Synthesis and biological evaluation of a new family of constrained azabicyclic homocholine analogues

Article abstract of DOI:10.1071/CH10024

A family of constrained acylated homocholine analogues have been synthesized, based on the azabicyclic ring scaffold derived from a double-Mannich annulation of cyclic ketones. The short synthetic route allows generation of structural diversity including, variation in the carbocyclic ring size, bridgehead substitution, nitrogen substitution and the ester sidechain. Biological assays on selected analogues demonstrate these compounds are nicotinic acetylcholine receptor (nAChR) antagonists. Several analogues also bind to other neuronal transporter and receptor targets. CSIRO 2010.

Full text of DOI:10.1071/CH10024

RESEARCH FRONT
Full Paper
CSIRO PUBLISHING
Aust. J. Chem. 2010, 63, 808–812
www.publish.csiro.au/journals/ajc
Synthesis and Biological Evaluation of a New Family
of Constrained Azabicyclic Homocholine Analogues
Jill I. Halliday,A Mary Chebib,B and Malcolm D. McLeodC,D
ASchool of Chemistry, F11, University of Sydney, NSW 2006, Australia.
BFaculty of Pharmacy, A15, University of Sydney, NSW 2006, Australia.
CResearch School of Chemistry, Australian National University,
Canberra, ACT 0200, Australia.
DCorresponding author. Email: malcolm.mcleod@anu.edu.au
A family of constrained acylated homocholine analogues have been synthesized, based on the azabicyclic ring scaffold
derived from a double-Mannich annulation of cyclic ketones. The short synthetic route allows generation of structural
diversity including, variation in the carbocyclic ring size, bridgehead substitution, nitrogen substitution and the ester
sidechain. Biological assays on selected analogues demonstrate these compounds are nicotinic acetylcholine receptor
(nAChR) antagonists. Several analogues also bind to other neuronal transporter and receptor targets.
Manuscript received: 14 January 2010.
Manuscript accepted: 11 February 2010.
Introduction
4 and N,N-bis(ethoxymethyl)alkylamines 5, in a single step,
double-Mannich annulation developed in our laboratory.[11] The
potential for variation in structure inherent in this sequence is
large.Thedouble-Mannichannulationhasbeenreportedtooccur
efficiently with a range of carbo- and hetero-cyclic ketones 4
(Xn = O, CH2, (CH2)2, (CH2)3; R2 = H, Me, 2-propenyl) and
N,N-bis(ethoxymethyl)alkylamine 5 reagents (R1 = Et, Prn, Pri,
But). This method significantly expanded the representation of
structurally related azabicyclic scaffolds 3 reported in the chem-
ical literature.[11] In this paper, we report an elaboration of these
scaffolds to a range of acylated homocholine analogues, repre-
sented by structure 2, containing additional variation in the ester
sidechain (R3) and nitrogen substituent (R1). We also report the
biological evaluation of a subset of these homocholine analogues
across a collection of neuronal receptors.
Biologically active natural products provide a rich source of
inspiration to the synthetic chemist, particularly for the develop-
ment of small molecules to target the extensive array of thera-
peutically important drug targets.[1–4] One important but poorly
understood family of targets are the neuronal nicotinic acetyl-
choline receptors (nAChRs), which influence brain function
through the release of a variety of neurotransmitters in many
regions of the brain. The nAChRs modulate neuronal function
associated with cognition, learning and memory, arousal, cere-
bral blood flow, and metabolism; and have been implicated
in a variety of pathological conditions such as Alzheimer’s
disease, Parkinson’s disease, nicotine addiction, pain, epilepsy,
and schizophrenia.[5] For this reason there is great interest
in the development of selective neuronal nAChR ligands as
therapies.[6,7] The potential of nAChR targeted therapies has
recently been highlighted by the release, in 2006, of the smok-
ing cessation aid varenicline (Champix®); a partial agonist at the
α4β2 nAChR.[8] Despite this interest, there are few potent and
selective nAChR ligands known, and the development and study
of these ligands is made difficult due to the limited structural
information of the target receptors.[9]
As part of our research, directed toward the development
of nAChR ligands, we have shown that small molecule ana-
logues of the selective α7 nAChR antagonist methyllycaconitine
1 (Scheme 1), act as antagonists in functional assays of nicotinic
receptors.[10] To further explore the structural and pharmacolog-
ical potential associated with these systems, we sought access
to the azabicyclic ring systems represented by structure 2, that
retain the putative pharmacophore present in 1,[10] albeit with a
different topology. The acylated homocholine analogues repre-
sented by structure 2, can be derived from azabicyclic ketone 3 in
two steps, by reduction of the carbonyl group and esterification.
Ketone 3 can, in turn, be derived from simple cyclic ketones
Results and Discussion
The azabicyclic ketones
6 were prepared by reported
methods,[11] and were subsequently reduced with sodium boro-
hydride to give the corresponding alcohols 7 selectively and in
good yield (Table 1).[12] For these azabicyclo[4.3.1]decanones
6a–e, and azabicyclo[5.3.1]undecanones 6f–g, reduction
occurred from the less sterically-encumbered face of the car-
bonyl group, and opposite to the tetra- or penta-methylene
bridge, respectively. The selectivity of reduction observed
for azabicyclo[4.3.1]decanones 6a–e is consistent with results
reported by House for the sodium borohydride reduction of 8-
methyl-8-azabicyclo[4.3.1]decanone (49:1).[13] In the present
examples, the presence of nitrogen substituents larger than
methyl, did not significantly change the observed stereo-
selectivity. The relative stereochemistry of the major (10s)-
alcohol 7c, derived from ketone 6c, was established in previous
1
work by H NMR 2D NOESY experiment and single crystal
X-ray crystallography.[11] The alcohols 7 were cleanly converted
© CSIRO 2010
10.1071/CH10024
0004-9425/10/050808

RESEARCH FRONT
New Family of Constrained Azabicyclic Homocholine Analogues
809
OMe
Table 2. Sodium borohydride reduction of 1,5-dimethylazabicyclo
[3.3.1]nonanones
MeO
N
OMe
H
Me
N
H
Me
O
N
R1
Xn
OH
OH
HO
O
O
H
R2
O
N
R1
OMe
10a–c
NaBH4
O
N
R1
R3
R2
O
O
9a–c
N
2
R1
Methyllycaconitine 1
HO
11a–c
(a) Reduction
(b) Esterification
R1
EtO
R2
N
OEt
R2
5
Entry
R1
RatioA [10:11]
Yield of 10 [%]
Yield of 11 [%]
Xn
Xn
a
b
c
Et
iPr
tBu
2.2:1
1.7:1
1.7:1
60
46
48
23
22
Double-Mannich
N
R2
R1
B
–
R2
O
O
ARatio determined by 1H NMR integration.
BNot isolated.
3
4
Scheme 1. Methyllycaconitine 1 and acylated homocholine analogues 2
derived from ketones 4.
H
H
8
N
9
HO
H
Table 1. Sodium borohydride reduction and acetylation of azabicyclo
[4.3.1]decanones and azabicyclo[5.3.1]undecanones
2
CH3
H
H
( )
( )
n
NOE
n
NaBH4
N
N
R1
R1
10a
RO
O
Fig. 1. Key nuclear Overhauser effect (NOE) correlations observed for
alcohol 10a.
6a–g
7a–g R ϭ H
8a–g R ϭ Ac
Ac2O, DMAP, Et3N
Ac2O, DMAP,
Entry
R1
Et
n
Reduction yield
[%]
Acetylation yield
[%]
Et3N
N
R1
N
R1
HO
O
a
b
c
d
e
f
1
1
1
1
1
2
2
81
98
61
85
96
51A
89
85
93
96
81
98
99
69
iPr
tBu
nPr
Bn
Et
O
10a R1 ϭ Et
10c R1 ϭ tBu
12a R1 ϭ Et (92%)
12c R1 ϭ tBu (92%)
Scheme 2. Acetylation of 1,5-dimethylazabicyclo[3.3.1]nonan-10-ols.
g
tBu
AThe (11r) epimer was also formed (3%).
H2B/H4B protons (δ 1.80, 2H, dd, J 11.8, 2.1) were consis-
tent with their 1,3-diaxial orientation in a chair-like piperidine
ring.[14]
The major alcohol diastereomers, 10a and 10c, were cleanly
converted into the corresponding acetates, 12a and 12c, with
acetic anhydride and 4-(dimethylamino)pyridine in good yield
(Scheme 2).
into the corresponding acetates 8 with acetic anhydride and
4-(dimethylamino)pyridine in good yield (Table 1).
Thereductionofthe1,5-dimethylazabicyclo[3.3.1]nonanones
9[11] provided a chromatographically separable mixture of
diastereomers, with a modest preference (10:11, 1.7–2.2:1)
for hydride addition to the face adjacent to the ring nitrogen
(Table 2). The selectivity of reduction observed for azabicyclo
[3.3.1]nonanones 9 was also consistent with results reported
by House for the sodium borohydride reduction of 3-methyl-3-
azabicyclo[3.3.1]nonanone (1.9:1).[13] In the current examples,
the presence of nitrogen substituents larger than methyl and
bridgehead methyl groups, did not significantly change the
observed stereoselectivity. The (9s) relative configuration of the
hydroxyl-bearing stereocentre of alcohol 10a, was confirmed by
a 1H NMR 2D NOESY experiment performed in deuteriochloro-
form solution (Fig. 1). Strong NOE cross peaks between the
oxymethine H9 proton (δ 3.00, 1H, d, J 5.5) and the upfield
In addition to acetylation, several alternative esters, 13–18,
were synthesized in good to high yield, through the reaction of
the alcohols 7b and 7c with carboxylic acid chlorides (Table 3).
As a further demonstration of the potential for variation of
the ring system, a two-step exchange of the nitrogen substituent
was performed on N-benzyl derivative 8e (Scheme 3). Transfer
hydrogenolysis of the benzyl amine was achieved with ammo-
nium formate and palladium on carbon catalyst, to afford the
ammonium formate salt 19.[15] This material was then subjected
to reductive amination with two differently substituted benzalde-
hyde derivatives, mediated by sodium triacetoxyborohydride, to
afford the new benzylamines 20 and 21.[16] The reductive ami-
nation has the advantage of permitting substitution on the ring
nitrogen that may not be compatible with the double-Mannich
annulation chemistry.

RESEARCH FRONT
810
J. I. Halliday, M. Chebib, and M. D. McLeod
Table 3. Acylation of azabicyclo[4.3.1]decan-9-ols
Compounds, representative of the structural diversity gen-
erated by the synthetic chemistry, were subjected to biological
evaluation using two methods. First, the compounds were evalu-
ated in a functional assay using nicotinic receptors expressed
in Xenopus laevis oocytes, to establish mode of action and
potency.[10] Second, compounds were screened for binding affin-
ity against a range of neuronal receptors and transporters, by
the National Institute of Mental Health’s Psychoactive Drug
Screening Program.[17]
R3COCl,
Et3N
N
N
R1
R1
HO
O
R3
7b, c
R1
13–18
O
Alcohol
Ester
R3
Acylation yield
[%]
The two-electrode, voltage-clamp electrophysiology func-
tional assay was conducted on esters 8c, 8e, 8g, 12c, 17, and
18 (Fig. 2), according to previously reported procedures.[10]
Although time consuming relative to radioligand binding assays
that probe the orthosteric site, electrophysiology has the advan-
tage of providing more information, including mode of action,
or effects that arise from binding to allosteric sites. Ligand
screening studies were made to determine the functional effect
of these novel homocholine ligands on oocytes expressing
either rat α3β4, α4β2, or α7 nAChRs. No agonist activity was
observed for any of the esters tested against the nAChR sub-
types. Antagonist activity was assessed by pre-incubation of
the oocytes with the ester (100 μM for 3 min), followed by
co-administration of acetylcholine (EC50 α3β4, 150 μM; α4β2,
100 μM; and α7, 300 μM) and the ester (100 μM). The esters
8c, 8e, 8g, 12c, 17, and 18 displayed no activity when applied
alone to each receptor subtype, but acted as antagonists when co-
administered with acetylcholine. Inhibitory concentration (IC50)
response curves were constructed to determine the potency of
each of the compounds at the α3β4, α4β2, and α7 nAChRs
(Table 4).[18] Each receptor was exposed to a constant concentra-
tion of acetylcholine, corresponding to the relevant EC50 value
(α3β4, 150 μM; α4β2, 100 μM; and α7, 300 μM), and increas-
ing concentrations of the compound of interest (0.03–600 μM).
7b
7b
7b
7c
7c
7c
iPr
13
14
15
16
17
18
Cyclohexane
tBu
2-MeO-C6H4
Cyclohexane
2-MeO-C6H4
4-MeO-C6H4
92
99
81
69
81
97
iPr
iPr
tBu
tBu
tBu
H
NH4OCHO,
ϩ
Pd/C
N
N
O
Bn
H
O
O
Ϫ
(72%)
O
H
O
O
8e
19
R1CHO,
NaBH(OAc)3
N
R1
O
20 R1 ϭ CH2(4-Cl-C6H4) (84%)
21 R1 ϭ CH2(4-OH-C6H4) (99%)
O
Scheme 3. Variation of N-substituent by transfer hydrogenolysis and
reductive amination.
N
N
N
N
HO
HO
HO
HO
7g
7e
10c
7c
Sigma-1 Ki 9.7 1.9 µM
Sigma-2 Ki 1.6 0.1 µM
DAT Ki 0.32 0.09 µM
N
N
N
N
O
O
O
O
O
O
O
O
8g
12c
8c
8e
Sigma-2 Ki 3.4 0.3 µM
DAT Ki 0.66 0.19 µM
DAT Ki 0.65 0.15 µM
KOR Ki 2.7 0.3 µM
OMe
O
N
N
MeO
O
O
O
17
18
Fig. 2. Compounds screened for biological activity and dissociation constants (Ki) at neuronal receptors and transporters: DAT, dopamine transporter;
KOR, κ-opioid receptor; sigma-1 and -2, sigma receptor.

RESEARCH FRONT
New Family of Constrained Azabicyclic Homocholine Analogues
811
Table 4. IC50 values [μM] for 8c, 8e, 8g, 12c, 17, and 18 in the presence
of acetylcholine (α3β4, 150 μM; α4β2, 100 μM; and α7, 300 μM) at rat
neuronal nAChRs
receptor. A more detailed study of the mode of action of related
azabicyclic ligands at nAChRs is the subject of ongoing work.
Compounds that displayed significant inhibition of radio-
ligand binding (>50%) in the primary assays were selected
for secondary screening to determine the dissociation constant
(Ki).[17] This subsequent testing identified several compounds
that displayed moderate binding affinity to other neuronal tar-
gets (Fig. 2). Compounds 7c (Ki 0.32 μM), 8c (Ki 0.66 μM),
and 8e (Ki 0.65 μM) showed affinity to the human dopamine
transporter (DAT), compound 8e (Ki 2.8 μM) showed affinity
to the rat κ-opioid receptor (KOR), compound 7g (Ki 9.7 μM)
showed affinity to the rat sigma-1 receptor (sigma-1) and com-
pounds 7g (Ki 1.6 μM) and 8g (Ki 3.4 μM) showed affinity to
the rat sigma-2 receptor (sigma-2). This activity across several
receptors and transporters is clearly undesirable in the search
for selective nAChR ligands, but this moderate binding affinity
opens avenues for the development of new ligands for alternate
neuronal targets.
Compounds
IC50 [μM]A
α4β2
α3β4
α7
8c
8e
8g
12c
17
18
5.8 0.9
39.5 0.7
4.0 1.2
5.4 1.1
7.2 2.8
8.9 1.2
9.5 1.0
4.6 1.6
6.7 1.3
15.2 1.9
B
–
5.1 1.0
92.9 7.3
40.2 1.1
49.0 2.5
B
–
15.9 2.0
45.9 12.7
AThe IC50 is the effective dose that inhibits 50% of the response to a
fixed concentration of acetylcholine. The IC50 values are expressed as mean
IC50 SEM in μM (n = 3 oocytes).
BNot determined.
A variable slope sigmoidal curve was fitted to each dose–
response curve and the IC50 value, the effective dose that
inhibits 50% of the receptor population, was obtained for each
compound.
In conclusion a range of new constrained acylated homo-
choline analogues have been synthesized, based on the azabi-
cyclic ring system derived from a double-Mannich annulation
of cyclic ketones. Functional assays have shown that the esters
8c, 8e, 8g, 12c, 17, and 18 are all antagonists at the rat α3β4,
α4β2, and α7 nAChRs expressed in Xenopus oocytes, with IC50
values in the low micromolar range. Screening against a panel of
neuronal receptors and transporters has identified several com-
pounds (7c, 7g, 8c, 8e, and 8g) with affinity for other neuronal
receptors. Additional synthesis and testing will assist in further
elucidating the structure activity relationships of this new class
of nAChR antagonists.
Several trends were apparent from the IC50 data (Table 4).
At the heteromeric α3β4 receptor, the N-tert-butyl analogues 8c,
8g, 12c, 17, and 18, demonstrated similar potency (4.0–8.9 μM),
whereas the N-benzyl analogue 8e was 4–10-fold less potent. At
the heteromeric α4β2 receptor, the acetate analogues 8c, 8e, and
8g were generally of equal potency (4.6–9.5 μM), whereas the
benzoate analogues 17 and 18 were less potent by 2–10-fold.The
unsubstituted acetate analogues, 8c and 8g, demonstrated similar
potency at both the homomeric α7 receptor and the heteromeric
nAChRs. Substitution of the ester sidechain to give the benzoate
analogues17 and 18, orthebridgeheadpositioncarbonsinligand
12c, reduced potency at this receptor by 3–18-fold.
Accessory Publication
The Accessory Publication includes experimental procedures,
1
spectroscopic data and H and 13C NMR spectra for all new
The IC50 values for these ligands can also be compared across
the nAChR subtypes tested (Table 4). In general, the unsubsti-
tuted acetyl ligands, 8c and 8g, showed no selectivity between
the nAChRs. Substitution at the bridgehead position for ligand
12c led to a 17-fold increase in selectivity for the heteromeric
nAChRs over the α7 receptor. This trend, although less pro-
nounced, was also observed for the benzoate analogues 17 and
18, which showed up to 5-fold increased selectivity between
hetero- and homomeric receptors, as well as a 2–5-fold selectiv-
ity for the α3β4 receptor over the α4β2 receptor. The N-benzyl
analogue 8e, was the only ligand to show reduced selectivity
for the α3β4 receptor, lower than the α4β2 receptor by 8-fold;
although data for the α7 receptor is not available to give a full
impression for the activity of this ligand at these nAChRs.
The alcohols 7c, 7e, 7g, and 10c and esters 8c, 8e, 8g, and
12c (Fig. 2) were also screened by the National Institute of Men-
tal Health’s Psychoactive Drug Screening Program, according
to reported procedures.[17] Primary assays involved displace-
ment of a suitable radioligand and were conducted at a substrate
concentration of 10 μM. None of the compounds tested showed
appreciable inhibition (≤11%) of [3H]-epibatidine radioligand
binding to a range of nicotinic receptor subtypes, including
cloned human α2β2, α2β4, α3β2, α3β4, α4β2, α4β4, and endo-
genous α4β2 from rat forebrain. The acetate analogues, 8c, 8e,
and 8g, which showed no significant inhibition (≤7%) of radio-
ligand binding to endogenous rat forebrain α4β2 nAChR in the
PDSP primary assay, act as antagonists (IC50 4.6–9.7 μM) at rat
α4β2 expressed in Xenopus oocytes (Table 4). This suggests that
these ligands may act as non-competitive ligands at the α4β2
compounds. Inhibitory concentration (IC50) response curves are
provided for compounds tested at each receptor subtype. This
material is available on the Journal’s website.
Acknowledgements
We thank The Australian National University, The University of Sydney and
the Australian Research Council Discovery Project Scheme (DP0663006)
for supporting this work.
References
[1] A. M. Clark, Pharm. Res. 1996, 13, 1133. doi:10.1023/
A:1016091631721
[2] G. M. Cragg, D. J. Newman, K. M. Snader, J. Nat. Prod. 1997, 60, 52.
doi:10.1021/NP9604893
[3] D. J. Newman, G. M. Cragg, K. M. Snader, J. Nat. Prod. 2003, 66,
1022. doi:10.1021/NP030096L
[4] D. J. Newman, G. M. Cragg, J. Nat. Prod. 2007, 70, 461. doi:10.1021/
NP068054V
[5] D. Paterson,A. Nordberg, Prog. Neurobiol. 2000, 61, 75. doi:10.1016/
S0301-0082(99)00045-3
[6] B. K. Cassels, I. Bermúdez, F. Dajas, J. A. Abin-Carriquiry, S.
Wonnacott, Drug Discov. Today 2005, 10, 1657. doi:10.1016/
S1359-6446(05)03665-2
[7] A. A. Jensen, B. Frølund, T. Liljefors, P. Krogsgaard-Larsen, J. Med.
Chem. 2005, 48, 4705. doi:10.1021/JM040219E
[8] H. Rollema, L. K. Chambers, J. W. Coe, J. Glowa, R. S. Hurst,
L. A. Lebel, Y. Lu, R. S. Mansbach, R. J. Mather, C. C. Rovetti,
S. B. Sands, E. Schaeffer, D. W. Schulz, F. D. Tingley III,
K. E. Williams, Neuropharmacology 2007, 52, 985. doi:10.1016/
J.NEUROPHARM.2006.10.016

RESEARCH FRONT
812
J. I. Halliday, M. Chebib, and M. D. McLeod
[9] H. A. Lester, M. I. Dibas, D. S. Dahan, J. F. Leite, D. A. Dougherty,
Trends Neurosci. 2004, 27, 329. doi:10.1016/J.TINS.2004.04.002
[10] D. Barker, D. H.-S. Lin, J. E. Carland, C. P.-Y. Chu, M. Chebib,
M. A. Brimble, G. P. Savage, M. D. McLeod, Bioorg. Med. Chem.
2005, 13, 4565. doi:10.1016/J.BMC.2005.04.054
[11] J. I. Halliday, M. Chebib, P. Turner, M. D. McLeod, Org. Lett. 2006,
8, 3399. doi:10.1021/OL061242S
[12] The experimental procedures, spectroscopic data and 1H and 13C
NMR spectra for all new compounds are reported in the Accessory
Publication.
[16] A. F. Abdel-Magid, C. A. Maryanoff, K. G. Carson, Tetrahedron Lett.
1990, 31, 5595. doi:10.1016/S0040-4039(00)97905-X
[17] The Ki determinations and receptor binding profiles were generously
provided by the National Institute of Mental Health’s Psychoac-
tive Drug Screening Program, Contract # HHSN-271–2008–00025-C
(NIMH PDSP). The NIMH PDSP is Directed by Bryan L. Roth MD,
PhD at the University of North Carolina at Chapel Hill and Project
Officer Jamie Driscol at NIMH, Bethesda MD, USA. For experimen-
tal details please refer to the PDSP website http://pdsp.med.unc.edu/
and click on “Binding Assay” on the menu bar (accessed January 7,
2010).
[13] H. O. House, H. C. Müller, C. G. Pitt, P. P. Wickham, J. Org. Chem.
1963, 28, 2407. doi:10.1021/JO01044A064
[14] M. S. Arias-Pérez, A. Alejo, A. Maroto, Tetrahedron 1997, 53, 13099.
doi:10.1016/S0040-4020(97)00832-6
[18] The inhibitory concentration (IC50) response curves for each receptor
subtype are reported in the Accessory Publication.
[15] S. Ram, L. D. Spicer, Tetrahedron Lett. 1987, 28, 515. doi:10.1016/
S0040-4039(00)95769-1