Dioxane and oxathiane nuclei: Suitable substructures for muscarinic agonists

Article abstract of DOI:10.1016/j.bmc.2006.10.040

Muscarinic agonists, bearing 1,4-dioxane and 1,4-oxathiane nuclei, were synthesized and tested to evaluate their potency at M₁-M₄ muscarinic receptor subtypes. The stereochemical relationship between the 2-side chain and the 6-methyl

Full text of DOI:10.1016/j.bmc.2006.10.040

Bioorganic & Medicinal Chemistry 15 (2007) 886–896
Dioxane and oxathiane nuclei: Suitable substructures
for muscarinic agonists
Alessandro Piergentili,a,* Wilma Quaglia,a Mario Giannella,a Fabio Del Bello,a
Bruno Bruni,b Michela Buccioni,a Antonio Carrieric and Samuele Ciattinid
a
`
Dipartimento di Scienze Chimiche, Universita degli Studi di Camerino, Via S. Agostino, 1, 62032 Camerino, Italy
`
b
Dipartimento di Chimica, Universita degli Studi di Firenze, Via della Lastruccia, 3, 50019 Sesto Fiorentino, Firenze, Italy
c
`
Dipartimento Farmaco-Chimico, Universita degli Studi di Bari, Via E. Orabona, 4, 70125 Bari, Italy
Centro Interdipartimentale di Cristallografia, Universita degli Studi di Firenze, Via della Lastruccia, 3,
d
`
50019 Sesto Fiorentino, Firenze, Italy
Received 31 August 2006; revised 16 October 2006; accepted 19 October 2006
Available online 21 October 2006
Abstract—Muscarinic agonists, bearing 1,4-dioxane and 1,4-oxathiane nuclei, were synthesized and tested to evaluate their potency
at M1–M4 muscarinic receptor subtypes. The stereochemical relationship between the 2-side chain and the 6-methyl group plays an
important role in drug–receptor interaction, since the cis isomers are more potent than the corresponding trans isomers. However,
the latter are able to discriminate between the muscarinic receptor subtypes. Among them compound 5b proves particularly inter-
esting, since it selectively activates the ileal M3 receptor subtype and is devoid of agonist activity at the others.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
levels.4 Recent studies suggest that M2, M3, and M4 sub-
types are expressed in staminal cells, whose proliferation
and neurogenesis are modulated by the endogenous li-
gand acetylcholine (ACh).5 Besides, the ability of ACh
to stimulate the DNA synthesis of neuronal cells had al-
ready been described 15 years ago in primary astrocytes.6
In the CNS, the muscarinic receptors are involved in mo-
tor control, cardiovascular and corporeal temperature
regulation, and memory. In the periphery, they mediate
smooth muscle contraction, glandular secretion, and reg-
ulation of cardiac rate and force.4 Therefore, the poten-
tial therapeutic applications of ligands interacting with
such receptors are numerous, though the clinical use is
limited by various side effects due to the lack of a marked
subtype-discrimination. The low selectivity, which is
rarely more than 100-fold higher for one subtype with re-
spect to the others, seems to be due to the high aminoac-
idic sequence homology of the five receptor subtypes.7,8
The study of muscarinic agonists has been arousing
great interest in recent years due to the numerous ther-
apeutic properties mediated by the activation of the cor-
responding receptors.1 On the basis of genetic and
pharmacological characterizations, these receptors,
belonging to the G-protein-coupled receptors (GPCRs)
superfamily,2 are subdivided into five different subtypes
(M1–M5).3
Due to the different transduction mechanism, these five
subtypes have been grouped into two series: the stimula-
tion of those designated by odd numbers (M1, M3, and
M5) causes, above all, the activation of phospholipase
C, leading to the formation of phosphatidyl-inositol
(IP3) and diacylglycerol (DAG) and intracellular Ca2+
mobilization, whereas the stimulation of those designat-
ed by even numbers (M2 and M4) produces adenyl cy-
clase inhibition and the consequent reduction of AMPc
The main applications of the muscarinic agonists con-
cern the treatment of cognitive disorders in Alzheimer’s
disease (AD), in which M1 selective agonists are able to
improve learning and memory.9 Moreover, the activa-
tion of this receptor subtype causes a significant reduc-
tion of b-amyloid peptide (Ab) release via the
stimulation of phosphatidyl-inositol hydrolysis.10
Keywords: Muscarinic agonists; Subtype selectivity; 1,4-Dioxane
nucleus; 1,4-Oxathiane nucleus.
*
Corresponding author. Tel.: +39 0737 402237; fax: +39 0737
637345; e-mail: alessandro.piergentili@unicam.it
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.10.040

A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
887
The beneficial effects in AD psychotic behaviors suggest
CH3
N
the use of muscarinic agonists in the treatment of schizo-
phrenia, in which dopamine release may be reduced by
the stimulation of M1 and M4 heteroreceptors or by
the inhibition of M2 autoreceptors, which, in their turn,
modulate ACh release.11 The observation that musca-
rinic analgesia is mediated by a combination of M2
and M4 subtypes at spinal and supraspinal levels makes
the corresponding ligands, endowed with agonist activi-
ty, promising anti-nociceptive agents.12,13 Moreover, the
activation of muscarinic receptors has been shown to be
neuroprotective in several different models of apopto-
sis;14 the exact action mechanism is not yet known,
but it has been demonstrated that the M3 muscarinic
receptor anti-apoptotic response is independent of calci-
um/phospholipase C signaling.15
CH3
4
2
H3C
5
O
3
1
CH3
Figure 2.
and 3a were synthesized and included in the pharmaco-
logical study (Fig. 1).17,23,25
All the compounds examined bear the archetypical
structure of classical muscarinic agonists, such as
ACh, muscarine, and 1,3-dioxolane 1a (Fig. 2). Howev-
er, though respecting Ing’s rule of the ‘5-atom
chain’,26,27 the different architecture of the backbone
of the fourth N-substituent might determine different
activities at the four muscarinic receptor subtypes and,
Most muscarinic agonists possess a conformationally
constrained structure with respect to the endogenous li-
gand ACh16 and, among them, one of the most potent is
the cis-trimethyl-(2-methyl-[1,3]dioxolan-4-ylmethyl)-
ammonium iodide (1a); this was prepared together with
its corresponding trans isomer (1b) over 40 years ago17
and, at that time, defined a ‘supermuscarinic agent’.
Subsequently, it was demonstrated that its cis-1,3-oxa-
thiolane analogue (2a) showed similar muscarinic
potency.18,19
consequently,
obtained.
a
subtype-discrimination might be
2. Chemistry
The isomers 4a, 4b and 5a, 5b were obtained according
to the reaction sequence reported in Scheme 1. The
treatment of alcohols 6a, 6b and 7a, 7b28,29 with tosyl
chloride and subsequent amination with dimethylamine
afforded the corresponding amines 10a, 10b and 11a,
11b, which, treated with methyl iodide, gave compounds
4a, 4b and 5a, 5b, respectively.
Some studies, among which one of ours,20 have reported
that ligands bearing a hexacyclic structure are endowed
with cholinergic activity.21,22 Also the higher homo-
logues of compound 1a, cis-trimethyl-(2-methyl-[1,3]di-
oxan-4-ylmethyl)-ammonium iodide (3a),23,24 and its
corresponding 1,4-dioxane regioisomer (4 as a cis/trans
mixture)21 have been reported to show a parasympatho-
mimetic action. However, in these papers, accurate
activity–stereochemistry relationships of the molecules
and a complete pharmacological study at all the four
muscarinic subtypes were not investigated. Therefore,
in the present study, the higher homologues of 1 and 2
were synthesized, and the cis and trans isomers were sep-
arated (compounds 4a, 4b and 5a, 5b, respectively).
Moreover, the cis form ( )-4a was resolved into the cor-
responding enantiomers. Finally, compounds 1a, 1b,
The cis stereochemical relationship between the 2-side
chain and the 6-methyl group of 4a, assigned according
to what has been reported in the literature for alcohols
6a, 6b and 7a, 7b,28,29 was confirmed through
X
6a: X = O (cis)
6b: X = O (trans)
7a: X = S (cis)
H3C
O
CH2OH
7b: X = S (trans)
a
X
8a: X = O (cis)
8b: X = O (trans)
9a: X = S (cis)
H3C
O
CH2OTs
O
O
S
9b: X = S (trans)
b
H3C
H3C
H3C
H3C
H3C
O
R
R
O
R
R
O
2a
R
1a
1b
X
10a: X = O (cis)
10b: X = O (trans)
11a: X = S (cis)
11b: X = S (trans)
O
O
O
H3C
O
CH2N(CH3)2
O
H3C
O
O
R
3a
4a
4b
c
S
S
X
4a: X = O (cis)
4b: X = O (trans)
5a: X = S (cis)
H3C
O
R
H3C
O
R
+
5a
5b
H3C
O
CH2N(CH3)3 I
5b: X = S (trans)
+
R = -CH2N(CH3)3 I
Scheme 1. Reagents: (a) p-TsCl, pyridine; (b) (CH3)2NH, benzene; (c)
CH3I, diethyl ether.
Figure 1.

888
A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
fect mediated by M1 receptor subtypes,32 even though
more recent studies attribute the same effect to M4-acti-
vation.37 Therefore, since the pharmacological charac-
terization of the receptor subtype involved does not
appear to be clearly established, in the present work
the rabbit vas deferens will be considered a putative
M1 subtype. The functional data of the new compounds,
reported in Table 1, are expressed as pD2 (ꢀlogED50,
agonist potency), or as pKb (antagonist potency) and
as a (intrinsic activity). Carbachol was included in the
study as reference compound and the two enantiomers
of ( )-2a [(2R,5R)-(+)-2a and (2S,5S)-(ꢀ)-2a]38 are
reported for useful comparison. Moreover, the musca-
rinic activity of compounds ( )-4a, its two enantiomers
(2R,6S)-(+)-4a and (2S,6R)-(ꢀ)-4a, ( )-4b, and carba-
chol was also evaluated by functional studies performed
with a cytosensor microphysiometry instrument (Molec-
ular Devices, Sunnyvale, CA, USA) on stably transfec-
ted CHO (Chinese hamster ovary) cells expressing the
human M1–M5 receptors according to the procedure
reported in the literature.39 The concentration–effect
curves were constructed as a peak acidification response,
seen at increasing concentrations of the agonist, and
pEC50 values (ꢀlogEC50) are reported in Table 2.
Figure 3. A view of the cation in the X-ray structure of the 10a picrate
salt; 30% probability ellipsoids are shown.
single-crystal X-ray diffraction analysis of the picrate
salt of the corresponding base 10a (Fig. 3).
The two enantiomers of ( )-4a were prepared by stereo-
selective synthesis starting from (R)- or (S)-3-benzyloxy-
propane-1,2-diol, whose reaction with chloroacetone led
to the corresponding diastereomeric forms (4S)- and
(4R)-4-benzyloxymethyl-2-chloromethyl-2-methyl-
[1,3]dioxolane [(4S)-12 and (4R)-12; cis/trans ratio = 1:1]
(Scheme 2). The hydrogenolysis with Pd(OH)2 followed
by the treatment with KOH and the subsequent
regiospecific opening30 of the bicyclo[3.2.1]octanes
(1S,5R)-14 and (1R,5S)-14 afforded mixtures of the cor-
responding diastereomeric forms (2R,6S)-6a, (2R,6R)-6b
and (2S,6R)-6a, (2S,6S)-6b, respectively. The diastereo-
mers were separated by column chromatography. The
trans enantiomeric forms (2R,6R)-6b and (2S,6S)-6b
were obtained in poor yields and, therefore, it was not
possible to proceed with the subsequent reaction and
to obtain the corresponding final compounds. Finally,
the two methiodides (2R,6S)-(+)-4a and (2S,6R)-(ꢀ)-
4a were prepared by subjecting the enantiomeric alco-
hols (2R,6S)-(+)-6a and (2S,6R)-(ꢀ)-6a to the same
reaction sequence as that shown in Scheme 1. Enantio-
meric purity of amines (2R,6S)-(+)-10a and (2S,6R)-
The analysis of data shows that all the compounds are
full agonists at M1–M3 muscarinic receptor subtypes,
except for compounds 3a and (2S,6R)-(ꢀ)-4a at M1
and compound 5b at M1 and M2, which are not able
to activate these subtypes and, when tested as antago-
nists, proved inactive up to 10 lM. Concerning the M4
subtype, the behavior of the tested compounds is vari-
ous: in fact, most of them are full agonists, ( )-4a and
(2S,6R)-(ꢀ)-4a have nearly full-intrinsic activity
(a = 0.81 and 0.75, respectively), 3a and 4b behave as
partial agonists, and 5a and 5b are weak antagonists.
Analogously to carbachol, all the compounds display
higher potency values toward M2 and M3 subtypes with
respect to the others, except for compound 3a, which is
more active toward the M4 subtype. Interestingly, the
biological behaviors of the two higher homologues of
1a (compounds 3a and 4a) are different. The enlarge-
ment of the 1,3-dioxolane nucleus of 1a by inserting a
methylene group between the 1-oxygen atom and the
4-side chain, affording 3a, alters both activity and selec-
tivity toward muscarinic receptors. In fact, compound
3a, not activating M1 subtype and being significantly
more active than 1a at M4, and, above all, 155-fold and
91-fold less active at M2 and M3, respectively, proves
to be selective for the M4 subtype. Instead, compound
4a, obtained by inserting a methylene group between
the 1-oxygen atom and the 2-carbon atom of 1a, shows
pD2 values not significantly different from those of 1a,
allowing the same interaction mechanism to be hypothe-
sized. In the absence of experimental data regarding the
accurate X-ray structure of any of the muscarinic recep-
tor subtypes which might help to formulate a sound
hypothesis on the receptor–ligand interaction pattern, a
plausible way to interpret the biological profiles of differ-
ent compounds is the use of the so-called active analogue
approach; this ranks the biological profile of similar
compounds also according to the comparison of some
basic three-dimensional features of molecules (i.e.,
1
(ꢀ)-10a, determined by H NMR spectroscopy in com-
parison with the spectrum of racemic compound 10a
and on addition of the chiral shift reagent (R)-(+)-a-
methoxy-a-(trifluoromethyl)phenylacetic acid [(+)-
MTPA],31 was found to be >98% (detection limit) for
both enantiomers. In fact, the spectrum of racemic 10a
in the presence of (+)-MTPA showed a double doublet
at d 1.05 ppm for the 6-methyl protons, whereas only
one doublet was observed for (2R,6S)-(+)-10a and
(2S,6R)-(ꢀ)-10a at d 1.02 and 1.06 ppm, respectively.
3. Results and discussion
The muscarinic activity of compounds 1–5 was evaluat-
ed by functional studies performed on classical prepara-
tions: rabbit stimulated vas deferens (M1),32 guinea-pig
stimulated left atria (M2),33 guinea-pig ileum (M3),34,35
and guinea-pig lung strips (M4).36 Concerning this, it
is appropriate to mention that the contraction of rabbit
vas deferens was for a long time considered to be an ef-

A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
889
HO
HO
HO
HO
CH2OCH2C6H5
CH2OCH2C6H5
a
b
O
O
H3C
H3C
O
CH2OCH2C6H5
O
CH2OCH2C6H5
ClH2C
ClH2C
(4S)-12
(4R)-12
O
O
O
H3C
H3C
O
CH2OH
CH2OH
ClH2C
ClH2C
(4R)-13
(4S)-13
c
H
H3C
O
H3C
O
H
1
5
1
5
O
O
O
O
(1R,5S)-14
(1S,5R)-14
d
O
O
O
O
O
O
+
+
H3C
CH2OH
H3C
O
CH2OH
H3C
O
CH2OH
H3C
CH2OH
(2S,6R)-6a
(2S,6S)-6b
(2R,6R)-6b
(2R,6S)-6a
e, f, g
O
e, f, g
O
+
+
H3C
O
CH2N(CH3)3 I
H3C
O
CH2N(CH3)3 I
(2S,6R)-4a
(2R,6S)-4a
Scheme 2. Reagents: (a) ClCH2COCH3, TsOH, toluene; (b) Pd(OH)2, methanol; (c) powdered KOH; (d), LiAlH4, AlCl3, diethyl ether; (e) p-TsCl,
pyridine; (f) (CH3)2NH, benzene; (g) CH3I, diethyl ether.
molecular volume, interatomic distances, partial charges,
dipole orientation, etc.) which can nowadays be easily
calculated and visualized through molecular modeling
tools.
atom of the 1,3-dioxolane ring in the case of 1a; on
the other hand, this cannot be equally observed for 3a,
probably due to the presence of an extra methylene
group. The presence of more than one polar residue in
the receptor site may be postulated; these residues might
interact with the two properly oriented lone pairs of the
1-oxygen atom of 1a, resulting in a stronger hydrogen
bond (Fig. 4c). A steric clash should hamper such inter-
action with the receptor counterpart in the case of 3a.
In our case, the pharmacological data of compounds 3a
and 4a might be analyzed by the overlapping of their
low-energy conformations with that of the highly active
compound 1a. In fact, the molecular overlay of 1a and
4a shows that similar spatial regions are occupied by
the presumed pharmacophore points, such as the basic
methylated nitrogen atoms, the polar oxygen atoms,
and the methyl substituents (Fig. 4a), whereas the intro-
duction of a methylene group between the 1-oxygen
atom and the 4-side chain of 1a, affording 3a, seems to
produce a slight increase of the steric hindrance in the
1-oxygen atom interaction area (Fig. 4b).
A similar interpretation might be given also in the case of
the thio isoster of 4a considered in this study. In fact, the
replacement of the 4-oxygen atom of 1,4-dioxane com-
pound 4a with a sulfur atom, affording 5a, significantly
reduces the potency at all the muscarinic receptor sub-
types, unlike what occurs by carrying out the same modi-
fication on the 1,3-dioxolane derivative 1a, whose sulfur
analogue2ashowsacomparablepharmacologicalprofile.
This hypothesis might also be supported by inspection
of the molecular interaction fields computed around
the molecular surface of the two compounds; these fields
highlight a more extended region around the 1-oxygen
The different effects produced by the same modification
carried out on compounds 1a and 4a, that is the
replacement of the oxygen atom in position 1 and 4,

890
A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
Table 1. Potency expressed as ꢀlogED50 (pD2), intrinsic activity (a)a and dissociation constant (pKb)b of compounds 1–5 at M1–M4 muscarinic
receptor subtypesc
Compound
Rabbit vas deferens
(M1)
Guinea-pig atrium
(M2)
Guinea-pig ileum
(M3)
Guinea-pig lung
(M4)
a
pD2 (pKb)
a
pD2 (pKb)
a
pD2 (pKb)
a
pD2 (pKb)
d
Carbachol
1a
—
—
—
1
—
—
1
1
1
1
7.33 0.08
7.43 0.11
6.73 0.15
7.40 0.05
7.54 0.10e
5.37 0.05e
5.24 0.29
7.57 0.11
6.66 0.19
6.80 0.10
7.35 0.21
6.26 0.16
(<5)f
1
6.68 0.01
7.69 0.13
7.47 0.15
7.75 0.02
8.06 0.04e
5.82 0.06e
5.73 0.15
7.34 0.13
6.02 0.16
7.65 0.12
6.53 0.16
6.46 0.23
5.21 0.10
1
>1
>1
1
5.43 0.03
5.80 0.19
5.71 0.05
6.15 0.04
>1
>1
1
1b
( )-2a
—
6.81 0.09
(2R,5R)-(+)-2a
(2S,5S)-(ꢀ)-2a
3a
( )-4a
(<5)f
1
1
1
1
1
1
1
1
1
1
1
1
1
0.54 0.06
0.81 0.07
0.53 0.07
1
6.41 0.04
5.90 0.14
6.01 0.03
5.87 0.15
4.34 0.20
(5.14 0.02)
(5.42 0.05)
1
1
1
5.65 0.06
5.08 0.24
6.78 0.04
(<5)f
4b
(2R,6S)-(+)-4a
(2S,6R)-(ꢀ)-4a
0.75 0.11
5a
5b
1
4.46 0.15
(<5)f
a Intrinsic activity was measured by the ratio between the maximum response of the agonist and the maximum response of McN-A-343 at M1, and
arecaidine propargyl ester (APE) at M2, M3, and M4 subtypes.
b Dissociation constants were calculated from the equation: log(DR-1) = log[ant] ꢀ logKb according to van Rossum.46
c The results are means ( SEM) of four to six independent experiments.
d Not determined.
e Data from Ref. 38.
f This compound showed no agonist activity and, when tested as antagonist, proved inactive up to 10 lM.
a
Table 2. Potency expressed as ꢀlogEC50 (pEC50
)
carbachol at human cloned M1–M5 receptor subtypes
and intrinsic activity (a)b of compounds ( )-4a, ( )-4b, (2R,6S)-(+)-4a, (2S,6R)-(ꢀ)-4a, and
Compound
hM1
pEC50
hM2
hM3
pEC50
hM4
hM5
a
a
pEC50
a
a
pEC50
a
pEC50
Carbachol
( )-4a
( )-4b
1
4,30 0.04
4.41 0.05
3.31 0.03
5.08 0.06
3.10 0.05
1
1
1
1
1
4.19 0.06
4.31 0.07
3.11 0.05
4.43 0.02
4.81 0.06
1
5.74 0.04
6.56 0.04
4.25 0.05
6.39 0.05
4.30 0.07
1
1
1
1
1
5.40 0.03
5.60 0.07
3.55 0.05
5.55 0.08
4.60 0.04
1
4.96 0.06
5.10 0.04
3.16 0.03
5.43 0.04
NAc
1
>1
1
1
1
1
(2R,6S)-(+)-4a
(2S,6R)-(ꢀ)-4a
1
0.8
1
1
—
1
a Determined by applying the cytosensor microphysiometry system to study the five human M1–M5 muscarinic subtypes expressed in CHO cells.
b Intrinsic activity was measured by the ratio between the maximum response of the agonist and the maximum response of carbachol.
c NA, not active (intrinsic activity <0.3).
Figure 4. (a) Molecular overlays of compounds 1a and 4a; (b) 1a and 3a. (c) Molecular interaction fields for compounds 1a (cyan) and 3a (yellow)
contoured at ꢀ3.0 kcal/mol.
respectively, with a sulfur one, might be explained by
overlapping 1a and 2a (Fig. 5a) and 1a, 4a, and 5a
(Fig. 5b), a good overlay being obtained only in the case
of 1a and 2a.
tionship between the side chain and the methyl group is
not important either for the activation of M4 subtype
(pairs 1a/1b and 4a/4b) or for its blocking (pair 5a/5b).
Different behaviors may be observed when M2 and M3
subtypes are considered. In fact, while the two isomers
1a and 1b show similar agonist activities at the M3 sub-
type and the cis isomer is only slightly more active than
The analysis of data relative to the three pairs of isomers
(1a/1b, 4a/4b, 5a/5b) shows that the stereochemical rela-

A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
891
with respect to the other muscarinic receptor subtypes is
noteworthy due to the evidence that the configurations
of its stereogenic centers are the opposite of those of
the eutomer of ( )-2a. The above results are also con-
firmed by the data of compounds ( )-4a, its two enanti-
omers [(2R,6S)-(+)-4a and (2S,6R)-(ꢀ)-4a], and ( )-4b
obtained on CHO cells (Table 2), though the pD2 values
are lower than those performed on tissues (Table 1).
These differences may not represent an anomaly because
in screening procedures a homogeneous population of
cloned receptors is used, which can be organized differ-
ently to native receptors in functional tissues and, conse-
quently, their biological behavior may not be coincident.
Figure 5. (a) Molecular overlays of compounds 1a and 2a; (b) 1a, 4a,
and 5a.
Therefore, clearly, all the muscarinic subtypes prove to
have stereospecific requirements, whose characterization
might be useful for the design of selective ligands.
the trans isomer at the M2, in the case of the pairs 4a/4b
and 5a/5b, the differences of activities are more evident,
since the cis isomers are significantly more active than
the corresponding trans isomers at both M2 and M3sub-
types. Therefore, for compounds bearing hexacyclic
nuclei, the stereochemical relationship between the
2-side chain and the 6-methyl group plays an important
role in drug–receptor interaction.
In conclusion, this study suggests that hexacyclic 1,4-di-
oxane and 1,4-oxathiane compounds might represent
suitable templates for the design of new muscarinic li-
gands and that, depending on ligand structure, a cis
and trans relationship between the 2-side chain and the
6-methyl group might play an important role for a pref-
erential muscarinic subtype activation.
Moreover, none of the cis compounds is able to discrim-
inate between the M2 and M3 subtypes. On the contrary,
in the case of the trans compounds a slight but signifi-
cant preferential activation of one of these two subtypes
may be observed. Among them, the behavior of com-
pound 5b, which exclusively activates the ileal M3 recep-
tor subtype, though with a not very high pD2 value, is
extremely interesting. Such a marked M3-selectivity,
which is particularly significant due to the complete lack
of activity toward M2, makes this compound a useful
tool for the characterization of the M3-mediated func-
tions and for the design of new muscarinic receptor ago-
nists selective for this subtype.
4. Experimental
4.1. Chemistry
Melting points were taken in glass capillary tubes on a
Buchi SMP-20 apparatus and are uncorrected. IR and
¨
1H NMR spectra were recorded on Perkin-Elmer 297
and Varian EM-390 instruments, respectively. Chemical
shifts are reported in parts per million (ppm) relative to
tetramethylsilane (TMS), and spin multiplicities are giv-
en as s (singlet), br s (broad singlet), d (doublet), t (trip-
let), q (quartet), or m (multiplet). Although the IR
spectral data are not included because of the lack of
unusual features, they were obtained for all compounds
reported and were consistent with the assigned struc-
tures. The elemental compositions of the compounds
were performed by the Microanalytical Laboratory of
our department and agreed to within 0.4% of the cal-
culated value. When the elemental analysis was not
included, crude compounds were used in the next step
without further purification. Optical rotation was mea-
sured at a 1 g/100 mL concentration (c = 1), unless
otherwise stated, with a Perkin-Elmer 241 polarimeter
(accuracy 0.002ꢁ). Chromatographic separations were
performed on silica gel columns (Kieselgel 40, 0.040–
0.063 mm, Merck) by flash chromatography. Com-
pounds were named following IUPAC rules as applied
by Beilstein-Institut AutoNom (version 2.1), a software
for systematic names in organic chemistry. Carbachol
and compound 2a were commercially available (RBI-
Sigma).
The comparison of the biological profiles of the two
enantiomers of the most active compound ( )-4a among
those bearing a hexacyclic nucleus is noteworthy. Also
in this case, a significant selectivity for M2 or M3 musca-
rinic subtypes, absent in all the cis racemic isomers, may
be observed for the two enantiomers (2R,6S)-(+)-4a and
(2S,6R)-(ꢀ)-4a. In fact, (2R,6S)-(+)-4a is significantly
more potent at M3 with respect to M1, M2, and M4 sub-
types, whereas its optical antipode (2S,6R)-(ꢀ)-4a shows
a different biological profile, since it is unable to activate
M1 subtype (pKb < 5) and is more active at M2 with re-
spect to M3 and M4 (7- and 1023-fold, respectively).
Moreover, (2R,6S)-(+)-4a is more potent than (2S,6R)-
(ꢀ)-4a at M3 and M4 subtypes with eudismic ratios
(ER) of 13 and 34, respectively, and, above all, behaves
as a full agonist at M1 subtype, which is not activated by
(2S,6R)-(ꢀ)-4a. In the case of the M2 subtype, the
eutomer is the levorotatory form (2S,6R)-(ꢀ)-4a
(ER = 4), an interesting reversal enantioselectivity being
obtained. These data are quite interesting since, in the
case of the oxathiolane agonist ( )-2a, the eutomer is
the enantiomer (2R,5R)-(+)-2a for both M2 and M3 sub-
types. Therefore, the selectivity of (2S,6R)-(ꢀ)-4a for M2
4.1.1. cis/trans-(4S)-4-Benzyloxymethyl-2-chloromethyl-
2-methyl-[1,3]dioxolane [(4S)-12]. A solution of chloro-
acetone (2.5 g, 27.0 mmol), (R)-3-benzyloxy-propane-1,
2-diol (5.0 g, 27 mmol), and p-toluenesulfonic acid

892
A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
(0.07 g, 0.44 mmol) in toluene (30 mL) was refluxed with
vigorous stirring and in a Stark apparatus for 6 h. After
cooling, the mixture was washed with Na2CO3 saturated
solution (50 mL) and dried over Na2SO4. Evaporation
of the solvent gave a residue, which was purified by col-
umn chromatography. Eluting with cyclohexane/EtOAc
(9:1) afforded (4S)-12 as a cis/trans mixture (ratio 1:1):
dried over Na2SO4. Removal of the solvent gave an oil,
which was purified by column chromatography eluting
with cyclohexane/EtOAc (7:3). The cis isomer (2R,6S)-
20
6a eluted first: 0.8 g (26% yield); ½aꢁ +13.40 (c 1,
D
1
CH3OH); H NMR (CDCl3) d 0.99 (d, 3, CH3), 2.95–
3.64 (m, 8, cyclo and CH2O), 4.62 (br s, 1, OH). The sec-
ond fraction was the trans isomer (2R,6R)-6b: 0.08 g (3%
1
1
6.03 g (87% yield); H NMR (CDCl3) d 1.44 and 1.51
yield); H NMR (CDCl3) d 1.04 (d, 3, CH3), 3.10–3.74
(m, 8, cyclo and CH2O), 4.62 (br s, 1, OH).
(two s, 6, CH3; cis and trans), 3.42–4.48 (m, 14,
CH2O, CH2Cl and OCHCH2; cis and trans), 4.60 (s, 4,
CH2Ar; cis and trans), 7.24–7.39 (m 10, ArH; cis and
trans).
4.1.8. (2S,6R)-(ꢀ)- and (2S,6S)-(6-Methyl-[1,4]dioxan-2-
yl)-methanol [(2S,6R)-(ꢀ)-6a and (2S,6S)-6b]. These
were prepared as described for (2R,6S)-6a and
(2R,6R)-6b starting from (1R,5S)-14 (3.5 g, 27.0 mmol).
4.1.2. cis/trans-(4R)-4-Benzyloxymethyl-2-chloromethyl-
2-methyl-[1,3]dioxolane [(4R)-12]. This was prepared as
a cis/trans mixture (ratio 1:1) as described for (4S)-12
starting from (S)-3-benzyloxy-propane-1,2-diol (5.0 g,
The cis isomer (2S,6R)-6a eluted first: 0.8 g (23% yield);
20
D
½aꢁ ꢀ13.29 (c 1, CH3OH). The 1H NMR spectrum was
identical to that of (2R,6S)-6a. The second fraction
was the trans isomer (2S,6S)-6b: 0.07 g (2% yield).
The 1H NMR spectrum was identical to that of
(2R,6R)-6b.
1
27.0 mmol): 6.0 g (87% yield). The H NMR spectrum
was identical to that of (4S)-12.
4.1.3. cis/trans-(4S)-(2-Chloromethyl-2-methyl-[1,3]dioxo-
lan-4-yl)-methanol [(4S)-13]. 20% Pd(OH)2/C (0.4 g,
0.57 mmol) was added to a solution of (4S)-12 (6.03 g,
23.5 mmol) in MeOH (40 mL) and the mixture was
mechanically shaken under 50 psi of H2 in a Parr appa-
ratus for 3 h. The catalyst was removed by Millipore fil-
tration, and the solvent was evaporated to give (4S)-13
4.1.9. Toluene-4-sulfonic acid (2S,6S)-(+)-6-methyl-[1,4]di-
oxan-2-ylmethyl ester [(2S,6S)-(+)-8a]. Tosyl chloride
(3.56 g, 18.7 mmol) was added to a stirred solution of
(2R,6S)-6a (1.7 g, 12.9 mmol) in pyridine (20 mL) at
0 ꢁC over 30 min. After being stirred for 3 h at 0 ꢁC,
the mixture was left for 20 h at 4 ꢁC in the freezer. Then
it was poured into ice and concentrated HCl (20 mL)
and extracted with CHCl3. The organic layers were
washed with HCl 2 N (60 mL), NaHCO3 saturated solu-
tion (60 mL), and H2O (60 mL) and then dried over
Na2SO4. The solvent was concentrated in vacuo to give
a residue, which was purified by column chromatogra-
1
as a cis/trans mixture (ratio 1:1): 3.8 g (97% yield); H
NMR (CDCl3) d 1.46 and 1.52 (two s, 6, CH3; cis and
trans), 1.73 (br s, 2, OH; cis and trans), 3.47–4.42 (m,
14, CH2O, CH2Cl and OCHCH2; cis and trans).
4.1.4. cis/trans-(4R)-(2-Chloromethyl-2-methyl-[1,3]dioxo-
lan-4-yl)-methanol [(4R)-13]. This was prepared as a cis/
trans mixture (ratio 1:1) as described for (4S)-13 starting
from (4R)-12 (6.0 g, 23.4 mmol): 3.5 g (90% yield). The
1H NMR spectrum was identical to that of (4S)-13.
phy. Eluting with cyclohexane/EtOAc (9:1) afforded
20
(2S,6S)-8a: 2.0 g (54 % yield); ½aꢁ +1.30 (c 1, CHCl3);
D
1H NMR (CDCl3) d 1.09 (d, 3, CH3), 2.23 (s, 3, ArCH3),
3.20–4.39 (m, 8, cyclo and CH2O), 7.39 (d, 2, ArH), 7.81
(d, 2, ArH).
4.1.5. (1S,5R)-(ꢀ)-5-Methyl-3,6,8-trioxa-bicyclo[3.2.1]oc-
tane [(1S,5R)-(ꢀ)-14]. A mixture of (4S)-13 (3.8 g,
22.8 mmol) and powdered KOH (1.90 g, 33.9 mmol)
was heated until an exothermic reaction developed.
4.1.10. Toluene-4-sulfonic acid (2R,6R)-(ꢀ)-6-methyl-
[1,4]dioxan-2-ylmethyl ester [(2R,6R)-(ꢀ)-8a]. This was
20
prepared as described for (2S,6S)-8a starting from
(2S,6R)-6a (0.82 g, 6.2 mmol): 1.25 g (70% yield); ½aꢁ
After cooling, the residue was distilled (61–62 ꢁC/
D
20
1
130 mmHg): 2.9 g (98% yield); ½aꢁ ꢀ31.41 (c 1, CHCl3);
ꢀ1.19 (c 1, CHCl3). The H NMR spectrum was identi-
D
1H NMR (CDCl3) d 1.38 (s, 3, CH3), 3.52–4.83 (m, 7,
trioxa-bicyclo).
cal to that of (2S,6S)-8a.
4.1.11. cis-Toluene-4-sulfonic acid 6-methyl-[1,4]dioxan-
2-ylmethyl ester (8a). This was prepared as described for
(2S,6S)-8a starting from 6a28 (1.1 g, 8.3 mmol): 1.92 g
(81% yield). The H NMR spectrum was identical to
that of (2S,6S)-8a.
4.1.6. (1R,5S)-(+)-5-Methyl-3,6,8-trioxa-bicyclo[3.2.1]oc-
tane [(1R,5S)-(+)-14]. This was prepared as described
for (1S,5R)-14 starting from (4R)-13 (3.5 g, 21.0 mmol):
1
20
2.5 g (92% yield); ½aꢁ +30.95 (c 1, CHCl3). The 1H
D
NMR spectrum was identical to that of (1S,5R)-14.
4.1.12. trans-Toluene-4-sulfonic acid 6-methyl-[1,4]diox-
an-2-ylmethyl ester (8b). This was prepared as described
for (2S,6S)-8a starting from 6b28 (0.55 g, 4.2 mmol):
0.96 g (81% yield); 1H NMR (CDCl3) d 1.09 (d, 3,
CH3), 2.45 (s, 3, ArCH3), 3.13–4.38 (m, 8, cyclo and
CH2O), 7.35 (d, 2, ArH), 7.80 (d, 2, ArH).
4.1.7. (2R,6S)-(+)- and (2R,6R)-(6-Methyl-[1,4]dioxan-2-
yl)-methanol [(2R,6S)-(+)-6a and (2R,6R)-6b]. A solution
of AlCl3 (3.0 g, 23.0 mmol) in diethyl ether (23 mL) was
added dropwise over 3 min to a stirred solution of
(1S,5R)-14 (3.0 g, 23.0 mmol) and LiAlH4 (0.9 g,
23.0 mmol) in diethyl ether (53 mL). The reaction mix-
ture was refluxed vigorously until the starting material
disappeared. Then it was cooled to 0 ꢁC and quenched
cautiously by the dropwise addition of Na2SO4 saturat-
ed solution. The solid was filtered off and the filtrate was
4.1.13. cis-Toluene-4-sulfonic acid 6-methyl-[1,4]oxathi-
an-2-ylmethyl ester (9a). This was prepared as described
for (2S,6S)-8a starting from 7a29 (1.9 g, 12.8 mmol):
3.55 g (91% yield); 1H NMR (CDCl3) d 1.12 (d, 3,

A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
893
CH3), 2.17–2.53 (m, 4, CH2SCH2), 2.41 (s, 3, ArCH3),
3.55–4.02 (m, 4, CHOCHCH2O), 7.31 (d, 2, ArH),
7.77 (d, 2, ArH).
2.22 (s, 6, N(CH3)2), 2.30–2.80 (m, 6, CH2SCH2 and
CH2N), 4.02 (m, 1, CHO), 4.10 (m, 1, CHO).
4.1.21. (2R,6S)-(+)-Trimethyl-(6-methyl-[1,4]dioxan-2-
ylmethyl)-ammonium iodide [(2R,6S)-(+)-4a]. A solution
of (2R,6S)-10a (0.4 g, 2.5 mmol) in diethyl ether (15 mL)
was treated with an excess of methyl iodide and left at rt
4.1.14. trans-Toluene-4-sulfonic acid 6-methyl-[1,4]oxa-
thian-2-ylmethyl ester (9b). This was prepared as de-
scribed for (2S,6S)-8a starting from 7b29 (1.4 g,
1
9.5 mmol): 1.6 g (56% yield); H NMR (CDCl3) d 1.22
in the dark for 24 h. The solid was filtered and recrystal-
20
D
(d, 3, CH3), 2.28–2.78 (m, 4, CH2SCH2), 2.46 (s, 3,
ArCH3), 3.82–4.37 (m, 4, CHOCHCH2O), 7.35 (d, 2,
ArH), 7.80 (d, 2, ArH).
lized from 2-PrOH; mp 200 ꢁC; ½aꢁ +38.91 (c 1,
1
CH3OH); H NMR (DMSO) d 1.07 (d, 3, CH3), 2.99–
3.10 (m, 2, CH2N), 3.12 (s, 9, N(CH3)3), 3.45 (m, 2,
cyclo), 3.60 (m, 2, cyclo), 3.80 (m, 1, cyclo), 4.21 (m,
1, cyclo). Anal. Calcd for C9H20INO2: C, 35.89; H,
6.69; N, 4.65. Found: C, 35.91; H, 6.81; N, 4.55.
4.1.15.
ylmethyl)-amine [(2R,6S)-(+)-10a].
(2R,6S)-(+)-Dimethyl-(6-methyl-[1,4]dioxan-2-
solution of
A
(2S,6S)-8a (0.7 g, 2.4 mmol) and dimethylamine
(5.5 mL) in dry benzene (10 mL) was heated in a sealed
tube for 60 h at 120 ꢁC. After evaporation of the solvent,
the residue was dissolved in CHCl3, which was washed
with NaOH 2 N and dried over Na2SO4. The solvent
was concentrated in vacuo to give a residue, which
was purified by column chromatography. Eluting with
CHCl3/CH3OH (9:1) afforded (2R,6S)-10a as the free
4.1.22. (2S,6R)-(ꢀ)-Trimethyl-(6-methyl-[1,4]dioxan-2-
ylmethyl)-ammonium iodide [(2S,6R)-(ꢀ)-4a]. This was
prepared as described for (2R,6S)-4a starting from
(2S,6R)-10a (0.5 g, 3.1 mmol): the solid was recrystal-
20
lized from 2-PrOH; mp 200 ꢁC; ½aꢁ ꢀ39.01 (c 1,
D
CH3OH). The 1H NMR spectrum was identical to
that of (2R,6S)-4a. Anal. Calcd for C9H20INO2: C,
35.89; H, 6.69; N, 4.65. Found: C, 35.78; H, 6.55;
N, 4.39.
20
1
base: 0.3 g (77% yield); ½aꢁ +12.24 (c 1, CHCl3); H
D
NMR (CDCl3) d 1.10 (d, 3, CH3), 2.12–2.43 (m, 2,
CH2N), 2.26 (s, 6, N(CH3)2), 3.08–3.24 (m, 2, cyclo),
3.56–3.81 (m, 4, cyclo); enantiomeric purity was >98%
(detection limit), determined with (+)-MTPA31 as the
chiral shift reagent.
4.1.23. cis-Trimethyl-(6-methyl-[1,4]dioxan-2-ylmethyl)-
ammonium iodide (4a). This was prepared as described
for (2R,6S)-4a starting from 10a (0.5 g, 3.1 mmol): the
solid was recrystallized from 2-PrOH: mp 200 ꢁC. The
1H NMR spectrum was identical to that of (2R,6S)-4a.
Anal. Calcd for C9H20INO2: C, 35.89; H, 6.69; N,
4.65. Found: C, 35.77; H, 6.88; N, 4.45.
4.1.16. (2S,6R)-(ꢀ)-Dimethyl-(6-methyl-[1,4]dioxan-2-
ylmethyl)-amine [(2S,6R)-(ꢀ)-10a]. This was prepared
20
as described for (2R,6S)-10a starting from (2R,6R)-8a
(0.8 g, 2.8 mmol): 0.4 g (89% yield); ½aꢁ ꢀ12.49 (c 1,
D
1
CHCl3). The H NMR spectrum was identical to that
of (2R,6S)-10a. Enantiomeric purity was >98% (detec-
tion limit), determined with (+)-MTPA31 as the chiral
shift reagent.
4.1.24. trans-Trimethyl-(6-methyl-[1,4]dioxan-2-ylmeth-
yl)-ammonium iodide (4b). This was prepared as de-
scribed for (2R,6S)-4a starting from 10b (0.2 g,
1.3 mmol): the solid was recrystallized from 2-PrOH;
1
mp 216–217 ꢁC; H NMR (DMSO) d 1.10 (d, 3, CH3),
4.1.17. cis-Dimethyl-(6-methyl-[1,4]dioxan-2-ylmethyl)-
amine (10a). This was prepared as described for
(2R,6S)-10a starting from 8a (1.92 g, 6.7 mmol): 0.74 g
3.12 (s, 9, N(CH3)3), 3.19–3.30 (m, 2, CH2N), 3.50 (m,
2, cyclo), 3.62 (m, 2, cyclo), 4.00 (m, 2, cyclo). Anal.
Calcd for C9H20INO2: C, 35.89; H, 6.69; N, 4.65.
Found: C, 35.81; H, 6.56; N, 4.48.
1
(69% yield). The H NMR spectrum was identical to
that of (2R,6S)-10a.
4.1.25. cis-Trimethyl-(6methyl-[1,4]oxathian-2-ylmethyl)-
ammonium iodide (5a). This was prepared as described
for (2R,6S)-4a starting from 11a (1.0 g, 5.7 mmol): the
solid was recrystallized from EtOH; mp 193–194 ꢁC;
1H NMR (DMSO) d 1.15 (d, 3, CH3), 2.29–2.50 (m, 4,
CH2SCH2), 3.10 (s, 9, N(CH3)3), 3.45 (m, 2, CH2N),
3.89 (m, 1, CHO), 4.22 (m, 1, CHO). Anal. Calcd for
C9H20INOS: C, 34.07; H, 6.35; N, 4.42; S, 10.11. Found:
C, 34.25; H, 6.21; N, 4.18; S, 10.23.
4.1.18. trans-Dimethyl-(6-methyl-[1,4]dioxan-2-ylmeth-
yl)-amine (10b). This was prepared as described for
(2R,6S)-10a starting from 8b (0.5 g, 1.7 mmol): 0.2 g
(71% yield); 1H NMR (CDCl3) d 1.20 (d, 3, CH3),
2.30 (s, 6, N(CH3)2), 2.35–2.62 (m, 2, CH2N), 3.30–
3.52 (m, 2, cycle), 3.68–3.73 (m, 2, cyclo), 3.83–4.11
(m, 2, cyclo).
4.1.19. cis-Dimethyl-(6-methyl-[1,4]oxathian-2-ylmethyl)-
amine (11a). This was prepared as described for (2R,6S)-
10a starting from 9a (1.4 g, 4.6 mmol): 0.6 g (74% yield);
1H NMR (CDCl3) d 1.20 (d, 3, CH3), 2.23 (s, 6,
N(CH3)2), 2.26 (m, 6, CH2SCH2 and CH2N), 3.75 (m,
2, CHOCH).
4.1.26. trans-Trimethyl-(6-methyl-[1,4]oxathian-2-ylm-
ethyl)-ammonium iodide (5b). This was prepared as de-
scribed for (2R,6S)-4a starting from 11b (1.4 g,
8.0 mmol); the solid was recrystallized from EtOH; mp
207–208 ꢁC; 1H NMR (DMSO) d 1.20 (d, 3, CH3),
2.40–2.80 (m, 4, CH2SCH2), 3.15 (s, 9, N(CH3)3), 3.35
(m, 2, CH2N), 4.18 (m, 1, CHO), 4.55 (m, 1, CHO).
Anal. Calcd for C9H20INOS: C, 34.07; H, 6.35; N,
4.42; S, 10.11. Found: C, 33.97; H, 6.38; N, 4.39; S,
10.35.
4.1.20. trans-Dimethyl-(6-methyl-[1,4]oxathian-2-ylmeth-
yl)-amine (11b). This was prepared as described for
(2R,6S)-10a starting from 9b (1.4 g, 4.6 mmol): 0.4 g
(49% yield); 1H NMR (CDCl3) d 1.36 (d, 3, CH3),

894
A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
4.2. Molecular modeling
Centre, 12 Union Road, Cambridge CB2 1EZ, UK;
fax: +44-1223-336033; e-mail: deposit@ccdc.cam.ac.uk.
Molecular superpositions of compounds 1a, 2a, 3a, 4a,
and 5a were performed by means of the QXP software40
as follows: initially molecular structures were built using
standard bond length and valence angles; then reference
compound 1a was submitted to a 5000-steps Monte-
Carlo conformational analysis run in order to assess
the proper spatial arrangement with respect to the 1,3-
dioxolane ring. Each of the other compounds was fitted
onto the low energy conformation of the selected tem-
plate according to the template fitting search algorithm
implemented within the same QXP package. This proce-
dure searches for the best superposition on the basis of
attractive forces between similar atoms or functional
groups in agreement with energetically permitted
three-dimensional states.
4.4. Pharmacology
General considerations. Male guinea pigs (200–300 g)
and male New Zealand white rabbits (3.0–3.5 kg) were
killed by cervical dislocation. The organs required were
set up rapidly under 1 g of tension in 20-mL organ baths
containing physiological salt solution (PSS) maintained
at an appropriate temperature and aerated with 5%
CO2–95% O2. Dose–response curves were constructed
by cumulative addition of the reference agonist. The
concentration of agonist in the organ bath was increased
approximately 3-fold at each step, with each addition
being made only after the response to the previous addi-
tion had attained a maximal level and remained steady.
After 30 min of washing, a cumulative dose–response
curve to the agonist under study was constructed. When
the compound in question behaved as an antagonist, fol-
lowing 30 min of washing, tissues were incubated with
the antagonist for 30 min, and a new dose–response
curve to the agonist was obtained. Responses were ex-
pressed as a percentage of the maximal response ob-
tained in the control curve. Contractions were
recorded by means of a force displacement transducer
connected to the MacLab system PowerLab/800. In
addition, parallel experiments, in which tissues did not
receive any antagonist, were run in order to check any
variation in sensitivity.
During the conformational search, the geometry of the
reference compound 1a was kept fixed while the internal
geometry of 3a and 4a was randomly perturbed for 5000
Monte-Carlo runs.
The GRID fields41 on the overlay showing the best
superposition according to the QXP scoring function
were in turn calculated, placing hydroxyl oxygen as
˚
atom probe on the nodes of a 86 · 76 · 76 0.2-A-spaced
cubic grid.
Images were rendered with the software PYMOL 0.99
(http://pymol.sourceforge.net).
In all cases, parallel experiments in which tissues re-
ceived only the reference agonist were run in order to
check any variation in sensitivity.
4.3. X-ray crystal structure analysis of 10a picrate salt
Crystals of 10a picrate salt were obtained by slow evapora-
tion from H2O/CH3OH (3:1) solution. The X-ray data col-
lection wasperformedat room temperaturewith anOxford
Diffraction Xcalibur PX Ultra CCD diffractometer,
equipped with Onyx optics, using Cu-Ka radiation
All animal testing was carried out according to the
European Community Council Directive of 24 Novem-
ber 1986 (86/609/EEC).
˚
(k = 1.5418 A). Intensity data were corrected for absorp-
4.4.1. Rabbit stimulated vas deferens. This preparation
was set up according to the method of Eltze.32 Vasa
deferentia were carefully dissected free of surrounding
tissue and were divided into four segments, two prostatic
portions of 1 cm and two epididymal portions of approx-
imately 1.5 cm length. The four segments were mounted
in PSS with the following composition (mM): NaCl
(118.4), KCl (4.7), CaCl2 (2.52), MgCl2 (0.6), KH2PO4
(1.18), NaHCO3 (25.0), and glucose (11.1); 10ꢀ6 M
yohimbine and 10ꢀ8 M tripitramine were included to
block a2-adrenoceptors and M2 muscarinic receptors,
respectively. The solution was maintained at 30 ꢁC, and
tissues were stimulated through platinum electrodes by
square-wave pulses (0.1 ms, 2 Hz, 10–15 V). Contrac-
tions were measured isometrically after tissues had been
equilibrated for 1 h, and then a cumulative dose–re-
sponse curve to p-Cl-McN-A-343 was constructed.
tion by a multi-scan procedure.42 Crystal data: formula
(C8H18NO2)+Æ(C6H2N3O7)ꢀ, Mw = 388.34, Monoclinic,
˚
Space group P21/n; a = 6.9911(6) A, b = 10.4852(7) A,
˚
3
˚
˚
c = 25.039(2) A, b = 90.048(6)ꢁ, V = 1835.4(2) A , Z = 4,
l = 1.024 mmꢀ1. 14,527 reflections measured, 2530 unique
(Rint = 0.0322), 2046 reflections with I > 2rI. R[F2 >
2r(F2)] = 0.0429, wR(F2) = 0.1268, GoF = 1.047, dqmax
=
0.252, dqmin = ꢀ0.136 e Aꢀ3. The structure was solved by
direct methods, with SIR-97,43 and refined by full-matrix
least-squares with SHELXL-97,44 applying a correction
for twinning by merohedry (monoclinic pseudo-ortho-
rhombic lattice; 0.60 major component). In the final
refinement cycles (249 parameters, no restraints) all non-
hydrogen atoms were refined anisotropically and hydro-
gens were in geometrically calculated positions, riding on
the respective carrier atoms, each with an isotropic temper-
ature factor linked to that of its carrier atom. For graphics
ORTEP-3 was employed.45
˚
4.4.2. Guinea-pig ileum. Two-centimeter-long portions
of terminal ileum were taken at about 5 cm from the ile-
um–cecum junction and mounted in PSS, at 37 ꢁC, of
the following composition (mM): NaCl (118.0), NaH-
CO3 (23.8), KCl (4.7), MgSO4–7H2O (1.18), KH2PO4
(1.18), CaCl2 (2.52), and glucose (11.7). Tension changes
CCDC-619120 contains the supplementary crystallo-
graphic data for this paper. These data can be obtained
free of charge at www.ccdc.cam.ac.uk/conts/retriev-
ing.html or from the Cambridge Crystallographic Data

A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
895
were recorded isotonically. Tissues were equilibrated for
30 min, and dose–response curves to arecaidine propar-
gyl ester were obtained at 30-min intervals, the first one
being discarded and the second being taken as control.
4.4.6. Statistical analysis. The results are expressed as
means SEM Student’s t test was used to assess the sta-
tistical significance of the difference between means.
Potency was expressed as ꢀlogED50 SEM derived
from dose–response curves and represents ꢀlog of the
concentration of agonists to produce 50% of the maxi-
mum contraction.
4.4.3. Guinea-pig stimulated left atrium. The heart was
rapidly removed, and the right and left atria were sepa-
rately excised. Left atria were mounted in PSS (the same
as was used for ileum) at 30 ꢁC and stimulated through
platinum electrodes by square-wave pulses (1 ms, 1 Hz,
5–10 V) (Tetra Stimulus, N. Zagnoni). Inotropic activity
was recorded isometrically. Tissues were equilibrated for
2 h and a cumulative dose–response curve to arecaidine
propargyl ester was constructed.
Acknowledgments
The authors thank Dr. Francesca Ghelfi for her techni-
cal support in assays performed with the cytosensor
microphysiometry instrument. This work was supported
by grants from the MIUR (Rome) and the University of
Camerino.
4.4.4. Guinea-pig lung strips. The lungs were rapidly re-
moved and strips of peripheral lung tissue were cut
either from the body of a lower lobe with the longitu-
dinal axis of the strip parallel to the bronchus or from
the peripheral margin of the lobe. The preparations
were mounted, with a preload of 0.3 g, in PSS with
the following composition (mM): NaCl (118.78), KCl
(4.32), CaCl2–2H2O (2.52), MgSO4–7H2O (1.18),
KH2PO4 (1.28), NaHCO3 (25.0), and glucose (5.55).
Contractions were recorded isotonically at 37 ꢁC after
tissues had been equilibrated for 1 h; then two cumu-
lative dose–response curves to arecaidine propargyl es-
ter (0.01, 0.1, 1, 10, and 100 mM) were obtained at
45-min intervals, the first one being discarded and
the second being taken as control.
References and notes
1. Eglen, R. M.; Choppin, A.; Watson, N. Trends Pharmacol.
Sci. 2001, 22, 409.
2. Lefkowitz, R. J. Trends Pharmacol. Sci. 2004, 25, 413.
3. Wess, J.; Duttaroy, A.; Zhang, W.; Gomeza, J.; Cui, Y.;
Miyakawa, T.; Bymaster, F. P.; McKinzie, L.; Felder, C.
C.; Lamping, K. G.; Faraci, F. M.; Deng, C.; Yamada, M.
Receptor Channel 2003, 9, 279.
4. Caulfield, M. P.; Birdsall, N. J. M. Pharmacol. Rev. 1998,
50, 279.
5. Ma, W.; Li, B.-S.; Zhang, L.; Pant, H. C. Drug News
Perspect. 2004, 17, 258.
4.4.5. Measurement of extracellular acidification rates.
CHO cells, stably expressing human mAChRs M1–M5,
were supplied by Professor Tom I. Bonner (National
Institute of Mental Health, Bethesda, MD, USA).
Cells were maintained in a-minimum essential medium
containing 10% fetal bovine serum at 37 ꢁC under 5%
CO2–95% O2; they were grown to confluence and
harvested by scraping in fresh medium. Changes in
extracellular acidification were determined using an
eight-channel cytosensor microphysiometry instrument
(Molecular Devices). Cells were seeded into cytosensor
cell capsules 20 h prior to experiments at a density of
3 · 105 cells per well. The cell capsules, mounted in
sensor chambers, were placed on the instrument, and
the cells were perfused with media via a peristaltic
pump, during which the pH of the microenvironment
surrounding the sensor was kept constant. The remov-
al of acid from the cells by the perfusate was period-
ically halted (pump turned off), allowing a build up of
acid metabolites and, therefore, a change in chamber
pH (acidification rate). This on–off cycle was repeated
throughout the experiment and the effect of com-
pounds was determined by adding the compound to
the chamber. The extracellular acidification rates were
measured in each 120-s pump cycle; flow on at
100 ll minꢀ1 for 90 s with a 20-s exposure to the test
compound, flow off for 30 s. The acidification rate was
determined between 98 and 118 s, using the cytosoft
program (Molecular Devices). Concentration–effect
curves were obtained by exposing the cells sequentially
to increasing concentrations of agonist at intervals of
21 min.
6. Ashkenazi, A.; Ramachandran, J.; Capon, D. J. Nature
1989, 340, 146.
7. Pedretti, A.; Vistoli, G.; Marconi, C.; Testa, B. Chem.
Biodiv. 2006, 3, 481.
8. Shi, L.; Javitch, J. A. Annu. Rev. Pharmacol. Toxicol.
2002, 42, 437.
9. Fisher, A. Expert Opin. Invest. Drugs 1997, 6, 1395.
10. DeLapp, N.; Wu, S.; Belagaje, R.; Johnstone, E.; Little,
S.; Shannon, H.; Bymaster, F.; Calligaro, D.; Mitch, C.;
Whitesitt, C.; Ward, J.; Sheardown, M.; Fink-Jensen, A.;
Jeppesen, L.; Thomsen, C.; Sauerberg, P. Biochem. Bio-
phys. Res. Commun. 1998, 244, 156.
11. Felder, C. C.; Porter, A. C.; Skillman, T. L.; Zhang, L.;
Bymaster, F. P.; Nathanson, N. M.; Hamilton, S. E.;
Gomeza, J.; Wess, J.; McKinzie, D. L. Life Sci. 2001, 68,
2605.
12. Duttaroy, A.; Gomeza, J.; Gan, J.-W.; Siddiqui, N.;
Basile, A. S.; Harman, W. D.; Smith, P. L.; Felder, C.
C.; Levey, A. I.; Wess, J. Mol. Pharmacol. 2002, 62,
1084.
13. Wess, J.; Duttaroy, A.; Gomeza, J.; Zhang, W.; Yamada,
M.; Felder, C. C.; Bernardini, N.; Reeh, P. W. Life Sci.
2003, 72, 2047.
14. De Sarno, P.; Shestopal, S. A.; Zmijewska, A. A.; Jope, R.
S. Brain Res. 2005, 1041, 112.
15. Budd, D. C.; Spragg, E. J.; Ridd, K.; Tobin, A. B.
Biochem. J. 2004, 381, 43.
16. Angeli, P. Il Farmaco 1995, 50, 565.
17. Triggle, D. J.; Belleau, B. Can. J. Chem. 1962, 40,
1201.
18. Pigini, M.; Brasili, L.; Giannella, M.; Gualtieri, F. Eur.
J. Med. Chem. 1981, 16, 415.
19. Grana, E.; Lucchelli, A.; Zonta, F.; Santagostino-Bar-
bone, M. G.; D’Agostino, G. Naunyn Schmiedeberg’s
Arch. Pharmacol. 1986, 332, 213.

896
A. Piergentili et al. / Bioorg. Med. Chem. 15 (2007) 886–896
20. Piergentili, A.; Angeli, P.; Gagliardi, R.; Gentili, F.;
Giannella, M.; Marucci, G.; Pigini, M.; Quaglia, W. Med.
Chem. Res. 2002, 11, 12.
21. Werner, L. H. U.S. Patent 2,903,464, 1959; Chem. Abstr.
1960, 54, 11544.
22. Smirnov, V. V.; Antonova, N. G.; Zotov, S. B.; Kvasnyuk-
Mudryi, F. V.; Sitanova, N. A. Khim. Geterotsikl. Soedin.
1970, 131, 318; . Chem. Abstr. 1970, 73, 66518.
23. Van Cauwenberghe, R.; Anteunis, M.; Valckx, L. Bull.
Soc. Chim. Belg. 1974, 83, 285.
24. Anteunis, M. J. O.; Van Cauwenberghe, R.; Valckx, L.;
Van Bever, W. F. M.; Van Nueten, J. M. Eur. J. Med.
Chem. 1975, 10, 360.
35. Eglen, R. M.; Harris, G. C. Br. J. Pharmacol. 1993, 109,
946.
36. Roffel, A. F.; Elzinga, C. R. S.; Zaagsma, J. Eur.
J. Pharmacol. 1993, 250, 267.
37. Budriesi, R.; Cacciaguerra, S.; Di Toro, R.; Bolognesi, M.
L.; Chiarini, A.; Minarini, A.; Rosini, M.; Spampinato, S.;
Tumiatti, V.; Melchiorre, C. Br. J. Pharmacol. 2001, 132,
1009.
38. Angeli, P.; Brasili, L.; Giannella, M.; Gualtieri, F.;
Picchio, M. T.; Teodori, E. Naunyn Schmiedeberg’s Arch.
Pharmacol. 1988, 337, 241.
39. Wood, M. D.; Murkitt, K. L.; Ho, M.; Watson, J. M.;
Brown, F.; Hunter, A. J.; Middlemiss, D. N. Br. J. Phar-
macol. 1999, 126, 1620.
40. McMartin, C.; Bohacek, R. S. J. Comput. Aided Mol. Des.
1997, 11, 333.
41. Goodford, P. J. J. Med. Chem. 1985, 28, 849.
42. Oxford Diffraction, 2006, CrysAlisPro (Version
1.171.29.2). Oxford Diffraction Ltd, Abingdon, Oxfordd-
hire, England.
43. Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G.
L.; Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.;
Polidori, G.; Spagna, R. J. Appl. Crystallogr. 1999, 32,
115.
25. Inch, T. D.; Williams, N. J. Chem. Soc. (C) 1970, 2, 263.
26. Ing, H. R. Science 1949, 109, 264.
27. Barlow, R. B. In Introduction to Chemical Pharmacology;
Barlow, R. B., Ed.; Methuen, Wiley: London, New York,
1964; p 203.
`
28. Gelas, J.; Veyssieres-Rambaud, S. Carbohydr. Res. 1974,
37, 293.
29. Gelas, J.; Veyssieres-Rambaud, S. Carbohydr. Res. 1974,
`
37, 303.
30. Duclos, R. I., Jr.; Makriyannis, A. J. Org. Chem. 1992, 57,
6156.
31. Villani, F. J., Jr.; Costanzo, M. J.; Inners, R. R.; Mutter,
M. S.; McClure, D. E. J. Org. Chem. 1986, 51, 3715.
32. Eltze, M. Eur. J. Pharmacol. 1988, 151, 205.
33. Kenakin, T. P.; Boselli, C. Naunyn Schmiedeberg’s Arch.
Pharmacol. 1991, 344, 201.
44. Sheldrick, G. M. SHELXL97, Program for Crystal Struc-
ture Refinement; University of Go¨ttingen, Go¨ttingen:
Germany, 1986.
45. Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837.
46. van Rossum, J. M. Arch. Int. Pharmacodyn. Ther. 1963,
143, 299.
34. Ringdahl, B. Mol. Pharmacol. 1987, 31, 351.