Bis(ammonio)alkane-type agonists of muscarinic acetylcholine receptors: Synthesis, in vitro functional characterization, and in vivo evaluation of their analgesic activity

lable at ScienceDirect

European Journal of Medicinal Chemistry 75 (2014) 222e232

Contents lists avai

European Journal of Medicinal Chemistry

journal homepage: http: / /www.elsevier .com/locate/ejmech

Original article

Bis(ammonio)alkane-type agonists of muscarinic acetylcholinereceptors: Synthesis, in vitro functional characterization, and in vivoevaluation of their analgesic activity

Carlo Matera a, Lisa Flammini b, Marta Quadri a, Valentina Vivo b, Vigilio Ballabeni b,Ulrike Holzgrabe c, Klaus Mohr d, Marco De Amici a, Elisabetta Barocelli b, Simona Bertoni b,Clelia Dallanoce a,*

aDipartimento di Scienze Farmaceutiche, Sezione di Chimica Farmaceutica “Pietro Pratesi”, Università degli Studi di Milano, Via L. Mangiagalli 25, 20133Milano, ItalybDipartimento di Farmacia, Università degli Studi di Parma, Parco Area delle Scienze 27/A, 43124 Parma, ItalycDepartment of Pharmaceutical Chemistry, Institute of Pharmacy, University of Würzburg, Am Hubland, 97074 Würzburg, Germanyd Pharmacology & Toxicology Section, Institute of Pharmacy, University of Bonn, Gerhard-Domagk-Straße 3, 53121 Bonn, Germany

a r t i c l e i n f o

Article history:Received 12 August 2013Received in revised form15 January 2014Accepted 20 January 2014Available online 25 January 2014

Keywords:Alkylbisammonio quaternary saltsMuscarinic receptor subtypesMuscarinic agonistsIn vitro pharmacologyIn vivo pharmacologyAnalgesic activity

* Corresponding author.E-mail address: [email protected] (C. Dalla

0223-5234/$ e see front matter � 2014 Elsevier Mashttp://dx.doi.org/10.1016/j.ejmech.2014.01.032

a b s t r a c t

In this study, we synthesized and tested in vitro and in vivo two groups of bis(ammonio)alkane-typecompounds, 6ae9a and 6be9b, which incorporate the orthosteric muscarinic agonist iperoxo into amolecular fragment of the M2-selective allosteric modulators W84 and naphmethonium. The agonistpotency and efficacy of these hybrid derivatives at M1, M2 and M3 muscarinic receptor subtypes and theiranticholinesterase activity were evaluated on isolated tissue preparations. Their analgesic action wasthen assayed in vivo in the acetic acid writhing test and the occurrence of peripheral and centralcholinergic side effects was also determined. The investigated hybrids behaved as potent muscarinicagonists and weak cholinesterase inhibitors. These effects were more pronounced for bisquaternary saltsbearing the naphmethonium moiety than for the W84-containing analogs, and resulted in a significantanalgesic activity in vivo. A promising profile was displayed by the naphmethonium-related compound8b, which combined the most potent antinociception among the test compounds with the absence ofrelevant cholinergic side effects.

� 2014 Elsevier Masson SAS. All rights reserved.

1. Introduction

Within the rather complex network of pathways thatcontribute to the modulation of pain transmission, heterotrimericG-protein-coupled receptors (GPCRs) exert a relevant role. Indeed,activation of different GPCRs, including opioid, cannabinoid, a2-adrenergic, muscarinic, g-aminobutyric acid (GABAB), groups IIand III metabotropic glutamate, and somatostatin receptor fam-ilies, may produce analgesic effects [1]. As far as the muscarinic-mediated antinociception is taken into account, two complemen-tary approaches have been applied, which are based on genetic orpharmacological evidence, in the attempt to better define thespecific role played by each muscarinic acetylcholine receptor(mAChR) subtype. The first strategy makes use of mutant mouse

noce).

son SAS. All rights reserved.

strains lacking specific mAChR populations, the second one utilizesmuscarinic ligands, in particular agonists, which however display apoor degree of subtype selectivity [2]. As a matter of fact, almost allthe different mAChR subtypes, M1-M5, have been found to beinvolved in the regulation of pain either in the spinal cord, with amajor contribution, and/or in other central and peripheral dis-tricts. The supraspinal mechanism of muscarinic analgesia in-volves the M1 subtype as observed after intracerebroventricularinjection of antisense oligonucleotides against these receptors [3].Studies with M2 and M4 knock-out mice revealed that spinalanalgesia is mainly mediated by the M2 subtype, but even the M4receptor makes a contribution to the analgesic activity [4,5]. M3receptors, present in the spinal cord, potentiate the effects ofmuscarinic agonists on GABA and glycine release in dorsal hornneurons in mice [6]. Little is known on the role of M5 mAChRs innociception. Systemic, intrathecal or intracerebroventricularadministration of cholinomimetics produces analgesia in severalacute nociceptive tests such as hot plate and tail flick tests

Fig. 2. Structures of the two sets of hybrid mAChR ligands 6ae9a and 6be9b inves-tigated in this study.

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232 223

(thermal algesia), paw pressure test (mechanical algesia) andwrithing test (chemical algesia), and in more complex systems,such as inflammatory and neuropathic pain models. VariousmAChR antagonists were found to block or counteract the abovecited analgesic effects [7e9].

Some years ago, we designed, prepared and tested a series ofderivatives related to the muscarinic agonists oxotremorine 1 andoxotremorine-M 2 (Fig. 1). On the whole, the novel compoundswere characterized as potent, nonselective muscarinic receptoragonists [10]. The quaternary ammonium salt iperoxo 3 (Fig. 1), inwhich both the nature of the heterocyclic ring and the attachmentpoint of the acetylenic side chain were changed with respect tomodel derivatives 1 and 2, behaved as an exceptionally potent,albeit unselective, muscarinic agonist [10]. Quite recently, thesupraphysiological agonist efficacy of 3 and analogs has beenstudied at label-free M2 mAChRs by means of whole cell dynamicredistribution, estimating G protein-activation and cell surfaceagonist binding, and performing computation of operational effi-cacies [11]. Furthermore, iperoxo and related analogs were assayedin vivo for their antinociceptive activity (formalin licking and aceticacid writhing tests) and their ability to induce tremor in mice [12].Peripheral cholinergic effects such as salivation, bradycardia, hy-potension and intestinal hypermotility were also evaluated inanaesthetized rats [12].

In the framework of a research project focused on the design,synthesis and pharmacological investigation of unprecedentedmuscarinic ligands [13,14], we made use of the orthostericmuscarinic agonist iperoxo as a building block to design novelhybrid derivatives incorporating the molecular fragments of twoarchetypal allosteric modulators of mAChRs, i.e., 4 (W84) [15,16]and 5 (naphmethonium) [17,18], whose structure is depicted inFig. 1. We demonstrated that the new molecular probes, i.e.,bis(ammonio)alkane-type derivatives such as 6a, 8a (W84-related)and 6b, 8b (naphmethonium-related) (Fig. 2), simultaneously bindthe orthosteric as well as the allosteric recognition site of the M2mAChR subtype [14,19,20]. As a result of this bitopic (“dualsteric”)interaction, these M2-preferring muscarinic agonists cause func-tional responses which are affected by both, the length of thepolymethylene spacer chain and the nature of the allosteric mo-lecular component part [20].

Fig. 1. Structures of model orthosteric (1, 2, and

In this paper, we synthesized the new derivatives 7a and 7b andthe partially rigidified analogs 9a and 9b, then evaluated thepharmacological profile of the whole set of compounds 6ae9a and6be9b at the M1, M2, and M3 mAChR subtypes by means of func-tional assays on isolated tissue preparations, and tested theirin vitro anticholinesterase activity. Next, similarly to our previousinvestigation on iperoxo and related analogs, we studied the in vivo

3) and allosteric (4 and 5) mAChR ligands.

Scheme 1. a): CH3CN, reflux.

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232224

analgesic properties of hybrid derivatives and the appearance ofmuscarinic side effects. In addition, we estimated the in vivo inci-dence of cardiovascular side effects caused by the muscarinicagonist 8b, that showed the most relevant analgesic activity.

2. Results and discussion

2.1. Chemistry

The synthesis of target compounds was accomplished followingthe reaction sequences illustrated in Schemes 1 and 2. Knownphthalimidopropylamine 10a [13] and 1,8-naphthalimidopropilamine 10b [13], containing the W84- and naphmethonium-related pharmacophoric fragments respectively, were treated

Scheme 2. a) DIAD, (C6H5)3P, THF, 0 �C to rt; b): 1,6-d

with a large excess of the appropriate commercially available alkyldibromide (1,6-dibromohexane 11, 1,6-dibromoheptane 12, and1,6-dibromoctane 13) in refluxing acetonitrile (Scheme 1). Theresulting monoquaternary intermediates 14ae16a and 14be16b,obtained in 65e90% yield, were refluxed in acetonitrilewith a slightexcess of the tertiary dimethylamine 17 [10,21], thereby affordingthe desired final bisquaternary ammonium salts 6ae8a and 6be8bin 55e92% yield (Scheme 1).

The preparation of 9a and 9b, the two spatially demandinganalogs of 6a and 6b, took advantage of the Mitsunobu reactionprotocol, whichwas applied on commercial 1-methylpiperidin-4-ol19 in the presence of phthalimide 18a (or 1,8-naphthalimide 18b),triphenylphosphine and diisopropyl azodicarboxylate (DIAD). Theuse of DIAD [22] improved the reaction yields of N-(1-

ibromohexane (large excess), rt; c) CH3CN, reflux.

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232 225

methylpiperidin-4-yl)phthalimide 20a and N-(1-methylpiperidin-4-yl)1,8-naphthalimide 20b (Scheme 2) [23], when compared withdiethyl azodicarboxylate (DEAD), which had been previouslyemployed in the preparation of 20a [24]. The desired mono-quaternary bromides 21a and 21bwere then obtained by treatmentof 20a and 20b with excess 1,6-dibromohexane 11, and the finalderivatives 9a and 9b by reaction with tertiary amine 17, similarlyto their analogs of Scheme 1.

2.2. Pharmacology

We evaluated the functional characteristics of the hybrid com-pounds at three different mAChR subtypes in isolated organ prep-arations. The reference allosteric derivatives 4 and 5, which werereported to behave as antagonists at mAChRs [13,25], were alsoincluded in this analysis. The rabbit vas deferens was taken as anM1model [26,27], although there is still some controversy in how farthe M4 subtype is involved [28,29]. The guinea pig left atrium waschosen to measure M2 receptor mediated actions, and guinea pigileum served as M3 model. Both groups of compounds 6ae9a and6be9bwere characterized by an overall agonist profile, being fromtwo to three orders of magnitude less potent than the referenceagonist 3, and showed a negligible to modest discrimination amongthe studied receptor subtypes with the exception of compounds 6aand 6b, which were endowed with an apparent M2 preferring ac-tivity (Table 1).

The super-potent muscarinic agonist iperoxo 3 directs the pro-file of the hybrid derivatives towards receptor activation, at vari-ance with the effect exerted by less potent orthosteric muscarinicbuilding blocks, which were found to be unable to counteract theantagonistic behavior of the allosteric fragments of 4 and 5 [13]. Asa general trend, the W84-rela monium salts (a group), showedlower pEC50 values than their naphmethonium-related counter-parts (b group) [i.e., pEC50 ¼ 7.79 (M1), 7.85 (M2), 7.16 (M3) for 8a,and pEC50¼ 8.19 (M1), 8.15 (M2), 7.94 (M3) for 8b as in Table 1], thusconfirming previous results obtained on some of these and relatedanalogs [13,14,19]. Of note, both potency and efficacy values of all

Table 1In vitro functional activity of compounds 6ae9a and 6be9b at muscarinic receptor subtyprat brain cholinesterase inhibitory potency.

Compound M1:a rabbit vas deferens M2: Guinea pig left atrium

pEC50b i.a.c pKB

d pEC50b i.ac

McN-A-343 6.43 (0.07) 1.00Bethanechol 5.85 (0.08) 1.003 9.87 (0.07)g 1.00g 10.10 (0.13)g 1.00g

4 0 5.45 (0.16) 06a 6.04 (0.26) 0.96 (0.02) 7.18 (0.08) 0.93 (0.03)7a 7.98 (0.10) 1.00 (0) 7.91 (0.13) 1.00 (0.02)8a 7.79 (0.08) 1.00 (0) 7.85 (0.07) 1.05 (0.05)9a 7.41 (0.03) 1.03 (0.01) 7.42 (0.16) 0.95 (0.05)5 0 6.24 (0.19) 06b 7.66 (0.20) 0.90 (0.04) 8.11 (0.13) 0.93 (0.01)7b 8.06 (0.05) 0.97 (0.04) 7.73 (0.07) 0.97 (0.04)8b 8.19 (0.02) 1.00 (0) 8.15 (0.09) 1.01 (0.03)9b 8.00 (0.11) 1.01 (0.03) 7.67 (0.06) 0.98 (0.04)

ND: not determined.a See comment [29].b pEC50 (SE) values are the negative logarithm of the agonist concentration that causec i.a. ¼ intrinsic activity (a) measured by the ratio between the maximum response od Apparent pKB values � SE were calculated according to Furchgott and Bursztyn [38]e pIC50 (SE) values are the negative logarithm of the concentration causing half-maximf Maximum percent inhibition of rat brain cholinesterase.g Data taken from Dallanoce et al. [10].h At a concentration of 100 mM.i At a concentration of 10 mM.

derivatives were found to be slightly lower at the M3 subtype whencompared with those measured at M1 and M2 subtypes. This ismore pronounced for the intrinsic activity of the octamethylene-spaced derivative 8b showing the lowest intrinsic activity at M3receptors (a ¼ 0.75) within the studied compounds.

Additionally, in the Ellman assay, group b derivatives exhibited ahigher, althoughmodest, anticholinesterase activity with respect togroup a analogs. Thus, hybrids 6be8b increased, from 4.38 to 5.10,the inhibitory potency (pIC50 value) of the allosteric parent com-pound 5, with derivative 8b exhibiting the most efficacious inhi-bition (Table 1). On the other hand, starting from the inactiveparent compound 4, only derivatives 7a and 8a, characterized bythe longest polymethylene middle-chain spacer, produced a weakcholinesterase inhibition (Table 1). In both series, the rigid analogs9a and 9b did not significantly enhance the ability either of theallosteric prototypical compounds or of the orthosteric ligandiperoxo, inactive up to 100 mM, to block the enzyme (Table 1).Derivatives 6a and 6b were also tested for their inhibitory potencyagainst purified acetylcholinesterase (AChE) [30], showing IC50values of 17.5 mM and 3.5 mM, respectively, which are in accordancewith those found in the Ellmann assay.

The compounds under study were then assayed in vivo for theirantinociceptive effect by means of the writhing test, a conventionalchemical method used to induce pain of peripheral origin byintraperitoneal injection of 0.6% acetic acid solution [31]. Part of thetarget ligands showed a dose-dependent analgesic activity withlower potency than the reference agonist 3, whose antinociceptivepotency had been previously estimated (ID50 ¼ 1 mg/kg) [12]. Theircapacity to prevent nociception was affected by the length of thepolymethylene chain. In fact, both 8a and 8b, i.e., the analogscharacterized by the longest chain (C-8), were the compounds withthe highest antinociceptive activity in each series [ID50 ¼ 339.6 mg/kg (8a), 23.8 mg/kg (8b)]. In addition, the nature of the allostericmolecular portion modulates the extent of the analgesic effect,since 6b (ID50 ¼ 248.3 mg/kg) and 8b, containing thenaphmethonium-like fragment, were more potent (about 30 and15 times, respectively) than their corresponding W84-like

es in rabbit vas deferens (M1), guinea pig left atrium (M2), guinea pig ileum (M3), and

M3: Guinea pig ileum AntiChE activity

pKBd pEC50

b i.a.c pKBd pIC50e Effmax (%)f

ND ND6.00 (0.06) 1.00 ND ND9.78 (0.10)g 1.00g Inactiveh

Inactivei 0 5.35 (0.17) ND 48.5 (6.5)6.06 (0.10) 0.94 (0.04) ND 44.0 (9.0)6.98 (0.28) 0.92 (0.05) 4.50 (0.05) 73.5 (1.5)7.16 (0.08) 0.90 (0.06) 4.71 (0.04) 83.5 (0.5)7.16 (0.04) 0.96 (0.06) ND 58.5 (4.5)

6.57(0.18) 0 5.49 (0.10) 4.38 (0.02) 69.5 (0.5)7.68 (0.11) 0.81 (0.07) 4.99 (0.04) 80.0 (2.0)7.30 (0.08) 0.87 (0.09) 5.10 (0.06) 87.3 (2.2)7.94 (0.09) 0.75 (0.12) 5.00 (0.03) 90.0 (1.0)7.33 (0.08) 1.05 (0.10) 4.64 (0.03) 81.0 (1.0)

d 50% of the maximum response.f the compound and the maximum response of the reference agonist..al inhibition of cholinesterase activity.

CTR 1 5 10 1 2.5 5 0.1 0.5 1 5 0.1 0.5 10

5

10

15

20

25

CTR 0.1 0.25 0.5 0.005 0.01 0.05 0.1 0.50

5

10

15

20

25

a

Num

ber o

f writ

hes

(5th-1

0thm

in)

Num

ber o

f writ

hes

(5th-1

0thm

in)

b

*

*

******

*

****

*

**

**

**

**

***

6a 7a 8a 9a

6b 8b

Fig. 3. Number of acetic acid-induced writhes in mice s.c. administered with vehicle (control group e CTR), W84-related hybrids 6ae9a (mg/kg) (a) and naphmethonium-relatedhybrids 6b, 8b (mg/kg) (b). Data are expressed as mean � SEM of 6e8 observations. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control group, one-way analysis of variance test followedby Dunnett’s post-test.

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232226

congeners 6a (ID50 ¼ 7.73 mg/kg) and 8a (Fig. 3). Interestingly, asconcerns the two partially rigidified ligands, while compound 9adisplayed the most potent analgesic activity among the W84-related derivatives (ID50 ¼ 249.0 mg/kg), its naphmethonium-likeanalog 9b was devoid of antinociceptive effects up to 1 mg/kgand caused lethal effects at higher doses, seemingly due tocholinergic hyperstimulation. Similarly, hybrid 7b was devoid ofanalgesic activity up to 1 mg/kg. The phenomenon will beaddressed in future research.

Since potential sedative effects could interfere with the datacollected in the writhing test, iperoxo and derivatives 6ae9a and6be9b were evaluated in mice in the open-field test [32], which isconventionally used to assess locomotor, exploratory and anxiety-like behaviors. We estimated the overall immobility time of micefor 90min after subcutaneous treatment with different doses of the

compounds under study (data not shown). None of the derivativesactive in the writhing test produced sedative effects when tested atantinociceptive doses; only iperoxo 3 and hybrid 8a exhibited astatistically significant prolongation of the immobility time(P < 0.01 vs. control group) at high doses, 0.01 mg/kg and 5 mg/kg,respectively. Also hybrid 7b (1 mg/kg), devoid of analgesic activityup to 1 mg/kg, showed significant sedative effects (P < 0.01 vs.control group).

The most active analgesic agent in the series of hybrid com-pounds, i.e., the naphmethonium-related agonist 8b, was assayedin the writhing test at 0.5 mg/kg s.c. (the dose giving 80% analgesiceffect) also in the presence of atropine, methylatropine, mecamyl-amine and naloxone. The antinociceptive effect of 8b is muscarinicin nature, since it was considerably prevented by pretreatmentwith the two non-selective muscarinic antagonists but was

0123456789

101112131415161718

Num

ber o

f writ

hes

(5th- 1

0thm

in)

**

** **

CTR S A MA Me N

8b

Fig. 4. Number of acetic acid-induced writhes in mice s.c. administered with vehicle(control group e CTR) or hybrid 8b (0.5 mg/kg s.c.) following i.p. pre-treatment withvehicle (saline e S), atropine (5 mg/kg e A), methylatropine (5 mg/kg e MA), meca-mylamine (1 mg/kg e Me) or naloxone (1 mg/kg e N). Data are expressed asmean � SEM of 6e8 observations. **P < 0.01 vs. control group, one-way analysis ofvariance test followed by Dunnett’s post-test.

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232 227

unaffected by both mecamylamine and naloxone administration(Fig. 4). Since 8b is a permanently charged compound, its analgesicactivity, which is comparably neutralized by both atropine andmethylatropine, appears to be mediated by activation of peripheralmAChR populations. Indeed, there is evidence for the expression ofmAChRs, predominantly of the M2 subtype, in the peripheral sen-sory nerve fibers [33], where their stimulation reduces the heat-induced release of Calcitonin gene-related peptide [2,34].

In a parallel investigation, to ascertain the putative therapeuticratio between the desirable analgesic activity and the unwantedmuscarinic side effects, we estimated the incidence of majorcholinergic side effects such as diarrhea, salivation, lachrymation,tremor and hypothermia following the subcutaneous administra-tion of derivatives at their antinociceptive doses in mice. As re-ported in Table 2, the tested compounds displayed a generally goodtolerability except for 7a, which produced all of the inspected sideeffects. On the contrary, the most potent compound 8b wascompletely devoid of the unwanted muscarinic effects at a dose(0.01 mg/kg) evoking a significant analgesic action. Hypothermia,diarrhea and lachrymation appeared only at 0.5 mg/kg, a doseinhibiting 80% acetic acid-induced writhing response, and were

Table 2Unwanted muscarinic side effects produced by s.c. administration of compounds 3,6ae9a, 6b and 8b at doses that inhibit significantly acetic acid-induced writhingresponses compared to the control group.

Compd Diarrhea Salivation Lachrymation Hypothermia % Analgesia

3 (0.001 mg/kg) 0 þa 0 þþb 50d

6a (10 mg/kg) þa 0 0 þþþc 807a (1 mg/kg) þþb þa þþb þþb 428a (1 mg/kg) þþb 0 0 0 609a (0.5 mg/kg) 0 0 0 0 676b (0.25 mg/kg) 0 0 0 0 458b (0.01 mg/kg) 0 0 0 0 47

a 25%.b 50%.c 75%: percent appearance of the side effect.d Data taken from Barocelli et al. [12].

completely prevented by themuscarinic antagonist methylatropine(data not shown). It is worth noting that none of the derivativesunder study evoked tremors of the treated animals: the absence oftremorigenic effects confirms the peripheral mechanism of actionof these bisquaternary ammonium salts, reinforcing the data ob-tained on 8b after pretreatment with the muscarinic antagonistmethylatropine. The hypothermic effect, detected for some of thecompounds under study, is congruous with the involvement ofboth central and peripheral muscarinic pathways in the regulationof body temperature [35].

Finally, considering the analgesic properties shown by 8b andthe absence of muscarinic side effects, we extended its investiga-tion by measuring blood pressure and heart rate changes inducedby 8b (0.01e0.5 mg/kg) on anaesthetized rats. No statistically sig-nificant alteration of the basal value of the two measured param-eters was evidenced at the effective antinociceptive dose of0.01 mg/kg 8b, and the cardiovascular unwanted effects, i.e., hy-potension and bradycardia, appeared at higher doses only (Fig. 5).

Fig. 5. Doseeresponse relationship for mean arterial blood pressure (a) and heart rate(b) changes induced in anaesthetized rats by i.p. administration of vehicle (saline 1 ml/kg) or of increasing doses of hybrid 8b (0.01e0.5 mg/kg). Data are expressed asmean � SEM of 4e6 observations. *P < 0.05; **P < 0.01 vs. vehicle, repeated-measuresone-way analysis of variance test followed by Bonferroni’s post-test.

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232228

3. Conclusion

In summary, the two sets of bis(ammonio)alkane-type musca-rinic ligands, 6ae9a and 6be9b, incorporating a common orthos-teric super-potent agonist moiety and different allostericfragments, were designed, prepared and tested in functional assaysfor their activity at M1, M2, and M3 mAChR subtypes as well as theiranticholinesterase potency. All of these derivatives behaved asmuscarinic agonists less potent than the orthosteric building block3 and, especially compound 8b, were endowed with a reducedintrinsic activity at the M3 subtype. At variance with the orthostericligand 3, they showed a weak cholinesterase inhibition. Muscarinicpotency as well as anticholinesterase activity appeared more pro-nounced for naphmethonium-related hybrids (b series) than forW84-analogs (a series). The compounds were then assayed fortheir antinociceptive activity in the writhing test and the resultscollected indicated that both, the length of the polymethylenemiddle-chain and the nature of the allosteric moiety, act as opti-mizing features for the analgesic properties. In fact, the combina-tion of a C-8 polymethylene chain with a naphmethonium-containing allosteric fragment engenders the most active in vivoanalgesic agent 8b that, in addition, does not evoke any remarkablemuscarinic side effect. By taking into account the structure ofpermanently charged compounds for the studied derivatives andthe properties shown by the most interesting analog 8b, themeasured analgesic activity seems to be accounted for by activationof peripherally mediated muscarinic responses. The low incidenceof peripheral undesirable M3-dependent cholinergic effects,exhibited by compound 8b, could ensue from its functional selec-tivity towards M2 receptors.

Our data do not allow a specific correlation between theobserved analgesic effects of the studied compounds and theiragonist profile at mAChR subtypes. Nonetheless, the reduced po-tency compared to the reference agonist iperoxo and the acquisi-tion of a lower efficacy at the M3 subtype appear to act ascooperative factors in bringing about for 8b a promising anti-nociceptive activity, which is dissociated from muscarinic side ef-fects. Thus, suitable structural modifications of a rather potentmuscarinic agonist have led to a favorable modulation of its anal-gesic properties together with a substantial gain in the safetyprofile, a result which could pave the way to a therapeutic exploi-tation of antinociception mediated by muscarinic acetylcholinereceptors.

4. Experimental section

4.1. Chemistry

The D2-isoxazolinyl tertiary base 17 [10,21], W84 [16], andnaphmethonium [17], and the two couples of bisquarternary li-gands 6a/6b [14] 8a/8b [20] were prepared according to previouslyreported procedures. Melting points were determined on a modelB540 Büchi melting point apparatus and are uncorrected. 1H NMRand 13C NMR spectrawere recordedwith a VarianMercury 300 (1H,300.063 MHz; 13C, 75.451 MHz) spectrometer at 20 �C. Chemicalshifts (d) are expressed in ppm and coupling constants (J) in Hz. TLCanalyses were performed on commercial silica gel 60 F254aluminum sheets; spots were further evidenced by spraying with adilute alkaline potassium permanganate solution and, for tertiaryamines, with the Dragendorff reagent. ESI mass spectra of the noveltested salts were obtained on a Varian 320 LC-MS/MS instrument.Data are reported as mass-to-charge ratio (m/z) of the corre-sponding positively chargedmolecular ions. Microanalyses (C, H, N)agreed with the theoretical value within �0.4%.

4.1.1. 7-Bromo-N-[3-(1,3-dioxoisoindolin-2-yl)propyl]-N,N-dimethylheptan-1-aminium bromide 15a

A stirred solution of 10a [13] (1 g, 4.30 mmol) and 3 mL 1,7-dibromoheptane 12 in 80 mL acetonitrile was refluxed for 3 days.After the reaction was completed (TLC monitoring, eluent a:CH2Cl2/CH3OH 9:1; eluent b: 0.2 M aqueous KNO3/CH3OH 2:3),about one-half of the solvent was evaporated, then the resultingsolid was collected by filtration, repeatedly washed with n-hexaneand dried in vacuo to give the desired monoquaternary bromide15a (1.84 g, 87% yield).

Compound 15a: Colorless solid (from abs. ethanol); mp 152e155 �C; Rf ¼ 0.70 (0.2 M aqueous KNO3/CH3OH 2:3); 1H NMR(CDCl3): 1.40 (m, 6H, þNeCH2eCH2eCH2eCH2), 1.72 (m, 2H, CH2e

CH2eCH2eBr), 1.84 (m, 2H, CH2eCH2eBr), 2.22 (m, 2H, Npht-CH2-CH2), 3.38 (t, 2H, CH2eBr, J ¼ 6.9), 3.43 (s, 6H, þN(CH3)2), 3.63 (m,4H, CH2e

þNeCH2), 3.84 (t, 2H, Npht-CH2, J ¼ 6.3), 7.73 (m, 2H,arom.), 7.85 (m, 2H, arom.). 13C NMR (CDCl3): 22.72, 22.85, 26.19,27.92, 28.37, 32.62, 34.11, 35.19, 51.60, 61.61, 64.44, 123.74, 132.00,134.56, 168.41. MS (ESI) m/z [M]þ Calcd for C20H30BrN2O2

þ: 411.15.Found: 411.1.

4.1.2. 2-{3-[1-(7-{1,1-Dimethyl-1-[4-(4,5-dihydroisoxazol-3-yloxy)but-2-ynyl]ammonium} heptyl)-1,1-dimethylammonio]propyl}isoindoline-1,3-dione dibromide 7a

A solution of intermediate bromide 15a (470 mg, 0.96 mmol)and tertiary amine 17 (220 mg, 1.21 mmol) in acetonitrile (15 mL)was refluxed for 2 days under stirring. After the reaction wascompleted (TLC monitoring, eluent: CH3OH/0.2 M aqueous KNO33:2), the solvent was evaporated and the residual solid was filteredand crystallized, providing the target compound 7a (380 mg, 59%yield).

Compound 7a: Yellowish, hygroscopic solid (from abs. ethanol/2-propanol); mp 177e178 �C; Rf ¼ 0.58 (0.2 M aqueous KNO3/CH3OH 2:3); 1H NMR (D2O): 1.37 (m, 6H, þNeCH2eCH2eCH2eCH2e

CH2), 1.74 (m, 4H, þNeCH2eCH2eCH2eCH2eCH2eCH2), 2.20 (m,2H, Npht-CH2-CH2), 3.08 (s, 6H, CH2e

þN(CH3)2eCH2), 3.11 (t, 2H, H-42-isox, J ¼ 9.4), 3.16 (s, 6H, þN(CH3)2eCH2eC^), 3.30 (m, 2H, CH2e

Nþ), 3.40 (m, 4H, CH2eNþ), 3.81 (t, 2H, Npht-CH2, J ¼ 6.3), 4.31 (s,2H, þNeCH2eC^), 4.45 (t, 2H, H-52-isox, J ¼ 9.4), 4.91 (s, 2H, ^CeCH2e), 7.9 (m, 4H, arom.). 13C NMR (D2O): 21.82, 21.96, 22.09, 23.92,25.32, 25.39, 27.85, 32.87, 34.88, 50.76, 51.00, 54.24, 57.85, 61.25,64.33, 70.49, 75.80, 85.77, 123.73, 131.44, 135.12, 169.03, 170.63. MS(ESI)m/z [M]þ Calcd for C29H44BrN4O4

þ: 592.26. Found: 592.2. Anal.Calcd for C29H44Br2N4O4 (672.49): C, 51.79; H, 6.59; N, 8.33. Found:C, 52.07; H, 6.41; N, 8.12.

4.1.3. 7-Bromo-N-[3-(1,3-dioxo-1H-benzo[de]isoquinolin-2(3H)-yl)-2,2-dimethylpropyl]-N,N-dimethylheptan-1-aminium bromide15b

A stirred solution of 10b [13] (500 mg, 1.61 mmol) and 1.2 mL1,7-dibromoheptane 12 in 30 mL acetonitrile was refluxed for 4days. The desired monoquaternary bromide 15b (650 mg, 71%yield) was isolated following the protocol adopted for 15a.

Compound 15b: Colorless solid (from 2-propanol); mp 146e148 �C; Rf ¼ 0.64 (CH2Cl2/CH3OH 9:1); 1H NMR (CDCl3): 1.30 (s, 6H,C(CH3)2), 1.37 (m, 6H, þNeCH2eCH2eCH2eCH2eCH2), 1.80 (m, 4H,þNeCH2eCH2eCH2eCH2eCH2eCH2), 3.35 (t, 2H, CH2eBr, J ¼ 6.2),3.56 (s, 6H, þN(CH3)2), 3.65 (s, 2H, CH2e

þN(CH3)2), 3.79 (t, 2H, þNeCH2, J ¼ 7.3), 4.30 (s, 2H, Nnapht-CH2), 7.72 (t, 2H, arom., J ¼ 7.6),8.20 (d, 2H, arom., J ¼ 8.2), 8.50 (d, 2H, arom., J ¼ 7.3). 13C NMR(CDCl3): 23.24, 26.20, 26.77, 28.43, 32.66, 34.19, 39.45, 48.38,53.22, 58.41, 68.33, 72.20, 122.18, 127.36, 128.20, 131.74, 132.03,134.82, 165.38. MS (ESI) m/z [M]þ Calcd for C26H36BrN2O2

þ: 487.20.Found: 487.2.

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232 229

4.1.4. 2-{3-[1-(7-{1,1-Dimethyl-1-[4-(4,5-dihydroisoxazol-3-yloxy)but-2-ynyl]ammonium} heptyl)-1,1-dimethylammonio]2,2-dimethylpropyl}benzo[de]isoquinoline-1,3-dione dibromide 7b

A solution of intermediate bromide 15b (350 mg, 0.62 mmol)and tertiary amine 17 (142 mg, 0.78 mmol) in acetonitrile (10 mL)was refluxed for 2 days under stirring. After the reaction wascompleted (TLC monitoring, eluent: 0.2 M aqueous KNO3/CH3OH2:3), the solvent was evaporated and the residual solid was filteredand crystallized, providing the target compound 7b (407 mg, 88%yield).

Compound 7b: Yellowish, hygroscopic solid (from 2-propanol);mp 150e152 �C; Rf ¼ 0.46 (0.2 M aqueous KNO3/CH3OH 2:3); 1HNMR (CDCl3): 1.32 (s, 6H, C(CH3)2), 1.52 (m, 6H, þNeCH2eCH2e

CH2eCH2eCH2), 1.90 (m, 4H, þNeCH2eCH2eCH2eCH2eCH2eCH2),2.98 (t, 2H, H-42-isox, J¼ 9.6), 3.43 (s, 6H, þN-(CH3)2), 3.56 (s, 6H, þN-(CH3)2), 3.65 (s, 2H, þNeCH2eC^), 3.91 (m, 4H, þNeCH2eCH2e

CH2eCH2eCH2eCH2eCH2), 4.33 (s, 2H, C(CH3)2eCH2eNþ), 4.39 (t,2H, H-52-isox, J ¼ 9.6), 4.81 (s, 2H, ^CeCH2eO), 4.85 (s, 2H, Nnapht-CH2), 7.76 (t, 2H, arom., J ¼ 7.4), 8.23 (d, 2H, arom., J ¼ 8.3), 8.54(d, 2H, arom., J ¼ 7.2). 13C NMR (CDCl3): 22.29, 22.56, 25.29, 25.59,26.78, 27.50, 33.18, 39.50, 48.34, 50.67, 53.09, 57.68, 64.22, 68.71,70.27, 72.39, 76.37, 86.39, 100.24, 122.21, 127.40, 128.16, 131.70,132.00, 134.76, 165.35, 167.00. MS (ESI) m/z [M]þ Calcdfor C35H50BrN4O4

þ: 670.30. Found: 670.0. Anal. Calcd forC35H50Br2N4O4 (750.60): C, 56.00; H, 6.71; N, 7.46. Found: C, 55.88;H, 6.95; N, 7.30.

4.1.5. 2-(1-Methylpiperidin-4-yl)isoindoline-1,3-dione 20aIn an atmosphere of dry argon, DIAD (1.18 mL, 6.0 mmol) was

added dropwise at 0 �C to a solution of 19 (705 mL, 6.0 mmol),phthalimide 18a (883 mg, 6.0 mmol) and triphenylphosphine(1.57 g, 6.0 mmol) in dry THF (10 mL). The reactionwas then stirredfor about 3 h at room temperature (TLC monitoring, eluent:acetone/CH3OH 85:15), then the solvent was evaporated at reducedpressure and the crude mixture underwent silica gel columnchromatography (eluent: acetone/CH3OH 9:1), which afforded762 mg of phthalimide 20a (52% yield, 15% yield using DEAD asfound in the literature [24]).

Compound 20a: Pale yellow prisms (from n-hexane); mp 152e153 �C dec. (lit. [23] 143 �C); Rf ¼ 0.22 (acetone/CH3OH 85:15); 1Hand 13C NMR data matched those known from the literature [24].

4.1.6. 1-(6-Bromohexyl)-4-(1,3-dioxoisoindolin-2-yl)-1-methylpiperidinium bromide 21a

A solution of 20a (530 mg, 2.17 mmol) in 1,6-dibromohexane 11(15 mL) was stirred at room temperature for 3 days. The desiredmonoquaternary bromide 21a (540 mg, 51% yield) was isolatedfollowing the protocol adopted for 15a. For this intermediate and itsrelated structural analogs, we have adopted the following notationon the piperidinium moiety, to indicate the proton chemical shiftsin the corresponding NMR spectra.

Compound 21a: Colorless solid (from 2-propanol); mp 229e232 �C; Rf ¼ 0.66 (0.2 M aqueous KNO3/CH3OH 2:3); 1H NMR(CDCl3): 1.53 (m, 4H, þNeCH2eCH2eCH2eCH2), 1.88 (m, 4H, þNeCH2eCH2eCH2eCH2CH2), 2.03 (m, 2H, H3,5,eq), 2.93 (m, 2H, H3,5,ax),3.41 (t, 2H, CH2eBr, J ¼ 6.2), 3.44 (s, 3H, þNeCH3), 3.91 (m, 4H,H2,6,ax/eq), 4.02 (m, 2H, þNeCH2), 4.62 (m, 1H, Npht-CH or H-4), 7.73(m, 2H, arom.), 7.80 (m, 2H, arom.). 13C NMR (DMSO-d6): 21.82,

23.21, 25.69, 27.73, 32.56, 35.77, 43.84, 44.94, 59.49, 67.35, 123.78,132.11, 135.25, 168.35. MS (ESI) m/z [M]þ Calcd for C20H28BrN2O2

þ:409.13. Found: 408.6.

4.1.7. 1-(6-{[4-(4,5-Dihydroisoxazol-3-yloxy)but-2-ynyl]dimethylammonio}hexyl)-4-(1,3-dioxoisoindolin-2-yl)-1-methylpiperidinium bromide 9a

A solution of intermediate bromide 21a (488 mg, 1.0 mmol) andtertiary amine 17 (237 mg, 1.30 mmol) in acetonitrile (15 mL) wasrefluxed for 6 days under stirring. After the reaction wascompleted (TLC monitoring, eluent: 0.2 M aqueous KNO3/CH3OH2:3), the solvent was evaporated and the residual solid was filteredand crystallized, providing the target compound 9a (530 mg, 79%yield).

Compound 9a: Yellow, hygroscopic solid (from 2-propanol);mp 162e164 �C; Rf ¼ 0.45 (0.2 M aqueous KNO3/CH3OH 2:3); 1HNMR (D2O): 1.33 (m, 4H, þNeCH2eCH2eCH2eCH2), 1.69 (m, 4H,þNeCH2eCH2eCH2eCH2eCH2), 1.88 (m, 2H, H3,5,eq), 2.72 (m, 2H,H3,5,ax), 2.98 (t, 2H, H-42-isox, J ¼ 9.4), 3.01 (s, 6H, þN-(CH3)2), 3.06(s, 3H, þNeCH3), 3.30 (m, 6H, (CH2)2þNeCH2), 3.51 (m, 2H, CH2eþNe(CH3)2), 4.17 (s, 2H, þNeCH2eC^), 4.31 (m, 1H, Npht-CH), 4.32(t, 2H, H-52-isox, J ¼ 9.4), 4.76 (s, 2H, ^CeCH2eO), 7.68 (m, 4H,arom.). 13C NMR (D2O): 21.49, 22.13, 22.40, 22.65, 25.21, 25.40,32.85, 44.00, 45.04, 50.67, 54.29, 57.82, 60.20, 64.13, 70.46, 75.69,85.74, 123.49, 131.28, 134.92, 169.04, 170.11. MS (ESI) m/z [M]2þ

Calcd for C29H42N4O42þ: 255.16. Found: 255.2. Anal. Calcd for

C29H42Br2N4O4 (670.48): C, 51.95; H, 6.31; N, 8.36. Found: C, 52.22;H, 6.53; N, 8.07.

4.1.8. 2-(1-Methylpiperidin-4-yl)-1H-benzo[de]isoquinoline-1,3(2H)-dione 20b

In an atmosphere of dry argon, DIAD (788 mL, 4.0 mmol) wasadded dropwise at 0 �C to a solution of 19 (470 mL, 4.0 mmol), 1,8-naphthalimide 18b (789 mg, 4.0 mmol) and triphenylphosphine(1.05 g, 4.0 mmol) in dry THF (8mL). The desired 1,8-naphthalimide20b (1.23 g, 57% yield), already characterized as the correspondinghydrochloride [23], was purified following the protocol applied tothe preparation of 20a.

Compound 20b: Yellow solid (from n-hexane); mp 195e196 �C;Rf ¼ 0.31 (acetone/CH3OH 85:15); 1H NMR (CDCl3): 1.65 (m, 2H,H3,5,eq), 2.18 (m, 2H, H2,6,ax), 2.28 (s, 3H, NeCH3), 2.98 (m, 2H,H3,5,ax), 3.06 (m, 2H, H2,6,eq), 5.05 (m, 1H, Npth-CH or H-4), 7.73 (t,2H, arom., J ¼ 7.4), 8.19 (d, 2H, arom., J ¼ 8.3), 8.57 (d, 2H, arom.,J ¼ 7.2). 13C NMR (CDCl3): 27.76, 45.69, 51.37, 55.87, 123.28, 127.16,128.40, 131.46, 131.68, 133.89, 164.74. MS (ESI) m/z [MH]þ Calcd forC18H19N2O2

þ: 295.14. Found: 295.1.

4.1.9. 1-(6-Bromohexyl)-4-(1,3-dioxo-1H-benzo[de]isoquinilin-2(3H)-yl)-1-methylpiperidinium bromide 21b

A solution of 20b (800 mg, 2.72 mmol) in 1,6-dibromohexane 11(18 mL) was stirred at room temperature for 3 days. The desiredmonoquaternary bromide 21b (1.04 g, 71% yield) was isolatedfollowing the protocol applied to 15a.

Compound 21b: Colorless solid (from ethanol); mp 247e249 �C; Rf ¼ 0.70 (0.2 M aqueous KNO3/CH3OH 2:3); 1H NMR(DMSO-d6): 1.40 (m, 4H, þNeCH2eCH2eCH2eCH2), 1.86 (m, 6H,þNeCH2eCH2eCH2eCH2CH2 and H3,5,eq), 2.97 (m, 2H, H3,5,ax),3.15 (s, 3H, NeCH3), 3.34 (t, 2H, CH2eBr, J ¼ 7.4), 3.57 (m, 6H,(CH2)2þNeCH2), 5.27 (m, 1H, Npht-CH or H-4), 7.87 (t, 2H, arom.,J ¼ 7.2), 8.46 (d, 2H, arom., J ¼ 8.5), 8.49 (d, 2H, arom., J ¼ 7.3). 13CNMR (DMSO-d6): 21.09, 22.10, 25.73, 27.76, 32.57, 35.79, 38.03,43.74, 47.28, 59.90, 67.28, 123.05, 128.00, 131.55, 131.87, 135.05,164.64. MS (ESI) m/z [M]þ Calcd for C24H30BrN2O2

þ: 457.15. Found:457.2.

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232230

4.1.10. 1-(6-{[4-(4,5-Dihydroisoxazol-3-yloxy)but-2-ynyl]dimethylammonio}hexyl)-4-(1,3-dioxo-1H-benzo[de]isoquinolin-2-(3H)-yl)-1-methylpiperidinium bromide 9b

Intermediate bromide 21b (900 mg, 1.67 mmol) and tertiaryamine 17 (395 mg, 2.17 mmol) were reacted following the proce-dure applied to the preparation of 9a, providing the desired bis-quaternary ammonium salt 9b (458 mg, 38% yield).

Compound 9b: Colorless solid (from 2-propanol); mp 198e202 �C; Rf¼ 0.40 (0.2 M aqueous KNO3/CH3OH 2:3); 1H NMR (D2O):1.42 (m, 4H, NþeCH2eCH2eCH2eCH2), 1.84 (m, 6H, NþeCH2eCH2e

CH2eCH2eCH2e and H3,5,eq), 2.92 (m, 2H, H3,5,ax), 3.00 (t, 2H, H-42-isox, J ¼ 9.4), 3.03 (s, 6H, þNe(CH3)2), 3.13 (s, 3H, þNeCH3), 3.47 (m,8H, (CH2)2þNeCH2 and CH2e

þNe(CH3)2), 4.20 (s, 2H, þNeCH2eC^),4.33 (t, 2H, H-52-isox, J ¼ 9.4), 4.78 (s, 2H, ^CeCH2eO), 5.11 (m, 1H,Npht-CH), 7.53 (t, 2H, arom., J¼ 8.0), 8.02 (d, 2H, arom., J ¼ 8.3), 8.15(d, 2H, arom., J ¼ 8.0). 13C NMR (D2O): 21.11, 21.46, 21.64, 22.15,25.23, 25.50, 32.76, 46.71, 47.38, 50.68, 54.27, 57.79, 60.50, 64.17,70.46, 75.68, 85.74, 121.36, 127.23, 130.99, 131.67, 135.29, 166.08,166.15, 169.06. MS (ESI) m/z [M]2þ Calcd for C33H44N4O4

2þ: 280.17.Found: 280.2. Anal. Calcd for C33H44Br2N4O4 (720.53): C, 55.01; H,6.16; N, 7.78. Found: C, 54.73; H, 6.40; N, 7.91.

4.2. Pharmacology

New Zealand male white rabbits (3.0e3.5 kg), male albinoguinea pigs (250e350 g), Swiss male mice (25e35 g) and Wistarrats (150e250 g) of both sexes (Charles River, Italy) were used.Animals were housed, handled and cared for according to the Eu-ropean Community Council Directive 86 (609) EEC, and theexperimental protocols were carried out in compliance with Italianregulations (DL 116/92) and with the local Ethical CommitteeGuidelines for Animal Research.

4.2.1. Functional studies on isolated tissuesThe tissues for in vitro experiments were removed from animals

fasted 18 h before the experiments and killed by CO2 inhalation.Isolated preparations were set up following the techniques previ-ously described [36].

4.2.1.1. Electrically-stimulated rabbit vas deferens. According toEltze [26], the prostatic portion of each vas deferens was mountedin a 10 mL organ bath, containing a modified Krebs solution (mMcomposition: NaCl 134, KCl 3.4, CaCl2 2.8, KH2PO4 1.3, NaHCO3 16,MgSO4 0.6, glucose 7.7) kept at 31 �C and bubbled with 95% O2 - 5%CO2. Yohimbine (1.0 mM)was present throughout the experiment toprevent prejunctional a2-adrenoceptors stimulation. For isometricrecordings, the tissues were left to equilibrate for 45 min under apassive load of 0.75 g before electrical field stimulation throughplatinum wire electrodes was applied by square-wave pulses(0.5 ms, 0.05 Hz, supramaximal intensity 450 mA; LACE ElettronicaMod. ES3, Ospedaletto PI, Italy).

4.2.1.2. Electrically-stimulated guinea pig left atrium. Heart wasrapidly dissected and the right and left atria were separated out.The left atriumwas suspended under 0.5 g tension in a 20mL organbath in a modified Krebs-Henseleit buffer solution (mM composi-tion: NaCl 118.9, KCl 4.6, CaCl2 2.5, KH2PO4 1.2, NaHCO3 25,MgSO4$7H2O 1.2, glucose 11.1) gassed with carbogen (95% O2 � 5%CO2) at 32 �C. The tissue was electrically-paced by rectangularsubmaximal impulses (2 Hz, 5 ms, 5 V) by using platinum wireelectrodes. After a 30 min stabilization period, inotropic responseswere measured as changes in isometric tension.

4.2.1.3. Guinea pig ileum. Segments of the terminal portion of theileum (2 cm in length) were transferred into a 10 mL organ bath

filled with Krebs-Henseleit solution gassed with carbogen at 37 �Cand loaded with a tension of 1 g. The preparation was allowed toequilibrate for 30 min before inducing isometric contractileresponses.

4.2.1.4. Protocols. Agonist concentrationeresponse curves wereconstructed in each tissue by cumulative application of concen-trations of the test compounds (1 nMe30 mM) [37]. The agonistpotency was expressed as pEC50 (�log EC50) calculated by linearregression analysis using the least square method. Intrinsic activity(a) was calculated as a fraction of the maximal response to thereference full agonist, Bethanechol or McN-A-343. Concentration-response curves of the agonists were reconstructed in the presenceof atropine 0.1 mM and hexamethonium 100 mM. When the com-pounds were tested as antagonists, a doseeresponse curve to thefull agonists Bethanechol or McN-A-343 was repeated after 30 minincubationwith the test compounds (1e30 mM). The potency of thecompounds acting as antagonists was expressed as pKB value, thecalculated molar concentration of the test compounds that causes atwofold increase in the EC50 values of the muscarinic agonists usedin the functional tests, calculated according to Furchgott’s method[38]. All data are expressed as means � SEM of 6e8 experiments.

4.2.1.5. Rat brain cholinesterase inhibition. The inhibition of braincholinesterase was determined spectrophotometrically using ace-tylthiocholine as substrate according to the method described byEllman [39]. Aliquots of rat brain homogenates were incubated inphosphate buffer 0.1 M (pH 8.0) with 5,50-dithiobis(2-nitrobenzoicacid) (DTNB) (5 mM) and test compounds at appropriate concen-trations (10 nMe100 mM). The reaction was started at 37 �C byadding 20 mL acetylthiocholine (75 mM). The reaction was stoppedafter 15 min by adding formalin (4%). The hydrolysis of acetylth-iocholine catalyzed by the enzyme was determined by monitoringthe formation of the yellow 5-thio-2-nitrobenzoate anion at412 nm (JenWaymod. 6505, Dunmow, Essex, England). The percentinhibition of cholinesterase was calculated as: [(Acontrol � Asample)/Acontrol] � 100. Results are provided as pIC50 (�log IC50, where IC50is the concentration causing half-maximal inhibition of cholines-terase activity) and maximum percent inhibition of rat braincholinesterase. All data are expressed as mean � SEM of 4e6experiments.

4.2.2. In vivo assays4.2.2.1. Behavioral assays in conscious mice. Behavioral tests werecarried out from 9 a.m. to 2 p.m. in mice fasted 18 h before theexperiments.

4.2.2.1.1. Writhing test. The writhing test was performed ac-cording to Koster et al. [31]. Briefly, 30 min after the compoundsunder study or the vehicle (saline 0.9%, 10 mL/kg) were subcuta-neously (s.c.) administered, groups of 6e8 mice were intraperito-neally (i.p.) injected with 0.6% acetic acid and placed inobservational chambers. The number of writhes of each mouse wascounted in a period of 5e10 min after algogen agent application, inorder to calculate the percentage of analgesia as difference inwrithing responses between the treated group and the control one.To characterize 8b antinociception, different sets of mice were i.p.administered with the opioid antagonist naloxone (1 mg/kg), themuscarinic antagonists atropine (5 mg/kg) or methylatropine(5 mg/kg), or the nicotinic antagonist mecamylamine (1 mg/kg)15 min before 8b treatment at 0.5 mg/kg s.c.

4.2.2.1.2. Cholinergic side effects. The incidence of cholinergicunwanted side effects were simultaneously estimated in miceduring the writhing test. Body temperature was recorded with arectal probe (Delta OhmHD8704, Padova, Italy) immediately beforeand 30 min after drug administration in an air conditioned room

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232 231

(20 �C); the hypothermic effect was scored as present if bodytemperature decreased by 2 �C with respect to basal temperature.The presence or absence of sialagogic, lachrymatory and tremori-genic responses and of intestinal hypermotility were recorded on aperiod of 1 h following s.c. administration of the test compounds.Salivation and lachrymation were scored as present if the areasurrounding the mouth or the eyes were wet, while tremor wasscored as present on the basis of the tremor provoked by handlingduring the temperature measurement. Results are expressed aspercentage of treated animals exhibiting the specific cholinergicside effect.

4.2.2.1.3. Locomotor activity. Spontaneous locomotor activitywas assessed in the open field apparatus, consisting in four squarechambers (height 40 cm, width 45 cm, depth 45 cm) with a videocamera fixed above them. Each animal, 30 min before the trial,received subcutaneously the compounds under study or thevehicle. Locomotor activity was recorded for 90 min by means of avideo-tracking software (Any-maze, Ugo Basile, Comerio (VA),Italy) measuring the total traveled distance (meters) and the totaltime of immobility (seconds).

4.2.2.2. Cardiovascular effects in anaesthetized rats. Rats underurethane anesthesia (1.25 g/kg i.p.) were tracheotomized to facili-tate respiration and maintained at approximately 37 �C throughoutthe experiment by means of an overhead heating lamp. After sur-gery, a period of 15 min was allowed to stabilize the parametersunder consideration. Blood pressure and heart rate were measuredfrom the left common carotid artery through a catheter connectedvia a pressure transducer (TDS104A Biopac Systems, 2BiologicalInstruments, Besozzo VA, Italia) to a MacLab digital data acquisitionsystem (PowerLab/4SP ADI Instruments, Ugo Basile, Comerio, VA,Italy). The vehicle or increasing doses of compound 8b wereinjected i.p. in a volume of 1 mL/kg body weight. Data wereexpressed as changes in mean blood pressure (mmHg, Fig. 5a) andheart rate (beats/min, Fig. 5b) by subtracting the baseline valuefrom each of the maximal values detected within 5 min after in-jection of the test compound.

4.2.2.3. Data analysis. All data are given as mean � SEM of 6e8observations. ID50 values of analgesic activity (i.e., the dose thatinhibits acetic acid writhing response by 50% compared to thecontrol group) were calculated by non-linear regression analysis(Prism 5.0, GraphPad Software, San Diego, CA, USA). Statisticalanalysis was performed by means of one-way analysis of variance,followed by Dunnett’s post test unless otherwise indicated. A Pvalue <0.05 was considered significant and a P value <0.01 highlysignificant.

Acknowledgments

Carlo Matera wishes to thank “Dote Ricerca”: FSE, RegioneLombardia, which co-financed his postdoctoral position. The au-thors are indebted to Jessica Klöckner for assaying the inhibitoryactivity of 6a and 6b on purified AChE and to Dr. Giacomo LucaVisconti for recording LC-MS/MS spectra.

References

[1] H.-L. Pan, Z.-Z. Wu, H.-Y. Zhou, S.-R. Chen, H.-M. Zhang, D.-P. Li, Modulation ofpain transmission by G-protein-coupled receptors, Pharmacol. Ther. 117(2008) 141e161 and references cited therein.

[2] D.F. Fiorino, M. Garcia-Guzman, Muscarinic pain pharmacology: realizing thepromise of novel analgesics by overcoming old challenges, in: A.D. Fryer,A. Christopoulos, N.M. Nathanson (Eds.), Muscarinic Receptors, Handbook ofExperimental Pharmacology, vol. 208Springer-Verlag, Berlin Heidelberg,2012, pp. 191e221 and references cited therein.

[3] C. Ghelardini, N. Galeotti, A. Bartolini, Loss of muscarinic antinociceptionby antisense inhibition of M1 receptors, Br. J. Pharmacol. 129 (2000) 1633e1640.

[4] J. Gomeza, H. Shannon, E. Kostenis, C.C. Felder, L. Zhang, J. Brodkin,A. Grinberg, H. Sheng, J. Wess, Pronounced pharmacologic deficits in M2muscarinic acetylcholine receptor knockout mice, Proc. Natl. Acad. Sci. 96(1999) 1692e1697.

[5] A. Duttaroy, J. Gomeza, J.-W. Gan, N. Siddiqui, A.S. Basile, W.D. Harman,P.L. Smith, C.C. Felder, A.I. Levey, J. Wess, Evaluation of muscarinic agonist-induced analgesia in muscarinic acetylcholine receptor knockout mice, Mol.Pharmacol. 62 (2002) 1084e1093.

[6] H.-M. Zhang, S.-R. Chen, M. Matsui, D. Gautam, J. Wess, H.L. Pan, Opposingfunctions of spinal M2, M3, and M4 receptor subtypes in regulation ofGABAergic inputs to dorsal horn neurons revealed by muscarinic receptorknockout mice, Mol. Pharmacol. 69 (2006) 1048e1055.

[7] G. Martino, C. Puma, X.-H. Yu, A.-K. Gilbert, M. Coupal, N. Markoglou,F.S. McIntosh, M.N. Perkins, J.M.A. Laird, The M1/M4 preferring agonist xano-meline is analgesic in rodent models of chronic inflammatory and neuropathicpain via central site of action, Pain 152 (2011) 2852e2860.

[8] A. Bartolini, L. Di Cesare Mannelli, C. Ghelardini, Analgesic and anti-neuropathic drugs acting through central cholinergic mechanisms, Recent Pat.CNS Drug Discovery 6 (2011) 119e140.

[9] J. Wess, R.M. Eglen, D. Gautam, Muscarinic acetylcholine receptors: mutantmice provide new insights for drug development, Nat. Rev. Drug Discovery 6(2007) 721e733.

[10] C. Dallanoce, P. Conti, M. De Amici, C. De Micheli, E. Barocelli, M. Chiavarini,V. Ballabeni, S. Bertoni, M. Impicciatore, Synthesis and functional character-ization of novel derivatives related to oxotremorine and oxotremorine-M,Bioorg. Med. Chem. 7 (1999) 1539e1547.

[11] R. Schrage, W.K. Seemann, J. Klöckner, C. Dallanoce, K. Racké, E. Kostenis,M. De Amici, U. Holzgrabe, K. Mohr, Agonists with supraphysiological efficacyat the muscarinic M2 ACh receptor, Br. J. Pharmacol. 169 (2013) 357e370.

[12] E. Barocelli, V. Ballabeni, S. Bertoni, M. De Amici, M. Impicciatore, Evidence forspecific analgesic activity of a muscarinic agonist selected among a new seriesof acetylenic derivatives, Life Sci. 68 (2001) 1775e1785.

[13] T. Disingrini, M. Muth, C. Dallanoce, E. Barocelli, S. Bertoni, K. Kellershohn,K. Mohr, M. De Amici, U. Holzgrabe, Design, synthesis and action ofoxotremorine-related hybrid-type allosteric modulators of muscarinicacetylcholine receptors, J. Med. Chem. 49 (2006) 366e372 and referencescited therein.

[14] J. Antony, K. Kellershohn, M. Mohr-Andrä, A. Kebig, S. Prilla, M. Muth,E. Heller, T. Disingrini, C. Dallanoce, S. Bertoni, J. Schrobang, C. Tränkle,E. Kostenis, A. Christopoulos, H.-D. Höltje, E. Barocelli, M. De Amici,U. Holzgrabe, K. Mohr, Dualsteric GPCR targeting: a novel route to binding andsignaling pathway selectivity, FASEB J. 23 (2009) 442e450.

[15] K. Jepsen, H. Lüllmann, K. Mohr, J. Pfeffer, Allosteric stabilization of 3H-N-methylscopolamine binding in guinea-pig myocardium by an antidote againstorganophosphorus intoxication, J. Pharmacol. Toxicol. 63 (1988) 163e168.

[16] C. Tränkle, E. Mies-Klomfass, H.M. Botero Cid, U. Holzgrabe, K. Mohr, Identi-fication of a [3H] ligand for the common allosteric site of muscarinic acetyl-choline M2 receptors, Mol. Pharmacol. 54 (1998) 139e145.

[17] M. Muth, W. Bender, O. Scharfenstein, U. Holzgrabe, E. Balaktova, C. Tränkle,K. Mohr, Systematic development of high affinity bis(ammonio)alkane-typeallosteric enhancers of muscarinic ligand binding, J. Med. Chem. 46 (2003)1031e1040.

[18] M. Muth, M. Sennwitz, K. Mohr, U. Holzgrabe, Muscarinic allosteric enhancersof muscarinic ligand: pivotal pharmacophoric elements in hexamethonio-type agents, J. Med. Chem. 48 (2005) 2212e2217.

[19] K. Mohr, C. Tränkle, E. Kostenis, E. Barocelli, M. De Amici, U. Holzgrabe,Rational design of dualsteric GPCR ligands: quests and promise, Br. J. Phar-macol. 159 (2010) 997e1008.

[20] A. Bock, N. Merten, R. Schrage, C. Dallanoce, J. Bätz, J. Klöckner, J. Schmitz,C. Matera, K. Simon, A. Kebig, L. Peters, A. Müller, J. Schrobang-Ley, C. Tränkle,C. Hoffmann, M. De Amici, U. Holzgrabe, E. Kostenis, K. Mohr, The allostericvestibule of a seven transmembrane helical receptor controls G protein-coupling, Nat. Commun. 3 (2012) 1044. http://dx.doi.org/10.10308/ncomms2028.

[21] J. Klöckner, J. Schmitz, U. Holzgrabe, Convergent, short synthesis of themuscarinic superagonist iperoxo, Tetrahedron Lett. 51 (2010) 3470e3472.

[22] M.L. Lolli, S.L. Hansen, B. Rolando, B. Nielsen, P. Wellendorph, K. Madsen,O. Miller Larsen, U. Kristiansen, R. Fruttero, A. Gasco, T.N. Johansen, Hy-droxy-1,2,5-oxadiazolyl moiety as bioisoster of the carboxy function. Syn-thesis, ionization constants, and pharmacological characterization of g-aminobutyric acid (GABA) related compounds, J. Med. Chem. 49 (2006)4442e4446.

[23] M. Langlois, J.L. Soulier, V. Rampillon, C. Gallais, B. Brémont, S. Shen, D. Yang,A. Giudice, F. Sureau, Synthesis of quinazoline-2,4-dione and naphthalimidederivatives as new 5-HT3 receptor antagonists, Eur. J. Med. Chem. 29 (1994)925e940.

[24] I. Iriepa, F.J. Villasante, E. Gálvez, A. Herrera, A. Sánchez, F.H. Cano, Synthesis,structural and conformational study of selected N-substituted phthalimides,Cent. Eur. J. Chem. 3 (2005) 683e704.

[25] C. Tränkle, I. Andresen, G. Lambrecht, K. Mohr, M2 receptor binding of theselective antagonist AF-DX 384: possible involvement of the common allo-steric site, Mol. Pharmacol. 53 (1998) 304e312.

C. Matera et al. / European Journal of Medicinal Chemistry 75 (2014) 222e232232

[26] M. Eltze, Muscarinic M1 and M2 receptors mediating opposite effects onneuromuscular transmission in rabbit vas deferens, Eur. J. Pharmacol. 151(1988) 205e221.

[27] A.L. Harvey, E. Kornisiuk, K.N. Bradley, C. Cervenansky, R. Durán, G. Sánchez,D. Jerusalinsky, Effects of muscarinic toxins MT1 and MT2 from green mambaon different muscarinic cholinoceptors, Neurochem. Res. 27 (2002) 1543e1554.

[28] M.P. Caulfield, N.J.M. Birdsall, International Union of Pharmacology. XVII.Classification of muscarinic acetylcholine receptors, Pharmacol. Rev. 50 (1998)279e290.

[29] This functional test is commonly used to characterize ligands for M1 re-ceptors. However, Melchiorre et al. suggested that inhibition of the neuro-genic contractions in this tissue preparation could be mediated by M4receptors [Br. J. Pharmacol. 132 (2001) 1009e1016].

[30] P. Kapková, V. Alptüzün, P. Frey, E. Erciyas, U. Holzgrabe, Search for dualfunction inhibitors for Alzheimer’s disease: synthesis and biological activity ofacetylcholinesterase inhibitors of pyridinium-type and their Ab fibril forma-tion inhibition capacity, Bioorg. Med. Chem. 14 (2006) 472e478 and refer-ences cited therein.

[31] R. Koster, M. Anderson, E.J. De Beer, Acetic acid for analgesic screening, Fed.Proc. 18 (1959) 412e418.

[32] T.D. Gould, D.T. Dao, C.E. Kovacsics, Mood and anxiety related phenotypes inmice: characterization using behavioral tests, Neuromethods 42 (2009) 1e20.

[33] N. Bernardini, A.I. Levey, G. Augusti-Tocco, Rat dorsal root ganglia express m1-m4 muscarinic receptor proteins, J. Peripher. Nerv. Syst. 4 (1999) 222e232.

[34] N. Bernardini, C. Roza, S.K. Sauer, J. Gomeza, J. Wess, P.W. Reeh, Muscarinic M2receptors on peripheral nerve endings: a molecular target of antinociception,J. Neurosci. 22 (12) (2002) RC229.

[35] C. Sánchez, E. Meier, Central and peripheral mediation of hypothermia, tremorand salivation induced by muscarinic agonists in mice, Pharmacol. Toxicol. 72(1993) 262e267.

[36] E. Barocelli, V. Ballabeni, S. Bertoni, C. Dallanoce, M. De Amici, C. De Micheli,M. Impicciatore, New analogues of oxotremorine and oxotremorine-M: esti-mation of their in vitro affinity and efficacy at muscarinic receptor subtypes,Life Sci. 67 (2000) 717e723.

[37] J.M. Van Rossum, Cumulative dose-response curves. II. Technique for themaking of dose-response curves in isolated organs and the evaluation of drugparameters, Arch. Int. Pharmacodyn. Ther. 143 (1963) 299e330.

[38] R.F. Furchgott, P. Bursztyn, Comparison of dissociation constants and ofrelative efficacies of selected agonists acting on parasympathetic receptors,Ann. N. Y. Acad. Sci. 144 (1967) 882e898.

[39] G.L. Ellman, K.D. Courtney, V. Andres, R.M. Feather-Stone, A new and rapidcolorimetric determination of acetylcholinesterase activity, Biochem. Phar-macol. 7 (1961) 88e95.