Cis-decahydroquinolines via asymmetric organocatalysis: Application to the total synthesis of lycoposerramine Z

Article abstract of DOI:10.1021/ol303257y

A concise synthesis of the Lycopodium alkaloid lycoposerramine Z is reported. Key to the strategy is a one-pot organocatalyzed Michael reaction followed by a domino Robinson annulation/intramolecular aza-Michael reaction promoted by LiOH, leading to enantiopure cis-decahydroquinolines.

Full text of DOI:10.1021/ol303257y

ORGANIC
LETTERS
2013
Vol. 15, No. 2
326–329
cis-Decahydroquinolines via Asymmetric
Organocatalysis: Application to the Total
Synthesis of Lycoposerramine Z
Ben Bradshaw,* Carlos Luque-Corredera, and Josep Bonjoch*
´
Laboratori de Quımica Organica, Facultat de Farmacia, IBUB,
Universitat de Barcelona, Av. Joan XXIII, s/n, 08028 Barcelona, Spain
ꢀ
ꢁ
benbradshaw@ub.edu; josep.bonjoch@ub.edu
Received November 27, 2012
ABSTRACT
A concise synthesis of the Lycopodium alkaloid lycoposerramine Z is reported. Key to the strategy is a one-pot organocatalyzed Michael reaction
followed by a domino Robinson annulation/intramolecular aza-Michael reaction promoted by LiOH, leading to enantiopure cis-decahydroquinolines.
Lycoposerramine Z (1), isolated by Takayama in 2006,1
belongs to a small class of phlegmarine-type Lycopodium
alkaloids (Figure 1) that serve as biogenetic intermediates
between pelletierine and all other alkaloids of this complex
family with around 250 congeners.2 Interestingly, 1 incor-
porates an unusual nitrone moiety,3 which has been pos-
tulated to act as a radical trap, halting destructive cascades
initiated by free radicals, and hence has potential applica-
tion in neurodegenerative diseases.4 In 2009, starting from
the chiral pool, Takayama’s group completed the first total
synthesis of 1,5 which remains to date the only synthesis of
(1) Katakawa, K.; Kitajima, M.; Yamaguchi, K.; Takayama, H.
Heterocycles 2006, 69, 223–229.
(2) (a) Ma, X.; Gang, D. R. Nat. Prod. Rep. 2004, 21, 752–772. (b)
Hirasawa, Y.; Kobayashi, J.; Morita, H. Heterocycles 2009, 77, 679–729.
(3) The nitrone group was later found in related compounds: Gao,
W.; Li, Y.; Jiang, S.; Zhu, D. Helv. Chim. Acta 2008, 91, 1031–1035.
(4) Sun, Y.; Yu, P.; Zhang, G.; Wang, L.; Zhong, H.; Zhai, Z.; Wang,
L.; Wang, Y. J. Neurosci. Res. 2012, 90, 1667–1669.
(5) Tanaka, T.; Kogure, N.; Kitajima, M.; Takayama, H. J. Org.
Figure 1. Phlegmarine-type Lycopodium alkaloids.
Chem. 2009, 74, 8675–8680.
(6) For the synthesis of related phlegmarines embodying the most
usual trans-ring fusionin the decahydroquinoline ring, see: (a) Leniewski,
a phlegmarine alkaloid with a cis-ring fused decahydro-
quinoline unit.6
We herein report a concise, enantioselective synthesis of
lycoposerramine Z (1). The cornerstone of our synthetic
A.; Szychowski, J.; MacLean, D. B. Can. J. Chem. 1981, 59, 2479–
2490. (b) Comins, D. L.; Libby, A. H.; Al-awar, R.; Foti, C. J. J. Org.
Chem. 1999, 64, 2184–2185. (c) Wolfe, B. H.; Libby, A. H.; Al-awar, R.;
Foti, C. J.; Comins, D. L. J. Org. Chem. 2010, 75, 8564–8570. (d)
Reference 5.
r
10.1021/ol303257y
Published on Web 12/27/2012
2012 American Chemical Society

were first examined in a nonasymmetric version. After
considerable experimention12 we found that the decahy-
droquinoline ring could be generated in a straighforward
Scheme 1. Retrosynthesis of Lycoposerramine Z
one-pot reaction using LiOH H2O (1 equiv) in iPrOH13 in
3
the presence of water (10 equiv).14 We were delighted to
find that under these conditions decahydroquinoline rac-8
was delivered in only one step and as a single diastereo-
isomer. As predicted, the retention of the ester group
stabilized the compound by forming the enolic tautomer,
which effectively acts as a locking group, preventing
the ring opening by a retro aza-Michael reaction in the
basic reaction medium or in the purification step (silica gel
chromatography).
Scheme 2. Synthesis of rac-8
approach is a rapid asymmetric assembly of the azabicyclic
core by a novel organocatalyzed diastereo- and enantio-
selective one-pot tandem synthesis of decahydroquino-
lines. In this process two CꢀC bonds and one CꢀN bond
(Scheme 1) and three stereogenic centers are created in a
single step.7,8
As outlined in Scheme 1, we envisaged that the piper-
idine appendage could be introduced by the coupling of
a methylpyridine 2 with a 5-oxodecahydroquinoline. The
β-keto ester 3 precursor of the latter would be formed via
a Robinson annulation of the simple acyclic keto ester 5
through an initial organocatalyzed Michael addition, fol-
lowed by an in situ intramolecular aza-Michael reaction
of the generated cyclohexenone 4. We speculated that the
retention of the ester group in 3 would be essential for the
success of our strategy, helping to stabilize the β-amino
ketone moiety and prevent side reactions.9,10
The synthesis of 1 began with the N-tosylation of the
commercially available 5-aminopentanoic acid, and the
resulting acid 6 was subjected to a homologation with
mono-tert-butylmalonate under Masamune-type condi-
tions11 to give 7 (77% over two steps, Scheme 2). The
reaction conditions for the key Robinson annulation/
intramolecular aza-Michael biscyclization process to
achieve decahydroquinoline 8 from 7 and crotonaldehyde
To carry out the reaction in asymmetric form, the Hayashi
catalyst15 was chosen to promote the initial organocatalyzed
Michael addition, after which the tandem cyclization con-
ditions (LiOH) were applied.16,17 A screening of solvents
indicated that toluene was the optimal choice. Several other
catalysts,18 mainly diaryl-prolinol silyl ethers,19 were then
screened, and the slight superiority of triphenylsilyl deriva-
tive 920 led to its selection (Table 1, entry 1).
The reaction was further refined by lowering the tem-
perature (entries 2 and 3) and the use of additives was
(12) For instance, t-BuOK in t-BuOH was not sufficiently effective,
despite its frequent use in Robinson reactions: Chong, B.; Ji, Y.; Oh, S.;
Yang, J.; Baik, W.; Koo, S. J. Org. Chem. 1997, 62, 9323–9325.
(13) These reaction conditions were adapted from those used by
Baran (LiOH (0.1 equiv), iPrOH, rt, 24 h) in the synthesis of cryptone:
(7) For organocatalytic cascade reactions in total synthesis, see: (a)
Grondal, C.; Jeanty, M.; Enders, D. Nat. Chem. 2010, 2, 167–178. (b)
Jones, S. B.; Simmons, B.; Mastracchio, A.; MacMillan, D. W. C.
Nature 2011, 475, 183–188. (c) Lu, L.-Q.; Chen, J.-R.; Xiao, W.-J.
Acc. Chem. Res. 2012, 45, 1278–1293.
ꢀ
Chen, K.; Ishihara, Y.; Galan, M. M.; Baran, P. S. Tetrahedron 2010, 66,
4738–4744. It should be noted that these previously published conditions
were unsuccessful when applied to 3 and crotonaldehyde.
(14) Water was necessary to drive the aza-Michael reaction to
completion. In its absence significant amounts of the ring-opened
Robinson annulation product were obtained (20ꢀ30%).
(15) Hayashi, Y.; Gotoh, H.; Hayashi, T.; Shoji, M. Angew. Chem.,
Int. Ed. 2005, 44, 4212–4215.
(8) For recent enantioselective synthesis of cis-decahydroquinolines,
see: (a) Ito, T.; Overman, L. E.; Wang, J. J. Am. Chem. Soc. 2010, 132,
3272–3273. (b) Kagugawa, K.; Nemoto, T.; Kohno, Y.; Yamada, Y.
Synthesis 2011, 2540–2548. (c) Amat, M.; Navio, L.; Llor, N.; Molins,
€
E.; Bosch, J. Org. Lett. 2012, 14, 210–213. (d) Gartner, M.; Qu, J.;
Helmchen, G. J. Org. Chem. 2012, 77, 1186–1190.
(16) The classical procedure to achieve cyclohexenones developed by
Jorgensen using an initial organocatalyzed Michael addition followed
by treatment with TsOH was unsuccessful when starting from 3:
Carlone, A.; Marigo, M.; North, C.; Landa, A.; Jørgensen, K. A. Chem.
Commun. 2006, 4928–4930.
(17) For an organocatalytic initial Michael reaction and Robinson
annulation that retains the ester group, see: Marigo, M.; Bertelsen, S.;
Landa, A.; Jørgensen, K. A. J. Am. Chem. Soc. 2006, 128, 5475–5479.
(18) The results of the initial screening of solvents and catalysts are
summarized in the Supporting Information.
(9) We have observed that direct azacyclization upon an enone to
give 5-oxodecahydroquinolines is troublesome since the cyclized com-
pound is in equilibrium with the ring-opened R,β-unsaturated ketone:
Borregan, M. Ph.D. Thesis, University of Barcelona, Spain, 2009.
(10) No intramolecular aza-Michael process leading to 5-oxodeca-
hydroquinolines has been described so far. For interesting results related
to this field, see: (a) Brosius, A. D.; Overman, L. E. J. Org. Chem. 1997,
62, 440–441. (b) Taber, D. F.; Joshi, P. V.; Kanai, K. J. Org. Chem. 2004,
69, 2268–2271.
(11) (a) Brooks, D. W.; Lu, L. D.-L.; Masamune, S. Angew. Chem.,
Int. Ed. Engl. 1979, 18, 72–73. (b) Hodgson, D. M.; Labande, A. L.;
Pierard, Y. T. M.; Castro, M. A. E. J. Org. Chem. 2003, 68, 6153–6159.
(19) Jensen, K. L.; Dickmeiss, G.; Jiang, H.; Albrecht, L.; Jørgensen,
K. A. Acc. Chem. Res. 2012, 45, 248–264.
Org. Lett., Vol. 15, No. 2, 2013
327

Withenantiopure decahydroquinoline (þ)-8 in hand, we
set about converting it into the natural product lycopo-
serramine Z. Removal of the tert-butyl ester locking
group with TFA gave ketoacid 10, which upon azeotropi-
cal removal of TFA with toluene by heating underwent
decarboxylation to ketone 11 (Scheme 3). The material
was used immediately in the next step without any
purification23 and added directly to a solution of the
lithium anion of phosphonate 1224 to give vinylpyridine
derivatives 13 in excellent yield (91%). A mixture of Z/E
isomers (∼1:4.2) was observed,25,26 which were separated
by chromatography. However, this turned out to be in-
consequential since hydrogenation of the mixture or each
isolated isomer gave the same all-cis-product 14 in which
the hydrogen wasdeliveredfrom the topface(seeScheme 4,
Table 1. Screening Conditions for the Organocatalyzed
Robinson Annulation/aza-Michael Reactiona
entry
additive (equiv)
none
temp (°C)
yield (%)
ee (%)
1
which depicts the major isomer 13a). Analysis of the H
1
2
3
4
5
6
7
8
9
rt
0
57
57
49
69
72
58
43
63
66
80
82
82
84
85
83
80
80
82
and 13C NMR spectra of both Z/E diastereomers of the
starting material (13a and 13b) allowed us to establish that
the preferred conformation of these compounds accom-
modates axial positioning of the methyl group, which
avoids the allylic 1,3-strain27 both for the exocyclic double
bond (with the C(4)ꢀC(4a) bond) and the N-tosyl group
(with the C(8)ꢀC(8a) bond).28 This preferred conforma-
tion is involved in steric interactions that influence the
diastereoselectivity of the hydrogenation process. The
reduction led exclusively to decahydroquinoline 14, in
which the substituents at C(5) and C(7) are axially located
according to NMR data. Thus, the crucial role of the axial
methyl group in the process was clearly established, as it
sterically impedes hydrogenation from the bottom face.
With all four introduced stereogenic centers now in
place, the tosyl group was replaced by a Teoc group prior
tothe installation of the sensitivenitronemoiety.29,30 Thus,
the tosyl group was removed (HBr, phenol) to give 15,
which was converted into the Teoc carbamate 16 (Scheme 5).
Reduction of the pyridine ring with PtO2/AcOH gave
the piperidine 17 as an inconsequential mixture of epimers
none
none
ꢀ20
0
H2O (10)
LiOAc (0.5)
KOAc (0.5)
BzOH (0.5)
H2O (10)/LiOAc (0.5)
LiOAc (1)
0
0
0
0
0
a 20% catalyst loading of 5 was used in a 0.1 M solution of 3. Toluene
was evaporated before proceeding with the tandem biscyclization.
investigated (entries 4ꢀ10). The addition of LiOAc21
(entry 5) was essential for obtaining a good conversion
yield (72% for three bond-forming reactions) and enan-
tiomeric ratio (>92:8). It is worth mentioning that the
combination of LiOAc and another good additive, water
(entry 4), were not synergistic, giving results (entry 8)
inferior to those with their individual use.
In general, in comparison with bulkier groups, the small
size of the methyl group sets a limit to the maximum
enantioselectivity (in our case 84ꢀ85% ee) in organocata-
lyzed Michael additions.16 However, wewere able toaccess
(þ)-8 in enantiopure form by recrystallization of 8 from
MeOH,22 which provided the product in >99% ee (first
crop, 65% recovery).
(20) For the first reported use of this catalyst, see: Wang, Y.; Li, P.;
Liang, X.; Ye, J. Adv. Synth. Catal. 2008, 350, 1383–1389. For its
preparation and subsequent application, see: Gomez-Bengoa, E.;
Landa, A.; Lizarraga, A.; Mielgo, A.; Oiarbide, M.; Palomo, C. Chem.
Sci 2011, 2, 353–357.
(21) For an interesting study on the effect of LiOAc in organocata-
lyzed Michael reactions, see: Duce, S.; Mateo, A.; Alonso, I.; Garcıa
Ruano, J. L.; Cid, M. B. Chem. Commun. 2012, 48, 5184–5186.
(22) The use of EtOH gave a certain amount of transesterification
products. For related phenomena, see: Witzeman, J. S.; Nottingham,
W. D. J. Org. Chem. 1991, 56, 1713–1718.
Scheme 3. Coupling of the Methylpyridine Fragment
(23) Fortunately, under the reaction conditions the product did not
undergo a retro aza-Michael reaction. However, it should be noted that
upon prolonged manipulation (e.g., silica gel chromatography, etc.) the
ring-opened product began to be observed.
(24) Gan, X.; Binyamin, I.; Rapko, B. M.; Fox, J.; Duesler, E. N.;
Paine, R. T. Inorg. Chem. 2004, 43, 2443–2448.
(25) Using the lithium anion of 2-(trimethylsilyl)methylpyridine (ref
6a) in a Peterson reaction, compounds 13 were isolated in the same ratio
but with a slightly lower yield (80%).
(26) Interestingly, this diastereoselectivity is the reverse of that ob-
served in the related 2,5-dioxodecahydroquinoline system: see ref 6a.
(27) Hoffmann, R. W. Chem. Rev. 1989, 89, 1841–1873.
(28) Booth, H.; Bostock, A. H. J. Chem. Soc., Perkin Trans 2 1972,
615–621.
(29) This protecting group was used by Takayama in his synthesis of
1, where it was shown that it could be removed in the presence of the
sensitive nitrone moiety.
328
Org. Lett., Vol. 15, No. 2, 2013

Scheme 4. Diasteroselective Reduction of Alkene 13
Scheme 5. Completion of the synthesis of lycoposerramine Z
(not shown). This was oxidized with Na2WO4/urea
peroxide31 to give nitrone 18, which was identical to the final
intermediate in Takayama’s synthesis.5 Although the depro-
tection of 18 to lycoposerramine Z has already been reported,
for the sake of a complete enantioselective total synthesis
of 1, the Teoc group was removed using TFA.32,33 The
resulting (þ)-lycoposerramine Z showed identical NMR
spectroscopic data to those reported for the natural product,1
as well as the same specific rotation value as synthetic 1.5
In summary, we have described a short synthesis of
lycoposerramine Z, which was completed in only 10 steps
(20% overall yield)34 from commercially available 5-ami-
nopentanoic acid. Key to the success of the synthesis was
a one-pot tandem organocatalyzed formation of decahy-
droquinolines, which allowed a rapid enantio- and diastero-
selective assembly of the alkaloid core structure. We
believe that the methodology presented here not only
has the potential to provide access to numerous other
Lycopodium alkaloids but also should prove applicable
in the asymmetric synthesis of other important nitrogen-
containing heterocyclic structures. Research in this direc-
tion is now underway.
(30) Attempts to avoid the exchange of the protecting groups via a
biomimetic chemoselective oxidation of diamine 20 were unsuccessful.
Diamine 20 was prepared via a one-pot hydrogenation of the vinylpyr-
idine moiety of 13 to give piperidine 19, from which the tosyl group was
removed by treatment with LiAlH4 (88% over two steps). Unfortu-
nately, we were unable to chemoselectively oxidize the latter with
Na2MoO431 (0.5ꢀ1 equiv) to directly deliver lycoposerramine Z (1).
Acknowledgment. Support of this research was pro-
vided by the Spanish Ministry of Economy and Competi-
tiveness (Project CTQ2010-14846/BQU).
Supporting Information Available. Experimental pro-
cedures, spectroscopic and analytical data, and copies
of NMR spectra of the new compounds. This mate-
rial is available free of charge via the Internet at http://
pubs.acs.org.
(31) Marcantoni, E.; Petrini, M.; Polimanti, O. Tetrahedron Lett.
1995, 36, 3561–3562. See also: Ohtake, H.; Imada, Y.; Murahashi, S.
Bull. Chem. Soc. Jpn. 1999, 72, 2737–2754.
(32) (a) For deprotection of N-Teoc carbamates with TFA, see:
Carpino, L. A.; Tsao, J.-H.; Ringsdorf, H.; Fell, E.; Hettrich, G. J.
Chem. Soc. Chem. Comm 1978, 358–350. (b) For the stability of a nitrone
moiety to TFA, see: Medina, S. I.; Wu, J.; Bode, J. W. Org. Biomol.
Chem. 2010, 8, 3405–3417.
(34) This compares favorably with the 24 steps reported in the first
synthesis of lycoposerramine Z (ref 5).
(33) This protocol greatly simplified the isolation procedure, avoid-
ing the use of TASF or TBAF for the N-Teoc cleavage in 18.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 2, 2013
329