N-Homolupinanoyl and N-(ω-lupinylthio)alkanoyl derivatives of some tricyclic systems as ligands for muscarinic M1 and M2 receptor subtypes

Article abstract of DOI:10.1016/S0014-827X(03)00104-6

A set of N-homolupinanoyl- and N-(ω-lupinylthio)alkanoyl derivatives of tricyclic systems (as phenothiazine, iminodibenzyl and dihydropyridobenzodiazepinone) has been prepared and tested for affinity for rat muscarinic M₁ and M₂ receptor subtypes labeled with [³H]pirenzepine and [³H]AF-DX 384. Good affinity for both M₁ and M₂ subtypes was displayed by most compounds, often with nanomolar K_i values, which for lupinylthiopropionyl- and lupinylthiobutyryl-phenothiazines (13-16) were comparable to those of pirenzepine and methoctramine, respectively. However, only moderate selectivity for one or the other subtype was seen.

Full text of DOI:10.1016/S0014-827X(03)00104-6

Il Farmaco 58 (2003) 669Á676
/
www.elsevier.com/locate/farmac
N-Homolupinanoyl and N-(v-lupinylthio)alkanoyl derivatives of
some tricyclic systems as ligands for muscarinic M1 and M2 receptor
subtypes
Bruno Tasso a, Anna Sparatore b, Fabio Sparatore a,*
a Dipartimento di Scienze Farmaceutiche, University of Genoa, Viale Benedetto XV, 3, I-16132 Genova, Italy
b Istituto di Chimica Farmaceutica e Tossicologica, University of Milan, Viale Abruzzi 42, I-20131 Milan, Italy
Received 1 October 2002; accepted 8 February 2003
Abstract
A set of N-homolupinanoyl- and N-(v-lupinylthio)alkanoyl derivatives of tricyclic systems (as phenothiazine, iminodibenzyl and
dihydropyridobenzodiazepinone) has been prepared and tested for affinity for rat muscarinic M1 and M2 receptor subtypes labeled
with [3H]pirenzepine and [3H]AF-DX 384. Good affinity for both M1 and M2 subtypes was displayed by most compounds, often
with nanomolar Ki values, which for lupinylthiopropionyl- and lupinylthiobutyryl-phenothiazines (13Á16) were comparable to
/
those of pirenzepine and methoctramine, respectively. However, only moderate selectivity for one or the other subtype was seen.
´
# 2003 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved.
Keywords: N-(Homolupinanoyl)phenothiazines; N-[(v-lupinylthio)alkanoyl]phenothiazines;
Quinolizidine derivatives; Muscarinic receptors ligands
N-(tert-Aminoalkanoyl)tricyclic
systems;
1. Introduction
A few years ago [7], we found that N-homolupinanoyl
phenothiazine 1 (Chart 1), studied as antiarrhythmic
agent along with other analogues, exhibited high affinity
Specific muscarinic receptor subtypes ligands, acting
either as agonists or antagonists, may show interest in
several kinds of peripheral and central nervous systems
for M1 subtype with IC50ꢀ30 nM, while the affinity for
/
M2 and M3 subtypes was more than two orders of
magnitude lower (IC50 around 10 mM), resulting in a
selectivity for M1 subtype superior to that of pirenze-
pine.
In compound 5, bearing the same homolupinanoyl
residue on the pyridobenzodiazepinone tricyclic system,
which characterizes pirenzepine, the affinity for M1
disorders [1Á3].
/
M1 and/or M3 receptor antagonists might be useful in
the treatment of gastroenteric apparatus dysfunctions
(such as gastric acid secretion and irritable bowel
syndrome) and chronic obstructive airways diseases,
while M2 antagonists might be used in the treatment of
bradycardic disorders. Moreover, central M4 receptor
antagonists could be applied as anti-Parkinson agents
[4], while central M2 receptor antagonists should in-
crease the release of acetylcholine and be useful in the
treatment of cholinergic-deficit depending pathologies,
such as Alzheimer’s disease [5,6]. The same use has been
suggested for central M1 receptor agonists, however, in
such a case, unacceptable side effects occurred.
subtype was lower (IC50ꢀ110 nM) and the selectivity
/
versus M2 and M3 subtypes was also reduced [8].
However, when the homolupinanoyl residue was
replaced with the longer lupinylthioacetyl moiety (6),
the high affinity for M1 subtype was restored, though
the selectivity versus M2 and M3 subtypes remained a
little lower than that of pirenzepine.
Therefore, both the nature of the tricyclic system and,
even more, the length of the chain connecting the
quinolizidine ring to the tricyclic system seem to be
important for the affinity to M1 subtype.
* Corresponding author.
E-mail address: sparator@unige.it (F. Sparatore).
´
0014-827X/03/$ - see front matter # 2003 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved.
doi:10.1016/S0014-827X(03)00104-6

670
B. Tasso et al. / Il Farmaco 58 (2003) 669Á676
/
Preliminary functional tests suggested M1 and M2
muscarinic antagonism.
epimeric analogue of compound 1 (1-epi) was also
prepared, starting from D-epi-lupinine.
The novel compounds 1-epi and 7Á
/
20, together with
On the basis of the above, we have prepared two sets
of quinolizidine derivatives which are interesting as
muscarinic ligands (Chart 1).
In the first, a lupinylthioacetyl residue is linked to
variously substituted iminodibenzyl or phenothiazine
previously described compounds 1Á6, were tested for
/
their affinity to M1 and M2 receptor subtypes, by
displacing specific radioligands from membrane pre-
parations: [3H]pirenzepine from rat brain M1 receptors
and [3H]AF-DX 384 from rat heart M2 receptors.
10-(tert-Aminoacyl)phenothiazines endowed with
outstanding spasmolytic properties in vitro have already
rings (7Á
anoyl moiety of increasing length is linked to phenothia-
zine or 2-methoxyphenothiazine (13Á20).
/
12), while in the second set, a lupinylthioalk-
/
been described in the past [10Á12]. Among these, 10-(4-
/
The folding of the longer side-chain might allow a
more suitable spatial placement of the terminal proto-
nated nitrogen in relation to the tricyclic moiety, which
could influence the affinity and selectivity for the
muscarinic receptor subtypes, as has already been
pointed out by Eberlein et al. [9] for their pyridobenzo-
diazepinone derivatives.
The quinolizidine moiety of these compounds retains
the absolute configuration of natural L-lupinine, from
which they derive.
diethylaminobutyryl)phenothiazine was as potent as
atropine against acetylcholine-induced spasms.
However, as far as we know, these compounds were
not tested for affinity to muscarinic receptor subtypes.
Several 10-(aminoacyl)phenothiazines display high
antinicotinic and antimuscarinic activity and are there-
fore able to protect animals up to 10 times better than
atropine from powerful acetylcholinesterase inhibitors
such as Tabun (dimethylamidoethoxyphosphoryl cya-
nide) [13,14]. Interestingly, 10-(2-diethylaminopropio-
nyl)phenothiazine (Astra 1397) is reported to be a
selective inhibitor of human plasma butyrylcholinester-
In a first attempt to investigate the effect of a
configuration change in the quinolizidine moiety, the
Chart 1.

B. Tasso et al. / Il Farmaco 58 (2003) 669Á
/676
671
ase with respect to human red cell acetylcholinesterase
[15].
The structures of all the prepared compounds were
supported by elemental analyses and spectral (IR and
NMR) data. NMR spectra were consistent to the
assigned structures and did not exhibit any peculiar
features, thus only a few are described in Section 3 as
examples.
2. Chemistry
Compounds 1Á
[7,8].
/
6 were prepared as already described
3. Experimental
Compound 1-epi was prepared in the same manner as
compound 1 [7] starting from epi-homolupinanoylchlor-
ide hydrochloride [16].
Melting points were determined by the capillary
method on a Buchi apparatus and are uncorrected.
¨
Elemental analyses were performed with CE EA 1110
CHNS-O instruments and the results obtained for C, H
The new compounds 7Á20 were prepared by reacting
/
the proper tricycle (substituted iminodibenzyl or phe-
nothiazine) with the suitable v-haloalkanoyl chloride in
refluxing toluene and subsequent reaction of the N-(v-
halo)alkanoyl tricycle with thiolupinine [17] in ethanol
or acetonitrile (Scheme 1).
and N were within 90.3% of the calculated values; for
/
several compounds, the sulfur content resulted lower
than the calculated value for more than 0.3%.
IR spectra were recorded on PerkinÁElmer Paragon
/
1
1000PC spectrophotometer; H NMR were taken on a
Varian Gemini 200 spectrometer, using CDCl3 or d6-
DMSO as solvent, with TMS as internal standard.
It is worth noting that, while N-chloroacetyl and N-
chloropropionyl-2-substituted phenothiazines react
smoothly with thiolupinine in ethanol, N-(4-chlorobu-
tyryl)phenothiazines do not react, but lose the aliphatic
chain to give N-unsubstituted phenothiazines, ethyl-
chlorobutyrate and butyrolactone. If the reaction be-
tween thiolupinine and chlorobutyrylphenothiazine is
run in dry benzene, the latter reactant is recovered
unchanged even after 70 h of heating at 100 8C in a
sealed tube. Finally, the required N-[(4-lupinylthio)bu-
tyryl]phenothiazines were obtained (with about 40%
yield) by reacting the N-(4-bromobutyryl)-2-substituted
phenothiazines with thiolupinine in acetonitrile. The
same experimental conditions were used to obtain
higher homologues.
3.1. N-(epi-Homolupinanoyl)phenothiazine (1-epi)
The epi-homolupinanoic acid hydrochloride (1 g, 4.3
mmol) was converted to the corresponding chloride by
means of oxalyl chloride in chloroform, as already
described [16].
A slurry of phenothiazine (0.86 g, 4.3 mmol) in 10 ml
of chloroform was added to the solution of the above
chloride in 10 ml of the same solvent and the mixture
was heated under reflux for 3 h. After removing the
solvent under reduced pressure, the residue was treated
Scheme 1.

672
B. Tasso et al. / Il Farmaco 58 (2003) 669Á676
/
with acidic water and the unreacted phenothiazine was
extracted with ether.
The acid solution was alkalized with 2 N NaOH
solution and extracted with ether. The solvent was
evaporated and the residue was crystallized from dry
ether.
times with ether to remove the unreacted haloalkanoyl-
tricycle. The acid solution was basified with 2 N NaOH
solution and the free base was worked as needed.
The solid compounds were collected and crystallized
from ethanol/water (9 and 10). Oily compounds were
extracted with ether; the ether solution was washed with
water, dried and evaporated to dryness. The residue was
rinsed with little dry ether and crystallized from the
same solvent (7, 11, 14). Compounds 8, 12 and 13 were
chromatographed on neutral alumina, eluting with
ether. After chromatography, compound 12 contained
a small amount of lupinyldisulfide. Therefore, it was
dissolved in ethanol and treated with 1 N HCl and zinc
dust; the solution was basified and extracted with ether.
The ether solution was chromatographed on basic
alumina, eluting with ether.
1H NMR (CDCl3): d 0.74Á
dine ringꢀQ), 1.30Á2.06 (m, 11H, Q), 2.15 (dd, Jꢀ
14.65 and 9.28 Hz; 1H of C(O)ÃCH2ÃQ), 2.65Á2.90 (m,
3H; 1H of C(O)ÃCH2ÃQ and 2H of Q), 7.20Á7.40 (m,
4H arom.), 7.40Á7.58 (m, 4H arom.).
/1.30 (m, 3H of quinolizi-
/
/
/
/
/
/
/
/
/
/
3.2. N-(v-Haloalkanoyl)-2-substituted phenothiazines
and other N-(v-haloalkanoyl)tricycles
A solution of the suitable v-haloalkanoyl chloride
(10% excess) in 5 ml of dry toluene was added dropwise
Characteristics of compounds 7Á
Table 1.
1H NMR (CDCl3) of compound 7: d 1.07Á
14H, Q), 2.61Á3.06 (m, 6H, 4H of CH2 of dibenzyl, 1H
of SÃCH2ÃQ and 1H of Q), 3.06Á3.68 (m, 4H, 2H of
C(O)ÃCH2ÃS, 1H of SÃCH2ÃQ and 1H of Q), 7.05Á
7.48 (m, 8H arom.) ppm.
1H NMR (CDCl3) of compound 9: d 1.08Á
14H, Q), 2.68Á2.89 (m, 3H, 1H of SÃCH2ÃQ and 2H of
Q), 2.99 (dd, Jꢀ4.8 and 12.9 Hz, 1H of SÃCH2ÃQ),
3.37 (s, 2H, C(O)ÃCH2ÃS), 7.16Á7.40 (m, 4H arom.),
7.46 (dd, Jꢀ1.6 and 7.4 Hz, 2H arom.), 7.60 (d, Jꢀ7.8
Hz, 2H arom.) ppm.
1H NMR (CDCl3) of compound 14: d 1.08Á
14H, Q), 2.50Á2.70 (m, 2H, 1H of SÃCH2ÃQ and 1H of
Q), 2.70Á2.94 (m, 6H, 4H of C(O)ÃCH2ÃCH2ÃS, 1H of
SÃCH2ÃQ and 1H of Q), 3.82 (s, 3H of OCH3), 6.83
(dd, Jꢀ2.7 and 8.7 Hz, 1H arom.), 7.12 (d, Jꢀ2.6 Hz,
1H arom.), 7.16Á7.40 (m, 3H arom.), 7.40Á7.55 (m, 2H
arom.) ppm.
/14 are collected in
to a solution of the proper tricyclic compound (2.6Á
mmol) in dry toluene (30 ml). The solution was then
heated at reflux for 3Á7 h; the completion of the
/
5
/
2.11 (m,
/
/
reaction was monitored by TLC on alumina. After
cooling, the toluene solution was either washed with
water, dried and evaporated under reduced pressure
(chloroacetyl and propionyl derivatives) or evaporated
directly under reduced pressure, rinsing the residues
with pentane (4-bromobutyrylphenothiazines) or dry
ether (v-bromohexanoylphenothiazines). The oily resi-
dues were chromatographed on alumina, eluting with
dichloromethane.
/
/
/
/
/
/
/
/
/2.13 (m,
/
/
/
/
/
/
/
/
/
/
/
Yields were in the 84Á
/
98% range.
/
2.09 (m,
Most compounds have already been described [11,18Á
/
/
/
/
24] and measured melting points were in agreement with
the published data.
The new N-(4-bromobutyryl)-2-methoxyphenothia-
/
/
/
/
/
/
/
/
zine melted at 132Á
pounds [N-(chloroacetyl)iminodibenzyl, N-(5-bromo-
pentanoyl)phenothiazine, N-(5-bromopentanoyl)-2-
/
134 8C, while the other new com-
/
/
methoxyphenothiazine and N-(6-bromohexanoyl)-2-
methoxyphenothiazine] were oils and after the chroma-
tography were used directly in the next step.
3.4. N-[(v-Lupinylthio)alkanoyl]-2-substituted
phenothiazines (15Á20)
/
All the N-(v-haloalkanoyl)tricycles exhibited a sharp
A solution of N-(v-bromoalkanoyl)-2-substituted
phenothiazine (3.2Á3.5 mmol) in 2Á3 ml of acetonitrile,
which was followed by a solution of thiolupinine (3.2Á
3.5 mmol) in 2 ml of the same solvent, was introduced
into an Aldrich pressure tube, flushed with nitrogen.
The tube was sealed and heated at 80 8C for 6 h
carbonyl band in the IR spectrum (KBr or film): 1661Á
/
/
/
1666 cmꢁ1 for iminodibenzyl derivatives; 1664Á
/1681
/
cmꢁ1 for phenothiazine derivatives.
3.3. N-[(2-lupinylthio)acetyl]3-substituted
iminodibenzyls (7 and 8) and N-[(2/3-lupinylthio)acetyl/
propionyl]-2-substituted phenothiazines (9Á14)
(butyryl derivatives), or at 1008 C for 16Á22 h (for the
/
/
remaining compounds). After cooling, the tube content
was diluted with acidic water and the unreacted
bromoalkanoylphenothiazine was removed by filtration
and extraction with ether. The acid solution was basified
and extracted with dichloromethane. After removing the
solvent, the residue was chromatographed on alumina
(1:25), eluting with ether (butyryl derivatives) or di-
chloromethane (the remaining compounds). Dichloro-
methane cannot be used to remove the unreacted
A solution of 2Á
absolute ethanol was rapidly added to a solution of N-
(2/3-chloroacetyl/propionyl)tricycle (2Á5 mmol) in ab-
solute ethanol (30 ml).
The solution was refluxed under nitrogen for 5Á
/
5 mmol of thiolupinine [17] in 5 ml of
/
/
7 h
and then evaporated under reduced pressure. The
residue was dissolved in acidic water and extracted three

B. Tasso et al. / Il Farmaco 58 (2003) 669Á
/676
673
Table 1
Characteristics of novel compounds 1-epi and 7Á
/
20 of Chart 1
a
Comp.
W
n
R
Formula
m.p. (8C) b
Yield (%) c
1-epi
7
C23H26N2OS, C23H26N2OS×
C26H32N2OS, C26H32N2OS×
/
HCl×
/
0.5H2O
121.5Á
104Á105
76Á78
131Á
127Á
96Á
/
122.5
70
73
70
67
61
61
40
43
50
40
44
45
41
52
43
CH2CH2
CH2CH2
1
1
1
1
1
1
2
2
3
3
4
4
5
5
H
Cl
/
HCl×
/
1.75H2O
/
8
C26H31ClN2OS d, C26H31ClN2OS×
C24H28N2OS2 e, C24H28N2OS2×
C24H27ClN2OS2, C24H27ClN2OS2×
C25H27F3N2OS2, C25H27F3N2OS2×
C25H30N2O2S2 e, C25H30N2O2S2×
HCl×
C25H30N2OS2 d, C25H30N2OS2×
HCl×
C26H32N2O2S2, C26H32N2O2S2×HCl×
C26H32N2OS2 d, C26H32N2OS2×
HCl×
C27H34N2O2S2 d, C27H34N2O2S2×
C27H34N2OS2, C27H34N2OS2×HCl×
C28H36N2O2S2 d, C28H36N2O2S2×
HCl×
28H36N2OS2, C28H36N2OS2×HCl×
1.5H2O e
C29H38N2O2S2 e, C29H38N2O2S2× 0.75H2O e
HCl×
/
HCl×
HCl×
0.5H2O e
HCl×2H2O
HCl×1.75H2O
H2O
/H2O
/
9
S
S
S
S
S
S
S
S
S
S
S
S
H
Cl
/
/
/
132
10
11
12
13
14
15
16
17
18
19
20
/
/
/
128
CF3
OCH3
H
/
/
/
97
/
/
Oil
/
/
0.75H2O
Oil
OCH3
H
/
/
0.5H2O e
106Á107
/
/
/
1.25H2O e
Oil
Oil
Oil
Oil
Oil
Oil
OCH3
H
/
HCl×
H2O e
0.5H2O e
/
H2O e
/
/
OCH3
H
/
/
C
/
/
OCH3
/
/
a
All compounds were analyzed for C, N, H, S and results were within 90.3% of the calculated values, unless otherwise stated.
/
b
c
m.p. of hydrochlorides are not indicated because they melt in a wide range of temperature.
The indicated yields refer to the free bases; the conversion to hydrochlorides was practically quantitative.
Not analyzed.
d
e
The sulfur content resulted lower than the calculated value for more than 0.3%.
bromoalkanoylphenothiazine since it also dissolves
certain amount of lupinylthioalkanoylphenothiazine
hydrochloride.
Compounds were tested at four or five concentrations
in the range 10 mMÁ1 nM. Ki values are the means of
duplicate experiments.
/
Characteristics of compounds 15Á/20 are collected in
Table 1.
3.5.1. Muscarinic M1 binding
Membranes from rat brain cortices were incubated
with 1 nM [3H]pirenzepine (Kdꢀ
6.7 nM) for 60 min at
/
3.5. Binding assays
22 8C. Non-specific binding was determined in the
presence of 1 mM atropine.
Binding assays were performed at the Receptology
Department of CEREP, Celle L’Evescault, France, or,
for central nicotinic receptor, at MDS Panlabs, Inc.,
Bothell, WA.
Membranes were prepared by standard techniques
and incubated in the conditions indicated below for each
receptor subtype. After incubation, membranes were
rapidly filtered under vacuum through glass-fiber filters
(GF/B, Packard). Filters were washed three times with
ice-cold buffer and bound radioactivity was measured
with a scintillation counter (Topcount, Packard or
Betaplate), using a liquid scintillation cocktail (Micro-
scint 0, Packard).
3.5.2. Muscarinic M2 binding
Membranes from rat heart were incubated with 2 nM
[3H]AF-DX 384 (Kdꢀ
2.1 nM) for 60 min at 22 8C.
Non-specific binding was determined in the presence of
/
1 mM atropine.
3.5.3. Central nicotinic acetylcholine receptor binding
Membranes from rat brain cortices were incubated
with [3H]cytisine (Kdꢀ
3.2 nM) at a concentration of 2
/
nM for 75 min at 4 8C. Non-specific binding was
estimated in the presence of 100 mM nicotine.

674
B. Tasso et al. / Il Farmaco 58 (2003) 669Á676
/
3.6. Inhibition of human acetylcholinesterase
On the whole, the N-homolupinanoyl derivatives 1Á5
/
exhibited a lower affinity for both M1 and M2 subtypes
in comparison to the N-[v-(S-lupinylthio)alkanoyl]der-
Compound 1 was tested as an acetylcholinesterase
inhibitor with the method described by Ellman et al. [25]
by MDS Panlabs, Inc. Human recombinant acetylcho-
linesterase (Sigma, C-1682) was used.
ivatives 6Á20 and affinity was generally higher for M1
/
than for M2 subtype.
It is worth noting that affinity for M1 and M2 receptor
subtypes was influenced in opposite directions (enhan-
cing the former and decreasing the latter), when the
connection of the chain to the quinolizidine ring was
changed from axial (1) to equatorial (1-epi).
Test compound and/or vehicle was incubated with
acetylthiocholine iodide (700 mM) and 5,5-dithiobis-2-
nitrobenzoic acid in sodium phosphate buffer (pH 7.4)
at 37 8C. The reaction was initiated by the addition of 2
ng acetylcholinesterase and the thiocholine generated
reacted continuously with dithiobis(nitrobenzoic)acid to
produce a yellow anion (5-thio-2-nitrobenzoic acid)
proportional to enzymatic activity. This was determined
after 20 min by spectrophotometry at 405 nm. The
compound was initially screened at 10 mM. Physostig-
In the set of N-homolupinanoyl derivatives, the
exchange of phenothiazine ring with the dihydropyr-
idobenzodiazepinone moiety (5) reduced the affinity for
both receptors. The introduction of substituents on
position 2 of phenothiazine ring (2 and 3) did not
modify significantly the affinity for both subtypes, with
the exception of the polar ethoxycarboamido group (4)
which resulted deleterious to the affinity for M1 subtype.
The comparison of affinities of compounds 6, 7 and 9,
bearing the same (S-lupinylthio)acetyl residue on the
cyclic nitrogen and lacking any substituent on the
aromatic nucleus indicates that the iminodibenzyl moi-
ety is less suitable for binding to both M1 and M2
subtypes, while the phenothiazine and the dihydropyr-
idobenzodiazepinone systems gave comparable results
in spite of their different polarity and lipophilicity.
mine was used as reference compound, with IC50ꢀ0.12
/
mM.
3.7. Muscarinic M1 functional response
The field stimulated (95% of maximum voltage, 0.1
Hz, 0.5 ms) prostatic portion of rabbit vas deferens,
bathed in physiological salt solution at 31 8C, was used
[26] (MDS Panlabs, Inc.). Test substance-induced re-
duction of resultant neurogenic twitch responses (ꢀ
/
The
2-unsubstituted
N-[v-(S-lupinylthio)alka-
50% related to control 3 mM McN-A-343 response)
indicated possible M1 receptor agonism. At a test
concentration where no significant agonistic activity
was seen, ability to inhibit the McN-A-343-induced
relaxation response indicated antagonist activity.
noyl]phenothiazines 9, 13 and 15 exhibited nanomolar
affinity for both M1 and M2 receptor subtypes, with Ki
comparable with those of pirenzepine (2.6 nM) and
methoctramine (4.46 nM).
Thus, the importance of the length of the chain
connecting the tricyclic system to the quinolizidine
nucleus is confirmed. However, further increasing the
number of carbon atoms of the alkanoyl residue (17 and
19), a moderate decrease of affinity was observed, which
is somewhat stronger for M1 than for M2 subtype, with
a consequent inversion of selectivity.
4. Results and discussion
Results of binding assay to M1 and M2 muscarine
receptor subtypes are collected in Table 2.
All tested compounds were highly capable of displa-
cing specific ligands ([3H]pirenzepine and [3H]AF-DX
384) from M1 and M2 receptor subtypes, with several of
them exhibiting Ki values in the nanomolar range.
However, only moderate selectivity for one or the other
subtype was observed, with a ratio of KiM2/KiM1 in the
Also in the set of N-(lupinylthio)alkanoyl derivatives,
the introduction of substituents on position 2 of the
phenothiazine ring has only moderate effects on the
affinity to receptors: chlorine and trifluoromethyl group
reduced the affinity for both M1 and M2 subtypes, while
the methoxy group generally reduced the affinity for
M2, but improved that for M1 subtype.
range 10.4Á0.36. The most selective compound was 10-
/
(epi-homolupinanoyl)phenothiazine (1-epi).
In order to gain some insight on the functional
characteristics of the relevant compounds which could
be useful for the future development of this preliminary
study, a few other assays were performed on single
selected compounds.
Compound 1 was assayed for inhibition of acetylthio-
choline hydrolysis produced by human recombinant
acetylcholinesterase; at a 10-mM concentration, this
compound failed to inhibit the enzyme.
These results might appear in conflict with the
previous observation of a very high selectivity of
compound 1 for M1 versus M2 receptor subtype [7].
Indeed, at that time affinity for M2 subtype was
determined through the displacement of [3H]N-methyls-
copolamine, a non-selective ligand that binds to this
subtype much more tightly than the selective ligand AF-
DX 384.
Concerning the structureÁ
following comments are relevant.
/
activity relationships, the
Compound 6 was tested for affinity to central
nicotinic receptor, but even at a 10-mM concentration,

B. Tasso et al. / Il Farmaco 58 (2003) 669Á
/676
675
Table 2
Affinity of compounds 1Á
/20 (Chart 1) for rat muscarinic M1 and M2 receptor subtypes
No.
Comp.
n
R
Ki a(nM)
Ratio Ki (M2/M1)
M1
M2
N-Homolupinanoyl and epi-homolupinanoyl
1
1-epi
phenothiazine
phenothiazine
H
31
9.8
30
82
102
65
2.64
10.4
2.16
2.24
0.71
1.41
H
Cl
2
3
4
5
phenothiazine
phenothiazine
OCH3
NHCOOEt
43
96.5
107
107.6
phenothiazine
dihydropyridobenzodiazepinone
150
76
N-(Lupinylthio)acetyl-
6
dihydropyridobenzodiazepinone
6.9
34
4.6
12.3
14.8
6.2
0.67
0.36
1.34
1.48
2.18
3.23
2.75
7
iminodibenzyl
iminodibenzyl
phenothiazine
phenothiazine
phenothiazine
phenothiazine
H
Cl
8
11
9
H
Cl
4.2
6.1
4.7
2.8
10
11
12
13.3
15.2
7.7
CF3
OCH3
N-(v-Lupinylthio)alkanoyl-
13
14
15
16
17
18
19
20
phenothiazine
phenothiazine
phenothiazine
phenothiazine
phenothiazine
phenothiazine
phenothiazine
phenothiazine
2
2
3
3
4
4
5
5
H
3.1
2.3
3.3
4.5
4.8
2.5
7.8
10
1.06
1.96
1.60
0.62
0.43
0.93
0.36
0.84
OCH3
H
3.0
4.0
OCH3
H
17.9
10.7
35.9
23.3
OCH3
H
OCH3
13
19.5
b
pirenzepine
methoctramine
2.6
17.8 b
206
79.2
0.25
4.46
a
Mean of duplicate experiments: each value differed from the mean by less than 10%.
Historical Ki values from CEREP.
b
it was unable to displace [3H]cytisine from a rat brain
preparation.
Finally, M1 antagonism was demonstrated for com-
pound 5, which inhibited the McN-A-343 (4-(3-chlor-
ophenylcarbamoyloxy)-2-butyryltrimethylammonium
chloride)-induced relaxation response of field stimulated
with nanomolar Ki’s which, for the best compounds N-
[(v-lupinylthio)propionyl/butyryl]phenothiazines (13Á
/
16), were comparable to those of pirenzepine and
methoctramine, respectively.
Only moderate selectivity for one or the other subtype
was seen in rat receptors labeled with [3H]pirenzepine
and [3H]AF-DX 384, respectively. The highest selectivity
for M1 versus M2 receptor subtype was seen in
compound 1-epi, the only one bearing the quinolizidine
moiety linked equatorially.
prostatic portion of rabbit vas deferens, with IC50ꢀ
/270
nM (pirenzepine IC50ꢀ120 nM).
/
Therefore, even on the basis of the previously
described [7] functional studies on isolated tissues
(guinea pig ileum and left atria), it can be argued that
our compounds behave as M1 and M2 antagonists,
however, the muscarinic receptor antagonism will be
unambiguously defined when the same functional in-
vestigations are carried out on the whole set of
compounds.
Acknowledgements
Financial support from the Italian ‘‘Ministero dell’Is-
truzione, dell’Universita` e della Ricerca scientifica’’ is
gratefully acknowledged.
5. Conclusion
References
The homolupinanoyl- and lupinylthioalkanoyl deri-
vatives of some tricyclic systems (such as phenothiazine,
iminodibenzyl and dihydropyridobenzodiazepinone)
display good affinity to both M1 and M2 receptors,
[1] R.M. Eglen, S.S. Hedge, N. Watson, Muscarinic receptor
subtypes and smooth muscle function, Pharmacol. Rev. 48
(1996) 531Á565.
/

676
B. Tasso et al. / Il Farmaco 58 (2003) 669Á676
/
[2] C.C. Felder, F.P. Bymaster, J. Ward, N. DeLapp, Therapeutic
opportunities for muscarinic receptors in the central nervous
system, J. Med. Chem. 43 (2000) 4333Á4353.
[3] R.M. Eglen, A. Choppin, N. Watson, Therapeutic opportunities
from muscarinic receptor research, Trends Pharmacol. Sci. 22
spasms of guinea pig and rat ileum, Acta Pharmacol. Toxicol. 9
(1953) 168Á178.
[13] R. Dahlbom, Studies on some phenothiazine derivatives, Sci.
Repts. Ist. Super Sanit. 2 (1962) 309Á319.
[14] R. Dahlbom, H. Diamant, T. Edlund, T. Ekstrand, B. Holmstedt,
Protective effect of phenothiazine derivatives against poisoning by
the irreversible cholinesterase inhibitor dimethylamido-ethoxy-
phosphoryl cyanide (Tabun), Acta Pharmacol. Toxicol. 9 (1953)
/
/
/
(2001) 409Á414.
/
[4] J.D. Salamone, M. Correa, B.B. Carlson, A. Wisniecki, A.J.
Mayorga, E. Nisenbaum, L. Nisenbaum, C. Felder, Neostriatal
muscarinic receptor subtypes involved in the generation of
tremulous jaw movements in rodents: implication for cholinergic
163Á167.
/
[15] K.B. Augustinsson, A titrimetric method for the determination of
involvement in Parkinsonism, Life Sci. 68 (2001) 2579Á
[5] H.N. Doods, Lipophylic muscarinic M2 antagonists as potential
drugs for cognitive disorders, Drugs Fut. 20 (1995) 157Á164.
[6] W. Greenlee, J. Clader, T. Asberom, S. McCombie, J. Ford, H.
Guzik, J. Kozlowski, S. Li, C. Liu, D. Lowe, S. Vice, H. Zhao, G.
Zhou, W. Billard, H. Binch, R. Crosby, R. Duffy, J. Lachowicz,
V. Coffin, R. Watkins, V. Ruperto, C. Strader, L. Taylor, K. Cox,
Muscarinic agonists and antagonists in the treatment of Alzhei-
/
2584.
plasma and red cell cholinesterase activity, Scand. J. Clin. Lab.
Invest. 7 (1955) 284Á
[16] A. Sparatore, M. Veronese, F. Sparatore, Derivati chinolizidinici
ad attivita` antimicrobica, Farmaco Ed. Sci. 42 (1987) 159Á174.
[17] F. Novelli, F. Sparatore, Thiolupinine and some derivatives of
pharmacological interest, Farmaco 48 (1993) 1021Á1049.
/290 [Chem. Abstr. 50 (1956) 11408b].
/
/
/
[18] F. Hafliger, W. Schindler, N-Acyl derivatives of iminodibenzyl,
¨
US Patent 2,666,051 (1954) [Chem. Abstr. 49 (1955) 9041f].
[19] T. Ekstrand, Phenothiazine derivatives of pharmacological inter-
mer’s disease, Farmaco 56 (2001) 247Á250.
/
[7] A. Sparatore, F. Sparatore, Preparation and pharmacological
activities of 10-homolupinanoyl-2-R-phenothiazines, Farmaco 49
est, Acta Chem. Scand. 3 (1949) 302Á303.
/
[20] H.G. Morren, Substituted N-acylphenothiazines, Belg. Patent
(1994) 5Á17.
/
586,033 (1960) [Chem. Abstr. 54 (1960) 18561e].
[8] F. Novelli, A. Sparatore, B. Tasso, F. Sparatore, Quinolizidinyl
derivatives of 5,11-dihydro-6H-pyrido[2,3-b][1,4]benzodiazepin-
6-one as ligands for muscarinic receptors, Bioorg. Med. Chem.
[21] Socie´te´ des usines chimiques Rhone-Poulenc, Phenothiazine
derivatives, Brit. Patent 740,932 (1955) [Chem. Abstr. 51 (1957)
500i].
Lett. 9 (1999) 3031Á3034.
/
[22] F. Kazutoshi, S. Atsuo, S. Kazuo, O. Kazuo, 10-Chloroacetyl-
phenothiazine, Japan Patent 74,27,191 [Chem. Abstr. 82 (1975)
140161w].
[9] W.G. Eberlein, W.W. Engel, K.M. Hasselbach, N. Mayer, G.
Mihm, K. Rudolf, H. Doods, Tricyclic compounds as selective
muscarinic antagonists: structureÁactivity relationships and ther-
/
[23] L. Toldy, I. Toth, M. Fekete, J. Borsy, Phenothiazine derivatives.
VI. Experiments for the preparation of phenothiazines with
selective coronary dilator activity, Acta Chim. Acad. Sci. Hung.
apeutic implications, in: P. Angeli, U. Gulini, W. Quaglia (Eds.),
Trends in Receptor Research, Elsevier, Amsterdam, 1992, pp.
231Á249.
/
44 (1965) 301Á325.
/
[10] R. Dahlbom, T. Ekstrand, 10-Aminoacylphenothiazines. I. Ami-
noacetyl and aminopropionyl derivatives, Acta Chem. Scand. 5
[24] O. Hromatka, L.H. Schlager, F. Sauter, Phenothiazine deriva-
tives. IV. Synthesis of 10-(piperazinoacyl)phenothiazines, Mon-
(1951) 102Á
[11] R. Dahlbom, T. Ekstrand, 10-Aminoacylphenothiazines. II.
Aminobutyryl derivatives, Acta Chem. Scand. 6 (1952) 1285Á
/
114.
atsh. 88 (1957) 234Á241.
/
[25] G.L. Ellman, K.D. Courtney, V. Andres, Jr., R.M. Featherstone,
/
A new and rapid colorimetric determination of acetylcholinester-
1288.
ase activity, Biochem. Pharmacol. 7 (1961) 88Á95.
[26] M. Eltze, Muscarinic M1- and M2-receptors mediating opposite
/
[12] R. Dahlbom, T. Edlund, T. Ekstrand, A. Lohi, The activity of 10-
aminoacylphenothiazine, basically substituted derivatives of phe-
nothiazine-10-carboxylic acid, atropine, papaverine and some
other spasmolytic drugs against barium and acetylcholine-induced
effects on neuromuscular transmission in rabbit vas deferens, Eur.
J. Pharmacol. 151 (1988) 205Á221.
/