.0
34.6
27.0
26.5
25.7
22.4
17.5
14.8
The structure of the less-polar isomer was established as
2 by NMR techniques. A remarkable NOE between H8 and
H11 evidences the configuration of the R-carbonyl stereo-
center at C11, whereas that of the east-side lactone is proved
by complementary NOEs from H14 or (CH3)18 to either proton
at C15. A comparison of the 13C NMR data of 2, [R]20
-16 (c 0.25, acetone), and natural stemonidine, [R]24
)
)
D
D
Acknowledgment. We acknowledge financial support
from DGI (project CTQ2004-2539) and CIRIT (project
2001SGR00178). We also acknowledge MCYT-ERDF for
a Ramo´n y Cajal contract (F.B.) and grants from DGR (P.B.)
and MECD (F.S.).
Scheme 7
Note Added after ASAP Publication: Compounds (-)-2
and (-)-epi-2 in Scheme 7 were lactams instead of amines
in the version published ASAP March 31, 2007; the corrected
version was published ASAP April 11, 2007.
Supporting Information Available: Experimental pro-
cedures, listing of physical data of new compounds, and
significant 1H and 13C NMR spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL070486P
1772
Org. Lett., Vol. 9, No. 9, 2007