Epibatidine isomers and analogues: Structure-activity relationships

Article abstract of DOI:10.1016/j.bmcl.2006.08.049

Binding affinities for a range of epibatidine isomers and analogues at the α4β2 and α3β4 nAChR subtypes are reported; compounds having similar N-N distances to epibatidine show similar, high potencies.

Full text of DOI:10.1016/j.bmcl.2006.08.049

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5493–5497
Epibatidine isomers and analogues: Structure–activity relationships
Richard White,a John R. Malpass,a,* Sandeep Handa,a S. Richard Baker,b
Lisa M. Broad,b Liz Follyb and Adrian Moggb
aDepartment of Chemistry, University of Leicester, Leicester LE1 7RH, UK
bEli Lilly and Co. Ltd, Erl Wood Manor, Sunninghill Road, Windlesham, Surrey GU20 6PH, UK
Received 6 July 2006; revised 9 August 2006; accepted 9 August 2006
Available online 28 August 2006
Abstract—Binding affinities for a range of epibatidine isomers and analogues at the a4b2 and a3b4 nAChR subtypes are reported;
compounds having similar N–N distances to epibatidine show similar, high potencies.
ꢀ 2006 Elsevier Ltd. All rights reserved.
The remarkable affinity of epibatidine (1) for the nico-
tinic acetylcholine receptor (nAChR) is now well
known.1 Epibatidine is a potent but non-selective ago-
nist, its pharmacological effects probably being medi-
ated by a variety of nAChR subtypes including a7,
a4b2, a3b4 and the a1b1cd receptor at the neuromuscu-
lar junction.2 The recognition of a wide variety of recep-
tor subtypes has encouraged the exploration of
epibatidine analogues in the hope of developing sub-
type-selective therapeutic compounds.2 These com-
pounds could then be employed as treatments for
several neurological and psychiatric disorders including
Parkinson’s and Alzheimer’s diseases, schizophrenia,
chronic pain and tobacco dependence.2 Greater subtype
selectivity is likely to be associated with lower toxicity
and fewer undesirable side effects.2 The a4b2 nAChR
is absent from the periphery but widely expressed within
the CNS and may be a useful therapeutic target in sev-
eral of the disorders described above. In contrast, activ-
ity at either the neuromuscular junction or the a3b4
ganglionic nAChR might be expected to lead to undesir-
able side effects. Work on defining the full role of each of
the nAChR subtypes is highly dependent on the avail-
ability of subtype-specific ligands.2
heterocyclic substituent; a rigid bicyclic framework
which is capable of providing an appropriate N–N dis-
tance in one or both of the minimum-energy conforma-
tions (based on rotation about the C-pyridyl bond).3
Our earlier work with tropanes (8-azabicyclo[3.2.1]oc-
tanes) and with 2- and 7-azabicyclo[2.2.1]heptanes (2-
and 7-azanorbornanes)4 has led us to synthesize a range
of analogues and isomers based on these azabicyclic
frameworks (Fig. 1). These include homoepibatidine
(2),5 dihomoepibatidine (3)5 and isoepibatidine (4)6 in
which the positions of the bicyclic nitrogen and the het-
erocycle have been reversed. The epibatidine isomers 5
and 6,7 in which the bicyclic nitrogen is now in the 2-po-
sition and the 5- (or 6-)heterocyclic substituents are
endo-, produce a calculated N–N distance in the appro-
priate range (Table 3).
The exo-isomers 7 and 8,7 having greater N–N distanc-
es, were included in these studies for comparison and
contrast. In addition, the syn- and anti-isoepiboxidines
10 and 11,8 in which the chloropyridyl substituent of 4
has been replaced by a methylisoxazole ring, were syn-
thesized in the hope of retaining potency but with lower
toxicity (as reported for epiboxidine (9)).9
We designed our target compounds to include: a bicyclic
secondary nitrogen centre (the source of the quaternary
nitrogen which is essential for binding); a freely rotating
The synthesis and spectroscopic properties of these com-
pounds have been published but full experimental de-
tails for syn-isoepibatidine (4) are recorded at the end
of this letter since our very recent report6 gave only a
general procedure.
Keywords: Epibatidine analogues; Epibatidine isomers; nAChR affin-
ity; Nicotinic receptor subtypes; Structure–activity relationships.
Activity data at the a4b2 and a3b4 receptors are listed in
Table 1. Some previously published a4b2 and a7 bind-
ing data for selected compounds7a are included for
*
Corresponding author. Tel.: +44 0116 252 2126; fax: +44 0116 252
3789; e-mail: jrm@le.ac.uk
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.08.049

5494
R. White et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5493–5497
H
H
H
N
N
N
Cl
N
Cl
N
Cl
N
2
1
3
Cl
N
Cl
N
NH
NH
NH
N
N
5
6
7
4
NH
Cl
Cl
N
N
O
O
H
N
N
Cl
N
O
NH
NH
NH
8
10 11
9
Figure 1. Epibatidine isomers and analogues.
Table 1. Inhibition of binding at human recombinant nicotinic receptors for compounds 1,2, 4–7,10 and 11
Compound
Kia (nM)
a3b4/a4b2
a4b2
a3b4
( )-1
(+)-2
0.0558 ( 0.0056)
0.127 ( 0.011)
0.343 ( 0.026)
0.0785 ( 0.00617)
0.182 ( 0.33)
0.191 ( 0.0075)
0.925 ( 0.032)
0.527 ( 0.053)
0.16 ( 0.00189)
0.944 ( 0.058)
8.73 ( 0.179)
2.35 ( 0.122)
ne at 1000 nM
9.51 ( 0.428)
ne at 1000 nM
525 ( 6.32)
3.42
7.28
1.54
2.05
5.19
7.1
(ꢀ)-2
( )-4
(+)-5
(ꢀ)-5
( )-6
( )-7
1.23 ( 0.025)
0.771 ( 0.067)
ne at 1000 nM
0.763 ( 0.092)
ne at 1000 nM
2.59 ( 0.202)
3.05
( )-10
( )-11
Cytisine
12.5
203
Preliminary data have been reported earlier.10
a Values are geometric means of at least three experiments, standard error is given in parentheses (ne = no effect).
comparison (Table 2). Differences between the Ki bind-
ing constants in Tables 1 and 2 (and early data recorded
in Ref. 5) may be due to the fact that in Table 1, binding
was at human recombinant nAChR, whereas the
nAChR used to generate the values in Table 2 were from
native rat tissue. These native receptors may be subtly
different from the human recombinant nAChR in terms
of their unit stoichiometry (ratio of a4 and b2 subunits)
and composition (inclusion of other subunits, for exam-
ple, a5).11
The most significant result is that isoepibatidine (4) is al-
most as potent as epibatidine (( )-1). It shows similar
activity to ( )-1 at both the a4b2 and a3b4 receptors.
This confirms earlier indications5 that the unusual chem-
ical properties4b of the nitrogen atom in the 7-azanor-
bornane ring system are not an essential factor in the
unusual biological properties of 1. Earlier results for
homoepibatidine (2) have shown that incorporation of
an extra methylene group into the 2-carbon bridge has
little effect on binding affinity5,12 (see also Table 2); thus
the 7-azanorbornane system is not a pre-requisite for
high activity although we have demonstrated that inser-
tion of a second additional methylene group in dihomo-
epibatidine (3) leads to an order of magnitude reduction
in activity, presumably because of the substantially
greater bulk of the tetramethylene portion.5 The new
data for the enantiomers of 2 (Table 1) confirm the high
Table 2. Inhibition of binding at native rat nicotinic receptors for
compounds 1, 2 and 5–87a
Compound
a4b2 Kia (nM)
a7 Kib (nM)
a7/a4b2
1
(+)0.019
(ꢀ)0.020
0.23
4.9 ( 0.7) (n = 3)
7.0 ( 1.8) (n = 4)
258
350
( )-2
( )-5
( )-6
( )-7
( )-8
13 (n = 2)
6.3
3.9
56.5
112.5
86.7
<87
<42
0.056
0.045
ca. 40
ca. 40
3300
1600
a [3H]Nicotine binding to rat cortical membranes.
b
[
125I-a]Bungarotoxin binding to rat hippocampal membranes.

R. White et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5493–5497
5495
potency; the two enantiomers shows similar a4b2
selectivity.
We recognise that calculated minimum energy N–N dis-
tances are a crude, indirect measure of the efficiency of
interaction between the ligand and receptor and that
many other factors are involved which are not yet fully
understood.3 Nevertheless, our work shows good corre-
lations between calculated N–N distances and binding
affinities for a homologue (2) of epibatidine, for epibati-
dine isomers (4–6) and for a variant (10) in which
methylisoxazole replaces chloropyridine. Whilst the
new compounds show similar high potency to epibati-
dine, there is no increase in selectivity between the
a4b2 and a3b4 receptor subtypes.
Of the other 2-azanorbornane isomers, the endo-5-
substituted compound 5 shows an almost 10-fold differ-
ence in the binding ability of the two enantiomers at
both the a4b2 and a3b4 subtypes (Table 1), with the val-
ues for the (+)-enantiomer at the a4b2 being only 3-fold
more than those for epibatidine (( )-1). The enantio-
mers of 5 show little subtype selectivity. The Ki values
for the racemic endo-6-substituted compound 6 are close
to the average values for the enantiomers of 5 despite the
difference in position of the heterocycle.
Membrane preparation. Cell pastes from large-scale pro-
duction of HEK-293 cells expressing cloned human
a4b2 or a3b4 nAChR were homogenized in 4 volumes
of buffer (50 mM Tris–HCl, 150 mM NaCl and 5 mM
KCl, pH 7.4). The homogenate was centrifuged twice
(40,000g, 10 min, 4 ꢁC) and the pellets re-suspended in
4 volumes of Tris–HCl buffer after the first spin and 8
volumes after the second spin. The re-suspended homog-
enate was centrifuged (100g, 10 min, 4 ꢁC) and the
supernatant kept and re-centrifuged (40,000g, 20 min,
4 ꢁC). The pellet was re-suspended in Tris–HCl buffer
supplemented with 10% w/v sucrose. The membrane
preparation was stored in 1 ml aliquots at ꢀ80 ꢁC until
required. The protein concentration of the membrane
preparation was determined using a BCA protein assay
reagent kit.
The previously accepted ‘ideal’ N–N distance of ca.
˚
5.5 A has been revised; recent X-ray studies suggest a
3b
˚
value of 4.4–4.5 A for the active conformation. This
is in agreement with calculations for the compounds in
Table 3. Interestingly, the methylisoxazole-substituted
10 is the only isomer which shows N–N distances close
to the ‘ideal’ in both minimum energy conformations.
Whatever the minor structural differences between the
epibatidine isomers 1, 4, 5 and 6, they all retain the
key pre-requisites, including a broadly similar N–N dis-
tance, and show high potency. Significant changes are
only observed when the two nitrogen centres are widely
separated, as in the exo-isomers 7 and 8. The exo-com-
pounds are essentially inactive; the a4b2 and a3b4 data
for 7 in Table 1 augment the earlier a4b2 and a7 results
(Table 2).7a
Nicotinic receptor radioligand binding scintillation prox-
imity assay (SPA). SPA radioligand binding assays
were performed in 96-well plates in a final volume of
250 ll Tris–HCl buffer (50 mM Tris–HCl, 150 mM
NaCl, 5 mM KCl, pH 7.4) using the following condi-
tions: [3H]epibatidine (53 Ci/mmol; Amersham)-
a4b2 = 1 nM, a3b4 = 2 nM; WGA-coated PVT SPA
beads (Amersham)-a4b2 = 1 mg/well, a3b4 = 1.5 mg/
well; membrane protein = 30 lg/well for both assay
types. Non-specific binding (<10% for both assay types)
was determined using 10 lM epibatidine. Reactions
were allowed to equilibrate for 2–4 h at room tempera-
ture prior to reading on a Trilux Scintillation counter
(Perkin Elmer). Data were analyzed using a standard
4-parameter logistic equation (Multicalc, Perkin Elmer)
to provide IC50 values that were converted to Ki values
using the Cheng–Prusoff equation.15
Variation of the heterocyclic substituent has been widely
investigated as a means of modifying the properties of
nicotinic agonists.2b One of the more successful bioisos-
teric replacements for the chloropyridyl ring is the
methylisoxazole ring, introduced by Daly in epiboxidine
(9)9 and, more recently, in the higher homologue homo-
epiboxidine.13 We have recently reported the synthesis
of the isoepiboxidine isomers 10 and 11 based on the
2-azanorbornane framework.8 The syn-isomer (10) has
approximately 13-fold weaker affinity than epibatidine
at the a4b2 receptor (Table 1); this modest reduction
mirrors that reported for epiboxidine (9).9 The potency
of 10 at the a3b4 receptor is lower but the discrimination
is modest. The inactivity of the anti-isomer 11, even at
1000 nM, is not surprising given the distance between
the secondary nitrogen and the heterocycle (Table 3).
Synthesis of isoepibatidine 4
Table 3. Calculated N–N distances for aminesa
anti-7-Bromo-2-azabicyclo[2.2.1]heptane (13). anti-7-
Bromo-2-benzyl-2-azabicyclo[2.2.1]heptane 126 (2.80 g,
0.012 mol) in dry MeOH (60 ml) was hydrogenolyzed
using a standard procedure;8 flash chromatography
(Et2O/MeOH; 9:1) gave 13 as a white crystalline solid
(32%). dH (300 MHz, CDCl3) 1.42–1.63 (m, 2H, H5n,
H6n), 1.97–2.17 (m, 2H, H5x, H6x), 2.40 (dd, 3.8,
3.8 Hz, 1H, H4), 2.68 (d, 9.7 Hz, 1H, H3n), 3.00 (ddd,
J ꢁ 9.7, 3.4, 3.4 Hz, 1H, H3x), 3.37 (dd, J = 3.1,
3.1 Hz, 1H, H1), 4.04 (dd, 1.5, 1.5 Hz, 1H, H7). dC
(75.5 MHz, CDCl3) 26.5, 29.3 (C5, C6), 42.3 (C4), 49.9
(C3), 55.0 (C7), 60.2 (C1). mmax 2976s, 2524 m, 1636 s,
1522m, 1421s cmꢀ1. m/z 176/178 (MH+). C6H11NBr
[MH+] requires 176.00749; observed 176.00751.
Compound
(unprotonated) conformation (A) the C-heterocycle bond (A)
Minimum energy After 180ꢁ rotation about
˚
˚
1
2
4.5
4.6
4.5
4.8
4.3
6.4
5.7
4.4
5.6
5.5
5.6
5.2
5.9
5.3
6.6
6.1
4.8
5.8
4
5
6
7
8
10
11
a Calculated using Spartan Pro; equilibrium geometry by Hartree–
Fock, 6-31G*.

5496
R. White et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5493–5497
N
Cl
Cl
N
Cl
N
Br
B(OH)
2
Ni(cod)
2
+
NR
NR
NR
bathophenanthroline
t
KO Bu, s-butanol
o
40 C 30%
15 R = Boc
17 R = H
16 R = Boc
12 R = Bn
13 R = H
4
R = H
14 R = Boc
0
0
0
anti-2-Boc-7-bromo-2-azabicyclo[2.2.1]heptane
(14).
138.2 (C4 ), 148.8 (C6 ), 149.3, 149.4 (C2 ), 154.2, 154.3
(Boc CO). Further chromatographic separation (Et2O)
allowed the isolation of a sample of the major (syn-) epi-
mer 16 as a yellow oil: dH [300 MHz, CDCl3; signal
descriptors as described above (rotamer ratio ꢂ 45:55).]
1.41, 1.50 (2· br s, 9H, Boc), 1.61–2.04 (m, 4H, H5, H6),
2.69 (br s 1H, H4), 2.95, 3.02 (br s, m, 3H, H3x, H3n,
The procedure described for compounds 19 and 23 in
Ref. 8 was followed using the anti-isomer 13 (344 mg,
1.95 mmol), Boc2O (694 mg, 3.18 mmol), NaHCO3
(595 mg, 6.77 mmol), THF (4 ml) and H2O (12 ml).
After reaction at rt for 72 h, 14 was obtained as a white
crystalline solid (448 mg, 1.62 mmol, 83%) Rf (Et2O/pet-
rol; 1:1) 0.57. dH [300 MHz, CDCl3; where there is signal
duplication because of slow N–CO rotation
(ratio ꢂ 45:55), the minor rotamer signal is shown in
italics.] 1.45 (br s, 9H, Boc), 1.04–1.88, 2.02–2.23 (2·
m, 4H, H5, H6), 2.57 (br s, 1H, H4), 3.05, 3.10 (2· d,
J = 9.6 Hz, 1H, H3n), 3.34 (ddd, J ꢁ 9.6, 3.1, 3.1 Hz,
1H, H3x), 4.05 (br s, 1H, H7), 4.05, 4.25 (2· br s, 1H,
H1). dC (75.5 MHz, CDCl3) 25.6, 28.1 (C5, C6), 28.4
(Boc CH3), 43.0, 43.5 (C4), 51.9 (C7), 51.3, 52.1 (C3),
59.7, 60.8 (C1), 79.7 (Boc C), 153.9 (Boc CO). mmax
2977s, 1692s, 1398s, 1331m, 1304m, 1246m, 1155s,
1111s cmꢀ1. m/z 276/278 (MH+). C11H19NO2Br [MH+]
requires 276.05992; observed 276.05987.
0
H7), 4.48, 4.61 (2· br s, 1H, H1), 7.25 (m, 1H, H5 ), 7.54
(m, 1H, H4 ), 8.31 (m, 1H, H2 ). dC (75.5 MHz, CDCl3)
28.1, 28.2, 30.6, 31.0 (C5, C6), 28.3, 28.5 (Boc CH3),
41.6, 42.4 (C4), 49.1, 49.5 (C7), 49.4, 50.0 (C3), 57.3, 58.4
0
0
0
(C1), 79.3, 79.5 (Boc C), 123.8, 123.9 (C5 ), 138.0, 138.2
0
0
0
(C5 ), 149.3, 149.4 (C2 ), 132.9, 133.0 (C3 ), 154.2 (Boc
CO). mmax 2970s, 2934s, 1696s, 1606m, 1488s, 1406s,
1284s cmꢀ1. m/z 309 (MH+). C16H22N2O2 [MH+] requires
309.13698; observed 309.13692.
syn-7-(6-Chloro-pyridin-3-yl)-2-azabicyclo[2.2.1]heptane
(syn-isoepibatidine) (4). A 75:25 mixture of epimers 15
and 16 (476 mg, 1.54 mmol) was dissolved in EtOAc
(50 ml); EtOH (10.4 ml) and CH3COCl (8.6 ml) were
added with cooling in ice and the reaction mixture was
allowed to reach rt and stirred for 4 h before being evap-
orated to dryness. The crude mixture of epimers of iso-
epibatidine–HCl salts was triturated with CH2Cl2 to
give a sample of pure 4 (94 mg) and a sample containing
17 and 4 (2:1, 107 mg) (53% total yield). Data for free
amine 4: dH (300 MHz, CDCl3) 1.24–1.50, 1.50–1.72,
1.72–1.93 (3· m, 4H, H5, H6), 2.63 (br s, 1H, H4), 2.68
(d, J = 10.1 Hz, 1H, H3n), 2.76 (ddd, J ꢁ 10.1, 2.9,
2.9 Hz, 1H, H3x), 3.71 (br s, 1H, H1), 2.89 (br s, 1H,
anti- and syn-2-Boc-7-(6-chloro-pyridin-3-yl)-2-azabicy-
clo[2.2.1]heptane (15 and 16). Procedure adapted from
Ref. 14. In a glove box, Ni(cod)2 (95 mg, 0.35 mmol)
was placed in a two-necked flask. Bathophenanthroline
(228 mg, 0.69 mmol), 4-chloro-3-pyridyl boronic acid
t
(148 mg, 0.94 mmol) and BuOK (136 mg, 1.22 mmol)
were added, the reaction vessel was evacuated and refilled
with N2 thrice. Dry s-BuOH (10 ml) was added and the
mixture stirred for 10 min at rt under N2; the colour chan-
ged to deep-purple, indicating the formation of the active
complex. A solution of anti-14 (210 mg, 0.76 mmol) in s-
BuOH (2 ml) was added and the resulting mixture stirred
under N2 at 40 ꢁC for 24 h, then cooled and passed
through a short pad of silica. Solvents were removed in
vacuo. Flash chromatography (Et2O) of the crude residue
gave a mixture of 15 and 16 as a pale yellow oil (ꢂ25:75;
anti-:syn-) (69 mg, 0.22 mmol, 30%) Rf 0.49. dH
[300 MHz, CDCl3; signals corresponding to the minor
(anti-) epimer are underlined; where there is signal dupli-
cation because of slow N–CO rotation (ratio ꢂ 45:55), the
minor rotamer signal is shown in italics.] 1.39, 1.48, 1.50
(3· br s, 9H, Boc), 1.44–2.04 (m, 4H, H5, H6), 2.69, 2.91
(2· br s, 1H, H4), 2.95, 3.02, 3.13–3.25, 3.44–3.53 (br s,
br s, m, m, 3H, H3x, H3n), 4.45, 4.48, 4.61 (4· br s, 1H,
0
H7), 7.26 (s, 1H, H2 ), 7.69 (dd, J = 3.1, 0.5 Hz, 1H,
0
0
H4 ), 8.36 (d, J = 2.4 Hz, 1H, H5 ). dC (75.5 MHz,
CDCl3) 29.1, 32.9 (C5, C6), 40.7 (C4), 48.9 (C3), 49.8
(C7), 58.4 (C1), 124.0, 138.5, 149.4 (pyridyl). m/z 209/
211 (MH+). mmax 2970s, 2750w, 2706w, 1622m, 1588w,
1561w cmꢀ1. m/z 209 (MH+). C11H14N2Cl [MH+] re-
quires 209.08455; observed 209.08454. Tosic acid salt
of 4: mp 186–190 ꢁC. Analysis: C18H21N2O3SCl re-
quires: C, 56.76; H, 5.56; N, 7.35; observed: C, 56.82;
H, 5.46; N, 7.26.
References and notes
1. Spande, T. F.; Garraffo, H. M.; Edwards, M. W.; Yeh, H.
J. C.; Pannell, L.; Daly, J. W. J. Am. Chem. Soc. 1992,
114, 3475.
2. For a review, see: (a) Jensen, A. A.; Frolund, B.; Liljefors,
T.; Krogsgaard-Larsen, P. J. Med. Chem. 2005, 7, 565;
For a review of epibatidine structure–activity relation-
ships, see: (b) Carroll, F. I. Bioorg. Med. Chem. Lett. 2004,
0
0
H1), 7.22–7.33 (m, 1H, H5 ), 7.47–7.60 (m, 1H, H4 ),
0
8.23–8.32 (m, 1H, H2 ). dC (75.5 MHz, CDCl3) 28.1,
28.2, 30.6, 31.0 (C5, C6), 28.3, 28.5 (Boc CH3), 39.0,
39.4, 41.5, 42.4 (C4), 49.1, 49.5 (C7), 53.3, 53.8, 49.4,
50.0 (C3), 57.2, 58.0, 58.4, 59.0 (C1), 79.3, 79.5 (Boc C),
0
123.8, 123.9, 124.0 (C5), 132.5, 132.8, 132.9 (C3 ), 138.0,

R. White et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5493–5497
5497
14, 1889, and 5713. For leading references to epibatidine
analogue synthesis see: Carroll, F. I.; Lee, J. R.; Navarro,
H. A.; Ma, W.; Brieaddy, L. E.; Abraham, P.; Damaj, M.
I.; Martin, B. R. J. Med. Chem. 2002, 45, 4755; Bunnelle,
W. H.; Dart, M. J.; Schrimpf, M. R. Curr. Top. Med.
Chem. 2004, 4, 299.
nitrogen in these systems:Belkacemi, D.; Davies, J. W.;
Malpass, J. R.; Naylor, A.; Smith, C. R. Tetrahedron 1992,
48, 10161; Malpass J. R.; Alkhuraiji, W. unpublished
work.
5. Malpass, J. R.; Hemmings, D. A.; Wallis, A. L.; Fletcher,
S.; Patel, S. J. Chem. Soc., Perkin Trans. 1 2001, 1044.
6. Malpass, J. R.; Handa, S.; White, R. Org. Lett. 2005, 7,
2759.
7. (a) Cox, C. D.; Malpass, J. R.; Rosen, A.; Gordon, J. J.
Chem. Soc., Perkin Trans. 1 2001, 2372. For work with
compound 6 see:(b) Hodgson, D. M.; Maxwell, C. R.;
Wisedale, R.; Matthews, I. R.; Carpenter, K. J.; Dicken-
son, A. H.; Wonnacott, S. J. Chem. Soc., Perkin Trans. 1
2001, 3150.
3. For leading references to pharmacophore models and
binding geometry, see: (a) Cashin, A. L.; Petersson, J.;
Lester, H. A.; Dougherty, D. A. J. Am. Chem. Soc. 2005,
127, 350; Glennon, R. A.; Dukat, M. Bioorg. Med. Chem.
Lett. 2004, 14, 1841. Crystal structure determinations of
proteins which bind acetylcholine; (b) Hansen, S. B.;
Sulzenbacher, G.; Huxford, T.; Marchot, P.; Taylor, P.;
Bourne, Y. EMBO J. 2005, 24, 3635.
4. (a) See key references to earlier work in Ref. 5 (tropanes),
and Ref. 7 (2-azanorbornanes); (b)The ‘bicyclic effect’ for
the 7-nitrogen in the 7-azanorbornyl system was first
proposed by Lehn, J. M.. Fortsch. Chem. Forsch. 1970, 15,
311. This effect was invoked to explain unusually high
nitrogen inversion barriers in this case. Barriers to
inversion at nitrogen can be almost as high in 7-azanor-
bornanes as in aziridines (where the angle strain is much
greater). The suggestion that contributions from ground-
state stabilization in 7-azanorbornanes might be impor-
tant was made by: Nelsen, S. F.; Ippoliti, J. T.; Frigo, T.
B.; Petillo, P. A. J. Am. Chem. Soc. 1989, 111, 1776;
Durrant, M. L.; Malpass, J. R. Tetrahedron 1995, 51, 7063
(footnote 4e). See also discussion in: Davies, J. W.;
Durrant, M. L.; Walker, M. P.; Malpass, J. R. Tetrahe-
dron 1992, 48, 4379. Additional delocalization of electron
density from nitrogen into the bicyclic framework is
supported by the unusual deshielding of the bridging
8. Malpass, J. R.; White, R. J. Org. Chem. 2004, 69, 5328.
9. Badio, B.; Garraffo, H. M.; Plummer, C. V.; Padgett, W.
L.; Daly, J. W. Eur. J. Pharmacol. 1997, 321, 189.
10. White, R.; Malpass, J. R. Abstracts of papers, 230th
National Meeting of the American Chemical Society,
Washington DC, 2005; American Chemical Society:
Washington, DC, 2005; abstract ORGN-645.
11. See: Sivilotti, L.; Colquhoun, D.; Millar, N. S. In Neuronal
Nicotinic Receptors; Clementi, F., Fornasari, D., Gotti, C.,
Eds.; Springer, 2000; Chapter 15, p 379 for discussion of
factors leading to variance in Ki values.
12. Bai, D. L.; Xu, R.; Chu, G. H.; Zhu, X. Z. J. Org. Chem.
1996, 61, 4600.
13. Fitch, R. W.; Pei, X.-F.; Kaneko, Y.; Gupta, T.; Shi, D.;
Federova, I.; Daly, J. W. Bioorg. Med. Chem. 2004, 12, 179.
14. Zhou, J.; Fu, G. C. J. Am. Chem. Soc. 2004, 126, 1340.
15. Cheng, Y.; Prusoff, W. H. Biochem. Pharmacol. 1973, 22,
3099.