Total synthesis of (±)-mersicarpine

Article abstract of DOI:10.1021/ol800259s

(Chemical Equation Presented) The first total synthesis of the indole alkaloid mersicarpine is reported. Key steps include a β-dicarbonyl radical cyclization, as well as an oxidation of the benzopyrrole moiety to establish the masked 1,2-dicarbonyl functionality. An X-ray crystal structure and discussion of the 1H NMR behavior of the natural product are also presented.

Full text of DOI:10.1021/ol800259s

ORGANIC
LETTERS
2
008
Vol. 10, No. 7
437-1440
Total Synthesis of (±)-Mersicarpine
1
Jakob Magolan, Cheryl A. Carson, and Michael A. Kerr*
Department of Chemistry, The UniVersity of Western Ontario, London,
Ontario, Canada N6A 5B7
makerr@uwo.ca
Received February 4, 2008
ABSTRACT
The first total synthesis of the indole alkaloid mersicarpine is reported. Key steps include a
oxidation of the benzopyrrole moiety to establish the masked 1,2-dicarbonyl functionality. An X-ray crystal structure and discussion of the
NMR behavior of the natural product are also presented.
â
-dicarbonyl radical cyclization, as well as an
1
H
The Kopsia species of flowering plants have provided the
yield secoleuconoxine itself, which could upon olefin reduc-
tion lactamize to leuconoxine (3). Aminal hydrolysis and
4
natural product isolation chemist with a cornucopia of indole
1
alkaloids. A subgroup of these alkaloids has piqued our
adjustment of the oxidation state of leuconoxine could yield
5
interest as synthetic chemists with its synthetically daunting
structural architectures and proposed biosynthetic nexus
the pyrrole-containing rhazinilam (4).
Mersicarpine is a particularly unusual natural product,
bearing a seven-membered cyclic imine and an intricately
oxidized indole moiety which is essentially a masked 1,2-
dicarbonyl. Herein we present the first total synthesis of
mersicarpine, the X-ray crystallographic analysis of its
structure, and clarification of its sensitive NMR behavior with
respect to solvent acidity.
(Figure 1). To illustrate their structural relation, one might
Retrosynthetically, we envisioned that mersicarpine would
be the product of imine formation from 5 (Scheme 1). This
keto hemiaminal would in turn be formed by the oxidation
6
of N-acylindole 6. It was our belief that the aminopropyl
and ethyl moieties of 6 could be installed via elaboration of
a suitable dicarbonyl species. We have recently disclosed a
malonic radical cyclization route which would be effective
in producing the substructure in 6 from a pendant malonic
Figure 1. Indole alkaloids isolated from Kopsia species.
2
envision hydrolysis of the imine in mersicarpine (1) and
7
recyclization onto the hemiaminal to yield the skeleton found
ester in 7.
3
in secoleuconoxine (2). Addition of an acetate unit would
(
3) Lim, K.-H.; Kam, T.-S. HelV. Chim. Acta 2007, 90, 31-35.
(
1) See (a) Lim, K.-H.; Hiraku, O.; Komiyama, K.; Koyano, T.; Hayashi,
(4) Abe, F.; Yamauchi, T. Phytochemistry 1994, 35, 169-171.
(5) Banerji, A.; Majumder, P. L.; Chatterjee, A. Phytochemistry 1970,
9, 1491-1493.
M.; Kam, T.-S. J. Nat. Prod. 2007, 70, 1302-1307. (b) Lim, S.-H.; Sim,
K.-M.; Abdulla, Z.; Hiraku, O.; Masahiko, H.; Komiyama, K.; Kam, T.-S.
J. Nat. Prod. 2007, 70, 1380-1383. (c) Zhou, H.; He, H.-P.; Kong, N.-C.;
Wang, Y.-H.; Liu, X.-D.; Hao, X.-J. HelV. Chim. Acta. 2006, 89, 515-519
and references therein.
(6) (a) Altinis Kiraz, C. I.; Emge, T. J.; Jimenez, L. S. J. Org. Chem.
2004, 69, 2200-2202 and references therein. (b) Wang, Z.; Jimenez, L. S.
Tetrahedron Lett. 1996, 37, 6049-6052. (c) Colandrea, V. J.; Rajaraman,
S.; Jimenez, L. S. Org. Lett. 2003, 5, 785-788.
(2) Kam, T.-S.; Subramaniam, G.; Lim, K.-H.; Choo, Y. M. Tetrahedron
Lett. 2004, 45, 5995-5998.
(7) Magolan, J.; Kerr, M. A. Org. Lett. 2006, 8, 4561-4564.
1
0.1021/ol800259s CCC: $40.75
© 2008 American Chemical Society
Published on Web 03/12/2008

Scheme 1. Retrosynthesis of Mersicarpine
3
Scheme 3. Mn(OAc) Mediated Cyclization of a Keto-Ester
The inauguration of our synthetic efforts toward mersi-
carpine is presented in Scheme 2. N-Acylindoline 9 was
Scheme 2. Radical Cyclization to Form a
Dihydropyridoindolone
3
). Indoline 13 was prepared in a similar fashion to malonic
derivative 9. As treatment of 13 with Mn(OAc) led to
3
decomposition of the starting material, the oxidation of 13
with DDQ to produce indole 14 was employed as an
alternative. The malonic radical cyclization of 14 proceeded
as expected when performed in acetic acid to yield 15 in
55% yield. We were disappointed to observe that all attempts
at decarboxylation of 15 yielded only decomposition prod-
ucts. Fortuitously, while attempting the malonic radical
cyclization of 14 in methanol, ester 11 was isolated, albeit
in low yield (Scheme 3). The apparent propensity of 15 to
undergo a retro-Claisen condensation was a useful observa-
tion given our inability to effect the desired decarboxylation
reaction.
efficiently prepared via the acylation of indoline 8 with
acryloyl chloride followed by the Michael addition of
dimethylmalonate in a high yielding, operationally simple
It now became apparent that the best route to mono-ketone
1
1
6 would be via a retro-Claisen condensation of diketone
8 (Scheme 4). Indole 18 was easily prepared in four steps
3
one-pot procedure. Subjection of 9 to Mn(OAc) in refluxing
from indoline 8. Gratifyingly, fragmentation of 18 through
a retro-Claisen condensation was indeed effective producing
methyl ketone 16 in 95% yield upon treatment with sodium
bicarbonate in methanol. Conjugate addition of 16 to
acrylonitrile yielded 19 (74%), which was quantitatively
reduced to secondary alcohol 20. Dehydration of 20 to 21
was exceptionally difficult and failed to succumb to a large
number of both standard and more elaborate measures.
Preparation of the methyl xanthate followed by Chugaev
elimination provided the most satisfactory results. Hydro-
genation of 21 over Adam’s catalyst saturated both the alkene
acetic acid promoted a tandem indoline oxidation/malonic
radical cyclization producing indole 10 in 67% yield. In
8
our experience, the direct N-acryloylation of indole is low
yielding and suffers from multiple side reactions. As a result,
the synthesis of substrates such as 10 by the aromatization
of indolines, either using DDQ or the one-pot procedure
depicted in Scheme 2, is advantageous.
All attempts to sequentially reduce or otherwise function-
alize the geminal diesters of 10 were unsuccessful, in part
due to the reductive and hydrolytic lability of the indole-
amide. Fortunately, Krapcho decarboxylation under micro-
wave conditions provided the monoester 11 in 65% yield.
We were pleased to find that conjugate addition of the ester
enolate of 11 into acrylonitrile could be effected to produce
and the nitrile producing, after Boc protection, carbamate
9
2
2. The eagerly anticipated oxidation of the N-acylindole
2 was extremely facile, proceeding in 93% yield upon
2
treatment with dimethyl dioxirane, prepared in situ from
12 bearing a latent aminopropyl group in the form of a
1
0
acetone and OXONE. Although 23 was obtained as a 1:1
mixture of diastereomers, we gambled that this would prove
inconsequential given the epimerizable nature of the hemi-
aminal. The final step, amine deprotection and presumed
imine formation, gave a single compound in 82% yield.11
cyanoethyl moiety. Unfortunately, once again the labile
nature of the indole-amide functionality made elaboration
of the remaining methyl ester unfeasible.
We reasoned that the use of an analogous keto-ester might
be a more effective strategy with the methyl ketone offering
facile options for conversion to an ethyl substituent (Scheme
(
9) Ackermann, J.; Aebi, J.; Dehmlow, H.; Hirth, G.; Maerki, H.-P.;
Morand, O.; Panday, N. U.S. Pat. Appl. Publ. 2005, 2005065210.
(10) (a) Curci, R.; Fiorentino, M.; Troisi, L. J. Org. Chem. 1980, 45,
4758-4760. (b) Waldemar, A.; Chantu, S.-M. R.; Zhao, C.-G. Org. React.
2002, 61, 219.
(11) (a) Sakaitani, M.; Ohfune, Y. Tetrahedron Lett. 1985, 26, 5543-
5546. (b) Sakaitani, M.; Ohfune, Y. J. Org. Chem. 1990, 55, 870-876.
(
8) For comprehensive reviews of Mn(OAc)3 chemistry see (a) Snider,
B. B. Chem. ReV. 1996, 96, 339-363. (b) Snider, B. B. Manganese(III)-
based oxidative free-radical cyclizations in Transition Metals for Organic
Synthesis, 2nd ed.; Beller, M., Bolm, C., Eds.; Wiley-VCH Verlag GmbH
&
Co. KGaA: Weinheim, Germany, 2004; Vol. 1, pp 483-490.
1438
Org. Lett., Vol. 10, No. 7, 2008

followed by treatment of the sample with incremental
amounts of trifluoroacetic acid (Figure 3). The changes in
Scheme 4. Total Synthesis of Mersicarpine
We were initially greatly disheartened to observe that
1
several H NMR resonances of the supposed natural product
were quite different from those reported by Kam. Although
we fully expected imine formation to be a trivial step, we
were cognizant of several potential alternative reaction
pathways. Diastereomer 24 (Figure 2) was one possible
1
Figure 3. Variation of H NMR resonances with added trifluoro-
acetic acid. Red lines indicate literature values for mersicarpine
(ref 2).
Figure 2. Alternative cyclization products.
chemical shift with the addition of acid were dramatic. Not
only did protons adjacent to the basic imine nitrogen show
significant variance (He and Hf), the four aromatic reso-
nances as well as aliphatic protons He and Hg all varied in
chemical shift with degree of protonation. Clearly, extended
resonance stabilization of the iminium cation of this natural
product has substantial effects on multiple proton environ-
reaction outcome as was regioisomer 25, in which cyclization
onto the hemiaminal would yield a six-membered ring.
Formation of 24 was intuitively unlikely given that this
unnatural diastereomer would have to be the thermodynamic
sink to be produced in 82% yield from a 1:1 mixture of
hemiaminal epimers. Cyclization to 25 appeared to be
plausible as it would form a structural subunit of the
3
ments. Being unaware of the degree CDCl decomposition
1
3
leuconoxine series of natural products, however the C and
(i.e., DCl content) associated with the initial characterization
of mersicarpine we were unable to precisely duplicate the
literature data. Furthermore, it follows that the identity of
the counteranion is also likely to have an effect on proton
2
-D NMR data did not support this structure.
Some clarification was obtained when H NMR analysis
1
3
was performed in CDCl passed through basic alumina,
Org. Lett., Vol. 10, No. 7, 2008
1439

resonances. The 13C NMR data for our sample matches the
reported data very well.
Finally, after the preparation of a sufficient amount of the
natural product, we were able to induce the formation of
X-ray quality crystals which offered definitive support of
the success of our synthetic effort.12 The structure obtained
via X-ray diffraction is shown in Figure 4. Note that the
It is possible, however (in our opinion) highly unlikely,
that the reported structure for mersicarpine is in fact incorrect.
We were unable to secure an authentic sample for direct
comparison. The fact that Kam and co-workers provide sound
evidence for the reported connectivity and stereochemistry,
as well as the fact that our NMR data is very close (almost
1
3
identical for C) lead us to be fully confident of that the
1
reported structure is correct. The only resonance in the H
NMR spectrum that we were not able to rectify with acid
titration (variable by 0.07 ppm) corresponds to the protons
He and Hf, which are adjacent to the imine nitrogen and
therefore expected to be sensitive to acid concentration.
In summary, we have reported the first total synthesis of
mersicarpine and confirmed the reported structure by crystal-
lographic analysis of our synthetic product. In addition, we
1
note the dramatic chemical shift variance in the H NMR
spectra with respect to solvent acidity.
Acknowledgment. We thank the Natural Sciences and
Engineering Research Council of Canada and Merck-Frosst
for funding. We thank Mr. Doug Hairsine of the University
of Western Ontario mass spectrometry facility for performing
MS analyses and Dr. Michael Jennings for X-ray diffraction
analysis. C.A.C. is an NSERC CGS-M and PGS-D awardee.
Figure 4. X-ray structure of synthetic mersicarpine.
Supporting Information Available: Full experimental
procedures and spectroscopic data for all new compounds
and crystal data for compound 1. This material is available
free of charge via the Internet at http://pubs.acs.org.
disposition of the aminal hydroxyl group and the quaternary
ethyl moiety are in fact trans, as required for correlation to
the reported structure.
(12) Note: there is disorder at the C6 position.
OL800259S
1440
Org. Lett., Vol. 10, No. 7, 2008