PdII-Mediated Oxidative Amination for Access to a 9-Azabicyclo[4.2.1]nonane Compound Library and Anatoxin-a

Article abstract of DOI:10.1002/ejoc.201801209

Intramolecular oxidative amination of readily accessible aminocyclooct-4-enes provides rapid and regioselective access to the 9-azabicyclo[4.2.1]nonane framework characteristic of the natural product anatoxin-a (1). This has enabled the synthesis of sp³-rich chemical scaffolds suitable for diversification. A library of 881 lead-like compounds is reported alongside a formal synthesis of anatoxin-a (1).

Full text of DOI:10.1002/ejoc.201801209

DOI: 10.1002/ejoc.201801209
Communication
Chemical Scaffolds
PdII-Mediated Oxidative Amination for Access to a
9-Azabicyclo[4.2.1]nonane Compound Library and Anatoxin-a
Rafid S. Dawood,[a,b] Suresh R. Chidipudi,[a] Daniel C. O'Connor,[a] William Lewis,[a]
Daniel Hamza,[c] Christopher A. Pearce,[c] Geraint Jones,[c] Ross P. Wilkie,[a] Irene Georgiou,[a]
Thomas E. Storr,[a] Jonathan C. Moore,*[a] and Robert A. Stockman*[a]
Abstract: Intramolecular oxidative amination of readily accessi- thesis of sp3-rich chemical scaffolds suitable for diversification.
ble aminocyclooct-4-enes provides rapid and regioselective ac-
cess to the 9-azabicyclo[4.2.1]nonane framework characteristic
of the natural product anatoxin-a (1). This has enabled the syn-
A library of 881 lead-like compounds is reported alongside a
formal synthesis of anatoxin-a (1).
Introduction
The structure of anatoxin-a (1) was unambiguously determined
in 1977 through the combination of X-ray crystallographic[1]
and spectroscopic[2] analysis. This potent neurotoxin is pro-
duced by certain strains of the freshwater blue-green algae
Anabena flos-aquae and has been responsible for fatal poison-
ing of animal life.[3] Also known as “Very Fast Death Factor”,
anatoxin-a (1) causes death by respiratory paralysis with an LD50
(intraperitoneal, mouse) of 0.2 mg/kg, with a survival time of
4–7 minutes.[4] The natural compound acts as a mimic for
acetylcholine and is a potent and irreversible agonist for the
nicotinic acetylcholine receptor (nAChR).[5] As acetylcholine de-
ficiency is implicated in diseases such as Alzheimer's, analogues
of anatoxin-a (1) possessing lower levels of toxicity may have
potential in the treatment of neurological disorders.[6] Paired
with its significant biological properties, the compound's
unusual structure has prompted much synthetic interest; a
number of distinct approaches to anatoxin-a (1) and its ana-
logues have been devised.[7] Since the topic was reviewed in
2006,[8] there have been a considerable number of total/formal
syntheses reported including our own two-directional ap-
proach[9] and a contemporary example (2016) that utilises a
novel formal amide insertion strategy.[10] One commonly em-
ployed approach is to construct the 9-azabicyclo[4.2.1]nonane
framework through the ring closure of an aminocyclooct-4-ene
(Figure 1), however there are drawbacks to each of the methods
reported to date.
Figure 1. Ring closure of aminocyclooct-4-enes for the synthesis of anatoxin-
a (1) and chemical scaffolds for library generation.
An aminocarbonylation strategy was reported by Ham in
1998,[11] however this heterocyclisation was observed not to be
regioselective. A series of reports by Parsons[12] saw the devel-
opment of a selenium-mediated ring closure, however this ap-
proach requires stoichiometric quantities of toxic phenylselenyl
chloride. Ho[13] reported a catalytic palladium mediated cyclisa-
tion of a corresponding trans-derivative, although yields were
hampered by the formation of an undesired tricyclic side-prod-
uct.
Results and Discussions
As part of our studies into the synthesis of chemical scaffolds
for drug discovery,[14] we set out to develop a catalytic and
selective protocol for the heterocyclisation of aminocyclooct-4-
enes. It was envisaged that a regioselective and high-yielding
protocol would expediate entry to series of 9-azabicyclo-
[4.2.1]nonane derived chemical scaffolds that could be further
derivatised to generate a lead-like compound library. Small
molecule libraries are highly desired in early-stage probe- and
drug-discovery screening programmes[15] and as part of our
work under the European Lead Factory[16] we have been en-
gaged in the development of facile synthetic routes to chemical
scaffolds that are suitable for diversification into high Fsp3 (frac-
[a] School of Chemistry, University of Nottingham,
University Park, Nottingham, NG7 2RD, UK
E-mail: pczjm@nottingham.ac.uk
robert.stockman@nottingham.ac.uk
[b] Department of Chemistry, College of Science, University of Baghdad,
Al-Jadriya campus, Baghdad 10071, Iraq
[c] Sygnature Discovery Limited, BioCity,
Pennyfoot Street, Nottingham, NG7 1GF, UK
Supporting information and ORCIDs from the authors for this article are
available on the WWW under https://doi.org/10.1002/ejoc.201801209.
Eur. J. Org. Chem. 0000, 0–0
1
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
tion of sp3-hybridised carbon atoms)17compound libraries. For
these purposes we were drawn to the simplicity and concise
nature of the heterocyclisation approach to the 9-azabicyclo-
[4.2.1]nonane framework. We set out to investigate ring closure
of the key precursor 2 under classical oxidative amination con-
ditions with Pd(OAc)2, using Cu(OAc)2 as the terminal oxidant
(Table 1).
Table 1. Catalyst/solvent optimisation. Reactions were performed on
a
0.1 mmol scale. Yields were determined by quantitative 1H NMR spectroscopy
with 1,3,5-trimethoxybenzene as an internal standard.
Entry
[M] (mol-%)
Cu(OAc)2 [equiv.]
Solvent Yield
1
2
3
4
5
6
7
8
Pd(OAc)2 (5)
Pd(OAc)2 (5)
Pd(OAc)2 (5)
Pd(OAc)2 (5)
Pd(OAc)2 (2.5)
–
2.1
2.1
2.1
2.1
2.1
2.1
2.5
2.5
THF
30%
37%
95%
99%
43%
–
PhMe
MeCN
DMF
DMF
DMF
DMF
DMF
Figure 2. Investigation into the substrate scope of the transformation. Yields
are for isolated material obtained following column chromatography.
[RuCl2(p-cymene)]2 (2.5)
[RhCp*Cl2]2 (2.5)
–
–
ised derivative 10, which could likely serve as a useful interme-
diate for the synthesis of side chain-modified anatoxin-a (1) de-
rivatives. Performing alkylation prior to ring closing was also
demonstrated to be a viable strategy, with the methyl and
phenyl ketones 11 and 12 being isolated in 88% and 77% re-
spectively.
Moderate yields of the desired product 3 were obtained us-
ing THF and PhMe as solvents (30% and 37% respectively).
Changing to MeCN (Entry 3) gave an excellent conversion but
DMF was found to be the superior solvent for this transforma-
tion, providing access to the 9-azabicyclo[4.2.1]nonane ring
system with complete selectivity and quantitative conversion
(Entry 4). Reducing the palladium loading to 2.5 mol% had a
significantly detrimental effect on the yield (43%) and when the
reaction was performed in the absence of palladium (Entry 6)
or in the presence of RuII and RhIII complexes (Entries 7–8), no
cyclisation products could be detected. With the optimised con-
ditions in hand, the substrate scope of the transformation was
investigated (Figure 2).
Initially we attempted exchanging the N-tosyl functionality
for the more readily de-protected carboxybenzyl (Cbz) group.
In this case, the desired product 4 was obtained in a moderate
yield of 39%. Benzoyl protection was considerably better toler-
ated, yielding the desired product 5 in a much improved 75%.
The electron rich N-benzyl substrate was observed to react with
poor selectivity and the targeted 9-azabicyclo[4.2.1]nonane 7
could not be isolated from the complicated mixture of prod-
ucts. We were however able to obtain the over-oxidised dienyl
derivative 8 and the bicyclo[3.3.1]nonane derivative 9, albeit in
modest yields in each case. Sulfonamides were found to be the
highest yielding and most selective substrates, which is consist-
ent with reports from the literature on many other palladium-
mediated alkene aminations.[18] The N-mesyl and -tosyl deriva-
tives 6 and 3 were obtained in 79% and 93% respectively. In
It was deemed prudent to focus the efforts of library synthe-
sis on the high yielding and selective N-sulfonyl substrates. Fur-
thermore, it was anticipated that the inherent stability of the
sulfonamide functionality would help safeguard against the ex-
pected toxicity of the N-H derivatives. It was decided that the
strategy for library synthesis would involve a series of sulfonyl-
ation reactions for decoration of the first point of diversity, with
R2 becoming the main diversity generating vector to be ex-
plored. As such, the ꢀ-amino esters 13/14 were prepared ac-
cording to a modified procedure of Choi[19] and then treated
with a range of sulfonyl chlorides. The selected examples repre-
sent alkyl, aryl (electron poor and rich) and heteroaryl sulfon-
amide substitutions (Scheme 1, A–E). In each case sulfonylation
proceeded smoothly and the products 2/15b–e were isolated
in 76–99% yield. Under our optimised conditions, the oxidative
amination of 2/15b–e gave excellent yields (79–93%) and regio-
selectivities (> 20:1).
Next, to increase the “lead-likeness” of the final library the
metabolically labile alkene functionality was removed via
hydrogenation over palladium on carbon. In the penultimate
step of library synthesis, base-mediated ester hydrolysis was
employed to yield the carboxylic acid functionalised scaffolds
17a–e in good to excellent yields (72–95% over two steps).
With the acids 17a–e in hand, the feasibility of a library syn-
light of these results, N-tosylated precursors were selected to thesis via HATU-mediated amidation in plate-format was inves-
further investigate the substrate scope of the transformation tigated. From the 5 scaffolds 17a–e, 15 test reactions were per-
with regard to the R2 substituent (Figure 2). We were delighted formed using a selection of cyclic, acyclic and aromatic amines
to obtain a good yield of 84% for the Weinreb amide functional- (See Supporting Information). The reaction mixtures were ana-
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
2
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Scheme 1. The synthesis and derivatisation of chemical scaffolds for the generation of a 9-azabicyclo[4.2.1]nonane compound library.
lysed by UPLC, filtered and purified by preparative HPLC, with
We were also compelled to apply the oxidative amination
purities of 85–100% as determined by LCMS. The stability of conditions to the synthesis of anatoxin-a (1). Parsons' 2000 syn-
1
the amidation products was assessed by comparison of the H thesis[12a] utilised a selenium-mediated ring closure of the
NMR spectra obtained before and after storage in DMSO-d6 at
room temp. for 7 days. In each case no significant change could
be observed. By utilising the general strategy for library synthe-
sis described herein, a library of 881 9-azabicyclo[4.2.1]nonane
derivatives were produced (ca. 0.1 mmol each) to be incorpo-
rated into the European Lead Factory's screening programme.
Computational assessment of the library indicates that it is
largely Lipinski's rule of five compliant, with an average molec-
ular weight of 445.6 Da and cLogP of 2.48. The library also has
a high degree of three-dimensionality and bond-saturation,
with an average Fsp3 of 0.56 (Figure 3). In recent years there
has been increased recognition of the importance of Fsp3 when
designing compound libraries for lead identification. Increased
bond saturation in general, as a measure of complexity, has
been shown to correlate well with clinical success as com-
pounds transition through the drug-discovery process.[17] With
an average Fsp3 of 0.56, this library compares favourably to
compounds in the discovery phase (average Fsp3 = 0.36) and
Scheme 2. Application of the optimised oxidative amination conditions to
the formal synthesis of anatoxin-a (1).
to marketed drugs (average Fsp3 = 0.47).[17]
Figure 3. cLogP vs. MW (left) and Fsp3 (right) plots of the produced 881 membered compound library.
Eur. J. Org. Chem. 0000, 0–0 www.eurjoc.org © 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3

Communication
[4] K. L. Swanson, H. Rapoport, R. S. Aronstam, E. X. Albuquerque, Acs Symp.,
Ser. 1990, 418, 107–118.
[5] C. E. Spivak, B. Witkop, E. X. Albuquerque, Mol. Pharmacol. 1980, 18, 384–
394.
N-Boc methyl ketone 18 to construct the 9-azabicyclo[4.2.1]-
nonane framework (Scheme 2). The intermediate selenide was
not isolated but instead treated with H2O2, promoting oxidation
and elimination to yield the key aza-bridged bicycle 19 in 55%
over two steps. Interestingly, despite the report that PdCl2,
Pd(OAc)2 and RhCl3 were unsuccessful at bringing about the
heterocyclisation of 18, we were delighted to observe that utili-
sation of our optimised oxidative amination conditions allowed
the isolation of 19 in a much improved 86% yield. This was
achieved in a single synthetic operation and constitutes a useful
refinement to the seminal work of Parsons[12a] and Rapoport.[7b]
[6] S. Wonnacott, T. Gallagher, Mar. Drugs 2006, 4, 228–254.
[7] For total/formal syntheses of anatoxin-a, see references 8–10 and further
references therein. For approaches to anatoxin derivatives please see: a)
N. A. Magnus, L. Ducry, V. Rolland, S. Wonnacott, T. Gallagher, J. Chem.
Soc. Trans. 1997, 1, 2313–2318; b) F. J. Sardina, M. H. Howard, A. M. P.
Koskinen, H. Rapoport, J. Org. Chem. 1989, 54, 4654–4660; c) P. Thomas,
P. A. Brough, T. Gallagher, S. Wonnacott, Drug Dev. Res. 1994, 31, 147–
156; d) D. Simoni, R. Rondanin, P. Marchetti, C. Rullo, R. Baruchello, G.
Grisolia, G. Barbato, R. Giovannini, C. Marchioro, A. M. Capelli, C. Virginio,
A. Bozzoli, P. A. Borea, S. Merighi, D. Donati, Bioorg. Med. Chem. Lett.
2011, 21, 5423–5427; e) D. B. Kanne, L. G. Abood, J. Med. Chem. 1988,
31, 506–509; f) F. I. Carroll, X. Hu, H. A. Navarro, J. Deschamps, G. R.
Abdrakhmanova, M. I. Damaj, B. R. Martin, J. Med. Chem. 2006, 49, 3244–
3250; g) E. Wright, T. Gallagher, C. G. V. Sharples, S. Wonnacott, Bioorg.
Med. Chem. Lett. 1997, 7, 2867–2870; h) H. Gohlke, S. Schwarz, D. Gün-
disch, M. C. Tilotta, A. Weber, T. Wegge, G. Seitz, J. Med. Chem. 2003, 46,
2031–2048; i) M. H. Howard, F. J. Sardina, H. Rapoport, J. Org. Chem.
1990, 55, 2829–2838.
Conclusions
In conclusion, an improved method for the intramolecular oxidative
amination of aminocyclooct-4-enes is reported for formation of the
biologically relevant 9-azabicyclo[4.2.1]nonane framework. The re-
action is generally high yielding, has good substrate scope and pro-
ceeds under a cheap and “ligand-free” catalytic manifold. The proto-
col was applied to the synthesis of a series of chemical scaffolds,
which were further derivatised to form a lead-like 881 membered
compound library. The oxidative amination methodology was then
applied to a formal synthesis of anatoxin-a (1), facilitating an im-
proved yield and increased operational simplicity.
[8] For reviews on synthetic approaches to anatoxin-a and its derivatives
see: a) H. L. Mansell, Tetrahedron 1996, 52, 6025–6061; b) V. Rodríguez,
S. Moura, E. Pinto, C. M. P. Pereira, R. C. Braga, Quim. Nova 2006, 29,
1365–1371.
[9] a) S. J. Roe, R. A. Stockman, Chem. Commun. 2008, 3432–3434; b) S. J.
Roe, D. L. Hughes, P. Aggarwal, R. A. Stockman, Synthesis 2009, 3775–
3784.
[10] M. Kono, S. Harada, Y. Hamada, T. Nemoto, Tetrahedron 2016, 72, 1395–
1399.
[11] C.-Y. Oh, K.-S. Kim, W.-H. Ham, Tetrahedron Lett. 1998, 39, 2133–2136.
[12] a) P. J. Parsons, N. P. Camp, N. Edwards, L. R. Sumoreeah, Tetrahedron
2000, 56, 309–315; b) P. J. Parsons, N. P. Camp, J. M. Underwood, D. M.
Harvey, Tetrahedron 1996, 52, 11637–11642; c) P. J. Parsons, N. P. Camp,
J. M. Underwood, D. M. Harvey, J. Chem. Soc., Chem. Commun. 1995,
1461–1462.
[13] T.-L. Ho, E. Zinurova, Helv. Chim. Acta 2006, 89, 134–137.
[14] a) M. J. Rawling, T. E. Storr, W. A. Bawazir, S. J. Cully, W. Lewis, M. S. I. T.
Makki, I. R. Strutt, G. Jones, D. Hamza, R. A. Stockman, Chem. Commun.
2015, 51, 12867–12870; b) T. E. Storr, S. J. Cully, M. J. Rawling, W. Lewis,
D. Hamza, G. Jones, R. A. Stockman, Bioorg. Med. Chem. 2015, 23, 2621–
2628; c) R. S. Dawood, I. Georgiou, R. P. Wilkie, W. Lewis, R. A. Stockman,
Chem. Eur. J. 2017, 23, 11153–11158; d) S. J. Cully, T. E. Storr, M. J. Rawl-
ing, I. R. Abeysena, D. Hamza, G. Jones, C. A. Pearce, A. Quddus, W. Lewis,
R. A. Stockman, Bioorg. Med. Chem. 2016, 24, 5249–5257.
[15] a) J.-L. Reymond, M. Awale, ACS Chem. Neurosci. 2012, 3, 649–657; b)
A. W. Hung, A. Ramek, Y. Wang, T. Kaya, J. A. Wilson, P. A. Clemons, D. W.
Young, Proc. Natl. Acad. Sci. USA 2011, 108, 6799–6804; c) J.-L. Reymond,
R. van Deursen, L. C. Blum, L. Ruddigkeit, Med. Chem. Commun. 2010, 1,
30–38.
Acknowledgments
This research was done within the European Lead Factory and
has received support from the Innovative Medicines Initiative
Joint Undertaking (grant no. 115489), with financial contribu-
tion from the European Union's Seventh Framework Pro-
gramme (FP7/2007-2013) and EFPIA companies' in-kind contri-
bution. The authors would like to thank the University of Not-
tingham and the Ministry of Higher Education and Scientific
Research-Iraq (MOHESR) for funding support, Professor Christo-
pher Moody and Tom McInally for useful insight and discussion,
Megan Thomsett and Pritesh Tailor for assistance with the li-
brary data and Thomas Ward for his synthetic contributions.
Keywords: Annulation · Chemical scaffolds · Medium-ring
compounds · Natural products · Regioselectivity
[16] European Lead Factory, https://www.europeanleadfactory.eu, (accessed
12 February 2018).
[17] F. Lovering, J. Bikker, C. Humblet, J. Med. Chem. 2009, 52, 6752–6756.
[18] E. M. Beccalli, G. Broggini, M. Martinelli, S. Sottocornola, Chem. Rev. 2007,
107, 5318–5365.
[1] C. S. Huber, Acta Crystallogr., Sect. B 1972, 28, 2577–2582.
[2] J. P. Devlin, O. E. Edwards, P. R. Gorham, N. R. Hunter, R. K. Pike, B. Stavric,
Can. J. Chem. 1977, 55, 1367–1371.
[19] W. Lee, S. Kwon, P. Kang, I. A. Guzei, S. H. Choi, Org. Biomol. Chem. 2014,
12, 2641–2644.
[3] a) W. W. Carmichael, D. F. Biggs, P. R. Gorham, Science 1975, 187, 542–
544; b) L. Krienitz, A. Ballot, K. Kotut, C. Wiegand, S. Pütz, J. S. Metcalf,
G. A. Codd, S. Pflugmacher, FEMS Microbiol. Ecol. 2003, 43, 141–148; c)
B. Puschner, B. Hoff, E. R. Tor, J. Vet. Diagnostic Investig. 2008, 20, 89–92.
Received: August 2, 2018
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
4
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Chemical Scaffolds
R. S. Dawood, S. R. Chidipudi,
D. C. O'Connor, W. Lewis, D. Hamza,
C. A. Pearce, G. Jones, R. P. Wilkie,
I. Georgiou, T. E. Storr,
J. C. Moore,* R. A. Stockman* ........ 1–5
PdII-Mediated Oxidative Amination
for Access to a 9-Azabicyclo[4.2.1]-
nonane Compound Library and Ana-
toxin-a
Biologically relevant
9
azabicyclo- ceeds under “ligand-free” catalytic
[4.2.1]nonanes can be synthesised conditions. The protocol was applied
through an intramolecular oxidative to the synthesis of anatoxin-a and a se-
amination of aminocyclooct-4-enes. ries of chemical scaffolds, which were
The reaction is generally high yielding, further derivatised to form an
has good substrate scope and pro- 881-membered compound library.
DOI: 10.1002/ejoc.201801209
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
5
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim