Mono- and disubstituted-3,8-diazabicyclo[3.2.1]octane derivatives as analgesics structurally related to epibatidine: Synthesis, activity, and modeling

Article abstract of DOI:10.1021/jm970427p

A series of 3,8-diazabicyclo[3.2.1]octanes substituted either at the 3 position compounds 1) or at the 8 position (compounds 2) by a chlorinated heteroaryl ring were synthesized, as potential analogues of the potent natural analgesic epibatidine. When tested in the hot plate assay, the majority of the compounds showed significant effects, the most interesting being the 3-(6-chloro-3-pyridazinyl)-3,8-diazabicyclo[3.2.1]octane (la). At a subcutaneous dose of 1 mg/kg, 1a induced a significant increase in the pain threshold, its action lasting for about 45 min. 1a also demonstrated good protection at a dose of 5 mg/kg in the mouse abdominal constriction test, while at 20 mg/kg it completely prevented the constrictions in the animals. Administration of naloxone (1 mg/kg ip) did not antagonize its antinociception while mecamylamine (2 mg/kg ip) did, thus suggesting the involvement of the nicotinic system in its action. Binding studies confirmed high affinity for the αβ₂ nAChR subtype (K(i) = 4.1±0.21 nM). nAChR functional activity studies on three different cell lines showed that 1a was devoid of any activity at the neuromuscular junction. Finally, due to the analogy in their pharmacological profile with that of epibatidine, compounds were compared from a structural and conformational point of view through theoretical calculations and high-field ¹H NMR spectroscopy. Results indicate that all of them present one conformation similar to that of epibatidine.

Full text of DOI:10.1021/jm970427p

674
J . Med. Chem. 1998, 41, 674-681
Mon o- a n d Disu bstitu ted -3,8-d ia za bicyclo[3.2.1]octa n e Der iva tives a s An a lgesics
Str u ctu r a lly Rela ted to Ep iba tid in e: Syn th esis, Activity, a n d Mod elin g
Daniela Barlocco,*,† Giorgio Cignarella,† Donatella Tondi,† Paola Vianello,† Stefania Villa,† Alessandro Bartolini,‡
Carla Ghelardini,‡ Nicoletta Galeotti,‡ David J . Anderson,§ Theresa A. Kuntzweiler,§ Diego Colombo,| and
Lucio Toma
Istituto di Chimica Farmaceutica e Tossicologica, Universita` Degli Studi di Milano, Viale Abruzzi 42, 20131 Milano, Italy,
Dipartimento di Farmacologia Preclinica e Clinica, Viale Morgagni 65, 50134 Firenze, Italy, Neurological and Urological
Diseases Research, D-47W, Pharmaceutical Products Division, Abbott Laboratories, Abbott Park, Illinois 60064,
Dipartimento di Chimica e Biochimica Medica, Via Saldini 50, 20133 Milano, Italy, and Dipartimento di Chimica Organica,
Via Taramelli 10, 27100 Pavia, Italy
Received J une 27, 1997
A series of 3,8-diazabicyclo[3.2.1]octanes substituted either at the 3 position (compounds 1) or
at the 8 position (compounds 2) by a chlorinated heteroaryl ring were synthesized, as potential
analogues of the potent natural analgesic epibatidine. When tested in the hot plate assay, the
majority of the compounds showed significant effects, the most interesting being the 3-(6-chloro-
3-pyridazinyl)-3,8-diazabicyclo[3.2.1]octane (1a ). At a subcutaneous dose of 1 mg/kg, 1a induced
a significant increase in the pain threshold, its action lasting for about 45 min. 1a also
demonstrated good protection at a dose of 5 mg/kg in the mouse abdominal constriction test,
while at 20 mg/kg it completely prevented the constrictions in the animals. Administration of
naloxone (1 mg/kg ip) did not antagonize its antinociception while mecamylamine (2 mg/kg ip)
did, thus suggesting the involvement of the nicotinic system in its action. Binding studies
confirmed high affinity for the R4â2 nAChR subtype (Ki ) 4.1 ( 0.21 nM). nAChR functional
activity studies on three different cell lines showed that 1a was devoid of any activity at the
neuromuscular junction. Finally, due to the analogy in their pharmacological profile with that
of epibatidine, compounds were compared from a structural and conformational point of view
1
through theoretical calculations and high-field H NMR spectroscopy. Results indicate that
all of them present one conformation similar to that of epibatidine.
Ch a r t 1
In tr od u ction
Epibatidine (Chart 1) was first isolated by Daly and
co-workers1 from the skin of the ecuadoran poison frog
Epipedobates tricolor and its structure determined as
exo-2-(6-chloro-3-pyridyl)-7-azabicyclo[2.2.1]heptane. The
alkaloid was shown to be 200-fold more potent than
morphine in the hot plate test with a non-opioid and
non-cholinergic muscarinic mechanism of action, as
evidenced by the inability of naloxone and scopolamine
to prevent it.1-3 On the contrary, its antinociception
was fully blocked by the nicotinic antagonists mecamyl-
amine and dihydro-â-erythroidine.2,3 Epibatidine dis-
placed [3H]cytisine from R4â2 nAChR in rat brain
membranes with high potency (Ki ) 0.043 nM),2,4 while
it exhibited much lower affinity for the muscarinic
ligand binding site ([3H]-N-methylscopolamine) in rat
cortical membranes.3
In the course of our studies on novel analgesics, struc-
turally unrelated to morphine, we developed a class of
3,8-disubstituted-3,8-diazabicyclo[3.2.1]octanes (I) pro-
vided with significant affinity and selectivity for the
µ-opioid receptors and potent analgesic activity in the
hot plate test.5 The presence in epibatidine of the 7-aza-
bicyclo[2.2.1]heptane system prompted us to verify if the
skeleton of the diazabicyclooctane could be a proper
substrate to develop a new series of epibatidine ana-
logues (1), provided with similar analgesic properties
and mechanism of action.
* To whom correspondence should be addressed.
† Istituto di Chimica Farmaceutica e Tossicologica.
‡ Dipartimento di Farmacologia Preclinica e Clinica.
§ Abbott Laboratories.
Compounds 1 and their isomers 2 were synthesized
and tested for their antinociceptive activity and nicotine
| Dipartimento di Chimica e Biochemica Medica.
Dipartimento di Chimica Organica.
S0022-2623(97)00427-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/30/1998

Novel Non-Morphine-like Analgesics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 5 675
Sch em e 1a
Sch em e 2a
a
(a) Di-tert-butyl dicarbonate/CH2Cl2; (b) 10% Pd-C/H2; (c)
ArCl/toluene/TEA/∆; (d) Et2O/HCl; (e) (CH3CH2CO)2O/CH2Cl2/∆;
(f) C2H5I/K2CO3/acetone/∆. Ar for 1 and 2: a , 6-chloro-3-pyridazi-
nyl; d , 6-chloro-2-(methylthio)-4-pyrimidinyl; e, 2-chloro-4-pyri-
midinyl; f, 4-chloro-2-pyrimidinyl; g, 6-chloro-4-pyrimidinyl; h ,
5-chloro-3-pyrazinyl; i, 3-chloro-2-pyridyl; j, 6-chloro-2-pyridyl.
receptor binding affinity. Moreover, they were inves-
tigated from a structural and conformational point of
1
view through theoretical calculations and high-field H
NMR spectroscopy.
Ch em istr y
The key intermediate in the synthesis of compounds
1 was the previously reported 3,6 which was first
protected at N8 with di-tert-butyl dicarbonate in dichlo-
romethane (compound 4) and then transformed into 5
by catalytic debenzylation. Treatment of the latter with
an equimolar amount of the appropriate heterocycle in
refluxing toluene and in the presence of equimolar
triethylamine gave the Boc derivatives 6, which were
finally deprotected by stirring with a solution of hydro-
chloric acid in diethyl ether to give the desired 1a ,d -j,
as their hydrochlorides. Compounds 1b,c could easily
be prepared from 1a by condensing with propionic
anhydride or iodoethane, respectively (Scheme 1).
To prepare the isomers 2a ,b, the known 77 was
condensed with 3,6-dichloropyridazine to form 2b, fol-
lowed by alkaline hydrolysis to give 2a (Scheme 2,
method a). Though in the case of 2a the chlorine in
position 6 of the pyridazine ring was resistant to the
alkaline medium,8,9 to avoid a possible replacement of
the halogen by the hydroxy group on a different sub-
strate, a more general synthetic pathway was devised,
as depicted in method b). Accordingly, 7 was reacted
with benzyl chloride to give 8, which was hydrolyzed in
acidic medium to 9, protected by Boc (10), debenzylated
(11), condensed with 3,6-dichloropyridazine (12), and
finally converted into 2a (as its hydrochloride) by
stirring overnight with a solution of hydrochloric acid
in diethyl ether. It should be noted that, though a much
greater number of steps were required in method b with
respect to method a, the final yields were almost com-
parable. Finally, 2l was easily obtained by condensing
3 with 3,6-dichloropyridazine in refluxing toluene. See
Table 1 for data of 1 and 2.
a
(a) 2,6-Dichloropyridazine/toluene/TEA; (b) 2 N NaOH/∆; (c)
BnCl/K2CO3/acetone/∆; (d) 4 N HCl/∆; (e) di-tert-butyl dicarbonate/
CH2Cl2; (f) 10% Pd-C/H2; (g) Et2O/HCl.
and Figures 1-3. Binding toward brain nicotinic recep-
tors as well as nAChR functional activity was also
examined.
Molecu la r Mod elin g
Epibatidine and compounds 1a ,d -j and 2a , in their
protonated form, were submitted to a modeling study
with the aid of theoretical calculations and high-field
1H NMR spectroscopy. The preferred conformations and
the heats of formation of all the compounds were
determined using a full geometry optimization carried
out at the RHF level with the semiempirical AM1
method10 implemented in the HyperChem program.11
All the degrees of conformational freedom were exam-
ined, in particular rotation of the aryl moiety and chair-
boat interconversion in the case of 1 and 2. Figure 4
shows the 3D plots and the calculated heats of formation
of the significant conformers of epibatidine and com-
pounds 1a and 2a , while Table C (see Supporting
Information) reports their selected geometrical and
conformational features.
P h a r m a cology
All the compounds were tested in vivo in the mouse
hot plate test (thermal stimulus) in order to determine
their potential analgesic potency and efficacy. They
were used at doses unable to modify normal behavior.
The integrity of motor coordination was evaluated by
the rota-rod test. Data are reported in Tables 2 and 3
Resu lts a n d Discu ssion
An a lgesic Activity. The data reported in Table 2
show that, at the highest tested concentration (25 mg/

676 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 5
Ta ble 1. Physicochemical Properties of Compounds 1 and 2
Barlocco et al.
compd
R
Ar
% yield
mp, °C
formula
1a
1b
1c
1d
1e
1f
1 g
1h
1i
H
6-chloro-3-pyridazinyl
6-chloro-3-pyridazinyl
6-chloro-3-pyridazinyl
6-chloro-2-(methylthio)-4-pyrimidinyl
2-chloro-4-pyrimidinylc
4-chloro-2-pyrimidinylc
6-chloro-4-pyrimidinyl
5-chloro-3-pyrazinyl
3-chloro-2-pyridyl
6-chloro-2-pyridyl
6-chloro-3-pyridazinyl
6-chloro-3-pyridazinyl
6-chloro-3-pyridazinyl
46
85b
60b
43
21
19
43
39
48
50
58d
50
48
>240a
C10H13ClN4
C13H17ClN4O
C12H17ClN4
C11H15ClN4S
C10H13ClN4
C10H13ClN4
C10H13ClN4
C10H13ClN4
C11H14ClN3
C11H14ClN3
C10H13ClN4
C13H17ClN4O
COEt
Et
H
H
H
H
H
H
H
132
210-213
>275a
>290a
>280a
165-168a
165-168a
220-225a
>250a
1j
2a
2b
2l
H
COEt
Bn
deca
171-172
192-194
C
17H19ClN4
a
b
d
As the hydrochloride. From 1a . c See ref 24. Method b, Scheme 2.
kg), the majority of the monosubstituted compounds 1
induced a significant increase in the mouse pain thresh-
old (hot plate test), when injected subcutaneously. At
lower doses (10-15 mg/kg) significant activity was still
seen in several derivatives (1a ,d -g,j). The rota-rod test
(see Table B in Supporting Information) showed that
none of them caused any visible change in the normal
behavior of the animals. The most interesting com-
pound was the 3-(6-chloro-3-pyridazinyl)-3,8-diazabicyclo-
[3.2.1]octane (1a ), which retained analgesic potency still
at 1 mg/kg (see Figure 1). Its analgesic action peaked
after 15-30 min and remained almost unchanged after
45 min. In addition, 1a exhibited the highest ratio (30)
between the maximal nontoxic dose (MNTD) and the
minimal analgesic dose (MAD). For the other active
drugs this value was never higher than 5 (see Table 3).
Insertion of an ethyl group at position 8 of 1a gave a
much weaker compound (1c), while substitution at the
same position by a propionyl group (1b) brought about
a complete loss of activity. Furthermore, shifting the
chloropyridazinyl ring from the 3 to the 8 position
(compound 2a ) led to a less-active compound. Also in
this case the insertion of a propionyl (2b) as well as of
a benzyl (2l) group caused complete loss of activity. This
indicates that no substituent or a small alkyl group on
the 3 (8) nitrogen, allowing the formation of an am-
monium ion, is an essential requirement for the activity
of this class. Compound 1a was also tested in the mouse
abdominal constriction test. It demonstrated good
protection already at a dose of 5 mg/kg, while at 20 mg/
kg no constrictions were observed in the animals (Figure
2). Compound 1a was then evaluated in comparison
with several known analgesics. The antinociceptive
effect (hot plate test) induced by icv injection of 1a (7.5
µg/mouse), nicotine (2 µg), epibatidine (7.5 ng), morphine
(7 µg), diphenydramine (10 µg), and clomipramine (30
µg) was greater for 1a than for the reference drugs,
which were all tested at the highest dose that did not
impair their rota-rod performance (see Figure 3).
prevention of analgesia by the nicotine antagonist
mecamylamine (2 mg/kg ip), together with binding
studies, suggests the involvement of a nicotinic system.
In Vitr o Assa ys. In binding studies 1a displayed a
Ki of 4.1 ( 0.21 nM (n ) 4; [3H]cytisine competition
assays). Although potent, it demonstrates a 100-fold
lower affinity for the R4â2 nAChR subtype compared to
(()-epibatidine with a Ki of 0.042 ( 0.036 nM (n ) 5)
(Table 4). nAChR functional activity was also investi-
gated. Three cell lines were examined based on their
phenotypic nAChR activities: IMR-32, ganglionic-like;
K177, central nervous system; and TE671, neuromus-
cular junction phenotype. From the Ca2+ dynamics
induced by 1a incubation with each cell line, the
following EC50 values and intrinsic activities (IA, rela-
tive to 100 µM (-)-nicotine) were calculated: IMR-32,
EC50 ) 12.2 ( 1.9 µM, IA ) 122.6 ( 5.2%; K177, EC50
) 24.0 ( 3.1 µM, IA ) 164.7 ( 20%; and TE671, EC50
> 1000 µM, IA ) 9.6 ( 1.6%. (()-Epibatidine exhibited
a high degree of potency and efficacy across all of the
cell lines tested (see Table 4). The functional activity
of both 1a and (()-epibatidine was completely blocked
by 100 µM mecamylamine, a specific nAChR antagonist
(data not presented).
Mod elin g. Epibatidine is a quite rigid molecule with
the rotation of the chloropyridyl group as the only
degree of freedom. This rotation presents a 2-fold
minimum; the corresponding conformers have a quite
similar energy and are separated by a very low barrier.
Compound 1a is more flexible due to the presence of
the six-membered piperazine ring which can assume a
chair or boat conformation and also to the fact that the
aryl substituent is linked to the bicyclic system through
a nitrogen atom instead of a carbon atom as in the case
of epibatidine. So, compound 1a presents four signifi-
cant conformations: two chair (1A,B) and two boat
(1C,D) conformations, with 1C being the global mini-
mum as indicated by calculations. The 3D plots in
Figure 4 show that conformer 1C is also the most
similar to epibatidine. Probably, the stability of this last
conformer is overestimated as calculations are made in
vacuum and a strong hydrogen bond is present between
an ammonium hydrogen atom and the 2′ nitrogen atom
Finally, attempts were made to elucidate its mecha-
nism of action. Administration of naloxone (1 mg/kg ip)
did not antagonize antinociception, thus ruling out any
involvement of opioid receptors. On the contrary, the

Novel Non-Morphine-like Analgesics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 5 677
Ta ble 2. Effects of Compounds 1 and 2 on Mouse Hot Plate Testa
treatment
licking latency (s)
compd
saline
dose mg/kg sc
beforeb treatment
15 minb
30 minb
45 minb
13.6 ( 0.9
(6)
14.7 ( 1.7
(6)
13.9 ( 1.4
(6)
15.0 ( 1.2
(6)
1a
1
5
13.5 ( 1.0
21.5 ( 2.5*
25.6 ( 2.1**
21.5 ( 2.2*
(5)
(5)
(5)
(5)
13.7 ( 1.3
(5)
27.8 ( 2.3**
(5)
28.5 ( 2.4**
(5)
26.5 ( 2.2**
(5)
10
20
20
40
10
25
5
14.2 ( 1.2
32.3 ( 2.5**
38.3 ( 2.6**
33.0 ( 2.2**
(7)
(7)
(7)
(7)
14.3 ( 0.9
(7)
13.8 ( 1.2
(6)
14.5 ( 0.8
(11)
13.2 ( 1.1
(5)
14.3 ( 1.0
(10)
14.7 ( 1.1
(6)
15.1 ( 0.6
(10)
14.5 ( 0.9
(5)
15.3 ( 0.8
(9)
15.0 ( 1.1
(6)
14.6 ( 0.8
(9)
13.6 ( 1.1
(7)
14.9 ( 1.3
(10)
15.1 ( 0.9
(8)
13.6 ( 0.6
(10)
13.7 ( 0.9
(10)
40.3 ( 2.7**
(7)
17.1 ( 1.9
(6)
17.0 ( 1.8
(11)
19.3 ( 1.9*
(5)
26.0 ( 2.5**
(10)
20.3 ( 2.4*
(6)
40.8 ( 2.8**
(10)
21.3 ( 2.0*
(5)
31.1 ( 3.4**
(9)
20.5 ( 1.7*
(6)
26.3 ( 2.6**
(9)
22.2 ( 1.4**
(7)
39.5 ( 5.3**
(10)
19.7 ( 2.2*
(8)
24.6 ( 3.1**
(10)
21.3 ( 2.5*
(10)
39.8 ( 4.1**
(7)
15.6 ( 2.7
(6)
16.4 ( 2.0
(11)
18.4 ( 1.4
(5)
20.5 ( 1.5*
(10)
20.5 ( 2.1*
(6)
43.8 ( 1.2**
(10)
17.4 ( 1.5
(5)
26.0 ( 2.5**
(9)
17.8 ( 1.8
(6)
21.2 ( 2.4**
(9)
19.7 ( 2.9*
(7)
38.5 ( 3.8**
(10)
15.5 ( 1.8
(8)
22.5 ( 1.5*
(10)
20.3 ( 2.1*
(10)
33.8 ( 3.0**
(7)
18.0 ( 1.8
(6)
18.4 ( 1.0
(11)
16.4 ( 1.5
(5)
20.8 ( 1.2*
(10)
18.7 ( 2.6*
(6)
36.5 ( 3.3**
(10)
16.2 ( 1.5
(5)
24.3 ( 1.9**
(9)
15.6 ( 2.2
(6)
22.1 ( 3.0**
(9)
19.6 ( 2.7*
(7)
39.1 ( 3.0**
(10)
15.7 ( 1.3
(8)
22.3 ( 2.2*
(10)
18.3 ( 2.0
(10)
1b
1c
1d
1e
1f
1g
1h
1i
25
3
10
5
10
10
25
5
10
5
25
15
25
10
15
25
10
50
50
15.3 ( 0.6
(10)
13.5 ( 0.8
(6)
14.6 ( 0.5
(12)
12.2 ( 0.4
(5)
15.3 ( 0.8
(5)
15.0 ( 0.9
(5)
15.8 ( 0.4
(7)
14.6 ( 0.6
(6)
39.2 ( 2.4**
(10)
23.6 ( 1.9**
(6)
31.9 ( 2.7**
(12)
17.0 ( 1.7
(5)
23.2 ( 1.6**
(5)
35.5 ( 2.5**
(5)
16.0 ( 0.7
(7)
17 ( 3 ( 0.8
(6)
29.7 ( 3.1**
(10)
22.1 ( 2.5**
(6)
37.0 ( 2.7**
(12)
15.4 ( 0.5
(5)
22.9 ( 2.7**
(5)
34.8 ( 5.8**
(5)
16.7 ( 1.2
(7)
13.0 ( 0.8
(6)
27.5 ( 3.2**
(10)
22.0 ( 2.4**
(6)
34.2 ( 2.9**
(12)
17.4 ( 1.2
(5)
20.6 ( 1.8*
(5)
26.4 ( 4.5**
(5)
16.1 ( 0.7
(7)
14.0 ( 0.6
(6)
1j
2a
2b
2l
14.8 ( 0.6
(12)
17.2 ( 1.0
(12)
15.3 ( 0.8
(12)
17.6 ( 0.8
(12)
a
b
Measured at 52.5 °C as licking latency (s) at various times after treatment. In parentheses is the number of animals; *P < 0.05, **P
< 0.01.
of the aromatic moiety. In solution, in particular in
polar solvents, the difference in energy between the boat
and chair conformers should be lower and the order of
stability could even reverse. Using experimental data
diagnostic of the conformation of the piperazine ring,
the two vicinal coupling constants between the hydrogen
atom at C-1 and the two hydrogen atoms at C-2 have
1A,B but not 1C,D. Thus in aqueous solution the
preferred conformers are 1A,B, though a small contri-
bution of conformer 1C cannot be excluded.
Compounds 1d -j are quite similar to 1a . In all the
cases, in fact, a nitrogen atom ortho with respect to the
carbon atom linked to the bicyclic moiety is present,
thus allowing the conformation of type 1C to become
the global minimum.
1
been measured in the H NMR spectrum of compound
1a in D2O (see Experimental Section). They resulted
in 1.5 and 2.8 Hz and are in agreement with the values
calculated with the Altona equation12 for the conformers
Compound 2a shows three allowed conformers: 2A-C
(see Figure 4); conformer 2B presents the same hydro-
gen bond as 1C and is the most similar to epibatidine.

678 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 5
Barlocco et al.
Ta ble 3. Comparison between the Minimal Analgesic Dose,
Maximal Nontoxic Dose, and Efficacy of 1, 2, and the Reference
Drugs Morphine and Epibatidinea
% analgesic efficacy
compared to
MAD,b
MNTD,
compd
mg/kg sc
mg/kg sc
morphine
epibatidine
morphine
epibatidine
2
20
0.007
30
>60
20
30
15
15
100
82
136
122
100
167
0.001
1
inactive
10
5
3
5
1a
1b
1c
1d
1e
1f
61
149
82
74
182
99
60
75
1g
1h
1i
10
5
10
15
15
inactive
inactive
30
15
40
30
30
>50
>50
127
57
124
116
107
156
70
151
142
130
1j
2a
2b
2lc
F igu r e 2. Dose-response curve of 1a in the mouse abdominal
constriction test. 1a was administered 15 min before the
test. Vertical lines show SEM. *P < 0.01 in comparison with
saline controls. Each point represents the mean of at least 8
mice.
a
Antinociceptive effect was evaluated on the mouse hot plate
test. The maximal analgesic effect of morphine or epibatidine is
b
indicated as 100%. Minimal dose able to induce a statistically
significant increase of the pain threshold. c Administered po.
F igu r e 1. Dose-response curve of 1a in the mouse hot plate
test. 1a was administered 15 min before the test. Vertical lines
show SEM; ∧P < 0.05, *P < 0.01 in comparison with saline
controls. Each point represents the mean of at least 8 mice.
F igu r e 3. Maximum antinociceptive effect of 1a in compari-
son with nicotine, epibatidine, morphine, diphenhydramine,
and clomipramine evaluated in the mouse hot plate test. The
nociceptive responses were recorded 5 min after nicotine icv
injection, 15 min after icv administration of epibatidine,
diphenhydramine and clomipramine, and 30 min after icv
morphine injection. Each column represents the mean of at
least 10 mice. Vertical lines show SEM.
The experimental values of the coupling constants
between H-1 and H2-2 are in agreement with the chair
conformation of the piperazine ring (2A) indicating that
also in this case calculations in vacuum overestimate
the hydrogen bond.
Con clu sion s
In conclusion, 3-(6-chloro-3-pyridazinyl)-3,8-diazabicy-
clo[3.2.1]octane (1a ) and several of its analogues proved
to be interesting analgesics, fully comparable to mor-
phine in their potency, though acting through a non-
opiate mechanism. In comparing the functional profiles
of 1a and (()-epibatidine, the following characteristics
have surfaced: (a) 1a binds with high affinity to the
central nAChR subtype, R4â2, albeit with 100-fold lower
potency than (()-epibatidine; (b) 1a is a full agonist at
both the ganglionic (IMR-32 cells) and central (K177
cells) nAChR subtypes exhibiting a narrow separation
between potencies at these receptors (2-fold difference
in EC50 values); and (c) unlike (()-epibatidine, 1a did
not stimulate the nAChR subtypes expressed in the
TE671 cell line. Thus, the antinociceptive properties
of 1a appear to be nAChR-mediated.13 Moreover, 1a
has the potential of producing antinociceptive effects in
vivo without the potential life-threatening side effects
presented by (()-epibatidine (i.e., cardiovascular pressor
effects and seizures)14 due to the lack of agonist activity
at the neuromuscular junction. Finally, the results of
the modeling indicate that 1a presents one conformation
(1C) similar to that of epibatidine. Thus, in the
hypothesis that 1a competes for the same nicotinic
receptor as epibatidine, 1C could be the bioactive
conformer, even if, according to NMR data, it is not
present in polar solvents.

Novel Non-Morphine-like Analgesics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 5 679
drogen ceased, the catalyst was filtered off, the solvent
evaporated, and the residue (8-(tert-butoxycarbonyl)-3,8-
diazabicyclo[3.2.1]octane, 5; 0.92 g, yield 95%; mp ) 58-61
°C) dissolved in toluene (20 mL). To the solution an equimolar
amount of the appropriate heteroaryl derivative was added
together with triethylamine (0.61 mL), and it was refluxed for
8 h. After cooling, the precipitate was filtered off, the solvent
evaporated, and the residue purified by flash chromatography
(eluent dichloromethane/ethyl acetate, 9/1). The so obtained
3-heteroaryl-8-(tert-butoxycarbonyl)-3,8-diazabicyclo[3.2.1]-
octanes 6 (yield 50-75%) were finally deprotected by treat-
ment with a solution of hydrochloric acid in diethyl ether to
give the desired 1a ,d -j, as their hydrochlorides. Compounds
1b,c were easily obtained from 1a , according to standard
procedures (see Table 1 and Scheme 1).
For 1a (as the hydrochloride): 1H NMR (D2O, 500 MHz) δ
1.90 (m, 2H), 2.01 (m, 2H), 3.39 (dd, 2H, J ) 14.5, 1.5 Hz),
4.01 (dd, 2H, J ) 14.5, 2.8 Hz), 4.17 (m, 2H), 7.41 (d, 1H, J )
9.5 Hz); 7.55 (d, 1H, J ) 9.5 Hz)
8-H et er oa r yl-Su b st it u t ed Dia za b icyclooct a n es 2a ,b .
Meth od a . An equimolar amount of 3-propionyl-3,8-diaza-
bicyclo[3.2.1]octane (7)7 (0.5 g, 3 mmol), 3,6-dichloropyridazine
(0.45 g), and triethylamine (0.41 mL) in toluene (20 mL) was
refluxed for 6 h. After cooling, the precipitate was filtered
off, the solvent evaporated, and the residue purified by
flash chromatography (cyclohexane/ethyl acetate, 95/5) to give
the 3-propionyl-8-(6-chloro-3-pyridazinyl)-3,8-diazabicyclo-
[3.2.1]octane (2b). 2b was then hydrolyzed by 4 N NaOH (5
mL) at 60 °C to 8-(6-chloro-3-pyridazinyl)-3,8-diazabicyclo-
[3.2.1]octane (2a ). The corresponding hydrochloride was
prepared by hydrochloric acid in diethyl ether (see Table 1 for
data)
F igu r e 4. Three-dimensional plots of the significant conform-
ers of epibatidine and of compounds 1a and 2a . In parentheses
are the heats of formation (∆Hf, kcal/mol).
For 2a (as the hydrochloride): 1H NMR (D2O, 500 MHz) δ
1.90 (m, 2H), 2.21 (m, 2H), 3.26 (dd, 2H, J ) 13.0, 1.7 Hz),
3.28 (dd, 2H, J ) 13.0, 2.3 Hz), 4.75 (m, 2H), 7.50 (d, 1H, J )
9.5 Hz);,7.62 (d, 1H, J ) 9.5 Hz).
Ta ble 4. Cholinergic Interactions of 1a and (()-Epibatidinea
assay
1a
(()-epibatidine
Met h od b . An equimolar amount of 3-propionyl-3,8-
diazabicyclo[3.2.1]octane (7)7 (0.7 g, 4.2 mmol), benzyl chloride
(0.53 g), and K2CO3 (0.57 g) in acetone (10 mL) was refluxed
overnight. After cooling, the solid was filtered off, the solvent
evaporated, and the residue purified by flash chromatography
(cyclohexane/ethyl acetate, 95/5) to give 3-propionyl-8-benzyl-
3,8-diazabicyclo[3.2.1]octane (8) (0.97 g, yield 90%). 8 was
hydrolyzed by refluxing with 4 N HCl (3 mL) for 3 h; the
mixture was cooled, brought to pH 10, and extracted with
CH2Cl2 (3 × 20 mL). After evaporation of the solvent the so
obtained 8-benzyl-3,8-diazabicyclo[3.2.1]octane (9) (0.73 g,
95%) was treated with an equimolar amount of di-tert-butyl
dicarbonate (0.79 g) in anhydrous dichloromethane (10 mL)
and stirred overnight at room temperature under nitrogen. The
solvent was evaporated and the residue purified by flash
chromatography (eluent cyclohexane/ethyl acetate, 98/2) to
give compound 10 (1.06 g, 97%). It was dissolved in ethanol
(10 mL). To this was added 10% Pd-C (0.1 g), and the mixture
was hydrogenated at room temperature and external pressure.
After the uptake of hydrogen ceased, the catalyst was filtered
off and the solvent evaporated to give 3-(tert-butoxycarbonyl)-
3,8-diazabicyclo[3.2.1]octane (11) in quantitative yield (0.74
g). Condensation of the latter with 3,6-dichloropyridazine was
performed as previously described for 2b , though a longer
time was required (24 h), to give 3-(tert-butoxycarbonyl)-8-(6-
chloro-3-pyridazinyl)diazabicyclo[3.2.1]octane (12) (85%), which
was purified by flash chromatography (eluent CH2Cl2/ethyl
acetate, 9/1) and finally converted into 8-(6-chloro-3-pyrid-
azinyl)diazabicyclo[3.2.1]octane (2a ) (as its hydrochloride) in
quantitative yield, by a solution of hydrochloric acid in diethyl
ether.
[3H]cytisine
Ki, µM
4.1 ( 0.21
0.042 ( 0.036
FLIPR,b IMR-32
EC50, µM
IA,c %
12.2 ( 1.9
122.6 ( 5.2
0.034 ( 0.003
145.6 ( 23
FLIPR, K177
EC50, µM
IA, %
FLIPR, TE671
EC50, µM
IA, %
24.0 ( 3.1
164.7 ( 20
0.35 ( 0.02
148.1 ( 17
>1000
9.6 ( 1.6
1.26 ( 0.11
79.5 ( 13
a
Rat cerebral cortical membranes. See also the Experimental
Section. Fluorescence imaging plate reader. c Intrinsic activity,
b
relative to 100 µM (-)-nicotine.
Exp er im en ta l Section
Ch em istr y. Melting points were determined on a Bu¨chi
510 capillary melting point apparatus and are uncorrected.
Analyses indicated by the symbols were within (0.4 of the
theoretical values. 1H NMR spectra were recorded on a Bruker
AC200 or Bruker AM500 spectrometer; chemical shifts are
reported as δ (ppm) relative to tetramethylsilane. TLC on
silica gel plates was used to check product purity. Silica gel
60 (Merck; 70-230 mesh) was used for column chromatogra-
phy.
Gen er a l P r oced u r e for th e Syn th esis of th e 3-Het-
er oa r yl-Su bstitu ted Dia za bicycloocta n es 1a -j. An equi-
molar mixture of 3-benzyl-3,8-diazabicyclo[3.2.1]octane 36 (1
g; 5 mmol) and di-tert-butyl dicarbonate (1.1 g) in anhydrous
dichloromethane (10 mL) was stirred overnight under nitrogen
at room temperature. The solvent was evaporated and the
residue purified by flash chromatography (eluent cyclohexane/
ethyl acetate, 98/2) to give 3-benzyl-8-(tert-butoxycarbonyl)-
3,8-diazabicyclo[3.2.1]octane (4) (1.39 g, 92%; mp ) 60-61 °C).
4 was dissolved in ethanol (10 mL), to the solution 10% Pd-C
(0.14 g) was added, and the mixture was hydrogenated at room
temperature and external pressure. After the uptake of hy-
3-Ben zyl-8-(6-ch lor o-3-pyr ida zin yl)dia za bicyclo[3.2.1]-
octane (2l) was easily obtained from 3 by refluxing with
equimolar amounts of 3,6-dichloropyridazine and triethyl-
amine in toluene for 6 h (yield 65%).
P h a r m a cology. Morphine (hydrochloride) was obtained
from USL 10D (Florence, Italy). Clomipramine (hydrochloride)
was purchased from Ciba Geigy (Basel, Switzerland) and

680 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 5
Barlocco et al.
diphenhydramine (hydrochloride) from De Angeli (Milan,
Italy). (()-Epibatidine was obtained from Research Biochemi-
cals International (Natick, MA). (-)-Nicotine (hydrogen tar-
trate salt) and mecamylamine (hydrochloride) were purchased
from Sigma Chemical Co. (St. Louis, MO). [3H]-(-)-Cytisine
was obtained from NEN Life Sciences (Boston, MA).
In Vivo Exp er im en ts. Analgesic activity was evaluated
in male Swiss-Webster mice (22-28 g; Morini, San Polo
d’Enza, Italy). Mice were kept at 23 ( 1 °C, with a 12-h light/
dark cycle, light at 7 a.m., with food and water ad libitum. All
experiments were carried out according to the guidelines of
the European Community Council on animal care.
HEK-293 cells (ATCC, Rockville, MD) stably expressing the
human R4â2 nAChR subunit combination, K177 cells, were
maintained as previously described.22 TE671 cells (ATCC,
Rockville, MD) were grown according to the protocol of Lukas.23
Cells were plated out at a density of 1 × 10-6 cells/well on
black-walled, clear-bottomed 96-well plates and used ap-
proximately 72 h after plating. All plates were coated with
poly(ethylenimine) to aid in the adherence of the cells to the
plate.
4. n ACh R F u n ction a l Activity: Changes in the intrac-
ellular free Ca2+ concentrations were measured using the
calcium-chelating dye Fluo-3 (Molecular Probes, Eugene, OR)
in conjunction with a FLIPR (Molecular Devices, Sunnyvale,
CA). The cell permeant acetoxymethyl (AM) ester form of
Fluo-3 was prepared to a concentration of 1 mM in anhydrous
DMSO and 10% pluronic acid. The dye was then diluted to a
final concentration of 4 mM in growth media and placed on
the cells for 1 h at 37 °C. Black-walled 96-well plates were
utilized to reduce light scattering. The unincorporated dye
was removed from the cells by excessive washing with Hepes-
salts buffer (composition, mM: Hepes, 20; NaCl, 120; KCl, 5;
MgCl2, 1; glucose, 5; and CaCl2, 5; pH 7.4 at 25 °C). After
addition of various concentrations of agonists, the Ca2+
dynamics were observed in the FLIPR apparatus equipped
with an argon laser (wavelength, 480 nm), an automated 96-
channel pipettor, and a CCD camera. The intensity of the
fluorescence was captured by the CCD camera every second
for the first minute following the agonist addition with
additional readings every 5 s for a total time period of 5 min.
These images were digitally transferred to an interfaced PC
and changes in fluorescence intensity processed for each
well. The exposure setting of the camera was 0.4 s with f
stop setting of 2 µm. The percent maximal intensity relative
to that induced by 100 µM nicotine was plotted against
the concentration of agonist, and EC50 values were calcu-
lated. Independent measurements of 100 µM nicotine (100%)
and unloaded cells (0%) were performed on each plate of
cells with an average range of 20 000 fluorescence units.
These controls allowed for normalization across several plates
of cells.
Hot plate method described by O’Callaghan and Holtzman15
was used to assess the potential analgesic activity of the test
compounds. The plate temperature was fixed at 52.5 ( 0.1
°C. Mice with a licking latency below 12 and over 18 s in the
test before drug administration (30%) were rejected. An
arbitrary cutoff time of 45 s was adopted. The number of mice
treated in each test varied from 5 to 12. To evaluate the
analgesic efficacy of the new products, their level of analgesia
reached after the injection of the maximal dose unable to
modify mouse rota-rod performance was compared to that of
morphine (8 mg/kg sc) or epibatidine (5 µg/kg sc), taken as
reference compounds. The integrity of motor coordination was
assessed on the basis of the endurance time of the animals on
the rotating rod, according to Vaugh.16 Analgesic efficacy of
each compound (X) is expressed as percentage of that of
epibatidine (5 µg/kg sc) or morphine (8 mg/kg sc). The
calculation was performed using the following formula:
% of analgesic efficacy of X ) (T1 - T2)/(T3 - T4) × 0.100
where T1 is maximum reaction time of X, T2 is pretest reaction
time of X, T3 is maximum reaction time of morphine or
epibatidine, and T4 is pretest reaction time of morphine or
epibatidine.
The maximal nontoxic dose is considered the highest dose
of X which does not cause any visible change in animal
behavior such as tremors, convulsions, hypomotility, etc., so
that the researchers were unable to distinguish between
treated and untreated mice. The evaluation of the behavioral
parameters was performed according to Irwing17 on a group
of at least 10 animals. The abdominal constriction test was
performed according to Koster et al.18 Standard errors on the
values expressed as percentage were not evaluated. Original
data, however, have been statistically processed by employing
Dunnett’s two-tailed test in order to verify the significance of
the differences between the means shown by treated mice at
the maximum reaction time and the pretest reaction time.
Differences were considered statistically significant when P
< 0.05. Percent values were calculated only for those differ-
ences that were statistically significant. In other case, drugs
were considered inactive. Intracerebroventricular (icv) ad-
ministration was performed under ether anesthesia using
isotonic saline as a solvent, according to the method described
by Haley and McCormick.19
Ack n ow led gm en t. The authors are deeply indebted
to Dr. Michael Williams and Dr. Michael Meyer (Abbott
Laboratories) for their support of the in vitro evaluation
of the compounds. Financial support by the Ministero
dell’Universita` e della Ricerca Scientifica e Tecnologica
(Rome) is also gratefully acknowledged.
1
Su p p or tin g In for m a tion Ava ila ble: Tables of H NMR
data, effects of 1 and 2 in the rota-rod test, and results of
modeling (3 pages). See any current masthead page for
ordering information.
Refer en ces
(1) Spande, T. F.; Garraffo, M.; Edwards, M. W.; Yeh, J . C.; Pannell,
L.; Daly, J . W. Epibatidine: A novel (chloropyridyl)azabicyclo-
heptane with potent analgesic activity from ecuadoran poison
frog. J . Am. Chem. Soc. 1992, 114, 3475-3478.
(2) Qian, C.; Li, T.; Shen, T. Y.; Libertine-Garahan, L.; Eckman, J .;
Biftu, T.; Ip, S. Epibatidine is a nicotinic analgesic. Eur. J .
Pharmacol. 1993, 250, R13-R14.
(3) Rupniak, N. M. J .; Patel, S.; Marwood, R.; Webb, J .; Traynor, J .
R.; Elliott, J .; Freedman, S. B.; Flechter, S. R.; Hill, R. G.
Antinociceptive and toxic effect of (+)epibatidine oxalate at-
tributable to nicotinic agonist activity. Br. J . Pharmacol. 1994,
113, 1487-1493.
(4) Sullivan, J . P.; Decker, M. W.; Brioni, J . D.; Donnelly-Roberts,
D.; Anderson, D. J .; Bannon, A. W.; Kang, C. H.; Adams, P.;
Piattoni-Kaplan, M.; Bukley, M. J .; Gopalakrishnan, M.; Wil-
liams, M.; Arneric, S. P. (()-Epibatidine elicits a diversity of in
vitro and in vivo effects mediated by nicotinic acetylcholine
receptors. J . Pharmacol. Exp. Ther. 1994, 271, 624-631.
(5) Cignarella, G.; Barlocco, D.; Tranquillini, M. E.; Volterra, A.;
Brunello, N.; Racagni, G. Interaction of 3,8-diazabyclo[3.2.1]-
octanes with µ and δ opiod receptors. Pharmacol. Res. Commun.
1988, 20, 383-393.
In Vit r o E xp er im en t s. 1. Mem br a n e P r ep a r a t ion s:
Rat cerebral cortical membranes were purchased from ABS
Inc. (Wilmington, DE). Prior to use, the frozen membrane
pellets were slowly thawed, washed, and resuspended in 30
volumes of assay buffer (composition, mM: Tris HCl, 50; NaCl,
120; KCl, 5; MgCl2, 1°, and CaCl2, 2.5; pH 7.4 at 4 °C). The
homogenate was centrifuged at 45000g for 20 min at 4 °C and
the pellet resuspended in ice-cold buffer.
2. [3H]-(-)-Cytisin e Bin d in g: Binding conditions were as
previously described.20 Samples containing 150-200 µg of
protein, 0.7 nM [3H]-(-)-cytisine (30 Ci/mmol), and the various
concentrations of the nAChR modulators were incubated in a
final volume of 500 µL for 75 min at 4 °C in triplicate.
Nonspecific binding was determined in the presence of 10 µM
(-)-nicotine.
3. Tissu e Cu ltu r e: Cells of the IMR-32 human neuroblas-
toma clonal line (ATCC, Rockville, MD) were maintained in a
log phase of growth according to established procedures.21

Novel Non-Morphine-like Analgesics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 5 681
(6) Cignarella, G.; Nathansohn, G. G. Bicyclic homologs of pipera-
zine. II. Synthesis of 3,8-diazabyclo[3.2.1]octane. New synthesis
of 8-methyl-3,8-diazabyclo[3.2.1]octane. J . Org. Chem. 1961, 26,
2747-2750.
(7) Cignarella, G.; Testa, E.; Pasqualucci, C. R. Thermal and
chemical intramolecular NfN acyl migration in 8-acyl-3,8-
diazabicyclo[3.2.1]octanes. Tetrahedron 1963, 19, 143-148
(8) Stek, E.; Brundage, R.; Fletcher, L. Pyridazine derivatives. II.
An improved synthesis of 3-aminopyridazine. J . Am. Chem. Soc.
1954, 76, 3225-3226.
(9) Druey, J .; Meier, K.; Eichenberger, K. Heilmittelchemische
Studien in der heterocyclischen Reihe. Pyridazine I. Derivate
des cyclischen Maleinsa¨ure- und citraconsa¨ure-hydrazids. Helv.
Chim. Acta 1954, 37, 121.
(10) Dewar, M. J . S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J . J . P.
AM1: A new general purpose quantum mechanical molecular
model. J . Am. Chem. Soc. 1985, 107, 3902-3909.
(17) Irwing, S. Comprehensive observational assessment: Ia. A
systematic, quantitative procedure for assessing the behavioural
and physiologic state of mouse. Psycopharmacology 1966, 13,
222-257.
(18) Koster, R.; Anderson, M.; De Beer, E. J . Acetic acid for analgesic
screening. Fed. Proc. 1959, 18, 412
(19) Haley, T. J .; McCormick, W. G. Pharmacological effects produced
by intracerebral injection of drugs in the conscious mouse. Br.
J . Pharmacol. Chemother. 1957, 12, 12-15.
(20) Anderson, D. J .; Arneric, S. P. Nicotinic receptor binding of [3H]-
cytisine, [3H]nicotine and [3H]methylcarbamylcholine. Rat Brain
Eur. J . Pharm. 1994, 253, 261-267.
(21) Lukas, R. J . Expression of ganglia-type nicotinic acetylcholine
receptors and nicotinic ligand binding sites by cells of the IMR-
32 human neuroblastoma clonal cell line. J . Pharmacol. Exp.
Ther. 1993, 265, 294-302.
(22) Gopalakrishnan, M.; Monteggia, L. M.; Anderson, D. J .; Arneric,
S. P.; Sullivan, J . P. Stable expression, pharmacologic properties
and regulation of the human neuronal nicotinic acetylcholine
R4â2 receptor. J . Pharmacol. Exp. Ther. 1996, 276, 289-297.
(23) Lukas, R. J . Characterization of curaremimetic neurotoxin
binding sites on membrane fractions derived from the human
medulloblastoma clonal line, TE671. J . Neurochem. 1986, 46,
1936-1941.
(11) HyperChem is a package from HyperCube, Inc., Canada.
(12) Haasnoot, C. A. G.; de Leeuw, F. A. A. M.; Altona, C. The
relatioship between proton-proton NMR coupling constants and
substituent electronegativities. I. An empirical generalization
of the Karplus equation. Tetrahedron 1980, 36, 2783-2792.
(13) Ghelardini, C.; Galeotti, N.; Barlocco, D.; Bartolini, A. Antinoci-
ceptive profile of the new nicotinic agonist DBO-83. Drug Dev.
Res. 1997, 40, 251-258.
(24) The structure of compounds 1e,f has been assigned on the basis
of NOE experiments, performed at 500 MHz in D2O on their
(14) Sullivan, J . P.; Bannon, A. W. Epibatidine: Pharmacological
properties of a novel nicotinic acetylcholine receptor agonist and
analgesic agent. CNS Drug Rev. 1996, 2, 21-39.
hydrochlorides, which showed for 1e the enhancement of H-2eq
/
(15) O’Callaghan, J . P.; Holtzman, S. G. Quantification of the
analgesic activity of narcotic antagonist by a modified hot-plate
procedure. J . Pharmacol. Exp. Ther. 1975, 192, 497-505.
(16) Vaught, J .; Pelley, K.; Costa, L. G.; Sether, P.; Enna, S. J . A
comparison of the antinociceptive responses to GABA-receptor
agonists THIP and baclofen. Neuropharmacology 1985, 24, 211-
216.
H-4eq (5%) on irradiation of the aromatic proton at δ 6.81 (H-5′)
and, vice versa, the enhancement of H-5′ (16%) on irradiation
of the signal at δ 4.32 (H-2eq/H-4eq). On the contrary, no
significant enhancement was observed on irradiation at the
aromatic hydrogen atoms at δ 6.75 and 8.10 of 1f.
J M970427P