Total Synthesis of (±)-Goniomitine via Radical Translocation

Article abstract of DOI:10.1021/acs.orglett.5b02277

The aspidosperma alkaloid goniomitine was synthesized in six steps from 2-ethyl-valerolactam. The convergent strategy features an Ullman coupling to assemble the required carbon atoms. A complexity-generating radical translocation reaction was used to bui

Full text of DOI:10.1021/acs.orglett.5b02277

Letter
pubs.acs.org/OrgLett
Total Synthesis of ( )-Goniomitine via Radical Translocation
Jessica K. Vellucci and Christopher M. Beaudry*
Department of Chemistry, Oregon State University, 153 Gilbert Hall, Corvallis, Oregon 97331, United States
S
* Supporting Information
ABSTRACT: The aspidosperma alkaloid goniomitine was synthe-
sized in six steps from 2-ethyl-δ-valerolactam. The convergent strategy
features an Ullman coupling to assemble the required carbon atoms. A
complexity-generating radical translocation reaction was used to build
the indole architecture.
oniomitine (1) is a monoterpene alkaloid that was isolated
Scheme 2. Tryptophol Functionalization Strategy
G
in 1987 by Husson and co-workers from the Madagascan
tree Gonioma malagasy.1 Goniomitine belongs to the family of
aspidosperma alkaloids, but it has an unusual molecular
architecture featuring four angularly fused rings and an aminal
functional group. Husson and co-workers forwarded an insightful
biogenetic hypothesis suggesting that goniomitine is derived
from vincadifformine, which has a more typical aspidosperma
alkaloid structure (Scheme 1).1 Intermediate 2 was hypothesized
carbonyl gives the aminal structure and completes the synthesis
of goniomitine.
The second general strategy for goniomitine5 is exemplified in
Scheme 3. An ortho-substituted aniline 8 is used as a starting
Scheme 1. Biosynthetic Hypothesis for Goniomitine
Scheme 3. Indole Functionalization Strategy
to undergo ester hydrolysis and decarboxylation to arrive at
tetracycle 3. Retro-Mannich reaction would produce iminium 4,
which could be captured by the indole to produce goniomitine.
This biosynthetic hypothesis was substantiated by subsequent
investigations by Husson,2 and by a biomimetic synthesis of
goniomitine from vincadifformine following this strategy by
Lewin and co-workers.3
The unusual molecular architecture of goniomitine and its
micromolar antiproliferative activity on several cancer cell lines
have captivated the attention of many synthetic chemists.4,5 In
addition to the biosynthetic studies by Husson and Lewin, eight
total syntheses of goniomitine have been reported. All of these
total syntheses follow two general strategies. In the first strategy4
(Scheme 2), a protected tryptophol starting material (5) is
functionalized at the C2 position of the indole in one or more
steps. Additional synthetic transformations are used to build the
δ-valerolactam C-ring found in goniomitine. Finally, reduction of
a tethered azide and reductive cyclization with the lactam
material for the construction of a 2,3-disubstituted indole 9.6
Several steps are used to append a tethered D-ring in the form of
a δ-valerolactam with the quaternary stereocenter intact.
Activation of the lactam carbonyl in 9 gives an iminium ion
(10) that cyclizes to form the aminal functional group in
goniomitine.
Our interest in aminal-containing molecules led us to
goniomitine.7 Initially, we explored strategies to use aminal
radical intermediates for the construction of the goniomitine
structure. Radical chemistry is well suited for the synthesis of
nitrogenous molecules, as carbon-centered radicals are generally
tolerant of the nitrogen lone pair and the rich acid−base
Received: August 5, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.5b02277
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
reactivity of the nitrogen atom.8 However, our initial efforts were
ultimately unsuccessful because of a general paucity of methods
for the preparation of aminals incorporating indole nitrogen
atoms.9 However, we continued to consider radical-based
approaches toward the goniomitine structure.
Goniomitine may arise from intermediate 11, which was
envisioned to be an immediate precursor to the natural product
by way of global hydride reduction (Scheme 4). In a departure
yields were disappointing, and an alternative route was
investigated.
Changing the order of steps proved to be auspicious and more
step-economical (Scheme 5, bottom). We found that the
Ullmann coupling of ethyl δ-valerolactam with 16 was a high-
yielding reaction producing N-aryl lactam 17. Moreover, the
installation of the quaternary carbon proceeded smoothly to give
12 in high yield, which avoided the use of the benzyl protecting
group altogether. Finally, bromination of the alkene gave key
cyclization intermediate 18 as a 2.8:1 mixture of alkene isomers.
Notably, this sequence begins from ethyl δ-valerolactam and
assembles 18 in just three high-yielding operations, and 18 is
complete with all the carbons of goniomitine and a functional
handle for the radical cyclization reaction.
Scheme 4. Retrosynthetic Analysis of Goniomitine
With 18 in hand, the key radical cyclization reaction was
investigated.13 Treatment of 18 with standard radical conditions
led to the formation of desired indolines 19 and 20 as a 2.4:1
mixture of indoline diastereomers. Presumably, the initial radical
species 21 arises from the homolytic cleavage of the C−Br bond.
Radical translocation then gives α-amino radical 22. Intra-
molecular cyclization gives radical 23, which abstracts a hydrogen
atom from Bu3SnH to give the products.
The combined chemical yield of 19 and 20 in the reaction was
quite good, and the production of two diastereomeric indolines
was acceptable because these stereocenters will be destroyed
upon oxidation to the corresponding indole. The pure geometric
alkene isomers of 18 were isolated and individually subjected to
the radical reaction; they gave an identical mixture of 19 and 20.
This result indicated there is no stereospecificity in this reaction,
and the alkene geometry of 18 is inconsequential.
from all previous synthetic strategies, we chose to unravel the
indole B-ring to give intermediate 12. Intermediate 12 could be
prepared from substituted δ-valerolactam 13 using standard
cross-coupling methods.
The previous syntheses of goniomitine involved 8−28
synthetic steps, and many of the transformations were devoted
to preparing the central C-ring of the target molecule. In fact, all
previous syntheses of goniomitine required construction of the
central C-ring. We suspected that beginning our synthesis with a
readily available δ-valerolactam starting material (as the C-ring)
could result in an expeditious synthesis of the target molecule.
A key consideration in the synthesis of goniomitine is creating
the quaternary stereocenter. We initially followed Pagenkopf’s
report that a one-step alkylation (or a more robust four-step
sequence) of 14 would give 15, which contains the requisite
quaternary carbon (Scheme 5, top).5b We were able to optimize
The stereochemistry of 19 and 20 is also inconsequential for
the goniomitine synthesis; however, the stereochemistry of the
major product diastereomer (19) was determined using NOESY
methods (see key correlations in Scheme 6). The similarity of the
coupling pattern in 19 and 20 led us to tentatively assign 20 as a
trans indoline as well.
Scheme 5. Preparation of Radical Reaction Substrate
Oxidation of 19 and 20 to indole 11 would convert both
diastereomers to one racemic product, and it would make
characterizing intermediates from the final steps of the synthesis
more convenient. However, the N-acyl group, substitution at
Scheme 6. Radical Translocation
the one-step conversion to a synthetically useful 63% yield.
However, we had substantial difficulty removing the benzyl
protecting group to give lactam 13. After many failed attempts at
this transformation with hydrogenolysis10 and other condi-
tions,11 we found that dissolving metal reduction cleanly
removed the benzyl group to give 13, which proved to be a
sensitive intermediate. Lactam 13 could be taken into the
subsequent coupling with iodide 1612 without purification, but
the Ullmann coupling to give 12 was low yielding. Ultimately,
although we were able to forge the desired bonds, the chemical
B
DOI: 10.1021/acs.orglett.5b02277
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
both C2 and C3, and the inductively withdrawing carbomethoxy
group in 19 and 20 conspired against this transformation, and no
conditions for the indoline oxidation to 11 were found. Most
standard reagents (MnO2, DDQ, Pd/C, S8, KMnO4) for indoline
oxidation14 were ineffective, returning unreacted starting
materials. More forcing conditions using CAN (Scheme 7, eq
1) or NBS (eq 2) gave overoxidation of the indole producing 24
and 25, respectively.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: christopher.beaudry@oregonstate.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by the National Science Foundation
under Grant Number 1465287.
■
Scheme 7. Attempted Indoline Oxidation
REFERENCES
■
(1) (a) Randriambola, L.; Quirion, J. − C.; Kan-Fan, C.; Husson, H. −
P. Tetrahedron Lett. 1987, 28, 2123.
(2) Hashimoto, C.; Husson, H. − P. Tetrahedron Lett. 1988, 29, 4563.
(3) (a) Lewin, G.; Bernadat, G.; Aubert, G.; Cresteil, T. Tetrahedron
2013, 69, 1622. (b) Lewin, G.; Schaeffer, C.; Lambert, P. H. J. Org. Chem.
1995, 60, 3282. (c) Lewin, G.; Schaeffer, C. Nat. Prod. Lett. 1995, 7, 227.
(4) (a) De Simone, F.; Gertsch, J.; Waser, J. Angew. Chem., Int. Ed.
2010, 49, 5767. (b) Jiao, L.; Herdtweck, E.; Bach, T. J. Am. Chem. Soc.
2012, 134, 14563. (c) Zhou, S.; Jia, Y. Org. Lett. 2014, 16, 3416.
(5) (a) Takano, S.; Sato, T.; Inomata, K.; Ogasawara, K. J. Chem. Soc.,
Chem. Commun. 1991, 462. (b) Morales, C. L.; Pagenkopf, B. L. Org.
Lett. 2008, 10, 157. (c) Zhou, B.; Du, J.; Yang, Y.; Li, Y. Chem. - Eur. J.
2014, 20, 12768. (d) Xu, Z.; Wang, Q.; Zhu, J. Angew. Chem., Int. Ed.
Of course, the goniomitine structure does not possess an N-
acyl indole, so we decided to postpone the indoline oxidation
step and perform the required hydride reduction first (Scheme
8). Previous goniomitine syntheses utilized a reductive
̀
2013, 52, 3272. (e) Wagnieres, O.; Xu, Z.; Wang, Q.; Zhu, J. J. Am.
Chem. Soc. 2014, 136, 15102. (f) Mizutani, M.; Inagaki, F.; Nakanishi,
T.; Yanagihara, C.; Tamai, I.; Mukai, C. Org. Lett. 2011, 13, 1796.
(6) The Pagenkopf synthesis begins with a nitrile (rather than an
aniline) to access intermediate 9. See ref 5b.
Scheme 8. Completion of the Goniomitine Synthesis
(7) (a) Schiedler, D. A.; Vellucci, J. K.; Lu, Y.; Beaudry, C. M.
Tetrahedron 2015, 71, 1448. (b) Schiedler, D. A.; Lu, Y.; Beaudry, C. M.
Org. Lett. 2014, 16, 1160. (c) Schiedler, D. A.; Vellucci, J. K.; Beaudry, C.
M. Org. Lett. 2012, 14, 6092.
(8) For examples, see: (a) Urry, W. H.; Juveland, O. O.; Stacey, F. W. J.
Am. Chem. Soc. 1952, 74, 6155. (b) Bongini, A.; Cardillo, G.; Orena, M.;
Sandri, S.; Tomasini, C. J. Org. Chem. 1986, 51, 4905. (c) Sibi, M. P.; Ji,
J.; Wu, J. H.; Gurtler, S.; Porter, N. A. J. Am. Chem. Soc. 1996, 118, 9200.
̈
(9) For examples of aminals involving the indole nitrogen atom, see:
(a) Swaminathan, S.; Narasimhan, K. Chem. Ber. 1966, 99, 889.
(b) Katritzky, A. R.; Rewcastle, G. W.; Fan, W.−Q. J. Org. Chem. 1988,
53, 5685. (c) Katritzky, A. R.; Rachwal, S.; Rachwal, B. J. Chem. Soc.,
Perkin Trans. 1 1987, 799. (d) Love, B. E.; Nguyen, T. Synlett 1998,
1998, 1123. (e) Love, B. J. Org. Chem. 2007, 72, 630. (f) Hellmann, H.;
Haas, G. Chem. Ber. 1957, 90, 53. (g) Guillard, J. Heterocycles 2003, 60,
865.
(10) Nishimura, S. Handbook of Heterogeneous Catalytic Hydrogenation
for Organic Synthesis; John Wiley & Sons: New York, 2001; pp 583−593.
(11) Williams, R. M.; Kwast, E. Tetrahedron Lett. 1989, 30, 451.
(12) Iodide 15 is commercially available; however, it can be prepared in
one step from o-iodobenzaldehyde: Grigg, R.; Inman, M.; Kilner, C.;
Koeppen, I.; Marchbank, J.; Selby, P.; Sridharan, V. Tetrahedron 2007,
63, 6152.
(13) For related radical translocation reactions, see: (a) Lathbury, D.
C.; Parsons, P. J.; Pinto, I. J. Chem. Soc., Chem. Commun. 1988, 81.
(b) Curran, D. P.; Kim, D.; Liu, H. T.; Shen, W. J. Am. Chem. Soc. 1988,
110, 5900. (c) Robertson, J.; Peplow, M. A.; Pillai, J. Tetrahedron Lett.
1996, 37, 5825. (d) Curran, D. P.; Abraham, A. C.; Liu, H. J. Org. Chem.
1991, 56, 4335. (e) Beaulieu, F.; Arora, J.; Veith, U.; Taylor, N. J.;
Chapell, B. J.; Snieckus, V. J. Am. Chem. Soc. 1996, 118, 8727.
(14) Handbook of Reagents for Organic Synthesis: Oxidizing and
Reducing Agents; Burke, S. D., Danheiser, R. L., Eds.; John Wiley & Sons:
Chichester, 1999.
cyclization sequence of compounds with structure 7 (Scheme
1) that was promoted by LiAlH4. Treatment of 19 and 20 with
LiAlH4 in hot THF for 12 h, followed by an aqueous workup,
gave a mixture of compounds with spectroscopic characteristics
consistent with the desired reduction products 26 as a complex
mixture of diastereomers. Gratifyingly, intermediates 26 under-
went smooth oxidation with MnO2 to give the natural product as
the only isolable product. The combined yield of goniomitine
from radical translocation products 19 and 20 was 51%.
In summary, we have completed a synthesis of the
aspidosperma alkaloid goniomitine. The synthesis requires six
synthetic transformations from ethyl δ-valerolactam and has an
overall yield of 29%, which makes it the most efficient synthesis
of ( )-goniomitine to date. Key features of this synthesis are the
strategic use of a δ-valerolactam starting material and a radical
translocation reaction to build the indole B-ring. Efforts to extend
this strategy in other alkaloid targets are underway in our
laboratory.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.5b02277.
Experimental procedures, spectroscopic data, depiction of
1H and 13C NMR spectra for all new compounds (PDF)
C
DOI: 10.1021/acs.orglett.5b02277
Org. Lett. XXXX, XXX, XXX−XXX