Intramolecular Acylal Cyclisation (IAC) as an Efficient Synthetic Strategy towards the Total Synthesis of Erythrina Alkaloid Derivatives

Article abstract of DOI:10.1002/chem.201502436

Compounds that comprise the erythrina alkaloid class of natural products are based on a tetracyclic spiroamine framework and exhibit a range of biological activities on the central nervous system. Herein, we report a new and efficient total synthesis of this multiple-ring system based on an intramolecular acylal cyclisation (IAC) approach. Using this methodology, the tetracyclic core was rapidly assembled over a two-step domino process catalysed by a Lewis acid. The effect of heteroatoms, substituents and ring size on the IAC has also been investigated, and the broad application of this procedure is demonstrated by the synthesis of a library of derivatives in good yields with excellent regioselectivity. Extending the diversity of the erythrina core: The tetracyclic core of the erythrina alkaloid family can be accessed through a Lewis acid mediated intramolecular acylal cyclisation (IAC) approach in two steps. Using this, heteroatoms and variation of the cyclohexyl ring produced a library of derivatives in good yields and with excellent regioselectivity.

Full text of DOI:10.1002/chem.201502436

DOI: 10.1002/chem.201502436
Communication
&
Natural Products
Intramolecular Acylal Cyclisation (IAC) as an Efficient Synthetic
Strategy towards the Total Synthesis of Erythrina Alkaloid
Derivatives
Alessandra Monaco,[a] Abil E. Aliev,[b] and Stephen T. Hilton*[a]
Abstract: Compounds that comprise the erythrina alkaloid
class of natural products are based on a tetracyclic spiro-
amine framework and exhibit a range of biological activ-
ities on the central nervous system. Herein, we report
a new and efficient total synthesis of this multiple-ring
system based on an intramolecular acylal cyclisation (IAC)
approach. Using this methodology, the tetracyclic core
was rapidly assembled over a two-step domino process
catalysed by a Lewis acid. The effect of heteroatoms, sub-
stituents and ring size on the IAC has also been investigat-
ed, and the broad application of this procedure is demon-
strated by the synthesis of a library of derivatives in good
yields with excellent regioselectivity.
Figure 1. Classification of erythrina alkaloids depending on D ring (aromatic,
heteroaromatic and unsaturated lactone types) and on the number of olefin-
ic bonds (dienoid and alkenoid types).
The erythrina alkaloid family is a structurally diverse class of
biologically active tetracyclic natural products that have been
isolated from a number of tropical plant sources (Figure 1).[1]
Many members of this family of compound display a potent
effect on the central nervous system (CNS) and, as such, have
been used in traditional medicine for their anxiolytic, anticon-
vulsant, sedative, antidepressive and antiepileptic effects.[2] The
hydroalcoholic extract of Erythrina mulungu stem bark prod-
uces a nonopioid-like analgesic effect,[3] whereas neuroethol-
ogical and neurochemical experiments have demonstrated
that extracts of the flowers of E. mulungu produce an anxiolytic
effect.[4] Several studies have also reported that oral administra-
tion of extracts (3, 10, 50, 100 and 200 mgkgÀ1) produced
anxiolytic effects in patients, which was analogous to the ef-
fects of diazepam.[5]
structural features, the synthesis of the erythrina alkaloid core
has attracted significant attention over a number of years
through a variety of approaches, which have included radical-
and Pummerer-mediated syntheses, intramolecular condensa-
tion and Diels–Alder reactions, to name but a few.[6,7]
Nearly all methods focus on one of two routes to generate
the three aliphatic rings, either through a single cyclisation, or
through a tandem cyclisation approach.[7,8] However, despite
the reported potent biological activities, there has been a sur-
prising lack of reports on structural variation of the tetracyclic
core. Herein, we now wish to report on our novel approach to-
wards the erythrina core and related structural derivatives.
Following our recent reports on the reactivity and use of
acylals, we reasoned that the erythrina core could be obtained
by a Lewis acid mediated intramolecular acylal cyclisation (IAC)
to generate rings B and C in a two-step domino process trig-
gered by acylal activation as outlined below (Scheme 1).
All members of the erythrina family possess a distinctive
tetracyclic spiroamine core and can be classified by variations
of the D ring, into three subclasses with aromatic, hetero-
aromatic or unsaturated lactone types as shown (Figure 1).[6]
As a result of their potent biological activity and challenging
We envisaged that the cyclisation precursor could be ob-
tained through condensation between a primary amine, cyclo-
hexanone and diacetoxyacetyl chloride (Scheme 1). Under
Lewis acid mediated IAC conditions, the enamine would cyclise
onto the oxonium species 11 to generate the intermediate imi-
nium ion 12, which would react with the aromatic ring to form
the tetracyclic erythrina core 13. Simple variation of the ketone
would therefore provide ready access to a range of analogues
in this manner.[9] We decided to use BF3 as the Lewis acid in
[a] Dr. A. Monaco, Dr. S. T. Hilton
UCL School of Pharmacy, University College London
29–39 Brunswick Square, London, WC1N 1AX (UK)
E-mail: s.hilton@ucl.ac.uk
[b] Dr. A. E. Aliev
Department of Chemistry, University College London
20 Gordon Street, London, WC1H 0AJ (UK)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201502436.
Chem. Eur. J. 2015, 21, 13909 – 13912
13909
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Figure 2. Determination of the relative configuration of the cyclised tetra-
cyclic core structure 18a showing trans configuration around the pyrrolidi-
none ring.
Scheme 1. Cyclisation to generate the tetracyclic core 18.
our cyclisations, due to its mild nature and our previous experi-
ence with its reactivity towards diacetoxyamides.[11–14]
confirmed these results where the two hydrogen atoms are
trans to each other, with a large J coupling of 10.0 Hz
(Figure 2).
To investigate our hypothesis, we carried out initial conden-
sation of cyclohexanone and 2-(3,4-dimethoxyphenyl)ethan-
amine 14 under Dean and Stark water-removal conditions.[10]
The intermediate imine was reacted immediately with 2-
chloro-2-oxoethane-1,1-diyl diacetate[11] 16 in the presence of
pyridine to afford the key cyclisation precursor 17a in 63%
yield. Pleasingly, on reaction with BF3·OEt2 (5 equiv) and micro-
wave heating of the reaction mixture at 658C for 15 min the
tetracyclic core 18a was obtained in good yield (61%) and as
a single diastereoisomer (Scheme 2).
Following the success in the formation of the erythrina core,
we turned our attention towards formation of a range of close-
ly related analogues through simple variation of the initial
ketone component in order to explore the scope of the reac-
tion by variation of ring size and incorporation of heteroatom
substituents. The desired cyclisation precursors 17b–17j were
prepared in good yields (17–67%) analogous to the formation
of the erythrina core as shown below (Table 1).
Compounds 17c and 17d were selected to explore the
effect of ring size on diastereoselectivity, whilst the various het-
eroatom and substituted cyclohexyl rings were chosen to ex-
plore the scope of the reaction for the formation of potentially
biologically active moieties. All compounds gave good yields
of the cyclisation precursors, with the exception of the tri-
fluoromethyl analogue 17j, which was a result of the volatility
of the starting ketone. Having the cyclisation precursors in
hand we focused on their cyclisation to generate erythrina
analogues, the results of which are shown below (Table 2).
All reactions were heated in the microwave at 658C for
15 min as per the initial reaction with the cyclohexenyl cyclisa-
tion precursor 17a. Following purification, it can be readily ob-
served that the cyclised products were obtained in good to ex-
cellent yield with good control over diastereoselectivity in
most cases. Examination of the results for both the cyclopentyl
18c and cycloheptyl 18d erythrina analogues demonstrated
that the calculations of their respective minimised energies are
an effective predictor of the outcome of the diastereoselectiv-
ity of cyclisation. In the case of the cyclopentyl analogue 18c,
the trans diastereoisomer is 3.9 kcalmolÀ1 more stable than the
corresponding cis diastereoisomer and, as such, the ratio of
trans/cis is 85/15. In the case of the cycloheptyl product 18d,
the trans diastereoisomer is 0.4 kcalmolÀ1 more stable than the
corresponding cis diastereoisomer and, as such, the ratio de-
teriorates to give a trans/cis ratio of 60/40, clearly highlighting
that the effect of reducing or increasing ring size leads to
a lowering of diastereoselectivity. Incorporation of heteroatom
six-membered analogues of the erythrina core was well-tolerat-
ed with compounds 18b, 18g and 18h all obtained in good
yields (30–61%) and as single diastereoisomers. Substitution of
Scheme 2. Mechanism of cyclisation to generate the tetracyclic core 18a.
Although it could be anticipated that out of the four poten-
tial diastereoisomers, one would be predominant, we were sur-
prised by obtention of a single diastereoisomer in this in-
stance. In order to understand this perhaps surprising selectiv-
ity, the relative configuration of the three contiguous chiral
centres was determined by NMR J coupling and NOE measure-
ments of the single diastereoisomer 18a. In combination with
computational analysis of the four potential diastereoiso-
mers,[15] the results from energy minimisation calculations and
the values from the NMR parameters demonstrated that the
hydrogen atoms at C2 and C3 of the pyrrolidine ring would be
configured in a trans relationship as this diastereoisomer is ap-
proximately 5.9 kcalmolÀ1 more stable than its closest related
congener (see Supporting Information for full details of NMR
and computational analysis). The results from the NOE studies
Chem. Eur. J. 2015, 21, 13909 – 13912
www.chemeurj.org
13910
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Table 1. Formation of erythrina cyclisation precursor analogues.
Table 2. Formation of erythrina tetracyclic analogues.
the cyclohexyl ring in the 4-position was also well-tolerated
with a phenyl group leading to a single diastereoisomer 18e
in reasonable yield (40%), whereas the methyl-substituted ana-
logue 18i and the trifluoromethyl analogue 18j afforded rea-
sonable yields of the cyclised products. In the case of 18j, the
trifluoromethyl group led to a 70/30 mixture of diastereoiso-
mers at the trifluoromethyl position. As per previous six-mem-
bered ring-containing examples, the diastereoselectivity at the
pyrrolidinyl ring was not affected. Pleasingly, we were also able
to generate pentacyclic derivatives of the erythrina core with
compound 18 f obtained as a single diastereoisomer in 48%
yield and no evidence of any regioisomers.
a range of substituted analogues with good stereocontrol. Our
approach is readily adaptable to incorporation of heteroatom
substituents and related pentacyclic derivatives. Further stud-
ies on variation of the D ring of the erythrina core are under
way in our laboratories and will be reported in due course.
Acknowledgements
This work was gratefully supported by an FNS visiting postdoc-
toral fellowship grant to A.M. [P2GEP2_151840]. We thank the
EPSRC UK National Mass Spectrometry Facility at Swansea Uni-
versity for spectroscopic services.
In summary, we have developed a novel two-step route to
the erythrina core, which is based on an intramolecular acylal
cyclisation (IAC) approach. In addition, we have shown that, by
simple variation of the starting ketone, we can generate
Keywords: acylal · erythrina alkaloids · intramolecular acylal
cyclisation · Lewis acid · tandem cyclisation
Chem. Eur. J. 2015, 21, 13909 – 13912
www.chemeurj.org
13911
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
[1] a) U. P. De Albuquerque, P. M. De Medeiros, A. L. S. DeAlmeida, J. M.
Monteiro, E. M. D. F. L. Neto, J. G. De Melo, J. P. Dos Santos, J. Ethnophar-
macol. 2007, 114, 325–354; b) Y. Tsuda, T. Sano in The Alkaloids, Vol 48
(Ed.: G. A. Cordell), Academic Press, New York, 1996, pp. 249–337.
[2] a) A. P. S. Balbani, D. H. S. Silva, J. C. Montovani, Expert Opin. Ther. Pat.
2009, 19, 461–473; b) M. E. Garín-Aguilar, J. E. R. Luna, M. Soto-Hernµn-
dez, G. V. del Toro, M. M. Vµzquez, J. Ethnopharmacol. 2000, 69, 189–
196; c) D. Santos Rosa, S. A. Faggion, A. S. Gavin, M. A. de Souza, H. A.
Fachim, W. F. Dos Santos, A. M. S. Pereira, A. O. S. Cunha, R. O. Beleboni,
Epilepsy Behav. 2012, 23, 205–212.
b) P. C. Stanislawski, A. C. Willis, M. G. Banwell, Org. Lett. 2006, 8, 2143–
2146; c) G. Kim, J. H. Kim, K. Y. Lee, J. Org. Chem. 2006, 71, 2185–2187;
d) Y. Yasui, K. Suzuki, T. Matsumoto, Synlett 2004, 619–622; e) D. Kalait-
zakis, T. Montagnon, E. Antonatou, G. Vassilikogiannakis, Org. Lett. 2013,
15, 3714–3717; f) S. Ogawa, N. Iida, E. Tokunaga, M. Shiro, N. Shibata,
Chem. Eur. J. 2010, 16, 7090–7095; g) L. F. Tietze, N. Tolle, D. Kratzert, D.
Stalke, Org. Lett. 2009, 11, 5230–5233; h) F. Z. Zhang, N. S. Simpkins,
A. J. Blake, Org. Biomol. Chem. 2009, 7, 1963–1979; i) A. Padwa, Q.
Wang, J. Org. Chem. 2006, 71, 7391–7402; j) H. Ishibashi, K. Sato, M.
Ikeda, H. Maeda, S. Akai, Y. Tamura, J. Chem. Soc. Perkin Trans. 1 1985,
605–609.
[3] G. M. Onusic, R. L. Nogueira, A. M. S. Pereira, M. B. Viana, Braz. J. Med.
Biol. Res. 2002, 35, 473–477.
[8] Y.-M. Zhao, P. Gu, Y.-Q. Tu, H.-J. Zhang, Q.-W. Zhang, C.-A. Fan, J. Org.
Chem. 2010, 75, 5289–5295.
[9] a) C. Bonauer, T. Walenzyk, B. Kçnig, Synthesis 2006, 1–20; b) S. P. Lu,
A. H. Lewin, Tetrahedron 1998, 54, 15097–15104.
[10] H. I. Lee, M. P. Cassidy, P. Rashatasakhon, A. Padwa, Org. Lett. 2003, 5,
5067–5070.
[11] A. E. Aliev, S. T. Hilton, W. B. Motherwell, D. L. Selwood, Tetrahedron Lett.
2006, 47, 2387–2390.
[12] B. R. Szulc, A. Ruiz, S. T. Hilton, 2014, WO2014181101A1.
[13] B. C. Sil, S. T. Hilton, Synlett 2013, 24, 2563–2566.
[14] K. M. Cook, S. T. Hilton, J. Mecinovic, W. B. Motherwell, W. D. Figg, C. J.
Schofield, J. Biol. Chem. 2009, 284, 26831–26838.
[15] a) A. E. Aliev, Z. A. Mia, H. S. Khaneja, F. D. King, J. Phys. Chem. A 2012,
116, 1093–1109; b) A. E. Aliev, Z. A. Mia, M. J. M. Busson, J. Org. Chem.
2012, 77, 6290–6295.
[4] S. M. M. Vasconcelos, N. M. Lima, G. T. M. Sales, G. M. A. Cunha, L. M. V.
Aguiar, E. R. Silveira, A. C. P. Rodrigues, D. S. Macedo, M. M. F. Fonteles,
F. C. F. Sousa, G. S. B. Viana, J. Ethnopharmacol. 2007, 110, 271–274.
[5] a) J. O. A. Flausino, Jr., A. M. Pereira, V. D. S. Bolzani, R. L. Nunes-de-
Souza, Biol. Pharm. Bull. 2007, 30, 375–378; b) P. Setti-Perdig¼o, M. A. R.
Serrano, O. A. Flausino, V. S. Bolzani, M. Z. P. Guimar¼es, N. G. Castro,
PloS ONE 2013, 8, e82726; c) J. Sarris, E. McIntyre, D. Camfield, CNS
Drugs 2013, 27, 207–219.
[6] For recent examples see: a) A. Padwa, H. I. Lee, P. Rashatasakhon, M.
Rose, J. Org. Chem. 2004, 69, 8209–8218; b) M. A. Le Dreau, D. Des-
maele, F. Dumas, J. d’Angelo, J. Org. Chem. 1993, 58, 2933–2935; c) C.
L’Homme, M. MØnard, J. Org. Chem. 2014, 79, 8481–8485; d) J. M. Joo,
R. A. David, Y. Yuan, C. Lee, Org. Lett. 2010, 12, 5704–5707; e) J. X.
Liang, J. B. Chen, J. P. Liu, L. Li, H. B. Zhang, Chem. Commun. 2010, 46,
3666–3668; f) T. Onoda, Y. Takikawa, T. Fujimoto, Y. Yasui, K. Suzuki, T.
Matsumoto, Synlett 2009, 1041–1046; g) Y. Yoshida, K. Mohri, K. Isobe,
T. Itoh, K. Yamamoto, J. Org. Chem. 2009, 74, 6010–6015.
Received: June 23, 2015
[7] For examples, see: a) B. Juma, M. Adeel, A. Villinger, H. Reinke, A. Span-
nenberg, C. Fischer, P. Langer, Adv. Synth. Catal. 2009, 351, 1073–1079;
Published online on August 18, 2015
Chem. Eur. J. 2015, 21, 13909 – 13912
www.chemeurj.org
13912
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim