Benzocondensed derivatives as rigid analogues of the μ-opioid agonist 3(8)-cinnamyl-8(3)-propionyl-3,8-diazabicyclo[3,2,1]octanes: Synthesis, modeling, and affinity

Article abstract of DOI:10.1016/S0014-827X(98)00084-6

A new series of rigid analogues (1a-g, 2a-g) of the previously reported analgesic 3-cinnamyl-8-propionyl-3,8-diazabicyclo[3.2.1]octane (I) and its reverted isomer 3-propionyl-8-cinnamyl (II) were synthesized, in which the cinnamyl substituent is incorporated in benzocondensed bicyclic systems. Binding assays for the affinity towards μ receptors indicated that, while in the reverted series 2 the β-naphthylmethyl (2d) and the benzocycloheptenylmethyl derivative (2g) favorably compared with II, all compounds 1 displayed a μ-affinity lower than that of the parent I. Modeling studies suggest that the flexibility of the cinnamyl side chain plays an important role for activity.

Full text of DOI:10.1016/S0014-827X(98)00084-6

Il Farmaco 53 (1998) 667–674
Benzocondensed derivatives as rigid analogues of the m-opioid
agonist 3(8)-cinnamyl-8(3)-propionyl-3,8-diazabicyclo[3.2.1]octanes:
synthesis, modeling, and affinityꢀ
G. Cignarella a,*, D. Barlocco a, P. Vianello a, S. Villa a, G.A. Pinna b, P. Fadda c,
W. Fratta c, L. Toma d, S. Gessi e
a Istituto di Chimica Farmaceutica e Tossicologica, Viale Abruzzi 42, 20131 Milan, Italy
b Dipartimento Farmaco Chimico Tossicologico, Via Muroni 23, 07100 Sassari, Italy
c Dipartimento di Neuroscienze, Via Porcell 4, 09100 Cagliari, Italy
d Dipartimento Chimica Organica, Via Taramelli 10, 27100 Pa6ia, Italy
e Istituto Farmacologia, Via Fossato di Mortara 19, 44100 Ferrara, Italy
Abstract
A new series of rigid analogues (1a–g, 2a–g) of the previously reported analgesic 3-cinnamyl-8-propionyl-3,8-diazabicy-
clo[3.2.1]octane (I) and its reverted isomer 3-propionyl-8-cinnamyl (II) were synthesized, in which the cinnamyl substituent is
incorporated in benzocondensed bicyclic systems. Binding assays for the affinity towards m receptors indicated that, while in the
reverted series 2 the b-naphthylmethyl (2d) and the benzocycloheptenylmethyl derivative (2g) favorably compared with II, all
compounds 1 displayed a m-affinity lower than that of the parent I. Modeling studies suggest that the flexibility of the cinnamyl
side chain plays an important role for activity. © 1998 Elsevier Science S.A. All rights reserved.
Keywords: Opioid receptors; Analgesic activity; Diazabicyclooctanes
1. Introduction
This research aimed at the evaluation of the effects
on m-affinity determined by the hampered rotation of
the phenyl group out of the plane of the cinnamyl
double bond. Previous studies [2] had shown that mod-
ifications of the cinnamyl substituent of DBO, like
hydrogenation of the double bond or insertion on it of
a methyl group left m-affinity almost unchanged. On the
contrary, it was markedly decreased by substitution of
the double bond with a cyclopropane ring or inversion
of its configuration from trans to cis.
The class of 3,8-diazabicyclo[3.2.1]octanes (DBO)
substituted by a N3 (N8)-arylpropenyl and by a N8
(N3)-propionyl group is provided by central analgesic
properties related to a selective affinity towards m-opiod
receptors. It is however to note that, though m-affinity
of the most active term is similar to or one order of
magnitude lower than that of morphine, their in vivo
potency was found from five to twenty times higher
than that of the reference drug [1].
In this paper we wish to report on a structure–activ-
ity study carried out taking as models 3-cinnamyl-8-
propionyl-DBO (I) and its reverted isomer 3-pro-
pionyl-8-cinnamyl-DBO (II) [1], based on the insertion
of the styrene moiety of the cinnamyl substituent
into benzocondensed systems, to give compounds 1a–g,
2a–g.
This paper describes the synthesis of compounds
1a–g and 2a–g, together with their modeling and the
evaluation of their affinity towards m-opioid receptors.
ꢀ Dedicated to Professor Antonio Maccioni.
* Corresponding author.
0014-827X/98/$ - see front matter © 1998 Elsevier Science S.A. All rights reserved.
PII: S0014-827X(98)00084-6

668
G. Cignarella et al. / Il Farmaco 53 (1998) 667–674
real solution of hydrochloric acid and reduction of the
so obtained by sodium-bis-(2-methoxyethoxy)-
6
aluminumhydride (Red-Al®) in toluene, gave the inter-
mediate 7, finally acylated by propionic anhydride to
the desired 1a,b. The reverted 2a,b were similarly ob-
tained by condensing the appropriate acyl chloride (4)
with 3-benzyl-DBO [4] to 8, which was debenzylated by
catalytic hydrogenation (9) then reduced (10) and
finally acylated as reported for the isomers 1a,b. The
benzothieno derivatives (1c, 2c) were synthesized as
shown in Scheme 2, starting from the known 2-
chloromethyl derivative 11c [5,6], which was condensed
with 8-propionyl-DBO [7] or 3-propionyl-DBO [8] to
give 1c and 2c, respectively. The 2-naphthyl derivatives
1d, 2d were analogously prepared from 11d, easily
obtained by bromination of the commercial 2-methyl-
naphthalene. For the synthesis of 1e–g, 2e–g (Scheme
3), the key chloromethyl derivative (13) was prepared
following literature methods from the appropriate bi-
cyclic ketone [9–11] and finally condensed with 8-propi-
onyl-DBO or 3-propionyl-DBO (see Table 1 for data).
2. Chemistry
Compounds were prepared following three different
methods, depending on the available starting material.
According to Scheme 1, the appropriate carboxylic acid
(3) was treated with thionyl chloride to give the corre-
sponding chloride (4), which was condensed with
8-Boc-3,8-diazabicyclo[3.2.1]octane [3] in refluxing
toluene to give the amide 5. Deprotection by an ethe-
Scheme 1.

G. Cignarella et al. / Il Farmaco 53 (1998) 667–674
669
Scheme 2.
3. Binding
single bonds N3–C9 and C9–C10 represent the only
degrees of freedom which could influence the conforma-
tional behavior of compounds 1 and 2.
Compounds 1a–g, 2a–g together with I and II were
submitted to binding studies on mouse brain ho-
The exploration of the conformational space of these
compounds gave in all the cases similar results. A main
difference between 1 and their isomers 2 could be
observed: while for the former only an equatorial orien-
tation of the N3-aralkyl substituent is allowed, for the
latter both the equatorial and the axial orientations are
permitted. However, no significant difference in the
conformational behavior is found among the seven
compounds belonging to the 1 series as well as among
those belonging to the 2 series: the freedom of rotation
around the two N3–C9 and C9–C10 single bonds is
high and quite similar for all the compounds. A com-
parison of 1 and 2 with the parent compounds I and II
shows that the insertion of the styrene moiety of the
cinnamyl substituent into a benzocondensed system
does not significantly influence the behavior of the two
above cited single bonds.
3
mogenates in the presence of H-DAMGO [(
D
-Ala2,N-
Me-Phe4,Gly-ol5)enkephalin] as m-selective ligand.
Morphine was used as the reference compound. Their
inhibition constants are reported in Table 2.
4. Modeling
The conformational properties of compounds 1 and
of their isomers 2 were investigated with the MM+
force field of the HyperChem™ package [12]. The
3,8-diazabicyclo[2.2.1]octane system is in both cases
quite rigid. Actually, only a chair conformation can be
assumed by the piperazine ring due to the presence of
the endoethylenic bridge. Nitrogen inversion at the
tertiary amine center and rotation around the two
Scheme 3.

670
G. Cignarella et al. / Il Farmaco 53 (1998) 667–674
Table 1
Physicochemical properties of compounds 1 and 2
Com-
pound
X
Yield
(%)
M.p. (°C)
88–90
Formula
1H NMR l
1a
1b
1c
1d
O
40
40
60
27
C18H22N2O2
C18H23N3O
1.2 (t, 3H); 1.7–2.1 (m, 4H); 2.2–2.5 (m, 4H); 2.6–2.9 (m, 2H); 3.5–3.8
(m, 2H); 4.0–4.2 (bs, 1H); 4.6–4.7 (bs, 1H); 6.6 (s, 1H); 7.1–7.4 (m, 2H);
7.4–7.6 (m, 2H)
1.2 (t, 3H); 1.8–2.1 (m, 4H); 2.1–2.4 (m, 4H); 2.6–2.8 (m, 2H); 3.6 (dd,
2H); 4.0–4.2 (bs, 1H); 4.6–4.7 (bs, 1H); 6.4 (s, 1H); 7.0–7.2 (m, 2H); 7.4
(d, 1H); 7.6 (d, 1H); 8.3 (bs, 1H exch. with D2O)
1.2 (t, 3H); 1.7–2.1 (m, 4H); 2.2–2.4 (m, 4H); 2.7–2.9 (m, 2H); 3.6–3.8
(m, 2H); 4.0–4.1 (bs, 1H); 4.6–4.7 (bs, 1H); 7.1 (s, 1H); 7.2–7.4 (m, 2H);
7.6–7.8 (m, 2H)
1.1 (t, 3H); 1.8–2.2 (m, 4H); 2.3–2.5 (m, 4H); 3.1–3.3 (m, 4H); 4.3–4.4
(bs, 1H); 4.5–4.7 (bs, 1H); 7.6–7.7 (m, 2H); 7.9– 8.1 (m, 4H);
8.2 (s, 1H); 10.9–11.0 (s, 1H exch. with D2O)
NH
63–64
S
208–209a
205–207a
C18H22N2OS·HCl
C20H24N2O·HCl
CHꢀCH
1e
1f
CH2
35
31
16
193–195a
217–218a
155–156a
C19H24N2O·HCl
C20H26N2O·HCl
C21H28N2O·HCl
1.17 (t, 3H); 1.7–2.5 (m, 8H); 2.72 (q, 2H); 3.35 (s, 2H); 3.38 (s, 2H);
4.08 (bs, 1H); 4.65 (bs, 1H); 6.66 (s, 1H); 7.10–7.50 (m, 4H)
1.2 (t, 3H); 1.7–1.9 (m, 4H); 2.1–2.4 (m, 6H); 2.6–2.9 (m, 4H);
3.02 (s, 2H); 4.1 (bs, 1H); 4.7 (bs, 1H); 6.3 (s, 1H) 7.0–7.1 (m, 4H)
1.2 (m, 3H); 1.8–2.1 (m, 6H); 2.2–2.4 (m, 6H); 2.8–2.9 (m, 4H);
3.0 (d, 1H); 4.1 (bs, 1H); 4.4 (d, 1H); 4.7 (bs, 1H); 6.4 (s, 1H); 7.1–7.4
(m, 4H)
(CH2)2
(CH2)3
1g
2a
2b
2c
2d
O
50
45
60
40
156–157a
119–121a
108–109a
195–200a
C18H22N2O2·HCl
C18H23N3O·HCl
1.1 (t, 3H); 1.6–2.1 (m, 2H); 2.2–2.5 (m, 2H); 3.2–3.7 (m, 3H); 3.6–4.1
(m, 4H); 4.2–4.3 (m, 1H); 4.5 (d, 2H); 7.2–7.5 (m, 3H); 7.6 (d, 1H);
7.8 (d, 1H): 11.5 (bs, 1H exch. with D2O)
1.1 (t, 3H); 1.6–2.0 (m, 2H); 2.2–2.4 (m, 2H); 3.2–3.3 (m, 1H); 3.4 (q,
2H); 3.6–3.8 (m, 4H); 4.2–4.4 (m, 1H); 7.0–7.3 (m, 2H); 7.4 (d, 1H);
7.6 (d, 1H); 11.4 (s, 1H exch. with D2O)
1.1 (t, 3H); 1.5–1.7 (m, 2H); 1.8–2.1 (m, 2H); 2.1–2.4 (m, 2H); 2.9–3.0
(m, 1H); 3.2–3.4 (m, 4H); 3.8 (s, 2H); 4.1–4.3 (m, 1H); 7.1 (s, 1H);
7.2–7.4 (m, 2H); 7.7 (d, 1H); 7.8 (d, 1H)
1.1 (t, 3H); 1.8–2.0 (m, 2H); 2.2–2.4 (m, 2H); 3.1–3.2 (m, 1H); 3.2–3.4
(m, 2H); 3.6–3.8 (m, 4H); 4.2–4.3 (m, 1H); 4.5 (s, 2H); 7.6–7.7 (m, 2H);
7.8–8.1 (m, 4H); 8.2 (s, 1H); 10.8 (bs, 1H exch. with D2O)
1.2 (t, 3H); 1.5–2.1 (m, 6H); 2.3 (q, 2H); 2.9 (d, 1H); 3.2 (s, 2H);
3.4–3.5 (m, 4H); 4.2 (d, 1H); 6.7 (s, 1H); 7.2–7.4 (m, 4H)
1.1 (t, 3H); 1.5–1.6 (m, 2H); 1.9–2.0 (m, 2H); 2.3–2.4 (m, 4H); 2.8–2.9
(m, 2H); 3.1 (s, 2H); 3.2–3.4 (m, 5H); 4.2 (d, 1H); 6.4 (s, 1H) 7.0–7.1
(m, 4H)
NH
S
C18H22N2OS·HCl
CHꢀCH
C20H24N2O·HCl
2e
2f
CH2
20
58
182–184a
183–185a
C19H24N2O·HCl
C20H26N2O·HCl
(CH2)2
2g
(CH2)3
40
192–193a
C21H28N2O·HCl
1.2 (t, 3H); 1.6–1.7 (m, 2H); 1.9–2.1 (m, 4H); 2.2–2.5 (m, 4H); 2.7–3.0
(m, 4H); 3.2–3.5 (m, 5H); 4.2 (d, 1H); 6.4 (s, 1H); 7.0–7.2 (m, 4H)
a As the hydrochloride.
However, this structural modification involves two
main differences with respect to the cinnamyl: the ham-
pered rotation of the phenyl group out of the plane of
the double bond and the modified orientation of the
bicyclic substituent. As examples, in Fig. 1 the superim-
positions of the most populated conformers of com-
pounds 1a, 1b, 1g are reported with respect to I. Besides
a complete correspondence of the diazabicyclooctane
moiety and of the N3–C9, C9–C10, C10–C11 bonds,
different orientations of the benzo ring of 1 with respect
to the phenyl ring of I can be observed.
5. Discussion
The data reported in Table 2 clearly show that none
of compounds 1 is as active as the model I. This
lowering of affinity could be due to the fact that the

G. Cignarella et al. / Il Farmaco 53 (1998) 667–674
671
Table 2
Inhibition constants of morphine, I, II and compounds 1 and 2
towards m-opioid receptors
Comp.
3H-DAMGOa
Comp.
3H-DAMGOa
Ki (nM)
Ki (nM)
Ib
1a
1b
1c
1d
1e
1f
1g
55
2500
120
2200
328
303
408
213
2.8
IIb
2a
2b
2c
2d
2e
2f
160
3050
404
2400
126
766
215
150
2g
Morphine
Fig. 2. Dose–response curve of 1b in the hot plate test. Mice received
1b sc at the indicated dose and were controlled every 15 min. Each
point represents the mean9SEM of eight mice per group. ** PB
0.01 vs. vehicle. ꢁ, vehicle; ꢂ, 1b 10 mg kg−1 sc; ꢃ, 1b 20 mg kg−1
sc; ꢄ, 1b 40 mg kg−1 sc.
a Ki values were calculated with the LIGAND program [15], based on
a Kd value of 1 nM for 3H-DAMGO. Values are the mean from two
experiments.
b See Ref. [1].
the benzofurane derivative (1a) as well as the benzo-
thiophene analog (1c) have Kis in the micromolar
range, the indolyl (1b) as well as the indenyl (1e) and
the benzocycloheptenyl derivative (1g) are by one order
of magnitude more potent. As far as compounds 2 are
concerned, they show a slightly better profile, when
compared to their corresponding model II. The b-naph-
thyl (2d), the dihydronaphthyl (2f), and the benzocyclo-
heptenyl derivative (2g) favorably compare to the
reference compound II. Also in this series, however, a
large range of potency was observed, Kis ranging from
150 to 3050 nM.
aromatic moieties of 1 cannot accommodate into the
proper pocket of the active site of the receptor as easily
as the phenyl ring of I. However, different degrees of
potency are shown along the series. In particular, while
The most significant derivatives (1b, 2d) were also
tested in vivo, in the hot plate method. Though their
binding values were fully comparable, when adminis-
tered subcutaneously 1b was found by about 7-fold
more potent than 2d (ED50 3.3 and 24.0 mg kg−1
,
respectively) (see Figs. 2 and 3). However, both 1b and
Fig. 3. Dose–response curve of 2d in the hot plate test. Mice received
2d sc at the indicated dose and were controlled every 15 min. Each
point represents the mean9SEM of eight mice per group. ** PB
0.01 vs. vehicle. ꢁ, vehicle; ꢂ, 2d 10 mg kg−1 sc; ꢃ, 2d 20 mg kg−1
sc; ꢄ, 2d 40 mg kg−1 sc.
Fig. 1. 3D plots of the most populated conformations of compounds
1a, 1b and 1g in comparison with the corresponding conformation
of I.

672
G. Cignarella et al. / Il Farmaco 53 (1998) 667–674
Table 3
Elemental analyses
Compound
Mol. wt.
Calc. (found) (%)
C
H
N
S
Cl
1a
1b
1c
1d
1e
1f
1g
2a
2b
2c
2d
2e
2f
298.38
297.38
350.91
344.87
332.86
346.90
360.93
334.85
333.87
350.84
344.87
332.86
346.90
360.93
72.45 (72.05)
72.69 (72.30)
61.61 (61.35)
69.65 (70.03)
68.55 (68.63)
69.25 (69.54)
69.88 (69.50)
64.56 (64.17)
64.75 (64.68)
61.61 (61.95)
69.65 (70.27)
68.55 (68.75)
69.25 (69.43)
69.88 (70.11)
7.43 (7.65)
7.79 (7.74)
6.60 (6.43)
7.31 (6.99)
7.57 (7.65)
7.85 (7.83)
8.10 (8.07)
6.92 (7.01)
7.25 (7.29)
6.60 (6.28)
7.31 (7.27)
7.57 (7.68)
7.84 (8.03)
8.10 (8.23)
9.39 (9.06)
14.13 (13.99)
7.98 (7.58)
8.12 (8.02)
8.42 (8.41)
8.08 (8.09)
7.76 (7.40)
8.37 (8.43)
12.59 (12.60)
7.98 (8.02)
8.12 (8.31)
8.42 (8.57)
8.07 (7.88)
7.76 (7.56)
9.13 (8.83)
10.11 (10.27)
10.28 (10.09)
10.65 (10.81)
10.22 (10.15)
9.82 (9.56)
10.59 (10.78)
10.62 (10.24)
10.11 (10.35)
10.28 (10.19)
10.65 (10.79)
10.22 (10.34)
9.82 (9.75)
9.13 (9.45)
2g
2d were less active than their corresponding models I
and II (ED50 1.1 and 16.0 mg kg−1) in this test [1].
In conclusion, the present study gives a further con-
tribution to the elucidation of the structural and geo-
metrical requirements of the cinnamyl chain of I and II.
In particular, it indicates that a precise orientation of
the phenyl ring is indispensable for a good affinity, as
already evidenced by the lowering of activity observed
when going from trans- to cis-cinnamyl [2]. Moreover,
it suggests that the phenyl ring and the double bond of
the cinnamyl moiety in I and II should not be coplanar
to allow a good interaction with the m-receptor.
of the obtained chloride 4 (0.01 mol), 8-Boc-3,8-diaza-
bicyclo[3.2.1]octane [3] (0.01 mol) and triethylamine
(0.01 mol) in toluene (20 ml) was stirred overnight at
r.t. The inorganic salts were filtered off, the solvent
evaporated to give 5a,b which were used without fur-
ther purification.
1
For 5a: yield 65%, m.p. 109°C, H NMR l: 1.5 (s,
9H); 1.7–2.1 (m, 4H); 2.9–3.3 (m, 1H); 3.4.–3.7 (m,
1H); 4.0–4.6 (m, 4H); 7.2–7.8 (m, 5H).
1
For 5b: yield 60%, m.p. 227°C, H NMR l: 1.5 (s,
9H); 1.6–2.0 (m, 4H); 3.0–3.8 (m, 2H); 4.0–4.5 (m,
4H); 6.9 (s, 1H); 7.1 (t, 1H); 7.2 (t, 1H); 7.5 (d, 1H); 7.7
(d, 1H); 11.7 (s, 1H, exch. with D2O).
6.1.2. 3-Aryloyl-3,8-diazabicyclo[3.2.1]octanes (6a,b)
To a cooled solution of the required 5a,b (0.003 mol)
in diethyl ether (10 ml) a solution of HCl in diethyl
ether was added until the pH was about 1 and the
mixture stirred overnight at r.t. Sodium hydroxide (6
N) was then added until pH 9, the layers separated and
the aqueous phase extracted twice by diethyl ether
(2×20 ml). The reunited organic layers where dried
over sodium sulfate and the solvent evaporated to give
6a,b, which were used without further purification.
6. Experimental
6.1. Chemistry
Melting points were determined on a Bu¨chi 510
capillary melting points apparatus and are uncorrected.
Analyses indicated by the symbols were within 90.4 of
1
the theoretical values. H NMR spectra were recorded
on a Bruker AC200 spectrometer; chemical shifts are
reported as l (ppm) relative to tetramethylsilane (TMS)
as internal standard. CDCl3 was used as the solvent,
unless otherwise noted. TLC on silica gel plates was
used to check product purity. Silica gel 60 (Merck;
70–230 Mesh) was used for column chromatography.
The structures of all compounds were consistent with
their analytical (see Table 3) and spectroscopic data.
1
For 6a: yield 70%, m.p. 125–127°C, H NMR l: 1.8
(m, 4H); 2.7 (s, 1H exch. with D2O); 3.0–3.2 (m, 1H);
3.4–3.8 (m, 3H); 4.0–4.6 (m, 2H); 7.2–7.4 (m, 3H); 7.5
(d, 1H); 7.6 (d, 1H).
For 6b: yield 75%, m.p. 186–188°C, 1H NMR l:
1.6–2.2 (m, 4H); 3.6–3.9 (m, 2H); 4.0–4.2 (bs, 2H);
4.3–4.5 (m, 2H); 6.9 (s, 1H); 7.1 (t, 1H); 7.3 (t, 1H); 7.5
(d, 1H); 7.7 (d, 1H); 9.4–10.0 (bs, 1H, exch. with D2O);
11.7 (s, 1H exch. with D2O).
6.1.1. 3-Aryloyl-8-Boc-3,8-diazabicyclo[3.2.1]octanes
(5a,b)
The appropriate acid 3 (0.01 mol) was treated with
an excess thionyl chloride in dichloromethane and
stirred overnight at room temperature (r.t.). A mixture
6.1.3. 3-Arylmethyl-3,8-diazabicyclo[3.2.1]octanes (7a,b)
To an ice-cooled solution of the required 6a,b (0.02
mol) in anhydrous toluene (40 ml) a 65% (w/w) solution

G. Cignarella et al. / Il Farmaco 53 (1998) 667–674
673
of bis-(2-methoxyethoxy)aluminum hydride (Red-Al®)
(0.04 mol) in toluene was added. The mixture was
brought to r.t. and monitored by TLC (CH2Cl2/MeOH
7/3) until the starting material disappeared. The mix-
ture was again cooled at 0°C and cautiously added of a
10% aqueous solution of sodium carbonate until pH 10,
the inorganic salts were filtered off and the aqueous
layer repeatedly extracted with dichloromethane. After
drying over sodium sulfate, the solvent was evaporated
to give the desired 7a,b, which were used without
further purification.
6.1.6. 8-Aryloyl-3,8-diazabicyclo[3.2.1]octanes (9a,b)
A mixture of the required 8a,b (0.01 mol) and 10%
Pd–C (10/1 w/w) in ethanol (35 ml) was hydrogenated
at r.t. After the hydrogen absorption ceased, the cata-
lyst was filtered off and the solvent evaporated to give
9a,b, which were purified by silica gel chromatography,
eluting with dichloromethane/methanol 95/5.
For 9a: yield 53%, m.p. 167–168°C, 1H NMR l:
2.1–2.3 (m, 5H); 3.1–3.3 (m, 2H); 4.1–4.3 (m, 2H);
4.9–5.1 (m, 2H); 7.3–7.5 (m, 4H); 7.6 (d, 1H).
For 9b: yield 40%, m.p. 162–164°C.
1
For 7a: yield 80%, oil, H NMR l: 1.8–2.1 (m, 4H);
2.6 (s, 1H, exch. with D2O); 2.8 (s, 2H); 2.9–3.2 (m,
1H); 3.5–3.8 (m, 3H); 4.0–4.6 (m, 2H); 7.2–7.35 (m,
3H); 7.5 (d, 1H); 7.6 (d, 1H).
6.1.7. 8-Arylmethyl-3,8-diazabicyclo[3.2.1]octanes
(10a,b)
The compounds were prepared as above reported for
the isomers 7, starting from the appropriate 9a,b.
For 10a: yield 45%, oil, 1H NMR l: 2.1–2.4 (m, 5H);
3.2–3.4 (m, 2H); 2.6–2.8 (m, 2H); 4.2–4.3 (m, 2H);
4.8–5.0 (m, 2H); 7.2–7.5 (m, 4H); 7.6 (d, 1H).
For 10b: yield 40%, oil, 1H NMR l: 1.8–2.2 (m, 5H);
2.6–2.7 (m, 2H); 2.8–3.0 (m, 2H); 3.0–3.3 (m, 2H);
4.8–5.0 (m, 2H); 6.7 (s, 1H); 7.0–7.1 (m, 1H); 7.2–7.3
(m, 1H); 7.5 (d, 1H); 7.8 (d, 1H); 9.8 (bs, 1H, exch. with
D2O).
1
For 7b: yield 90%, oil, H NMR l: 1.8–2.2 (m, 5H);
2.9 (s, 2H); 3.1–3.4 (m, 2H); 3.5 (s, 1H); 3.5–3.7 (m,
2H); 4.3–4.5 (m, 2H); 6.8 (s, 1H); 7.1 (d, 1H); 7.2–7.4
(m, 1H); 7.45 (d, 1H); 7.6 (d.1H); 9.5 (s, 1H exch. with
D2O).
6.1.4. 3-Arylmethyl-8-propionyl-3,8-diazabicyclo-
[3.2.1]octanes (1a,b)
To an ice cooled solution of the required 7a,b (0.02
mol) in CH2Cl2 (20 ml) a solution of propionic anhy-
dride (0.2 mol) in CH2Cl2 (10 ml) was added and the
mixture refluxed for 1 h. After cooling, the suspension
was made alkaline by 20% NaOH and stirred until
excess propionic anhydride was destroyed. The aqueous
layer was extracted with dichloromethane (2×10 ml),
the solvent dried (Na2SO4) and evaporated to give 1a,b.
Compounds were purified by silica gel chromatogra-
phy, eluting with CH2Cl2/MeOH 98/2 (see Table 1 for
data).
6.1.8. 3-Propionyl-8-arylmethyl-3,8-diazabicyclo[3.2.1]-
octanes (2a,b)
The compounds were prepared as above reported for
the isomers 1, starting from the required 10a,b (see
Table 1 for data).
6.1.9. 2-Bromomethyl-naphthalene (11d)
To a solution of 12 (2.5 g, 0.018 mol) in CCl4 (25 ml),
bromine (2.8 g, 0.018 mol) in CCl4 was added. During
addition the mixture was irradiated with a 500 W lamp.
The mixture was stirred for 1 h at r.t., the solvent
evaporated and the residue purified by silica gel chro-
matography, eluting with petroleum ether/diethyl ether
(9/1 w/w) to give 11d.
6.1.5. 8-Aryloyl-3-benzyl-3,8-diazabicyclo[3.2.1]octanes
(8a,b)
A mixture of the appropriate 4 (0.01 mol), 3-benzyl-
3,8-diazabicyclo[3.2.1]octane [4] (0.01 mol) and triethyl-
amine (0.01 mol) in diethyl ether (20 ml) was stirred for
2 h at r.t. The suspension was added of water (10 ml)
and stirred for further 10 min then extracted with
dichloromethane (3×10 ml). The organic layer was
dried over sodium sulfate and the solvent evaporated to
give 8a,b which were purified by silica gel chromatogra-
phy, eluting with cyclohexane/ethyl acetate 7/3.
For 8a: yield 75%, m.p. 108–111°C, 1H NMR l:
2.0–2.3 (m, 5H); 2.8 (s, 2H); 3.0–3.2 (m, 2H); 4.1–4.3
(m, 2H); 4.8 (s, 2H); 4.9–5.0 (m, 2H); 7.3–7.7 (m,
10H).
1
Yield 71%, oil, NMR l: 4.65 (s, 2H); 7.5–7.6 (m,
3H); 7.8–7.9 (m, 4H).
6.2. 3-Arylmethyl-8-propionyl-3,8-diazabicyclo[3.2.1]-
octanes (1c–g)—general method
An equimolar mixture of the required halide 11 [5,6],
13 [9–11] (0.001 mol), 8-propionyl-3,8-diazabicy-
clo[3.2.1]octane [7] (0.16 g), and K2CO3 (0.14 g) in
acetone (10 ml) was refluxed overnight. After cooling,
the salts were filtered off, the solvent evaporated and
the residue purified by flash chromatography, eluting
with dichloromethane/methanol (98/2 w/w). If required,
the so obtained 1c–g could be transformed into the
corresponding hydrochloride by treatment with a solu-
tion of hydrochloric acid in diethyl ether (see Table 1
for data).
For 8b: yield 80%, m.p. 152–154°C, 1H NMR l:
1.8–2.1 (m, 5H); 2.7–2.9 (m, 2H); 3.0–3.2 (m, 2H); 4.7
(s, 2H); 4.8 (s, 2H); 4.9–5.1 (m, 2H); 6.7 (s, 1H);
7.0–7.6 (m, 8H); 7.7 (d, 1H); 9.7 (bs, 1H exch. with
D2O).

674
G. Cignarella et al. / Il Farmaco 53 (1998) 667–674
6.3. 8-Arylmethyl-3-propionyl-3,8-diazabicyclo[3.2.1]-
octanes (2c–g)
maintained at 65°C were determined. The time at which
mice displayed a nociceptive response, licking the front
paws, fanning the hind paws or jumping was recorded
and the animal was removed from the hot plate. In
each case, post treatment hot plate latencies were deter-
mined at the indicated times for every experiment and
14 s was set as the cut off time to avoid damage. To
establish the dose–response curve, at least four drug
doses were used on eight to ten mice per each dose. The
50% antinociceptive doses (ED50) and their 95% confi-
dence intervals were determined by the method of
Litchfield and Wilcoxon [17].
Compounds 2c–g were obtained according to the
procedure above reported for their isomers 1, starting
from 3-propionyl-3,8-diazabicyclo[3.2.1]octane [8].
6.4. Binding studies
Male Sprague–Dawley rats (Charles River, Italy)
weighing 180–200 g were used. Rat brain membrane
binding studies were carried out as described by Gillan
and Kosterlitz with slight modifications [13]. Whole
brain minus cerebellum was homogenized with Poly-
tron in 50 volumes (w/v) of 50 mM Tris–HCl pH 7.7,
centrifuged at 48 000×g for 20 min at 4°C, resus-
pended in 50 volumes of the same buffer and incubated
for 45 min at 37°C. After centrifugation at 48 000×g
for 20 min at 4°C, the final pellet was resuspended in
the same buffer to final concentration of 0.8–1.0 mg
References
[1] G. Cignarella, D. Barlocco, M.E. Tranquillini, A. Volterra, N.
Brunello, G. Racagni, Pharm. Res. Commun. 20 (1988) 383.
[2] D. Barlocco, G. Cignarella, S. Villa, XI Convegno Nazionale-Di-
visione di Chimica Farmaceutica, Bari, Oct. 1994.
[3] D. Barlocco, G. Cignarella, D. Tondi, P. Vianello, S. Villa, A.
Bartolini, C. Ghelardini, N. Galeotti, D. Colombo, L. Toma, J.
Med. Chem. 41 (1998) 674.
[4] G. Cignarella, G.G. Nathansohn, J. Org. Chem. 26 (1961) 2747.
[5] F.F. Blicke, US Patent 2,533,086, 1950; Chem. Abstr. 45 (1949)
2508e.
[6] F.F. Blicke, US Patent 2,533,087, 1950; Chem. Abstr. 45 (1949)
2508g.
[7] G. Cignarella, E. Occelli, E. Testa, J. Med. Chem. 8 (1965) 326.
[8] G. Cignarella, E. Testa, C.R. Pasqualucci, Tetrahedron 19 (1963)
143.
[9] H.O. House, C.B. Hudson, J. Org. Chem. 35 (1979) 647.
[10] J. Vebrel, R. Carrie, Bull. Soc. Chem. Fr. II 3-4 (1982) 116.
[11] R.B. Miller, J.M. Frincke, J. Org. Chem. 45 (1980) 5312.
[12] HyperChem™ is a package from HyperCube, Inc., Canada.
[13] M.G.G. Gillan, H.W. Kosterlitz, J. Pharmacol. 77 (1982) 461.
[14] O.H. Lowry, N.J. Rosenbrough, A.L. Farr, R.J.J. Randall, Biol.
Chem. 193 (1951) 265.
[15] P.J. Munson, D. Rodbard, Anal. Biochem. 107 (1980) 220.
[16] D.L. Oden, K.L Oden, Life Sci. 31 (1982) 1245.
[17] J.T. Litchfield, F. Wilcoxon, J. Pharmacol. Exp. Ther. 96 (1949)
99.
3
protein ml−1. H-DAMGO (2 nM) (New England Nu-
clear, Germany) was used to label m-receptors. Mem-
brane suspensions were incubated with the ligand for 60
min at 0°C in the presence or the absence of 10 mM of
naloxone. Final protein concentrations were determined
by the method of Lowry et al. [14]. Ki values were
calculated with the LIGAND program [15], from dis-
placement curves of each compound at a concentration
range between 10−10 and 10−4 M. Values are the mean
from two assays.
6.5. Antinociception
Male Albino-Swiss mice weighing 20–25 g (Charles
River, Italy) were used. Antinociception was estimated
by means of the hot plate method described by Oden
and Oden [16]. The effect of compounds on the reaction
time of mice placed on a hot plate, thermostatically
.