The total synthesis of (±)-fortucine and a revision of the structure of kirkine

Article abstract of DOI:10.1002/anie.200704996

(Chemical Equation Presented) A molecular zipper: The total synthesis of the natural product fortucine relies on a radical cascade process initiated by the generation of a nitrogen-centered (amidyl) radical (see picture). The procedure is concise, and tin

Full text of DOI:10.1002/anie.200704996

Communications
DOI: 10.1002/anie.200704996
Natural Product Synthesis
The Total Synthesis of (ꢀ )-Fortucine and a Revision of the Structure of
Kirkine**
AurØlien Biechy, Shuji Hachisu, BØatrice Quiclet-Sire, Louis Ricard, and Samir Z. Zard*
In memory of Charles Mioskowski
Lycorine alkaloids are isolated from the Amaryllidaceae
description given for the spectrum of fortucine was incom-
plete and could be subject to confusion. We therefore
embarked on a total synthesis of structure 2 to clarify the
matter on one hand, and on the other, to provide sufficient
quantities to explore more broadly the biological activity
profile. Our study would also serve to establish a general
strategy to access galanthan frameworks with a cis B/C-ring
junction.
[
1]
species of plants, and their medicinal value has an old and
[
2]
rich history dating back to ancient Greece. Members of the
lycorine alkaloids possess antineoplastic, antimitotic, insect
antifeedant, and antiviral activities;they also inhibit growth
and cell division in higher plants, protein and DNA synthesis
in murine cells, and in vivo growth of a murine transplantable
[
3]
ascite tumor.
The tetracyclic pyrrolo[d,e]phenanthridine (galanthan)
framework that characterizes the lycorine alkaloids has
been the subject of intense synthetic scrutiny ever since the
Our synthetic approach to structure 2, henceforth called
fortucine, is based on a radical cascade process (5!4) that is
initiated by the generation of a nitrogen-centered (amidyl)
radical, which undergoes cyclization onto the alkene followed
by an oxidative cyclization to afford the aromatic ring
(Scheme 2). If successful, this cascade process would simulta-
neously form rings B and D in one step. Furthermore, it was
[
4]
structure of lycorine itself was established in 1955. While the
great majority of the these alkaloids have a trans B/C-ring
junction, there are a few compounds with a cis B/C-ring
junction;for example, g-lycorane (1), fortucine (2), and
[
3a,5]
siculinine (3;Scheme 1).
Scheme 1. Lycorine-type alkaloids.
Fortucine (2) was isolated from the Fortune variety of
[
5a]
narcissus by Tokhtabaeva et al. in 1987
and unaware of the earlier work, Bastida et al.
from Crinum kirkii, a common grassland plant of East Africa
A few years later,
[
5b]
isolated
Scheme 2. Retrosynthetic analysis of of rtucine ( 2). Bn=benzyl,
TBS=tert-butyldimethylsilyl.
[
6]
used as a purgative, a rat poison, and a treatment for sores,
a
substance they called kirkine and for which they attributed
the same structure 2. The reported NMR spectra showed
some discrepancies between the two compounds, even if the
hoped that the cyclizations would be stereoselective, to
furnish the requisite cis B/C-ring junction, and regioselective,
to give the para (as opposed to ortho) cyclized product with
respect to the benzyloxy group. The key precursor would be
constructed by the union of p-benzoquinone ethylene mono-
ketal (7) with the azaenolate of hydrazone 8 . It is worth
noting the convergent manner by which rings A and C are
appended upon hydrazone 8, which in turn provides the
foundation of ring D.
[
*] A. Biechy, Dr. S. Hachisu, Dr. B. Quiclet-Sire, Prof. S. Z. Zard
Laboratoire de Synth›se Organique associØ au CNRS
Ecole Polytechnique
[
7]
[8]
Route de Saclay, 91128 Palaiseau (France)
Fax: (+33)1-6933-5972
E-mail: zard@dcso.polytechnique.fr
Dr. L. Ricard
The synthesis started by trapping the azaenolate derived
from hydrazone 8 (3 equiv) by treatment with LDA (3 equiv)
and p-monoketal 7 (1 equiv) to give alcohol 9 in quantitative
yield (Scheme 3). Pleasingly, the desired 1,2-addition occur-
red exclusively without formation of the undesired 1,4-
adduct. Furthermore, the efficiency of this process with
respect to hydrazone 8 was surprising, because cyclization of
Laboratoire “HØtØroØlØments et Coordination”
Ecole Polytechnique
Route de Saclay, 91128 Palaiseau (France)
[
**] We wish to thank Prof. J. Bastida (Barcelona) for a sample of kirkine
and Ecole Polytechnique for fellowships to A.B. and S.H.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
1
436
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1436 –1438

Angewandte
Chemie
the creation of an axial alcohol at C1. Thus, the protected
ketone in 4 was unmasked to give enone 10 in 88% yield
by treatment with TFA/H O (Scheme 4). All our efforts
2
to induce reductive elimination of the TBS group failed
and attempts to replace the TBS motif with better leaving
groups were frustrated by the reactivity of the free
alcohol.
In contrast, Michael addition of 2-methoxythiophenol
(ArSH;Scheme 4) onto enone 10 gave thioether 11 as
one diastereoisomer in 80% yield, which was subse-
quently treated with LiAlH to reduce both the amide
4
and the ketone to give the tertiary amine and alcohol
groups, respectively. As expected, reduction of the
ketone occurred stereoselectively from the convex face
to exclusively deliver the axial alcohol. Oxidation of the
sulfide group with DMDO (in the presence of TFA to
Scheme 3. Synthesis ofthe intermediate pentacycle 4. LDA=lithium diiso-
propylamide, Tf=trifluoromethanesulfonyl, DLP=1,2-dichloroethane with
lauroyl peroxide.
the azaenolate onto the thiocarbonyl group was feared to
cause major difficulties. The azaenolate of 8 was a relatively
stable species at ꢁ788C that could be trapped by a variety of
a,b-unsaturated ketones. These coupled products could in
principle be used to construct dihydropyrroles via iminyl
[
9]
[10]
radicals, and galanthan frameworks via amidyl radicals
through cyclizations onto neighboring unsaturated moieties.
At this point it proved necessary to protect alcohol 9 as
the silyl ether by treatment with TBSOTf. Attempts were
made to directly transform alcohol 9 into the radical precursor
5
without introduction of a protecting group, but the
instability of the free diallylic tertiary alcohol made it
impossible to execute the necessary benzoylation reaction.
Hydrazone 6 was converted into radical precursor 5 by
reduction with NaBH followed by immediate acylation, of
4
the somewhat unstable intermediate hydrazide, with
3
-benzyloxy-4-methoxybenzoyl chloride to give amidyl rad-
ical precursor 5. Treatment of 5 in refluxing 1,2-dichloro-
ethane with lauroyl peroxide (1.4 equiv) effected the radical
cascade to ultimately give pentacycle 4 in 60% yield
Scheme 4. Synthesis of of rtucine ( 2) and revision ofthe structure of
kirkine. TFA=trifluoroacetic acid, DMDO=2,2-dimethyldioxirane,
ArSH=2-methoxythiophenol.
[
10]
(
Scheme 3).
occurred both stereoselectively and regioselectively to give
as the major product. Isomers arising from the trans B/C-
To our delight, the cascade process had
4
ring junction were not observed, and the desired regioisomer
was formed in a 14:1 ratio (para/ortho with respect to the
benzyloxy group). The all-cis cyclization is inherent to this
type of skeleton, and the steric clash between the ethyl-
eneketal and benzyloxy functionalities disfavors the ortho
regiochemistry.
Interestingly, hydrazone 8 is a synthetic building block
capable of reacting through ionic and radical pathways on
each of its termini. The efficiency of the radical cascade to
produce densely functionalized pentacycle 4 in five steps is
remarkable. The structural diversity of lycorine-type alkaloids
generally originates from functional groups located on ring C.
The strategic positioning of the alkene, the protected ketone,
and the protected alcohol in 4 can potentially be deployed to
target other complex members of the lycorine family.
The remaining major task at hand was to carry out a
reductive cleavage of the allylic silyl ether in 4 with
protect the tertiary amine) furnished sulfone 12 in 92% yield.
Hydrogenolysis over Pd/C quantitatively unmasked phenol
13, which was subjected to Julia-type olefination using 6%
Na/Hg amalgam to finally give the target molecule fortucine
(2) in 50% yield. Some material arising from desulfonylation
without elimination of the TBS group was observed in the
crude product mixture, and this explains the modest yield in
the last step.
[
10]
Our synthetic final product was unambiguously confirmed
by X-ray crystallography as having structure 2 (racemic). The
NMR spectrum of our product was in accord with the data of
Tokhtabaeva et al. but not with those of Bastida et al. Indeed,
an authentic sample of kirkine, kindly supplied by Prof.
Bastida, proved different from our synthetic material (by
TLC and NMR spectroscopy). Interestingly, certain parts of
1
the H NMR spectrum in our case were ill resolved, ap-
parently due to slow conformational changes in ring C as
indicated by molecular models. A good resolution was
concomitant migration of the double bond (D2,3!D ), and
3,4
Angew. Chem. Int. Ed. 2008, 47, 1436 –1438
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1437

Communications
obtained by recording the spectrum at ꢁ278C, where the
[3] a) S. F. Martin in The Alkaloids, Vol. 30 (Ed.: A. Brossi),
Academic Press, New York, 1987, pp. 251 – 376;b) S. Ghosal,
K. S. Saini, S. Razdan, Phytochemistry 1985, 24, 2141;c) T.
Mineshita, K. Yayaguchi, K. Takeda, K. Kotera, Annu. Rep.
Shionogi Res. Lab. 1956, 6, 119;d) U. Chattopadhyay, L.
Chaudhuri, S. Das, Y. Kumar, S. Ghosal, Pharmazie 1984, 39,
compound was essentially frozen in its most stable conforma-
tion (see the Supporting Information). It appeared therefore
that kirkine and fortucine were either diastereoisomers or
double-bond isomers. In order to distinguish between these
two possibilities, we subjected our synthetic product and
authentic kirkine to catalytic hydrogenation and obtained the
same compound 15, as confirmed by NMR spectroscopy,
TLC, and HRMS (the only difference is that our sample is
racemic). Therefore, kirkine has structure 14 in which the
855;e) P. Leo De, G. Dalessandro, A. De Santis, O. Arigoni,
Plant Cell Physiol. 1973, 14, 487.
[4] a) S. Takagi, W. I. Taylor, S. Uyeo, H. J. Yajima, J. Chem. Soc.
1955, 4003;b) H. Fales, E. M. Warnhoff, W. C. Wildman, J. Am.
Chem. Soc. 1955, 77, 5885.
[
5] a) G. M. Tokhtabaeva, V. I. Sheichenko, I. V. Yartseva, O. N.
Tolkachev, N. Khim. Prir. Soedin. 1987, 727;b) J. Bastida, C.
Codina, P. Peeters, M. Rubiralta, M. Orozco, F. J. Luque, S. C.
Chharbra, Phytochemistry 1995, 40, 1291;c) P. Richomme, V.
Pabuccuoglu, T. Gozler, A. J. Freyer, M. Shamma, J. Nat. Prod.
[
5c]
olefin has the same position as in siculinine (3). It is the only
isomer that is compatible with the NMR spectra and that
would lead to 15 upon catalytic reduction.
In summary, we have devised a concise (12 steps, 9%
overall yield) first total synthesis of (ꢀ )-fortucine (2) and
corrected the structure of kirkine, which should be revised to
1989, 52, 1150.
[
6] a) J. M. Watt, M. G. Breyer-Brandwijk in Medicinal and Poison-
ous Plants of Souther and Eastern Africa, E. & S. Livingstone
Ltd., Edinburgh, 1962;b) J. O. Kokwaro in Medicinal Plants of
East Africa, East Africa Literature Bureau, Kampala, 1976.
7] M. E. Trân-Huu-Dâu, R. Wartchow, E. Winterfeldt, Y. S. Wong,
Chem. Eur. J. 2001, 7, 2349.
1
4. The synthetic route is noteworthy for its stereo- and
regioselective radical cascade process as well as the multi-
faceted use of hydrazone 8.
[
Received: October 29, 2007
Published online: January 17, 2008
[8] Formed by the condensation of acetaldehyde with 1-(methyl-
thio)thiocarbonyl-1-phenyl-hydrazine, see Ref. [9a].
[
9] a) A. C. Callier-Dublanchet, B. Quiclet-Sire, S. Z. Zard, Tetra-
hedron Lett. 1995, 36, 8791;b) A. C. Callier-Dublanchet, B.
Quiclet-Sire, S. Z. Zard, Tetrahedron Lett. 1997, 38, 2463;c) M.
Depature, J. Grimaldi, J. Hatem, Eur. J. Org. Chem. 2001, 941.
Keywords: alkaloids · cascade reactions · radical reactions ·
.
radicals
[
10] a) X. Hoang-Cong, B. Quiclet-Sire, S. Z. Zard, Tetrahedron Lett.
999, 40, 2125;b) G. L. Barclay, B. Quiclet-Sire, G. Sanchez-
Jimenez, Org. Biomol. Chem. 2005, 3, 823.
1
[
1] D. R. Dalton, The Alkaloids: The Fundamental Chemistry—A
Biogenetic Approach, Marcel Dekker, New-York, 1979.
2] J. L. Hartwell, Lloydia 1967, 30, 379.
[
1
438
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1436 –1438