Rational design of partial agonists for the muscarinic M1 acetylcholine receptor

Article

pubs.acs.org/jmc

Rational Design of Partial Agonists for the Muscarinic M₁

Acetylcholine Receptor

Xinyu Chen,^†,‡Jessika Klockner,^†Janine Holze,^§Cornelia Zimmermann,^§Wiebke K. Seemann,^∥

̈

Ramona Schrage,^§Andreas Bock,^⊥Klaus Mohr,^§Christian Trankle,*^,§Ulrike Holzgrabe,*^,†

̈

and Michael Decker*^,†,‡

^†Institute of Pharmacy and Food Chemistry, University of Wurzburg, Am Hubland, 97074 Wurzburg, Germany

̈

^‡Institute of Pharmacy, University of Regensburg, Universitatsstrasse 31, 93053 Regensburg, Germany

̈

^§Pharmacology and Toxicology Section, Institute of Pharmacy, University of Bonn, Gerhard-Domagk-Strasse 3, 53121 Bonn,

Germany

^∥Department of Pharmacology, University of Cologne, Gleueler Strasse 24, 50931 Cologne, Germany

^⊥Institute of Pharmacology and Toxicology, University of Wurzburg, Versbacher Strasse 9, 97078 Wurzburg, Germany

̈

S

* Supporting Information

ABSTRACT: Aiming to design partial agonists for a G-protein-

coupled receptor based on dynamic ligand binding, we synthesized

three diﬀerent series of bipharmacophoric ligands composed of the

orthosteric building blocks iperoxo and 1 linked to allosteric

modulators (BQCA-derived compounds, BQCAd; TBPB-derived

compound, TBPBd). Their interactions were studied with the

human muscarinic acetylcholine M₁-receptor (hM₁) with respect to

receptor binding and G_q-protein signaling. Results demonstrate that

iperoxo/BQCAd (2, 3) and 1/BQCAd hybrids (4) act as M₁partial

agonists, whereas 1/TBPBd hybrids (5) did not activate M₁-

receptors. Among the iperoxo/BQCAd-hybrids, spacer length in

conjunction with the pattern of substitution tuned eﬃcacy. Most

interestingly, a model of dynamic ligand binding revealed that the

spacer length of 2a and 3a controlled the probability of switch

between the inactive purely allosteric and the active bitopic orthosteric/allosteric binding pose. In summary, dynamic ligand

binding can be exploited in M₁receptors to design partial agonists with graded eﬃcacy.

the hM₁receptor might have therapeutic relevance because

they appear to improve symptoms of schizophrenia and

impaired cognitive functions such as in Alzheimer’s disease

(AD) as well as to positively inﬂuence other AD-related

pathophysiological processes.^2,10,13−15

The aim of this study was to design rationally as a proof-of-

concept bipharmacophoric partial agonists with aﬃnity for the

M₁receptor in three diﬀerent series of hybrids and to elucidate

their signaling behavior qualitatively. Signaling behavior might

be of particular relevance because M₁stimulation is considered

to have more than symptomatic procognitive eﬀects in AD¹⁵in

that it inﬂuences pathophysiology by reducing β-amyloid 42

(Aβ₄₂).^10,13

INTRODUCTION

■

Recently, dualsteric (bitopic orthosteric/allosteric) ligands of G

protein-coupled receptors (GPCRs) have attracted much

interest not only as pharmacological tools for the elucidation

of receptor activation-induced conformational transitions but

also as a concept for the design of new drugs.^1,2Dualsteric

compounds composed of an orthosteric ligand (either agonist

or antagonist) and an allosteric moiety have been found to

substantially enhance the aﬃnity to the receptor³but may as

well reduce aﬃnity or leave it unchanged depending on the

individual building blocks.¹Additionally, the dualsteric

approach may generate receptor subtype selectivity⁴and biased

signaling.⁵⁻⁷The ﬁve M receptor subtypes (mAChR₁₋₅or M₁−

M₅) are metabotropic GPCRs, with M₁being most abundantly

expressed in the central nervous system (cortex, striatum, and

hippocampus).⁸M₁signaling is mediated preferentially by G_q/

G₁₁-proteins activating phospholipase C and via other G-

proteins initiating additional signaling systems.^9,10Furthermore,

the M₁receptor inﬂuences ion channels that possess high

relevance for cognition such as voltage-gated Ca²⁺channels and

the N-methyl-^D-aspartate (NMDA) receptor.^11,12Agonists of

Two subtypes of muscarinic receptors are structurally well

characterized by means of their crystal structures, i.e., the

tiotropium-bound M₃receptor¹⁶and the quinuclidinyl

benzilate (QNB)-occupied M₂receptor,¹⁷both captured in

their inactive conformations. Very recently, the M₂receptor has

Received: June 7, 2014

A

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 1. Structures of compounds aimed at targeting orthosteric or allosteric sites of the M₁receptor (R = 1-(2-methylbenzyl)piperidin-4-yl for

TBPB).

Figure 2. (A) Iperoxo/BQCAd hybrids composed of derivatives of the prototypical M₁-selective BQCA, spacer, and the M₂-superagonist iperoxo.

(B) 1/BQCAd hybrids consisting of compound 1 and a BQCA analogue. (C) 1/TBPBd hybrids composed of compound 1 and N-piperidine

benzimidazolone derivative.

been cocrystallized with iperoxo in the active receptor

conformation.¹⁸Iperoxo represents a muscarinic agonist with

supraphysiological eﬃcacy at this subtype.¹⁹These muscarinic

receptor subtypes exhibit a high structural similarity consisting

of a deeply buried orthosteric binding site for agonists and

antagonists connected with a allosteric vestibule formed by

solvent accessible extracellular loops and the extracellular parts

of transmembrane helices (TM) 3, 5, and 7. These features also

apply to other class A receptors.²⁰Because of the sequence

identity between M₁and M₃receptors and M₁and M₂subtypes

amounting to 49.7 and 45.0%,²¹respectively, Ilien et al. were

able to develop a valid homology model of the M₁receptor.^22,23

Using this model, they could characterize the bitopic behavior

of hybrid ligands composed of the M₁-selective antagonist

pirenzepine and the allosteric ﬂuorescent label Bodipy; these

two moieties were connected via an alkylene, polyethylene

glycol, or amidoalkylene linker.

The orthosteric binding site of muscarinic receptors is well

characterized by computational and molecular biology methods

(for reviews see Davie 2013 and Goodwin 2007),^2,24unraveling

the essential role of a negatively charged aspartate residue in

TM3 that interacts with the positively charged nitrogen of the

endogenous ligand acetylcholine or antagonists like protonated

atropine.^16,17Currently, the presence of more than one

allosteric binding site is discussed: the “common allosteric

site” relates to the above-mentioned allosteric vestibule being

located next to the aforementioned extracellular loops (ECL2

and 3; vestibule) including negatively charged residues in

ECL3. The topography of this site is conserved across the

subtypes. Because of their positive charges, archetypal

modulators such as alcuronium and gallamine bind to the

common allosteric site.^25,26In addition, a second allosteric

binding site was postulated which seems to be occupied by the

compounds (9S,10S,12R)-2,3,9,10,11,12-hexahydro-10-hy-

droxy-9-methyl-1-oxo-9,12-epoxy-1H-diindolo[1,2,3-fg:3′,2′,1′-

kl]pyrrolo[3,4-i][1,6]benzodiazocine-10-carboxylic acid hexyl

ester (KT5720)²⁷and 17-β-hydroxy-17-α-ethynyl-δ-4-

androstano[3,2-b]pyrimido[1,2-a]benzimidazole

(WIN62,577).²⁸

By designing bipharmacophoric ligands consisting of the M₂-

superagonist iperoxo¹⁹and a phthalimido- or naphthalimido-

propylammonium moiety, which were found to bind to M₂

receptors with moderate selectivity, we had learned that

subtype selectivity of ligand binding can be achieved by

concomitantly addressing the allosteric and orthosteric site.²⁹

Compared to conventional agonists, including acetylcholine,

these ligands are able to preferentially use one signaling

pathway over others, hence displaying biased signaling.^4,5In

addition and most interestingly with respect to the current

study, receptor binding of a bipharmacophoric compound may

theoretically switch between a dualsteric and a purely allosteric

orientation.^4,7,30Regarding bipharmacophoric M₂receptor

ligands, the dynamic switch between dualsteric active and

allosteric inactive binding orientations (R_pose) has recently been

employed for the rational design of partial agonists.³⁰Partial

agonist activity could be a beneﬁcial feature of potential M₁

selective anti-Alzheimer drugs because it is known that partial

agonism can reduce agonist-induced receptor desensitization.³¹

Furthermore, partial agonism is a currently debated concept in

the design of new antipsychotic GPCR drugs.³²

B

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 1. Synthesis of Iperoxo/BQCAd Hybrids

a

Reagents and conditions: (i) toluene, reﬂux; (ii) diphenyl ether, 210 °C (microwave); (iii) benzyl chloride, K₂CO₃, DMF, 80 °C, 72%; (iv)

H₂N(CH₂)_nOH, 130−150 °C, 57%; (v) HBr, H₂SO₄, reﬂux, 50%; (vi) iperoxo base, K₂CO₃, CH₃CN, 55 °C, 44%.

In the current study, we synthesized derivatives of highly

subtype selective allosteric ligands and linked them to

orthosteric agonists aiming to design bipharmacophoric M₁

receptor agonists. From the group of allosteric M₁ligands

developed by diﬀerent companies such as Acadia, Merck,

Lundbeck, or research institutes and groups like Vanderbilt

University, Christopoulos, and others,^2,10,33,34we have chosen

the benzyl quinolone carboxylic acids (BQCAs), being a group

of ligands with positive allosteric modulator (PAM) properties

with respect to agonist (not antagonist) binding and function in

M₁receptors.³⁴⁻³⁷Additionally, BQCA has been shown to

have allosteric agonist properties.^34,38Recently, we applied the

message-address concept successfully to the design of dualsteric

M₂ligands which displayed G_ibiased signaling.^4,5In the current

study, following a similar approach, derivatives of the prototypal

M₁selective BQCA (BQCAd) served as the M₁address which

was linked via a spacer to an agonist, either compound 1

(AF292), which is claimed to be a M₁selective agonist,³⁹or the

orthosteric M₂-superagonist iperoxo (Figure 1).¹⁹Two points

of attachment were chosen due to the structure−activity

relationships (SARs) reported in the literature:⁴⁰the carboxylic

group and the benzyl group. Because the carboxylic group is a

suitable contact point for improving allosteric modulation, we

herewith describe the synthesis of iperoxo/BQCAd-prototype

hybrids (2,3) utilizing the carboxylic functional group to

connect the spacer (Figure 2A). Of note, that these quarternary

compounds (2,3) are unlikely to penetrate the blood−brain

barrier and were not designed in an attempt to perform drug

development but as a proof-of-concept regarding bipharmaco-

phoric partial M₁agonists. Furthermore, the M₁agonist 1^39,41

was attached to the naphthalene residue of a BQCA-derived

compound (4, Figure 2B). With regard to compound 1, the

lactam nitrogen atom was chosen for connection because its

modiﬁcation seemed the least susceptible to a loss of aﬃnity

(Figure 2B). Beside the substitution pattern on the BQCAd

part, the length of the linker was varied in both series. In

addition to using BQCA derived compounds for the allosteric

binding moieties in these ligands, the N-piperidine benzimida-

zolone moiety of an allosteric agonist 1-(1′-(2-methylbenzyl)-

1,4′-bipiperidin-4-yl)-1H-benzo[d]imidazol-2(3H)-one

(TBPB) was also applied to connect it to 1 via linkers of

diﬀerent lengths (5, Figure 2C). The N-piperidine benzimida-

zolone moiety has been identiﬁed as an example for a key

pharmacophoric structure of allosteric agonists (for review see

ref 14).¹⁴

M₁receptor activation and signaling behavior of the

compounds were studied by monitoring dynamic mass

redistribution (DMR) in live Chinese hamster ovary cells

(CHO-hM₁) stably transfected with the human M₁recep-

tor.^42,43A qualitative comparison at selected high concen-

trations in the absence and presence of speciﬁc blockers of G_i-

and G_q-protein-mediated pathways revealed that the M₁

receptor eﬀects of all hybrids were predominantly transmitted

by G_qsignaling. Concentration eﬀect curves revealed partial

agonism for M₁receptor activation going along with

submicromolar potency in case of hybrids 2, 3, and 4 but not

5. The functional DMR data of all iperoxo/BQCAd hybrids

(2,3) were analyzed by means of the operational model of

agonism for dynamic ligands which had been introduced

recently.³⁰This model accounts for a switch of binding

topography between an active pose and an inactive pose

resulting in partial activation of the receptor population, i.e.,

partial agonism. The resulting “overall” coupling eﬃciency is

quantiﬁed by the dynamic transduction coeﬃcient τ_dyn.

³⁰In the

present work, analysis of this functional data was extended by

including binding data from separate experiments, applying a

simultaneous global curve ﬁtting procedure. This analytical one-

step method allowed us to characterize conveniently the two

selected iperoxo/BQCAd hybrids 2a and 3a regarding their

respective functional aﬃnities of the active (pK_active) and the

inactive (pK_inactive) pose. This revealed a switch in favor of the

ligand active binding pose from 2a to 3a by the elongation of

the spacer length from C4 to C6. In addition, ε_max, i.e., the

system-independent maximum intrinsic eﬃcacy of the dynamic

C

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 2. Synthesis of M₁Orthosteric Agonist Compound 1

a

Reagents and conditions: (i) NaNO₂(1.2 equiv), KBr (3 equiv), HBr (47% soln), water, 0 °C to rt, 62%; (ii) (1) KOH (0.95 equiv), potassium

ethyl xanthate (1.1 equiv), water, 50 °C, 2 h, (2) diethylenediamine dihydrochloride (2 equiv), KOH soln (4 equiv), 50 °C, 2 h, 61%; (iii) 1-Boc-4-

piperidinone (0.7 equiv), (NH₄)₂CO₃(1 equiv), benzene, reﬂux, 48 h, 63%; (iv) HCl (5 M in iPrOH), MeOH, rt, 24 h, 76%.

a

Scheme 3. Synthesis of Compound 1/BQCAd Hybrids

a

Reagents and conditions: (i) conc H₂SO₄(cat.), MeOH, reﬂux, 24 h; (ii) LiAlH₄(2 equiv), dry THF, 0 °C to rt, 74% over two steps; (iii) NaOH

(1.1 equiv), 1-chloro-3-iodopropane or 1,8-dibromooctane (1.1 equiv), CH₃CN/H₂O, 65−70 °C, 16 h, 72%; (iv) TBDMSCl (1.2 equiv), Et₃N (1.5

equiv), DMAP (cat.), CH₂Cl₂, rt, 16 h; (v) NaI (1.2 equiv), acetone, reﬂux, 16 h; (vi) NaH (60% in paraﬃn oil, 2 equiv), 18 (1.2 equiv), dry DMF, 0

°C to rt, 23% over two steps; (vii) CH₃CO₂H/H₂O/THF 3:1:1, rt, 24 h, 78%; (viii) Ph₃P (1.6 equiv), imidazole (2.2 equiv), I₂(1.5 equiv), CH₂Cl₂,

0 °C to rt, 68%; (ix) 9a (1.1 equiv), Na₂CO₃(2 equiv), 50 °C, 16 h, 42%; (x) HCl (5 M in iPrOH), MeOH, rt, 24 h, 50%; (xi) LiOH (2 equiv),

EtOH/H₂O, rt, 24 h, 37%.

ligand in the active binding pose, has been estimated. The

maximum intrinsic eﬃcacy of 3a was smaller compared to 2a

and, in view of the inverse pose ratios, thus explaining the

resulting equal dynamic transduction coeﬃcients τ_dynof these

hybrids. Thus, the rational design of bipharmacophoric partial

agonists with a prevalent active binding pose was achieved for

the ﬁrst time in the case of muscarinic M₁acetylcholine

receptors.

RESULTS AND DISCUSSION

Chemistry. The quinolone skeleton (9) ﬁtted with varying

■

halogen substituents was built up using the Gould−Jacobs

synthetic pathway making use of microwave assistance to

facilitate conversion and provide high yields.⁴⁴Benzylation of

the quinolone nitrogen was achieved with benzyl chloride in

the presence of K₂CO₃and catalytic amounts of KI in DMF.

The ester function (10) was amidated with the aminoalkyl

D

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 4. Synthesis of M₁Allosteric Modulators Based on BQCA Naphthalene Derivatives

a

Reagents and conditions: (i) NaOH (2 equiv), iodomethane or 1-bromooctane (1.2−1.5 equiv), CH₃CN/H₂O, 65−70 °C, 16 h, 98%; (ii) Ph₃P

(1.6 equiv), imidazole (2.2 equiv), I₂(1.5 equiv), CH₂Cl₂, 0 °C to rt, 42−50%; (iii) 9a, Na₂CO₃(2 equiv), DMF, 50 °C, 16 h, 46%; (iv) LiOH (2

equiv), EtOH/H₂O, rt, 24 h, 48%.

a

Scheme 5. Synthesis of M₁TBPBd and the Corresponding 1/TBPBd Hybrids

a

Reagents and conditions: (i) NaH (60% in paraﬃn oil, 2.5 equiv), alkyl iodide (1.5 equiv), dry DMF, 0 °C to rt, 40%; (ii) Na₂CO₃(2 equiv), DMF,

50 °C, 16 h, 45%; (iii) HCl (5 M in iPrOH), MeOH, rt, 24 h, 76%; (iv) dimethyl sulfate or 1-bromooctane (1 equiv), Na₂CO₃(1.5 equiv), CH₃CN,

reﬂux, 24 h, 45%.

alcohol of the corresponding spacer length (C4 and C6) and

the hydroxyl function (11,12) substituted by a bromine atom

using HBr/H₂SO₄. Finally, iperoxo, synthesized according to a

method previously described,⁴⁵was connected to the spacer

(Scheme 1).

Because compounds 13 and 14 were subsequently applied as

reference substances in the pharmacological assays, their

inertness toward proteins were evaluated by incubation with

glutathione under the same conditions as applied in the

pharmacological assay system and proved no reaction as

determined by LC-MS (data not shown).

The orthosteric agonist 1 was synthesized in a modiﬁed way

as described in the literature:³⁹2-Bromobutyric acid (16) was

synthesized from 2-aminobutyric acid (15) by diazotization and

bromination in a Sandmeyer-like reaction. As described in the

literature, stereochemical inversion took place using potassium

ethyl xanthate. The conﬁguration of the absolute stereocenter

had been postulated based upon the mechanism of the reaction

and the results reported in the literature.^39,46Ring-closing was

performed using ammonium carbonate and 1-Boc-4-piper-

idone.³⁹After removing the BOC-group in acidic condition,

compound 1 was obtained (Scheme 2).

Synthesis of bipharmacophoric ligands from compound 1

and BQCA-derived compounds started from the synthesis of

the naphthalene-based linker intermediate (Scheme 3). 6-

Hydroxy-2-naphthoic acid (19) was esteriﬁed (20) and reduced

to the primary alcohol (21). Alkylation of this naphthol was

achieved via Williamson ether synthesis with an α,ω-

dihalogenoalkane followed by protection of the primary OH

group with TBDMSCl. BOC protected compound 1 (18)

could be alkylated at the lactam-N using alkyl iodide (24)

obtained after Finkelstein reaction of the naphthol derivative

(23). The TBDMS group was selectively removed in acetic

acid/water/THF 3:1:1. Appel reaction was applied to replace

the primary OH by an iodine atom.^47,48The alkyl iodide 27 was

in turn used to alkylate the ﬂuoroquinolone carboxylate 9a.

BOC and ethyl ester groups were sequentially removed in

acidic and basic conditions, respectively, to obtain the

bipharmacophoric ligands (4, Scheme 3). The corresponding

control compounds (33a,b), i.e., with the linker structures but

without the orthosteric moiety, were also synthesized following

the same synthetic strategy in order to evaluate the possible

contribution of the linker structure to the receptor binding

mode (Scheme 4).

E

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

In addition to connecting BQCA-derived compounds to the

M₁agonist compound 1, the N-piperidine benzimidazolone

structure representing the core pharmacophore of allosteric

agonist TBPB was also used to form a third series of

bipharmacophoric hybrids (Scheme 5). Allosteric agonists by

deﬁnition activate the receptor by binding to an allosteric site

without concomitant binding of an orthosteric agonist being

necessary for receptor activation.¹⁴First, alkylation of the

thiazolidinone-N 18 was achieved by using NaH in dry DMF

and an alkyl iodide, followed by reaction with N-piperidine

benzimidazolone structure 35, which was synthesized according

to a literature protocol.⁴⁹After BOC deprotection, the hybrids

(5a−c) were obtained. The corresponding control compounds

(37a,b), e.g., linker derived from the N-piperidine benzimida-

zolone structure, were synthesized by reacting the amine 35

with dimethyl sulfate or 1-bromooctane, respectively (Scheme

5).

DMR Measurements in hM₁-CHO Cells. Signaling of the

compounds was evaluated by means of DMR measurements

using live hM₁-CHO cells. The maximal system response was

deﬁned by the maximum response to saturating concentrations

of the endogenous ligand acetylcholine (ACh). ACh and the

M₂-superagonist iperoxo showed full agonism, whereas

pilocarpine could be categorized clearly as a partial agonist

(Figure 3C, Table 1). The functional aﬃnities of the

conventional agonists obtained according to the operational

model⁵⁰were in line with M₁aﬃnity measures published by

others, i.e., pK_ACh= 4.76 0.29 with pK_i= 4.90 0.11⁵¹and

pK_Pilo= 5.57

0.11 with pK_i,Pilo= 5.1.⁵²Applying the

operational model of agonism⁵⁰to the DMR data of iperoxo,

the transducer slope n was less than unity (F-test, P < 0.05).

The functional aﬃnity and the estimated operational eﬃcacy of

iperoxo at M₁receptors amounted to pK_iperoxo= 5.67 0.65

and log τ_iperoxo= 2.99

0.65, respectively (Table 1). The

Figure 3. Iperoxo/BQCAd hybrids (2a−c, 3a−d) and 1/BQCAd (4a,

4b) ligands act as partial M₁agonists. Concentration−eﬀect curves of

positive DMR after 2400 s generated by the indicated orthosteric

agonists and test compounds in live CHO-hM₁cells. The data points

in (A,B) and of 4b in (C) were ﬁtted to the operational model of

agonism for dynamic ligands,³⁰and the data of iperoxo, acetylcholine,

and pilocarpine in (C) were ﬁtted to the operational model of

agonism.⁵⁰The bottom and E_maxvalues were constrained to 0 and

100%, respectively. Logistic curve ﬁtting with variable slope was

applied to the data points of 37b. The DMR of solvent controls were

set to 0% and the maximal positive DMR response in the presence of

100 μM acetylcholine to 100%. Data are means SEM from 3−12

independent experiments conducted in quadruplicate.

numerical estimate of τ_iperoxoindicates how far the concen-

tration−eﬀect curve of iperoxo in functional experiments is

shifted to the left compared with its corresponding receptor

binding curve. log τ_iperoxowas considerably though not

signiﬁcantly higher than log τ_{acetylcholine}= 1.64 0.28 (Table

1; t test, P > 0.05). Thus, more-than-physiological agonism

(“superagonism”), ﬁrst demonstrated for iperoxo at the even-

numbered, G_i-coupled M₂-subtype,¹⁹was not encountered at

the odd-numbered, G_q-coupled M₁subtype.

Potency values pEC₅₀of acetylcholine and iperoxo (Table 1)

were far higher than their functional binding aﬃnities pK. This

ﬁnding reﬂected signal ampliﬁcation by the assay system, also

addressed as “receptor reserve”. In the case of pilocarpine,

however, pEC₅₀and pK were rather similar, and activation of

the test system was far below the 100% level deﬁned by the

maximum eﬀect of endogenous activator acetylcholine. Thus,

pilocarpine behaved as a partial agonist.

EPIC system applied in the current study may act as a full

agonist in another system with higher signal ampliﬁcation. In

the classical operational model of agonism⁵⁰and the opera-

tional model of agonism for dynamic ligands,³⁰coupling

Structure−Signaling Relationships in Iperoxo/BQCAd-

Type 2 and 3 Hybrids. Concentration−eﬀect curves for

DMR of live hM₁-CHO cells as induced by the newly

synthesized hybrid compounds 2 and 3 are shown in parts A

and B of Figure 3, respectively. Data were analyzed by a four-

parameter logistic equation (allowing a model-independent

comparison of all compounds studied), and by means of the

operational model of agonism for dynamic ligands,³⁰the

parameter values are summarized in Table 1. Like pilocarpine,

the new test compounds 2 and 3 were partial agonists with E_max

values of about 60%. Only compound 3b reached a higher E_max

of 80%. However, we are aware that a partial agonist in the

eﬃciency is expressed by the parameters τ and τ_dyn

,

respectively. The dynamic eﬃcacies τ_dynof the test compounds

ranged between log τ_dyn,2b= 0.09 0.03 and log τ_dyn,3b= 0.59

0.04 (Table 1). All iperoxo/BQCAd compounds (2, 3)

showed submaximum eﬀects (Figure 3A,B), and their τ_dyn

values were similar to τ of pilocarpine (Table 1). Thus, as

intended by the rational design, the hybrid ligands 2, 3 behaved

as partial agonists. The diﬀerence in potencies (pEC₅₀) among

members of the subgroup iperoxo/BQCAd hybrids 2a−c

amounted to 0.3 log units only. Obviously, the substitution

pattern at the quinolone moiety had no relevant impact on

hybrid potency (one-way ANOVA with Newman−Keuls

F

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Table 1. Measures of Potency and Eﬃcacy of DMR Induced by Muscarinic Agonists and Hybrid Compounds in Live CHO-hM₁

a

Cells

compd

N

R¹

R²

R³

pEC₅₀

%E_max

100

n_op

pK

log τ_dyn

n_op

N

acetylcholine

iperoxo

6.43 0.04

8.69 0.06

6.08 0.08

2

0.82 0.07*

0.54 0.04*

1

4.76 0.29^†

5.67 0.65^†

5.57 0.11^†

1.64 0.28^#

2.99 0.65^#

0.35 0.06^#

0.80 0.07*

0.52 0.03*

1

5

6

3

102

69

pilocarpine

3

2a

2b

2c

4

6

F

H

F

7.24 0.09

7.20 0.10

7.49 0.11

6.75 0.09

7.81 0.13

6.52 0.20

7.23 0.09

6.27 0.04

nd^‡

62

55

64

80

64

61

63

2

6

5

2

1

6.82 0.10

6.85 0.11

7.14 0.13

6.31 0.11

7.13 0.11

6.07 0.25

6.82 0.27

nd^‡

0.21 0.03

0.09 0.03

0.25 0.05

0.59 0.04

0.26 0.11

0.20 0.09

nd^‡

1

3

5

4

3

4

8

4

H

F

Br

Cl

H

1

3a

3b

3c

H

F

1

H

Br

Cl

H

1

3d

37b

4a

4b

F

1

3.64 0.59*

nd^‡

1

nd^‡

1

nd^‡

7.22 0.09

64

3

6.78 0.11

0.25 0.05

BQCA

5.08 0.14

91 14

1

4.26 1.32

0.95 1.13

1.33 0.55

5

a

pEC₅₀, concentration of the indicated compounds inducing a half-maximal DMR eﬀect (−log EC₅₀values); %E_max, maximum eﬀect as a percentage

of E_ACh(100 μM); n, slope factor obtained by curve ﬁtting to data from individual experiments shown in Figure 3A,B using a four-parameter logistic

equation; n_op, transducer slope applying the operational model of agonism;^30,50*, signiﬁcantly diﬀerent from unity (F-test, P < 0.05); log τ_dyn

,

operational eﬃcacy; pK = −log K_D(see Experimental Section, Data Analysis, eq 1) measures of functional aﬃnities were derived from DMR data

according to the operational model of agonism for dynamic ligands;³⁰† value is pK = − log K; # value is log τ, according to the operational model of

agonism;⁵⁰nd, not determined; ‡, ﬁt was ambiguous. Data are means

determinations.

SEM of N independent experiments conducted as quadruplicate

multiple comparison test, P > 0.05). In contrast, among the

iperoxo/BQCAd hybrids 3a−d containing the elongated C6-

linker, diﬀerences in pEC₅₀values of 1.3 log units were

observed, indicating that the substitution pattern aﬀected ligand

potency. Furthermore, potencies of hybrids with a C4- (2a−c)

versus a C6-spacer (3a−d) were lower or higher depending on

the pattern of substitution in the benzyl quinolone moiety. The

diﬀerence in pEC₅₀reached statistical signiﬁcance within the C6

series with 3a vs 3b and 3d, 3b vs 3c and 3d, 3c vs 3d and

between the C4 and C6 series with 2a vs 3a, 3b, and 3c as well

as 2b vs 3a, 3b, and 3c and 2c vs 3a and 3c. With respect to the

diﬀerences in E_max, statistical signiﬁcance was observed within

the C6 series with 3b vs 3a, 3c, and 3d, and between the C4 and

C6 series with 2a vs 3b, 2b vs 3b, and 2c vs 3b (one-way

ANOVA with Newman−Keuls multiple comparison test, P <

0.05). It is worth mentioning that test compound-induced

DMR was blocked by the muscarinic antagonist atropine, 10

μM, and did not occur in “empty” live CHO K1 cells lacking

the M₁receptor, thus demonstrating that eﬀects were mediated

by the M₁receptor (data not shown). Taken together, changing

the length of the hybrid spacer in conjunction with the pattern

of substitution allowed showing that dynamic ligand binding

tunes the eﬃcacy of iperoxo/BQCAd-type 2 and 3 hybrids in

hM₁-CHO cells.

Structure−Signaling Relationships in 1/BQCAd-Type

Hybrids 4. Compared to DMR induced by ACh (see above),

1, which had been claimed to be an M₁agonist,³⁹only showed

a weak maximum eﬀect of 12 3% at a concentration of 10 μM

(Figure 4C, Supporting Information, Figure 1). Nevertheless,

we used 1 as an orthosteric moiety in our hybrid molecules. As

can be seen in Figure 4C carried out as a comparison of the

eﬀects E at a single concentration (the respective highest

dissolvable compound concentration with less than 1.3%

DMSO in the assay, cf. Materials and Methods for further

details) linkage of 1 with the ineﬃcacious allosteric moieties

33a (0.3 μM: E = 1 1%) and 33b (1 μM: E = 1 1%)

yielded the 1/BQCAd hybrids 4a (10 μM: E = 55 11%) and

4b (10 μM: E = 61

9%). The concentration−eﬀect

relationship of 4b (Figure 3C, Table 1) showed “normal”

partial agonism with a submaximum eﬀect and a τ_dynvalue

similar to τ of pilocarpine (Figure 3C, Table 1). Thus, as

intended by the rational design, the hybrid ligand 4b behaved as

a partial agonist. The 1/BQCAd hybrid 4a at a rather high

concentration of 10 μM induced a brisk M₁receptor mediated

DMR response (Figure 3C) which could be abolished by 10

μM atropine (data not shown). As illustrated in Figure 3C, 4a

had a steep concentration−eﬀect relationship, suggesting a

peculiar, unknown mechanism of agonism and could not be

ﬁtted unambiguously by applying the operational model of

agonism (Table 1). Compound 4a had a very low potency with

an eﬀect only at 10 μM (higher concentrations were not

applicable due to poor solubility).

Structure−Signaling Relationships in 1/TBPDd-Type 5

Hybrids. The 1/TBPBd-type hybrids 5a, 5b, and 5c lacked

any DMR eﬀect; they were inactive (Supporting Information,

Figure 1). Interestingly, the TBPBd 37b triggered a very steep

response, which, however, leveled oﬀ at about 50%, but three

data points deﬁned an upper plateau, thus allowing derivation

of a potency, a maximum eﬀect, and a slope factor. (Figure 3C,

Table 1). The steep concentration−eﬀect relationship of 37b

suggested a peculiar, unkown mechanism of agonism and could

not be ﬁtted unambiguously applying the operational model of

agonism (Table 1). The eﬀect of 37b was likely to occur

allosterically because 37b retarded the dissociation of the

orthosteric radioligand [³H]NMS from M₁receptors in live

CHO-hM₁cells (Supporting Information, Figure 2). The half-

maximal inhibitory action of 37b with regard to allosteric

inhibition of [³H]NMS dissociation, pEC_0.5,diss,37b= 5.94

0.08, N = 3, occurred in the micromolar concentration range.

The maximum percent reduction of k₋₁by 37b was observed at

36 3% (Supporting Information, Figure 2). The latter was

equivalent to a 3-fold retardation of [³H]NMS dissociation

G

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

the radioligand.⁵³In conclusion, the agonist action of 37b

involves the allosteric site in hM₁receptors. Taken together, the

series of 1-derived bipharmacophoric compounds revealed

complexities which are currently not fully understood on the

molecular level and will be investigated further.

Ineﬀectivity of Iperoxo/BQCAd-Type 2, 3 Hybrids Lacking

Their Iperoxo Moiety. The partial- to full-agonistic action that

had been reported for the parent compound BQCA³⁴⁻³⁷was

also found in the current study (Figure 3C, Table 1). To rule

out the possibility that the respective allosteric building blocks,

i.e., bromoalkylamides of the benzyl quinolones, were

responsible for receptor activation by the iperoxo/BQCAd

hybrids (2, 3), we checked whether they had a DMR eﬀect on

their own. The compounds 13a, 13b, 13c, 14a, 14b, and 14c

and the prototypal BQCA at a concentration of 1 μM did not

produce a DMR signal (Figure 4A, one sample t test, P > 0.05).

13a and 14a, the allosteric building blocks of the hybrids 2a and

3a, respectively, which were studied in greater detail (see

below), both at 1 μM reached the maximum level of allosteric

inhibition of [³H]NMS binding (Figure 5B,D). This bottom

level is above zero, suggesting allosteric ternary complex

formation including receptor, radioligand, and test compound.

A closer look at the concentration-dependent eﬀects of 13b,c

and 14b,c included in Figure 4A revealed slightly smaller

binding aﬃnities pK_inactivecompared to 13a and 14a but a

similar submaximum [³H]NMS displacement at high concen-

trations of 3 and 10 μM, respectively (Supporting Information,

Figure 3A,B). This reﬂects moderately negative cooperativity

between 13a,b,c and 14a,b,c and the orthosteric [³H]NMS

with cooperativity factors ranging between log α′_14a= −0.82

0.04 and log α′_14c= −1.35

0.57 (13a, 14a, Table 2, and

13b,c, 14b,c, Supporting Information, Figure 3, respectively).

Since it might be argued that 13a,b,c and 14a,b,c nevertheless

may bear the allosteric agonistic potential of the prototypal

BQCA but at higher concentrations, we aimed at increasing the

concentrations studied in the EPIC system. However, 13a,b,c

and 14a,b,c had to be dissolved in DMSO, and the DMSO

content of their dilutions in buﬀer prevented studying higher

concentrations than 1 μM in the EPIC system due to a solvent

eﬀect of DMSO (for further details see Materials and

Methods). Because M₁receptor-mediated IP1 accumulation

in CHO-hM₁cells is less sensitive to DMSO compared to the

signal generated in the EPIC system, it allowed us to increase

the concentration of 13a,b,c and 14a,b,c up to a concentration

of 10 μM. In the IP1 assay, iperoxo, acetylcholine, 2a, and 3a,

but not BQCA showed somewhat higher potencies but very

similar eﬃcacies (Supporting Information, Figure 4) compared

to the respective parameters obtained in the EPIC sytem

(Table 1). Thus, the potency and eﬃcacy proﬁles of these

ligands in both assay systems were similar. BQCA was a full

agonist in the EPIC system and a partial agonist in the IP-1

assay. In Supporting Information, Figure 5, it can be seen that

in the IP1-assay, BQCA (10 μM) was signiﬁcantly active but

13a,b,c and 14a,b,c remained inactive. Taken together, the

agonistic action that had been reported for the parent

compound BQCA³⁴⁻³⁷was not shared by our BQCA

derivatives. Most importantly in the present context, the

absence of an agonist action of 13a,b,c and 14a,b,c allowed us

to apply the operational model of agonism for dynamic

ligands³⁰in order to check whether eﬃcacy is related to binding

pose. This was studied exemplarily with the two selected

iperoxo/BQCAd hybrids 2a and 3a (see below).

Figure 4. M₁agonist eﬃcacy does not depend on (A) the allosteric

moieties (13, 14) in iperoxo/BQCAd-prototype hybrids (2, 3),

whereas it depends (C) on “hybridization” in 1/BQCAd hybrids (4)

and is (B,C) mainly mediated via G_qproteins. DMR in live CHO-hM₁

cells generated after 2400 s by the indicated orthosteric agonists and

test compounds under (A) control conditions and, additionally (B,C),

after pretreatment with PTX (100 ng/mL, overnight) for inhibition of

G_i- or UQ (1 μM, 1.5 h prior to the DMR measurement) for

inhibition of G_qcoupling, respectively. The DMR were solvent-

corrected and expressed as percent of the maximal positive DMR

response in the presence of 100 μM acetylcholine. n.s., not signiﬁcant,

one sample t-test, P > 0.05. Data are means

independent experiments conducted in quadruplicate.

SEM from 3−12

relative to control conditions (t_1/2,diss= 8.9 0.3 min, N = 21)

by 37b. This behavior is undoubtedly indicative of an allosteric

action because the orthosteric M₁receptor site is occupied by

H

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 5. (A) Allosteric eﬀects of the hybrids (2a, 3a) and their corresponding allosteric moieties (13a, 14a) as reﬂected by the inhibition of

[³H]NMS dissociation from M₁receptors in live CHO-hM₁cells. The allosteric eﬀect is expressed as the ratio of the rate of [³H]NMS dissociation in

the presence of compound (k_obs) relative to the rate of dissociation of [³H]NMS in the absence of test compound (k₀), expressed in percent. Iperoxo

at a high concentration did not retard [³H]NMS dissociation (B−F). Switch of the fractional active binding pose of the iperoxo/BQCAd-hybrid 2a

and 3a derived from global data analysis by the operational model of agonism for dynamic ligands. (B,D) Eﬀect of increasing concentrations of ACh

and of the respective hybrid in percent DMR (left ordinate, data taken from Figure 3A) and of the respective hybrid and its allosteric moiety on

speciﬁc percent equilibrium binding of [³H]NMS (right ordinate) in live CHO-hM₁cells. (C,E) Concentration-dependent fractional receptor

occupancy of the receptor active and inactive binding poses of 2a and 3a. Global curve ﬁtting to sets of four curves included in (B,D) eq 3, for details,

cf. text and Experimental Section. Data in A−E represent mean SEM from 3−4 independent experiments, conducted in quadruplicate. (F) Note

the signiﬁcant switch of the fractional receptor occupancy to the active hybrid binding pose (means 95% CI) upon spacer elongation between 2a

and 3a (t-test, P < 0.001, see text for discussion).

Eﬀects of Iperoxo/BQCAd-Type 2, 3 and 1/BQCAd-Type 4

Hybrids Are Predominantely Mediated by G_qProteins. In the

presence of the G_qprotein inhibitor UBO-QIC (alternatively

called UQ, FR900359, structure cf. Supporting Information)

(Figure 4B,C), no positive DMR signal could be detected for

any of the compounds; merely a tiny negative DMR signal

remained which may represent a G_s-coupling component.

Thus, all active hybrid molecules behaved as partial agonists

characterized by a %E_maxvalue between 55 and 80% (Figure 3,

Table 1) and activated predominantly the G_q-signaling pathway

(Figure 4B,C). In the presence of the G_iprotein inhibitor

pertussis toxin (PTX), M₁receptor-mediated signaling of

I

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Table 2. Measures Obtained by Global Simultaneous Analysis of Hybrid Functional and Binding Data with the Operational

a

Model of Agonism for Dynamic Ligands in Live CHO-hM₁Cells

compd

N

pK_active

pK_inactive

log τ_dyn

R_pose

log ε_max*

2a

3a

4

6

6.64 0.15

6.98 0.06

6.92 0.04

6.30 0.05

0.15 0.02

0.08 0.01

−0.28 0.19

0.68 0.09

0.61 0.13

0.17 0.02

a

N, spacer length between the orthosteric and the allosteric moiety; pK_active, pK_inactive, (−log) equilibrium dissociation constants of the dynamic

ligand for the active pose and for the inactive receptor binding pose, respectively. pK_activeequals pK_A; note that pK_inactiveequals the equilibrium

dissociation constant pK_Bof the respective allosteric moiety (13a, 14a) at the free receptor; log τ_dyn, log dynamic transduction coeﬃcient of the

ligand; R_pose, −log K_active/K_inactive, i.e., the −log ratio of the equilibrium dissociation constants of the active and the inactive receptor binding pose;

ε_max*, system-independent maximum intrinsic eﬃcacy of the dynamic ligand at 100% occupancy in the active pose in the absence of receptor reserve.

Data are parameter estimates SEM of global ﬁts to CECs (1−4) (cf. Experimental Section) characterizing the respective hybrid each carried out as

3−4 independent experiments conducted as quadruplicate determinations. log ε_max* was determined in separate analyses (involving eq 7, cf.

Discussion, not included in Figure 5) in which log τ_dynwas replaced by ε_maxand R_T/K_E(involving eq 6, cf. Experimental Section). For 2a and 3a,

global curve ﬁtting with R_T/K_E= 1 = constant yielded ε_max*. For further details, see Results, Discussion, and Experimental sections.

receptor saturating concentrations of ACh, iperoxo, and

pilocarpine (100 μM, Figure 4B) as well as high concentrations

of the hybrids (10 μM, Figure 4B,C) was hardly aﬀected.

Interestingly, the DMR signal of the allosteric moiety 37b

was likewise mainly G_q-mediated (Figure 4C); this may suggest

that this signaling pathway can also be activated through the

allosteric binding area of M₁receptors.

hybrids and their allosteric moieties but not iperoxo bound to

the allosteric site of M₁receptors in live CHO-hM₁cells.

Because the allosteric moieties 13a and 14a on their own did

not show an agonistic action (Figure 4A and Supporting

Information, Figure 5), whereas the orthosteric iperoxo did

(Figure 3B and Supporting Information, Figure 4), we

concluded a dualsteric M₁receptor interaction of 2a and 3a.

Finally, we looked for a potential PAM eﬀect of 2a, 3a, 13a, and

14a on acetylcholine binding to hM₁receptors. From the

inspection of the insets of Supporting Information, Figure 6A−

D, it can be seen that neither of them induced a left shift of the

inhibitory eﬀect of acetylcholine on speciﬁc [³H]NMS

equilibrium binding to hM₁-receptors; instead, at compound

concentrations indicating their strong negative cooperativity

with NMS by the reduction of the upper plateau of the

respective [³H]NMS control curves (Supporting Information

Figure 6A−D), 2a, 3a, 13a, and 14a signiﬁcantly reduced the

pIC_{50,acetylcholine}, suggesting a negative cooperativity with

acetylcholine (F-test, P < 0.05). Therefore, we checked for

dynamic ligand binding by applying a corresponding model

reported recently.³⁰Sets of four curves as shown in Figure 5B,D

(i.e., ACh-induced DMR, hybrid-induced DMR, [³H]NMS

equilibrium displacement by the hybrid and by its respective

allosteric moieties) were subjected to the “ﬁrst” global

nonlinear regression analysis including the operational model

of agonism for dynamic ligands (for ﬁtting details see

Experimental Section). This yielded numerical estimates for

the parameters pK_activeand pK_inactiveof hybrid binding in the

active and inactive binding pose, respectively and, moreover, for

τ_dyn, n_OP(transducer slope of eq 1), α′_hybr, α′_frag, and n_frag(slope

of eq 5) The novel global approach introduced here joins

functional and binding data, allowing pK_activeto be estimated

from function and binding simultaneously and does neither

predeﬁne pK_inactivenor the cooperativity α′ of the hybrid with

[³H]NMS by that of its allosteric moiety (for further details cf.

Materials and Methods, data analysis, and Supporting

Information, Figure 7). The rationale of this approach was

that global ﬁtting with shared aﬃnities pK_activeand pK_inactivebut

with individual cooperativities log α′_hybrand log α′_fragof the

dynamic ligand and its allosteric fragment with [³H]NMS,

respectively, would come closer to reality than ﬁxing pK_inactive

and log α′_hybrto those values obtained in [³H]NMS

displacement experiments with the allosteric fragment. Curve

ﬁtting revealed the curve slopes n_fragof 13a and 14a (global

approach, Figure 5B,D, Supporting Information, Table 1) and

of 13b and 14b but not of 13c and 14c (cf. single curve

analyses, Supporting Information, Figure 3) to be signiﬁcantly

Fractions of Iperoxo/BQCAd-Type 2, 3 Hybrid Binding

Poses Estimated by Means of the Operational Model of

Agonism for Dynamic Ligands. To quantify the percentage of

active receptor pose of selected iperoxo/BQCAd hybrids, we

additionally performed binding experiments with both hybrids

and their allosteric moieties (13a and 14a, respectively) in live

CHO-hM₁cells using [³H]NMS as the orthosteric radioactive

probe. We chose the pair 2a and 3a that diﬀer structurally in

spacer length and pharmacologically solely in potency (i.e.,

pEC₅₀) (one-way ANOVA with Newman−Keuls multiple

comparison test, P < 0.05, cf. Table 1) but not with respect

to E_maxand log τ_dyn. Preconditions for the applicability of the

operational model of agonism for dynamic ligands³⁰to hybrid

data are an agonistic eﬀect of its orthosteric but not of its

allosteric moiety. Furthermore, to be capable of adopting a

dualsteric receptor binding pose, the hybrid and its

corresponding allosteric moiety should be able to interact

allosterically with orthosterically blocked receptors; on the

other hand, iperoxo should not. To check for an interaction

with the allosteric site in [³H]NMS-occupied M₁-receptors, we

measured the eﬀect of the hybrids 2a and 3a, of their allosteric

moieties 13a and 14a, and of iperoxo on the apparent rate

constant k₋₁of radioligand dissociation (Figure 5A). The half-

maximal inhibitory concentration with regard to the rate

constant k₋₁of [³H]NMS dissociation (pEC_0.5,diss) of both the

hybrids 2a (pEC_0.5,diss= 6.04 0.08, N = 3) and 3a (pEC_0.5,diss

= 6.10

0.07, N = 3) and their allosteric moieties 13a

(pEC_0.5,diss= 5.90 0.13, N = 3) and 14a (pEC_0.5,diss= 6.15

0.21, N = 3) occurred at micromolar concentrations. The

corresponding maximum percent reduction of k₋₁by the

compounds were observed at 2a = 18 3, 3a = 15 3, 13a =

35 4, and 14a = 47 5 (cf. Figure 5A). The latter were

equivalent to a 7- and 2-fold retardation, by the hybrids or

allosteric moieties, respectively, of [³H]NMS dissociation under

control conditions. The curves showed slopes not diﬀerent

from unity (F-test, P > 0.05). No compound blocked [³H]NMS

dissociation completely. Yet, a complete block of [³H]NMS

dissociation is not a prerequisite to identify an allosteric

interaction of a compound with a receptor site because an

allosteric ligand may reduce but not fully inhibit the probability

of orthosteric radioligand dissociation. Taken together, these

J

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

larger than unity (F-test, P < 0.05). Currently, we have no

explanation for the high steepness.

muscarinic receptor subtype selectivity of eﬀects shown by the

BQCAd-hybrid compounds. In M₁receptors, employing a

bitopic orthosteric/allosteric design concept the appropriate

choice of building blocks allows to set signaling to the desired

level of eﬃcacy.

The estimates of pK_activeand pK_inactive(Table 2) served to

calculate the fractional receptor occupancy in the dualsteric and

the allosteric pose, respectively (Figure 5C,E) and to derive

corresponding R_posevalues (for details see Experimental

Section). A complete compilation of the parameters, their

numerical estimates and treatment in the global ﬁtting

procedure is given in Supporting Information, Table 1. It

became obvious that increasing the spacer length from C4 in 2a

to C6 in 3a reduced binding in the inactive pose and, vice versa,

augments binding in the active receptor pose (cf. Figure 5C,E).

The diﬀerence in R_posevalues of 2a and 3a was statistically

signiﬁcant (t-test, P < 0.05, Figure 5F, Table 2).

To show that the resulting R_posevalues are robust, we

analyzed the dynamic ligands 2a and 3a with a “second” version

of the global ﬁt where α′_hybrand α′_fragwere treated as shared

variables and n_fragwas constrained to unity (Supporting

Information, Table 2 and Figure 7C,E). Most importantly,

the signiﬁcant switch in R_posebetween compounds 2a and 3a

was maintained (t-test, P < 0.05), although the resulting values

were slightly diﬀerent (Supporting Information, Figure 7F).

The Intrinsic Receptor Activity of 2a and 3a Bound in the

Active Pose Plus the Fraction of Active Pose Govern Their

Maximum Eﬀect. In view of the diﬀerence in R_posevalues

between 2a and 3a (Table 2), i.e., the diﬀering fractions of

active pose, the question arose why both compounds 2a and 3a

CONCLUSIONS

■

Three series of hybrid compounds have been synthesized and

studied, and the ﬁndings reported here clearly demonstrate that

bipharmacophoric iperoxo/BQCAd (2, 3) and 1/BQCAd (4)

but not 1/TBPBd (5) hybrid ligands are able to act as partial

hM₁receptor agonists. The eﬀects of the former have

pilocarpine-like eﬃcacy and functional aﬃnity and are mediated

mainly through the physiologically relevant G_q-signaling

pathway. We conclude that the hybrid approach to generate

partial M₁receptor agonists is sound as the ﬁndings suggested

that hybrid potency and eﬃcacy depended on the spacer length

between the orthosteric and allosteric moieties in conjunction

with the pattern of substitution in the allosteric moieties. In

addition, increasing the spacer length between the orthosteric

and allosteric moieties of selected hybrids in the iperoxo/

BQCAd series (2, 3) enhanced the fractional receptor

occupancy in the receptor active binding pose and lowered

the system-independent maximum intrinsic eﬃcacy of the

dynamic ligand ε_maxin the active binding pose while preserving

partial agonist activity. Thus, M₁partial agonism can be

designed by bipharmacophoric hybrid formation and ﬁne-tuned

with respect to the fraction of the M₁receptor active hybrid

binding pose and its eﬃcacy. The combination of bipharmaco-

phoric partial agonist design and active receptor binding pose

tuning represents an important new step of current GPCR drug

design and may, in muscarinic M₁receptors, become

therapeutically exploitable, e.g., in states of AD.

shared the same maximum eﬀect (cf. Table 1: %E_max62

2

and 64 2, respectively, corresponding to operational eﬃcacies

τ_dyn0.21 0.03 and 0.25 0.05). In this context, it is necessary

to note that signaling of the bitopic compounds does not only

depend on the probability of binding in the signaling-

competent pose but also on the intrinsic eﬃcacy generated

by a ligand in the active pose.⁵⁴

Therefore, the parameter τ_dynwas resolved into the system-

independent maximum intrinsic eﬃcacy of the dynamic ligand

ε_max(corresponding to the activity of the individual ligand/

receptor complex with agonist bound in the active pose) and

R_t/K_Eas a measure of the receptor reserve (for further details,

see Experimental Section). Because the cells and the conditions

applied in the experiments with hybrids 2a and 3a were

identical, we constrained for a comparison of their ε_maxvalues

the parameter R_t/K_Eto one same value in the respective global

ﬁts. ε_maxobtained with R_t/K_E= 1 = constant equals τ_maxin the

absence of a receptor reserve (i.e., R_t/K_E= 1; cf. Experimental

Section) and was termed ε_max*. These ε_max* values allowed a

comparison of 2a and 3a with regard to the system-

independent maximum intrinsic eﬃcacy of the dynamic ligand

at 100% occupancy of receptors in the active pose (Table 2).

Because the diﬀerence in ε_max* is signiﬁcant (t test, P < 0.05),

the ﬁndings indicated that compound 2a surpasses 3a with

respect to intrinsic eﬃcacy in the signaling competent

dualsteric binding pose. This explained why the overall dynamic

transduction coeﬃcient τ_dynfor DMR did not diﬀer between 2a

and 3a (Table 1). Furthermore, analysis of ligand eﬃcacy

beyond the E_maxof the measured response, i.e., the

consideration of the agonist−receptor complex related eﬃcacy

ε_max* in combination with the probability of ligand receptor

binding in the signaling-competent pose oﬀers an additional

way to diﬀerentiate ligand eﬃcacy in a graded fashion.

EXPERIMENTAL SECTION

Chemistry. General Methods for Synthesis. Melting points are

■

uncorrected and were measured in open capillary tubes, using a

Barnstead Electrothermal IA9100 melting point apparatus. ¹H and ¹³

C

NMR spectral data were obtained from a Bruker Advance

spectrometer (300 and 75 MHz, respectively). TLC was performed

on silica gel on aluminum foils with ﬂuorescent indicator 254 nm

(Fluka) or aluminum oxide on TLC-PET foils with ﬂuorescent

indicator 254 nm (Fluka). For detection iodine vapor or UV light (254

nm) were used. ESI-MS samples were analyzed using electrospray

ionization ion-trap mass spectrometry in nanospray mode using a

Thermo Finnigan LCQ Deca. For column chromatography, silica gel

60, 230−400 mesh (Merck) was used.

The purities of the ﬁnal new compounds (2a−c, 3a−d) were

determined using capillary electrophoresis and were found to be

≥95%. The CE analyses were performed on a Beckman Coulter P/

ACE System MDQ (Fullerton, CA, USA), equipped with an UV-

detector measuring at 210 nm, using a fused silica capillary (eﬀective

length 40 cm, total length 50.2 cm, diameter 50 μm) and as a running

buﬀer 50 mM aqueous sodium borate, pH 10.5. Prior to use, aqueous

solutions were ﬁltered through a 0.22 μm pore-size CME (cellulose

mix ester) ﬁlter (Carl Roth GmbH, Karlsruhe, Germany). The

aqueous buﬀer was prepared using ultrapure Milli-Q water (Millipore,

Milford, MA, USA). Purity of all other target compounds (4a,b, 5a−c)

was evaluated either by elemental analysis or by high-resolution mass

following HPLC system (conﬁrming purity ≥95%). Analytical HPLC

using a VWR Hitachi L-2130 pump coupled to VWR Hitachi column

oven L-2350 and L-2455 diode array detector. The solvents were as

follows: (A) water + 0.05% triﬂuoroacetic acid and (B) acetonitrile +

0.05% triﬂuoroacetic acid; ﬂow, 0.4 mL/min. Column Hibra 125-4

Purospher STAR RP-18e (3 μm) at 20 °C, detecting at 249 nm;

To summarize, the feasibility of dynamic ligand binding to

achieve designed partial agonism at muscarinic M₁receptors is

shown for the ﬁrst time. Future studies will address the

K

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

solvent A from 90% to 10% for 30 min, then 10% for 15 min, from

10% to 90% for 10 min, 90% for 5 min.

h. The mixture was cooled to less than 10 °C using ice/water bath, and

water (30 mL) was added to the mixture dropwise to quench the

reaction. The mixture was then extracted with ethyl acetate (3 × 30

mL). The organic phases were combined, dried over sodium sulfate

anhydrous, and concentrated under vacuum to get 21 as oﬀ-white solid

(3.4 g, 74%).

(6-(3-Chloropropoxy)naphthalen-2-yl)methanol (22a). The mix-

ture of 21 (300 mg, 1.72 mmol), 1-chloro-3-iodopropane (204 μL,

1.90 mmol), sodium hydroxide (40 mg, 1.9 mmol), water (2 mL), and

acetonitrile (5 mL) was heated at 65 °C overnight. Ethyl acetate (20

mL) and water (20 mL) were added to the reaction mixture, the

organic phase was separated, and the aqueous phase was extracted with

(2 × 20 mL) ethyl acetate. The organic phases were combined, dried

over sodium sulfate anhydrous, and concentrated under vacuum. The

residue was puriﬁed via column chromatography using petroleum

ether:ethyl acetate 3:1 as eluent system. The target compound 22a was

yielded as yellow solid (310 mg, 72%).

tert-Butyl((6-(3-chloropropoxy)naphthalen-2-yl)methoxy)-

dimethylsilane (23a). To the solution of 22a (310 mg, 1.2 mmol),

triethylamine (258 μL, 1.8 mmol), and catalytic amount of DMAP (15

mg, 0.12 mmol) in diethyl ether (20 mL) was added tert-

butyldimethylchlorosilane (217 mg, 1.44 mmol), and the mixture

was agitated overnight at room temperature. Water (30 mL) and

diethyl ether (20 mL) were added to the reaction mixture. The organic

phase was separated, and the aqueous phase was extracted with diethyl

ether (2 × 20 mL). The organic phases were combined, washed with

saturated ammonia chloride solution, dried over sodium sulfate

anhydrous, and concentrated under vacuum. The residue was used

directly in the next step without further puriﬁcation.

(S)-1-((6-(3-(2-Ethyl-3-oxo-1-thia-4,8-diazaspiro[4.5]decan-4-yl)-

propoxy)naphthalen-2-yl)methyl)-8-ﬂuoro-4-oxo-1,4-dihydroqui-

noline-3-carboxylic Acid (4a). To the solution of 29a (40 mg, 0.06

mmol) in tetrahydrofuran (10 mL) and water (5 mL) was added

lithium hydroxide (5 mg). The resulting solution was stirred at room

temperature for 24 h. The mixture was frozen under liquid nitrogen

and concentrated under high vacuum to yield the target compound 4a

as yellow grease (30 mg, 60%).

General Procedure for the Synthesis of the Diethyl Anilinome-

thylenemalonates 8a−d. According to the method described by

Leyva et al.⁴⁴and de la Cruz et al.,⁵⁵30 mmol of the corresponding

aniline derivative and 36 mmol of diethyl ethoxymethylenemalonate

were suspended in toluene and reﬂuxed for 6−24 h. Then, the solvent

was evaporated in vacuo, and 200 mL of pentane was added to the

residue. The solution was kept at 4 °C for several hours. The crystals

were collected, washed with pentane, and dried in vacuo.

General Procedure for the Synthesis of Compounds 9a−d. First,

40 mmol of the diethyl anilinomethylenemalonates 8a−d, respectively,

were dissolved in 100 mL of diphenylether and reﬂuxed for 2 h. After

cooling to room temperature, the precipitate was then ﬁltered several

times, washed with petroleum ether, and dried in vacuo.

General Procedure for N¹-Benzylation of 10a−d. First, 20 mmol

of the quinolone derivatives 9 were suspended in 10 mL of DMF, 48

mmol of K₂CO₃and 100 mmol of benzyl chloride were added, and the

solution was kept at 80 °C for 20 h. After cooling to room

temperature, the surplus of K₂CO₃was then ﬁltered and the solvents

evaporated. The oily residue was crystallized from ethanol and the

precipitate collected by ﬁltration and dried in vacuo.

(S)-tert-Butyl 4-(3-(6-((tert-Butyldimethylsilyloxy)methyl)-

naphthalen-2-yloxy)propyl)-2-ethyl-3-oxo-1-thia-4,8-diazaspiro-

[4.5]decane-8-carboxylate (25a). A solution of compound 23a and

sodium iodide (216 mg, 1.44 mmol) in acetone (20 mL) was reﬂuxed

overnight. The precipitate was ﬁltered, and the ﬁltrate was

concentrated under vacuum to get compound 24a. To the suspension

of sodium hydride (60% in paraﬃn oil, 86 mg, 2.16 mmol) in dry

N,N′-dimethylformamide (5 mL) under ice/water bath was added

compound 18 portionwise (432 mg, 1.44 mmol). The resulting

mixture was agitated for 0.5 h. Then the solution of compound 24a in

dry N,N′-dimethylformamide (5 mL) was added dropwise, and the

resulting mixture was allowed to warm to room temperature and

stirred overnight. The mixture was concentrated under vacuum. The

residue was diluted with ethyl acetate (20 mL) and washed with citric

acid 5% aqueous solution (10 mL) and brine (10 mL). The organic

phase was separated, dried over sodium sulfate anhydrous, and

concentrated under vacuum. The residue was puriﬁed via column

chromatography using petroleum ether:ethyl acetate 3:1 as eluent

system. The target compound was yielded as yellow oil (170 mg, 23%

from 22a).

(S)-tert-Butyl 2-Ethyl-4-(3-(6-(hydroxymethyl)naphthalen-2-

yloxy)propyl)-3-oxo-1-thia-4,8-diazaspiro[4.5]decane-8-carboxy-

late (26a). To the starting material 25a (170 mg, 0.27 mmol) was

added the mixed solution of acetic acid/water/tetrahydrofuran (3:1:1,

10 mL), and the resulting solution was stirred overnight at room

temperature. The solution was then cooled to ≤5 °C with ice/water

bath, and the pH value of was regulated to ≥10 with 25% ammonia

aqueous solution. Then the mixture was extracted with ethyl acetate (3

× 20 mL). The organic phases were combined, washed with brine,

dried over sodium sulfate anhydrous, and concentrated under vacuum.

The residue was puriﬁed via column chromatography using petroleum

ether:ethyl acetate 1:1 as eluent system. The target compound 26a was

yielded as yellow oil (120 mg, 76%).

General Procedure for the Amidation of 11a−c and 12a−d.

First, 2.0−5.0 mmol of the corresponding esters 10 were levigated with

a 2- or 8-fold surplus of 2-aminobutanol and 4-aminohexanol,

respectively, and heated to 150 °C to melt the esters. In the case a

solid developed, the mixture was then cooled down and the solid was

recrystallized from ethanol. If no solid developed, the mixture was

heated for another 6 h. After cooling to room temperature, the

obtained precipitate was ﬁltered and recrystallized from ethanol.

General Procedure for the Conversion of Alcohol to Bromine,

Compounds 13a−c and 14a−d. First, 2.0−5.6 mmol of the

hydroxyalkyl substituted 4-oxo-quinolinecarboxamides 11 and 12,

respectively, were dissolved in a huge excess of conc HBr. Then a

tenth of conc sulfuric acid (of HBr) was carefully added, and the

solution was reﬂuxed for 4−6 h. After cooling to room temperature the

solution was poured into water. This solution was extracted with

chloroform several times. The combined organic phases were

neutralized with K₂CO₃and the solvent evaporated. The so obtained

solid was crystallized in ethanol and recrystallized from methanol.

General Procedure for the Synthesis of the Quinolone-iperoxo

Hybrids 2a−c and 3a−d. First, 0.4 mmol of the bromoalkyl 4-oxo-

quinoline-3-carboxamides 13/14 were dissolved in 20 mL of

acetonitrile, and 2−4 equiv of iperoxo-base⁴³and a catalytic amount

of K₂CO₃were added. The mixture was then heated to 55 °C in a

sealed container. The reaction was followed up by TLC (silica gel,

MeOH/NH₄NO₂= 3:2, R_f= 0.54−0.68). When the reaction was

completed (after 6−20 days) the mixture was very slowly poured

(dropped) into diethyl ether. The so-obtained precipitate was ﬁltered

through a frit. The solid was washed with diethyl ether and dried in

vacuo. If necessary, the solid precipitate can be recrystallized from

methanol.

6-(Hydroxymethyl)naphthalen-2-ol (21). To a solution of 6-

hydroxy-2-naphthoic acid 19 (1 g, 5.32 mmol) in methanol (30 mL)

was added 10 drops of 98% sulfuric acid, and the solution was reﬂuxing

at 90 °C for 24 h. After the reaction, the solution was concentrated

under vacuum and the ester 20 was dissolved in freshly distilled

tetrahydrofuran anhydrous (30 mL). The solution was cooled to 0 °C

using ice/NaCl bath, and lithium aluminum hydride (350 mg, 9.21

mmol) was added to the above solution portionwise while maintaining

the temperature lower than 0 °C. The mixture was then allowed to rise

to room temperature gradually and stirred at room temperature for 24

(S)-tert-Butyl 2-Ethyl-4-(3-(6-(iodomethyl)naphthalen-2-yloxy)-

propyl)-3-oxo-1-thia-4,8-diazaspiro[4.5]decane-8-carboxylate

(27a). To the solution of triphenyl phosphine (92 mg, 0.35 mmol)

and imidazole (32 mg, 0.47 mmol) in dichloromethane (10 mL) was

added iodine (89 mg, 0.35 mmol). The resulting mixture was agitated

at room temperature for 1 h. Then the solution of 26a (120 mg, 0.23

mmol) in dichloromethane (5 mL) was added to the above mixture

dropwise, and the mixture was stirred for 1−2 h. The saturated

L

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

solution of sodium dithionite (5 mL) was added to the reaction

mixture and was agitated vigorously for 10 min. Then water (20 mL)

and dichloromethane (20 mL) were added, the organic phase was

separated, and the aqueous phase was extracted with dichloromethane

(2 × 20 mL). The organic phases were combined, dried over sodium

sulfate anhydrous, and concentrated under vacuum. The residue was

then puriﬁed via column chromatography using petroleum ether:ethyl

acetate 3:1 as eluent system to yield the product as yellow oil (90 mg,

62%).

(S)-tert-Butyl 4-(3-(6-((3-(Ethoxycarbonyl)-8-ﬂuoro-4-oxoquino-

lin-1(4H)-yl)methyl)naphthalene-2-yloxy)propyl)-2-ethyl-3-oxo-1-

thia-4,8-diazaspiro[4.5]decane-8-carboxylate (28a). The mixture of

9a (21 mg, 0.09 mmol), 27a (50 mg, 0.08 mmol), and sodium

carbonate (13 mg, 0.12 mmol) in DMF (5 mL) was heated to 50 °C

and stirred vigorously overnight. Water (20 mL) and ethyl acetate (20

mL) was added to the mixture, and the organic phase was separated.

The aqueous phase was extracted with ethyl acetate (2 × 20 mL). The

organic phases were combined, dried over sodium sulfate anhydrous,

and concentrated under vacuum. The residue was puriﬁed via column

chromatography (ethyl acetate:methanol, 25:1 as eluent system) to

yield the product as yellow oil (20 mg, 34%).

(S)-Ethyl 1-((6-(3-(2-Ethyl-3-oxo-1-thia-4,8-diazaspiro[4.5]decan-

4-yl)propoxy)naphthalen-2-yl)methyl)-8-ﬂuoro-4-oxo-1,4-dihydro-

quinoline-3-carboxylate (29a). To the solution of 28a (110 mg, 0.15

mmol) in dichloromethane (5 mL) was added the solution of HCl in

diethyl ether (3 mL). The resulting solution was stirred at room

temperature for 24 h. The reaction mixture was basiﬁed with 25%

ammonia solution under ice/water bath until the pH value ≥10. The

aqueous phase was extracted with dichloromethane (3 × 30 mL). The

organic phases were combined, dried over sodium sulfate anhydrous,

and concentrated under vacuum. The residue was puriﬁed via column

chromatography by using dichloromethane:methanol:ammonia

10:1:0.1 as eluent system. The target compound was obtained (40

mg, 42%) as yellow oil.

Discussion). In the IP1-accumulation assay DMSO concentrations

varied between 0.5 and 5%.

Cell Culture. Chinese hamster ovary cells (CHO) stably expressing

the hM₁receptor (CHO-hM₁cells) were cultured in Ham’s Nutrient

Mixture F-12 (HAM- F12) supplemented with 10% (v/v) FCS (FCS),

100 U/mL penicillin, 100 μg/mL streptomycin, 0.2 mg/mL G418, and

2 mM ^L-glutamine at 37 °C in a 5% CO₂humidiﬁed atmosphere.

Binding Experiments. Radioligand binding experiments using live

CHO-hM₁cells were performed in 96-well microtiter plates (ABgene,

Germany) applying 75000 cells/well as described previously.^5,19

Experiments were conducted at 28 °C using [³H]NMS as the

radioactive probe (2 nM for [³H]NMS dissociation and 0.2 nM for

equilibrium binding experiments).⁵Incubation time necessary to reach

binding equilibrium was checked by quantiﬁcation of the retarding

action of the compounds on [³H]NMS dissociation; incubation time

amounted to 3 h in case of the allosteric moieties 13a and 14a and to 8

h in the case of 2a and 3a. Radioligand [³H]NMS binding

characteristics B_maxand K_Din live CHO-hM₁cells did not change

over 8 h.

Binding experiments addressing the eﬀects of selected test

compounds on the inhibition of speciﬁc [³H]NMS binding to

membranes from CHO-hM₁cells by the endogenous agonist

acetylcholine were carried out as described earlier⁵⁶with two

diﬀerences. First, the incubations were carried out in 10 mM

HEPES, 10 mM MgCl₂, 100 mM NaCl, and 0.2 mM GTP, pH 7.4,

at 30 °C, and second, the analysis of the binding data was independent

from a model and based on curving ﬁtting applying a four parameter

logistic equation.

Dynamic Mass Redistribution (DMR) Assay. A general protocol

has been published recently.⁴³DMR assays using live CHO-hM₁cells

were carried out as outlined elsewhere.⁵For a chemical “knockout” of

selected G proteins, cells were pretreated either with 100 ng/mL PTX

for 16−24 h (G_iproteins)⁴or with 1 μM UBO-QIC for 1.5 h (G_q

proteins).^57,58

Determination of Inertness of Compound 13a toward

Glutathione. The buﬀer used was identical to the one applied in

the pharmacological assays, consisting of HBSS (purchased from Life

Technologies GmbH) and HEPES (20 mM). Solutions of compound

13a in DMSO (1 mM) and glutathione in water (1 mM and 0.1 mM)

were prepared. Two ratios (1:1 and 10:1 of compound 13a:gluta-

thione) were investigated. Solutions of compound 13a (100 μL) and

glutathione (100 μL) were added to the buﬀer (800 μL) to form a

solution (1000 μM), with the concentration of 13a being identical to

the highest concentration applied in the pharmacological experiments.

The solution was mixed and after 4 h of incubation at 28 °C analyzed

by LC-MS (method: water/methanol, 1 mL/min ﬂow rate, gradient

5−90%, 18 min, MS scan 100−1000). No products apart from the

starting materials could be detected.

Pharmacology. Materials and Methods. Reagents and cell

culture media were purchased from Sigma-Aldrich (Taufkirchen,

Germany) and Life Technologies (Darmstadt, Germany), and all of

the laboratory reagents were from Sigma-Aldrich (Taufkirchen,

Germany) unless speciﬁed otherwise. Commercially obtained

compounds had >99% purity. HEPES was purchased from AppliChem

(Darmstadt, Germany), and pertussis toxin (PTX) was purchased

from Biotrend (Cologne, Germany). UBO-QIC (FR900359) was

IP1 Assay. Changes in the concentration of the second messenger

IP1 in CHO-hM₁cells were quantiﬁed with the HTRF-IP1 kit (Cisbio

International) on a Mithras LB 940 plate reader (Berthold

Technologies, Bad Wildbad, Germany) according to the manufac-

turer’s instructions and as described previously in detail.⁵⁹In brief,

cells were resuspended in assay buﬀer (HEPES, 10 mM, CaCl₂, 1 mM,

MgCl₂, 0.5 mM, KCl, 4.2 mM, NaCl, 146 mM, glucose, 5.5 mM, and

LiCl, 50 mM, pH 7.4) and dispensed in 384-well microplates at a

density of 100000 cells/well. After preincubation in assay buﬀer for 30

min at 37 °C, 5% CO₂, cells were stimulated with the test compounds

for 30 min at 37 °C. The reactions were stopped by the addition of 50

mM Na₂HPO₄/KH₂PO₄phosphate buﬀer (pH 7.0), 1 mM KF, and

1.25% Triton X-100 containing HTRF assay reagents. The assay was

incubated 60 min at room temperature, and time-resolved FRET

signals were measured after excitation at 320 nm using the Mithras LB

940 multimode plate reader. Data analysis was based on the

ﬂuorescence ratio emitted by labeled IP1 (665 nm) over the light

emitted by the europium cryptate-labeled anti-IP1 (620 nm). Levels of

IP1 were normalized to the amount generated in the presence 100 μM

acetylcholine.

Data Analysis. Data analysis in general was performed using Prism

5.03 (GraphPad Software, San Diego, CA). [³H]NMS equilibrium

binding data from homologous inhibition curves applying CHO-hM₁

cells, and the eﬀect on DMR of all compounds listed in Table 1 were

analyzed by a four-parameter logistic function. The resulting IC₅₀

values from [³H]NMS homologous inhibition data were used to

calculate equilibrium dissociation constants K_Daccording to De Blasi

et al.⁶⁰Data from [³H]NMS dissociation experiments in the absence

and in the presence of test compounds were analyzed as described

recently.⁵⁶All DMR responses were solvent-corrected. The

quantiﬁcation of probe-induced DMR signals was based on the signal

value (upward- or downward-deﬂected) at 2400 s and normalized on

the corresponding value elicited by ACh (100 μM), which was set at

100%. Concentration−eﬀect curves of DMR measurements in the

presence of ACh, iperoxo, and pilocarpine were ﬁtted to the

operational model of agonism.⁵⁰In the analyses applying the

kindly provided by Profs. G. M. Konig and E. Kostenis, University of

̈

Bonn. [³H]NMS (speciﬁc activity 85.4 Ci/mmol) was purchased from

PerkinElmer (Rodgau, Germany). The buﬀer applied in functional and

binding experiments comprised Hank’s Balanced Salt Solution (HBSS;

Life Technologies, Darmstadt, Germany) with 20 mM HEPES, pH

7.0. Stock solutions of water-soluble compounds were prepared and

further diluted in incubation buﬀer, and poorly water-soluble

compounds were prepared in pure DMSO and further diluted in

buﬀer if possible without precipitating. In experiments applying the

EPIC system, the ﬁnal DMSO concentration per well never exceeded

1.3%, as this concentration did not induce a signal of its own; in

binding experiments, the highest DMSO concentration applied was

10%. Thus, very poorly soluble compounds such as 33a were

applicable only at a very low concentration of 0.3 μM (cf. Results and

M

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

operational model, the parameter E_max, i.e., the maximum response of

the system, was deﬁned and constrained to the eﬀect observed in the

presence of 100 μM ACh, E_max,ACh= constant = 100%; the parameter

basal, i.e., the basal response of the system, was deﬁned and

+ L_hybrin)/L_hybrin; B₀indicates the speciﬁc equilibrium binding of a ﬁxed

radioligand concentration in the absence of the dynamic ligand AB.

As the ﬁnal component of the global ﬁt, [³H]NMS equilibrium

binding data (CEC 4) of the allosteric moieties 13a or 14a were

analyzed according to the allosteric ternary complex model^4,30,61and

attributed to the following equation:⁶²

constrained to the response in the absence of ACh, basal_ACh

=

constant = 0%. The gobal ﬁts provided parameter estimates of K_Aand

τ to reﬂect functional aﬃnity and eﬃcacy of the agonists, respectively.

The concentration−eﬀect curves of hybrid-induced DMR were

analyzed according to the operational model of agonism for dynamic

ligands as recently described by Bock et al.³⁰

(1 + K_L·[L_fragin])(1 + α′_frag·(K_B·[B])ⁿ

)

frag

Y = B₀

1 + (K_B·[B])ⁿ+ K_L·[L_fragin](1 + α′_frag·(K_B·[B])ⁿ

)

frag

(5)

E_max

E =

B₀indicates the speciﬁc equilibrium binding of a ﬁxed radioligand

concentration in the absence of the allosteric moiety B, K_Lis the

equilibrium dissociation constant of the radioligand [³H]NMS, K_Bis

the equilibrium dissociation constant of the allosteric moieties 13a or

14a at the free receptor, and α′_fragis the cooperativity factor of the

interaction between the allosteric moiety B and the radioligand L.

[L_fragin] indicates the concentration of the radioligand [³H]NMS in the

equilibrium binding experiments characterizing the allosteric frag-

ments, and n_fragis a slope factor describing the steepness of the curve.

In the case of 13a and 14a, for which the slope factors of the

[³H]NMS equilibrium binding curves were signiﬁcantly diﬀerent from

unity (Figure 5B,D), eq 5, as had already been reported,⁶²does not

express interactions that can be visualized by a molecular scheme.

Therefore, to analyze these data under conditions in which the

mechanical implications are clear, the global analyses were also carried

out with slope factors n_fragconstrained to unity (Supporting

Information, Figure 7B,D). For the sake of consistency throughout

the global analysis, the equilibrium association constants K_Band K_L

from the original equation⁶²were treated as equilibrium dissociation

constants by replacing them with K_B= 1/K_Band K_L= 1/K_L,

respectively.

Overall, global ﬁtting yielded the parameter values K_A, K_B, τ_dyn, n_op

(transducer slope of eq 1), α′_hybr, α′_frag, and n_frag(slope of eq 5). Two

versions of the global ﬁt procedure were applied. In both cases, the

parameters L_hybrin, L_fragin, K_L, and B₀were constrained to ﬁxed values,

and K_Aand K_Bwere always treated as shared variables. In the ﬁrst

version of the global ﬁt (cf. Table 2 and Figure 5), the parameter

values α′_hybr, α′_frag, and n_frag(slope of eq 5) were treated as individual

variables. In contrast, in the second version of the global ﬁt (cf.

Supporting Information, Table 2 and Figure 7A,C), the parameters

α′_hybrand α′_fragwere additionally treated as shared variables and n_frag

was constrained to unity.

n

op

([AB] + K_D)

1 +

n

op

n

op

[AB] ·τ_dyn

(1)

E_MAXdenotes the maximum response of the system deﬁned as the

maximum eﬀect in the presence of a saturating concentration of ACh.

AB is the concentration of the dynamic ligand. τ_dynrepresents the

dynamic transduction coeﬃcient of the ligand. n_opindicates the

transducer slope. K_Dindicates the functional aﬃnity of the ligand AB

determined with functional data and can be expressed as³⁰

K_A·K_B

K_A+ K_B

K_D

=

(2)

K_Aand K_Bare taken to reﬂect the aﬃnity of the ligand AB for the

active pose and for the inactive receptor binding pose, respectively.³⁰

Global Fitting Analysis. To simultaneously quantify the active and

inactive ligand binding poses of each of the dynamic ligands 2a, 3a

using both binding and functional data, we, on the basis of the

mathematical framework pubished recently,³⁰applied a new unbiased

global ﬁtting approach to four diﬀerent sets of concentration−eﬀect

curves (CECs) for each dynamic ligand: (CEC 1) functional change in

DMR induced by the reference agonist acetylcholine, (CEC 2)

functional change in DMR induced by the dynamic ligands 2a or 3a,

(CEC 3) [³H]NMS displacement by the dynamic ligands 2a or 3a,

and (CEC 4) [³H]NMS displacement by the corresponding allosteric

moieties 13a or 14a, respectively.

The global ﬁts of every set of four curves was performed using the

following equations: Concentration−eﬀect curves of the dynamic

ligands 2a or 3a (CEC 2) together with the reference agonist

acetylcholine (CEC 1) were ﬁtted using eq 3 (derived from replacing

K_Din eq 1 with eq 2)

E_max

E =

n

To further ramify the mechanism of partial agonism of 2a and 3a, a

separate global ﬁt was applied to the data sets of the dynamic ligands

2a and 3a. In this regard, τ_dynin eq 3 was further resolved to³⁰

op

⎛

⎜

⎞

⎟

K

·K

B

A

[AB] +

⎝

K

+ K

⎠

A

B

op

1 +

n

op

[AB] ·τ_dyn

(3)

[³H]NMS equilibrium binding data (CEC 3) of the dynamic ligands

2a or 3a were analyzed according to a model of bivalent ligand binding

to the orthosteric and allosteric site (i.e., an extended allosteric ternary

complex model)^4,30,61and attributed to the following equation:

K_B

K_A+ K_B

ε_max·R_T

K_E

K_B

K_A+ K_B

τ

= τ_max

·

=

·

dyn

(6)

τ_dynis the dynamic transduction coeﬃcient of the ligand and can be

expressed as the product of τ_max, the maximum transduction coeﬃcient

of the dynamic ligand at 100% occupancy of receptors in the active

pose and the fraction of the ligand−receptor complexes in the active

pose (K_B/K_A+ K_B). τ_maxis deﬁned by the system-independent

maximum intrinsic eﬃcacy of the dynamic ligand ε_maxand a system-

dependent part R_T/K_E. R_Tis the total number of receptors. K_E

indicates the level of stimulus that elicits the half maximal system

response. Replacing τ_dynin eq 3 with eq 6 yielded:

[L_hybrin]·R_T

Y =

⎛

⎞

⎟

⎠

⎛

⎜

⎞

⎟

K

+ K

B

A

⎜

[L_hybrin] + K_L·^⎝^{1 + [AB]·}

⎝

·K

⎠

A

B

⎛

α ′

^·^[AB]⎞

hybr

K

⎜1 +

⎟

⎠

⎝

(4)

B

Y is the speciﬁc binding of the radioligand L [i.e., [³H]NMS].

[L_hybrin] indicates the concentration of the radioligand L, R_Tis the total

number of receptors. K_Ldenotes the equilibrium dissociation constant

of the [³H]NMS−receptor complex and represents the aﬃnity of the

radioligand L. AB is the concentration of the dynamic ligand 2a or 3a.

K_Aand K_Bare the equilibrium dissociation constants of the active and

inactive dynamic ligand−receptor complexes, respectively. They are

considered to represent the aﬃnities of the dynamic ligand in the

active and inactive pose, respectively. α′_hybrdenotes the cooperativity

between L and AB, and positive cooperativity is reﬂected by α′ > 1.

Equation 4 has been previously applied to ﬁt binding data of

bipharmacophoric, dualsteric ligands.^4,5,30To reduce the number of

variables during global curve ﬁtting, R_Tin eq 4 was replaced by B₀*(K_L

E_max

E =

n

op

⎛

⎜

⎞

⎟

K

·K

B

A

[AB] +

⎝

+ K

⎠

A

B

1 +

n

op

⎛

⎞

⎟

ε

·R

K

n

max

T

B

op

⎜

⎝

[AB]

·

K

A

K

+ K

⎠

(7)

E

B

During global ﬁtting of the 2a/13a and 3a/14a data sets, R_T/K_Ewas

treated as single variable and taken to represent the receptor reserve;

ε_maxand n_opwere treated as variables (further details, cf. text and Table

2 and Supporting Information, Table 1). In global ﬁts, K_Aand K_Bwere

treated as shared variables and R_T/K_Eset constant to unity allowing

N

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

estimation of ε_max* (equivalent to τ_maxin the absence of a receptor

reserve).

Notes

The authors declare no competing ﬁnancial interest.

Estimation of Fractional Occupancy. The set of occupancy curves

of 2a and 3a, depicted in parts C and E of Figure 5, respectively, were

calculated as described previously.³⁰In brief, we used the following

equations from the Supporting Information, Note 1 of ref 30: eq 7 for

the total binding of a dynamic ligand, eq 12 for its fractional occupancy

in the active pose, and eq 14 for the fractional occupancy in the

inactive pose. Equations 7, 12, and 14 from Supporting Information,

Note 1 in ref 30 were set up and used in a single user-deﬁned equation

in Prism 5.03 (GraphPad Software, San Diego, CA); with known

numerical estimates of K_A, K_B, and R_T, obtained as a result of the

global analysis of CECs (1−4), see above, the software was then able

to calculate and draw simultaneously the occupancy curves for the

three binding situations shown.

ACKNOWLEDGMENTS

■

We gratefully acknowledge the ﬁnancial support by the

Deutsche Forschungsgemeinschaft given to U.H. (HO 1368/

12-1), M.D. (DE 1546/4-1), and K.M. (MO 821/2-1). We

thank Corning Inc. for their support on the Epic system and

Professor E. Kostenis for allowing us to perform measurements

with it. The skillful technical assistance of Mechthild Kepe and

Iris Jusen is gratefully acknowledged.

ABBREVIATIONS USED

■

R_pose, the ratio of the active and inactive receptor binding pose was

Aβ, β-amyloid protein; ACh, acetylcholine; AD, Alzheimer’s

disease; BQCA, benzyl quinolone carboxylic acid; BQCAd,

BQCA derived compound; CHO, Chinese hamster ovary cell;

EC, extracellular loops; DMR, dynamic mass redistribution;

GPCRs, G-protein coupled receptors; M₁, muscarinic acetyl-

choline subtype 1; NMDA, N-methyl-^D-aspartate; NMS, N-

methylscopolamine; PAM, positive allosteric modulator; Pilo,

pilocarpine; PTX, pertussis toxin; SARs, structure−activity

relationships; TBPBd, TBPB derived compound; TM, trans-

membrane helix; QNB, quinuclidinyl benzilate

calculated as previously reported:³⁰

⎛

⎜

⎝

⎞

⎟

⎠

⎛

⎜

⎝

⎞

⎟

⎠

K_A

K_active

K

inactive

R_pose= −log

= −log

K

(8)

B

and implemented into the global ﬁt as a transform to report. The

software was able to report either the corresponding standard errors

(Table 2, Supporting Information, Tables 1 and 2) or the 95%

conﬁdence intervals (Figure 5F and Supporting Information, Figure

7F).

Statistical Analysis. Experimental data are presented as means

SEM of N observations. Single comparisons were performed using

Student’s t-test and multiple comparisons using One-Way ANOVA

with Newman−Keuls multiple comparison test. Nonlinear regression

data analyses based on diﬀerent models were compared by a F-test. In

all cases, P < 0.05 (*) was considered as the level of statistical

signiﬁcance, P < 0.01 (**) and P < 0.001 (***) was indicated for

selected comparisons.

REFERENCES

■

(1) Mohr, K.; Trankle, C.; Kostenis, E.; Barocelli, E.; De Amici, M.;

̈

Holzgrabe, U. Rational design of dualsteric GPCR ligands: Quests and

promise. Br. J. Pharmacol. 2010, 159, 997−1008.

(2) Davie, B. J.; Christopoulos, A.; Scammells, P. J. Development of

M1 mAChR allosteric and bitopic ligands: prospective therapeutics for

the treatment of cognitive deficits. ACS Chem. Neurosci. 2013, 17,

1026−1048.

ASSOCIATED CONTENT

(3) Steinfeld, T.; Mammen, M.; Smith, J. A.; Wilson, R. D.; Jasper, J.

R. A novel multivalent ligand that bridges the allosteric and orthosteric

binding sites of the M2 muscarinic receptor. Mol. Pharmacol. 2007, 72,

291−302.

(4) Antony, J.; Kellershohn, K.; Mohr-Andrae, M.; Kebig, A.; Prilla,

S.; Muth, M.; Heller, E.; Disingrini, T.; Dallanoce, C.; Bertoni, S.;

■

S

* Supporting Information

More detailed synthetic procedures of intermediates as well as

compound structure characterization. This material is available

free of charge via the Internet at http://pubs.acs.org.

Schrobang, J.; Trankle, C.; Kostenis, E.; Christopoulos, A.; Hoeltje, H.-

̈

AUTHOR INFORMATION

D.; Barocelli, E.; De Amici, M.; Holzgrabe, U.; Mohr, K. Dualsteric

GPCR targeting: a novel route to binding and signaling pathway

selectivity. FASEB J. 2009, 23, 442−450.

■

Corresponding Authors

*For M.D.: phone, 0049-931-3189676; E-mail, michael.

decker@uni-wuerzburg.de.

*For U.H.: phone, 0049-931-3185461; E-mail, u.holzgrabe@

pharmazie.uni-wuerzburg.de.

*For C.T.: phone, 0049-228-739104; E-mail, traenkle@uni-

bonn.de.

(5) Bock, A.; Merten, N.; Schrage, R.; Dallanoce, C.; Batz, J.;

̈

Klockner, J.; Schmitz, J.; Matera, C.; Simon, K.; Kebig, A.; Peters, L.;

̈

Muller, A.; Schrobang-Ley, J.; Trankle, C.; Hoffmann, C.; De Amici,

̈

M.; Holzgrabe, U.; Kostenis, E.; Mohr, K. The allosteric vestibule of a

seven transmembrance helical receptor controls G-Protein coupling.

Nature Commun. 2012, 3, 1044 DOI: 10.1038/ncomms2028.

(6) Lane, J. R.; Sexton, P. M.; Christopoulos, A. Bridging the gap:

bitopic ligands of G-protein-coupled receptors. Trends Pharmacol. Sci.

2013, 34, 59−66.

Author Contributions

The manuscript was written through contributions of all

authors. All authors have given approval to the ﬁnal version of

the manuscript. X.C., J.K., and J.H. contributed equally. M.D.,

U.H., and C.T. conceived the project. X.C. designed and

synthesized compounds 4a,b and 5a−c. J.K. performed the

syntheses of compounds 2a−c and 3a−d. J.H. conducted DMR

and IP1 assays, all of the binding experiments in live CHO-hM₁

cells and membranes, analyzed data, and contributed to the

discussion. C.Z. and W.S. conducted DMR assays, and R.S.

conducted DMR and IP1 assays. A.B. provided important ideas.

K.M. and A.B. contributed to the discussion. C.T. developed

and applied the global curve ﬁtting procedure, allowing to

analyze simultaneously functional and binding data with the

mathematical framework for dynamic ligands, which had been

published recently.³⁰

(7) Bock, A.; Mohr, K. Dualsteric GPCR targeting and functional

selectivity: the paradigmatic M(2) muscarinic acetylcholine receptor.

Drug Discovery Today: Technol. 2013, 10, 245−252.

(8) Bymaster, F. P.; McKinzie, D. L.; Felder, C. C.; Wess, J. Use of

M1−M5 muscarinic receptor knockout mice as novel tools to

declineate the physiological roles of the muscarinice cholinergic

system. Neurochem. Res. 2003, 28, 437−442.

(9) Lanzafame, A. A.; Christopoulos, A.; Mitchelson, F. Cellular

signaling mechanisms for muscarinic acetylcholine receptors. Recept.

Channels 2003, 9, 241−260.

(10) Fisher, A. M1 muscarinic agonists target major hallmarks of

Alzheimer’s diseasethe pivotal role of brain M1 receptor. Neuro-

degener. Dis. 2008, 5, 237−240.

(11) Liu, L.; Zhao, R.; Bai, Y.; Stanish, L. F.; Evans, J. E.; Sanderson,

M. J.; Bonventre, J. V.; Rittenhouse, A. R. M1 muscarinic receptors

O

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

inhibit L-type Ca²⁺current and M-current by divergent signal

transduction cascades. J. Neurosci. 2006, 26, 11588−11598.

(12) Marino, M. J.; Rouse, S. T.; Levey, A. I.; Potter, L. T.; Conn, P.

J. Activation of the genetically defined M1 muscarinic receptor

potentiates N-methyl-^D-aspartate (NMDA) receptor currents in

hippocampal pyramidal cells. Proc. Natl. Acad. Sci. U. S. A. 1998, 95,

11465−11470.

(13) Fisher, A. Cholinergic treatments with emphasis on M1

muscarinic agonists as potential disease-modifying agents for

Alzheimer’s disease. Neurotherapeutics 2008, 5, 433−442.

(14) Decker, M.; Holzgrabe, U. M₁muscarinic acetylcholine receptor

allosteric modulators as potential therapeutic opportunities for treating

Alzheimer’s disease. Med. Chem. Commun. 2012, 3, 752−762.

(15) Anagnostaras, S. G.; Murphy, G. G.; Hamilton, S. E.; Mitchell, S.

L.; Rahnama, N. P.; Nathanson, N. M.; Silva, A. J. Selective cognitive

dysfunction in acetylcholine M1 muscarinic receptor mutant mice.

Nature Neurosci. 2003, 6, 51−58.

(16) Kruse, A. C.; Hu, J.; Pan, A. C.; Arlow, D. H.; Rosenbaum, D.

M.; Rosemond, E.; Green, H. F.; Liu, T.; Chae, P. S.; Dror, R. O.;

Shaw, D. E.; Weis, W. I.; Wess, J.; Kobilka, B. K. Structure and

dynamics of the M3 muscarinic acetylcholine receptor. Nature 2012,

482, 552−556.

(17) Haga, K.; Kruse, A. C.; Asada, H.; Yurugi-Kobayashi, T.;

Shiroishi, M.; Zhang, C.; Weis, W. I.; Okada, T.; Kobilka, B. K.; Haga,

T.; Kobayashi, T. Structure of the human M2 muscarinic acetylcholine

receptor bound to an antagonist. Nature 2012, 482, 547−551.

(18) Kruse, A. C.; Ring, A. M.; Manglik, A.; Hu, J.; Hu, K.; Eitel, K.;

(29) Matera, C.; Flammini, L.; Quadri, M.; Vivo, V.; Ballabeni, V.;

Holzgrabe, U.; Mohr, K.; De Amici, M.; Barocelli, E.; Bertoni, S.;

Dallanoce, C. Bis(ammonio)alkane-type agonists of muscarinic

acetylcholine receptors: synthesis, in vitro functional characterization,

and in vivo evaluation of their analgesic activity. Eur. J. Med. Chem.

2014, 75, 222−232.

(30) Bock, A.; Chirinda, B.; Krebs, F.; Messerer, R.; Batz, J.; Muth,

̈

M.; Dallanoce, C.; Klingenthal, D.; Trankle, C.; Hoffmann, C.; De

̈

Amici, M.; Holzgrabe, U.; Kostenis, E.; Mohr, K. Dynamic ligand

binding dictates partial agonism at a G protein-coupled receptor.

Nature Chem. Biol. 2014, 10, 18−20.

(31) January, B.; Seibold, A.; Whaley, B.; Hipkin, R. W.; Lin, D.;

Schonbrunn, A.; Barber, R.; Clark, R. B. Beta2-adrenergic receptor

desensitization, internalization, and phosphorylation in response to full

and partial agonists. J. Biol. Chem. 1997, 272, 23871−23879.

(32) Mailman, R. B.; Murthy, V. Third generation antipsychotic

drugs: partial agonism or receptor functional selectivity? Curr. Pharm.

Des. 2010, 16, 488−501.

(33) Kuduk, S. D.; Beshore, D. C. Novel M(1) allosteric ligands: a

patent review. Expert Opin. Ther. Pat. 2012, 22, 1385−1398.

(34) Mistry, S. N.; Valant, C.; Sexton, P. M.; Capuano, B.;

Christopoulos, A.; Scammells, P. J. Synthesis and pharmacological

profiling of analogues of benzyl quinolone carboxylic acid (BQCA) as

allosteric modulators of the M₁muscarinic receptor. J. Med. Chem.

2013, 56, 5151−5172.

(35) Yeatman, H. R.; Lane, J. R.; Choy, K. H.; Lambert, N. A.;

Sexton, P. M.; Christopoulos, A.; Canals, M. Allosteric modulation of

M1 muscarinic acetylcholine receptor internalization and subcellular

trafficking. J. Biol. Chem. 2014, 289, 15856−15866.

(36) Ma, L.; Seager, M. A.; Wittmann, M.; Jacobson, M.; Bickerl, D.;

Burno, M.; Jones, K.; Graufelds, V. K.; Xu, G.; Pearson, M.;

McCampbell, A.; Gaspar, R.; Shughrue, P.; Danziger, A.; Regan, C.;

Flick, R.; Pascarella, D.; Garson, S.; Doran, S.; Kreatsoulas, C.; Veng,

L.; Lindsley, C. W.; Shipe, W.; Kuduk, S.; Sur, C.; Kinney, G.;

Seabrook, G. R.; Ray, W. J. Selective activation of the M₁muscarinic

acetylcholine receptor achieved by allosteric potentiation. Proc. Natl.

Acad. Sci. U. S. A. 2009, 106, 15950−15955.

Hubner, H.; Pardon, E.; Valant, C.; Sexton, P. M.; Christopoulos, A.;

̈

Felder, C. C.; Gmeiner, P.; Steyaert, J.; Weis, W. I.; Garcia, K. C.;

Wess, J.; Kobilka, B. K. Activation and allosteric modulation of a

muscarinic acetylcholine receptor. Nature 2013, 504, 101−106.

(19) Schrage, R.; Seemann, W. K.; Klockner, J.; Dallanoce, C.; Racke,

́

̈

K.; Kostenis, E.; De Amici, M.; Holzgrabe, U.; Mohr, K. Agonists with

supraphysiological efficacy at the muscarinic M2 ACh receptor. Br. J.

Pharmacol. 2013, 169, 357−370.

(20) Venkatakrishnan, A. J.; Deupi, X.; Lebon, G.; Tate, C. G.;

Schertler, G. F.; Babu, M. M. Molecular signatures of G-protein-

coupled receptors. Nature 2013, 494, 185−194.

(37) Shirey, J. K.; Brady, A. E.; Jones, P. J.; Davis, A. A.; Bridges, T.

M.; Kennedy, J. P.; Jadhav, S. B.; Menon, U. N.; Xiang, Z.; Watson, M.

L.; Christian, E. P.; Doherty, J. J.; Quirk, M. C.; Snyder, D. H.; Lah, J.

J.; Levey, A. I.; Nicolle, M. M.; Lindsley, C. W.; Conn, P. J. A selective

allosteric potentiator of the M₁muscarinic acetylcholine receptor

increases activity of medial prefrontal cortical neurons and restores

impairments in reversal learning. J. Neurosci. 2009, 29, 14271−14286.

(21) Hulme, E. C.; Birdsall, N. J.; Buckley, N. J. Muscarinic receptor

subtypes. Annu. Rev. Pharmacol. Toxicol. 1990, 30, 633−673.

(22) Daval, S. B.; Valant, C.; Bonnet, D.; Kellenberger, E.; Hibert,

M.; Galzi, J. L.; Ilien, B. Fluorescent derivatives of AC-42 to probe

bitopic orthosteric/allosteric binding mechanisms on muscarinic M1

receptors. J. Med. Chem. 2012, 55, 2125−2143.

(23) Daval, S. B.; Kellenberger, E.; Bonnet, D.; Utard, V.; Galzi, J. L.;

Ilien, B. Exploration of the orthosteric/allosteric interface in human

m1 muscarinic receptors by bitopic fluorescent ligands. Mol.

Pharmacol. 2013, 84, 71−85.

(38) Abdul-Ridha, A.; Lopez, L.; Keov, P.; Thal, D. M.; Mistry, S. N.;

́

Sexton, P. M.; Lane, J. R.; Canals, M.; Christopoulos, A. Molecular

determinants of allosteric modulation at the M1 muscarinic acetylcho-

line receptor. J. Biol. Chem. 2014, 289, 6067−6079.

(24) Goodwin, J. A.; Hulme, E. C.; Langmead, C. J.; Tehan, B. G.

Roof and floor of the muscarinic binding pocket: variations in the

binding modes of orthosteric ligands. Mol. Pharmacol. 2007, 72, 1484−

1496.

(25) Gnagey, A. L.; Seldenberg, M.; Ellis, J. Site-directed mutagenesis

reveals two epitopes involved in the subtype selectivity of the allosteric

interactions of gallamine at muscarinic acetylcholine receptors. Mol.

Pharmacol. 1999, 56, 1245−1253.

(39) Fisher, A.; Bar-Ner, N.; Karton, Y. Methods and compositions

for treatment of central and peripheral nervous system disorders and

novel compounds useful therefor. U.S. Patent 7,439,251, 2008.

(40) Kuduk, S. D.; Chang, R. K.; Di Marco, C. N.; Ray, W. J.; Ma, L.;

Wittmann, M.; Seager, S. A.; Koeplinger, K. A.; Thompson, C. D.;

Hartman, G. D.; Bilodeau, M. T. Quinolizidinone carboxylic acids as

CNS penetrant, selective M₁allosteric muscarinic receptor modulators.

ACS Med. Chem. Lett. 2010, 1, 263−267.

(26) Dror, R. O.; Green, H. F.; Valant, C.; Borhani, D. W.; Valcourt,

J. R.; Pan, A. C.; Arlow, D. H.; Canals, M.; Lane, J. R.; Rahmani, R.;

Baell, J. B.; Sexton, P. M.; Christopoulos, A.; Shaw, D. E. Structural

basis for modulation of a G-protein-coupled receptor by allosteric

drugs. Nature 2013, 503, 295−299.

(27) Lazareno, S.; Popham, A.; Birdsall, N. J. M. Allosteric

interactions of staurosporine and other indolocarbazoles with N-

[methyl-³H]scopolamine and acetylcholine at muscarinic receptor

subtypes: identification of a second allosteric site. Mol. Pharmacol.

2000, 58, 194−207.

(41) Fisher, A. Cholinergic modulation of amyloid precursor protein

processing with emphasis on M1 muscarinic receptor: Perspective and

challenges in treatment of Alzheimer’s disease. J. Neurochem. 2012, 120

(Suppl 1), 22−33.

(42) Schroder, R.; Janssen, N.; Schmidt, J.; Kebig, A.; Merten, N.;

̈

Hennen, S.; Muller, A.; Blattermann, S.; Mohr-Andra, M.; Zahn, S.;

̈

̈ ̈

Wenzel, J.; Smith, N. J.; Gomeza, J.; Drewke, C.; Milligan, G.; Mohr,

K.; Kostenis, E. Deconvolution of complex G protein-coupled receptor

signaling in live cells using dynamic mass redistribution measurements.

Nature Biotechnol. 2010, 28, 943−949.

(28) Lazareno, S.; Popham, A.; Birdsall, N. J. M. Analogs of WIN

62,577 define a second allosteric site on muscarinic receptors. Mol.

Pharmacol. 2002, 62, 1492−1505.

(43) Schroder, R.; Schmidt, J.; Blattermann, S.; Peters, L.; Janssen,

̈ ̈

N.; Grundmann, M.; Seemann, W.; Kaufel, D.; Merten, N.; Drewke,

C.; Gomeza, J.; Milligan, G.; Mohr, K.; Kostenis, E. Applying label-free

P

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

dynamic mass redistribution technology to frame signaling of G-

protein-coupled receptors noninvasively in living cells. Nature Protoc.

2011, 6, 1748−1760.

(60) De Blasi, A.; O’reilly, K.; Motulsky, H. J. Calculating receptor

number from binding experiments using same compound as

radioligand and competitor. Trends Pharmacol. Sci. 1989, 10, 227−229.

(61) May, L. T.; Leach, K.; Sexton, P. M.; Christopoulos, A.

Allosteric modulation of G protein-coupled receptors. Annu. Rev.

Pharmacol. Toxicol. 2007, 47, 1−51.

́

(44) Leyva, E.; Monreal, E.; Hernandez, A. Synthesis of fluoro-4-

hydroxyquinoline-3-carboxylic acids by the Gould−Jacobs reaction. J.

Fluor. Chem. 1999, 94, 7−10.

(62) Trankle, C.; Weyand, O.; Voigtlander, U.; Mynett, A.; Lazareno,

̈

(45) Klockner, J.; Schmitz, J.; Holzgrabe, U. Convergent, short

̈

S.; Birdsall, N. J.; Mohr, K. Interactions of orthosteric and allosteric

ligands with [3H]dimethyl-W84 at the common allosteric site of

muscarinic M2 receptors. Mol. Pharmacol. 2003, 64, 180−190.

synthesis of the muscarinic superagonist iperoxo. Tetrahedron Lett.

2010, 51, 3470−3472.

(46) Walz, A. J.; Miller, M. J. Synthesis and biological activity of

hydroxamic acid-derived vasopeptidase inhibitor analogues. Org. Lett.

2002, 4, 2047−2050.

(47) Zhu, S.-F.; Song, X.-G.; Li, Y.; Cai, Y.; Zhou, Q.-L.

Enantioselective copper-catalyzed intramolecular O−H insertion: an

efficient approach to chiral 2-carboxy cyclic ethers. J. Am. Chem. Soc.

2010, 132, 16374−16376.

(48) Nair, S. K.; Matthews, J. J.; Stephan, J. C.; Ma, C.; Dovalsantos,

E. Z.; Grubbs, A. W.; Sach, N. W.; ten Hoeve, W.; Koster, H.; Flahive,

E. J.; Tanis, S. P.; Renner, M.; van Wiltenburg, J. Novel synthesis of

CP-734432, an EP4 agonist, using Sharpless asymmetric dihydrox-

ylation. Tetrahedron Lett. 2010, 51, 1451−1454.

(49) Henning, R.; Lattrell, R.; Hermann, J. G.; Leven, M. Synthesis

and neuroleptic activity of a series of 1-[1-(benzo-1,4-dioxan-2-

ylmethyl)-4-piperidinyl]benzimidazolone derivatives. J. Med. Chem.

1987, 30, 814−819.

(50) Black, J. W.; Leff, P. Operational models of pharmacological

agonism. Proc. R. Soc. London B: Biol. Sci. 1983, 220, 141−162.

(51) Lazareno, S.; Birdsall, N. J. M. Detection, quantitation, and

verification of allosteric interactions of agents with labeled and

unlabeled ligands at G-protein-coupled receptors: Interaction of

strychnine and acetylcholine at muscarinic receptors. Mol. Pharmacol.

1995, 48, 362−378.

́ ́ ̌

(52) Jakubík, J.; Bacakova, L.; El-Fakahany, E. E.; Tucek, S. Positive

cooperativity of acetylcholine and other agonists with allosteric ligands

on muscarinic acetylcholine receptors. Mol. Pharmacol. 1997, 52, 172−

179.

(53) De Amici, M.; Dallanoce, C.; Holzgrabe, U.; Trankle, C.; Mohr,

̈

K. Allosteric ligands for G protein-coupled receptors: a novel strategy

with attractive therapeutic opportunities. Med. Res. Rev. 2010, 30,

463−549.

(54) Bock, A.; Kostenis, E.; Trankle, C.; Lohse, M. J.; Mohr, K. Pilot

̈

the pulse: controlling the multiplicity of receptor dynamics. Trends

Pharmacol. Sci. 2014, 35, 630−638.

(55) De la Cruz, A.; Elguero, J.; Goya, P.; Martínez, A. Tautomerism

and Acidity in 4-Quinolone-3-carboxylic Acid Derivatives. Tetrahedron

1992, 29, 6135−6150.

(56) Fang, L.; Jumpertz, S.; Zhang, Y.; Appenroth, D.; Fleck, C.;

Mohr, K.; Trankle, C.; Decker, M. Hybrid molecules from xanomeline

̈

and tacrine: enhanced tacrine actions on cholinesterases and

muscarinic M1 receptor. J. Med. Chem. 2010, 53, 2094−2103.

(57) Nesterov, A.; Hong, M.; Hertel, C.; Jiao, P.; Brownell, L.;

Cannon, E. Screening a plant extract library for inhibitors of

cholecystokinin receptor CCK1 pathways. J. Biomol. Screening 2010,

15, 518−527.

(58) Hennen, S.; Wang, H.; Peters, L.; Merten, N.; Simon, K.;

Spinrath, A.; Blattermann, S.; Akkari, R.; Schrage, R.; Schroder, R.;

̈

Schulz, D.; Vermeiren, C.; Zimmermann, K.; Kehraus, S.; Drewke, C.;

Pfeifer, A.; Konig, G. M.; Mohr, K.; Gillard, M.; Muller, C. E.; Lu, Q.

̈

R.; Gomeza, J.; Kostenis, E. Decoding signaling and function of the

orphan G protein-coupled receptor GPR17 with a small-molecule

agonist. Sci. Signaling 2013, 6, ra93 DOI: 10.1126/scisignal.2004350.

(59) Schroder, R.; Merten, N.; Mathiesen, J. M.; Martini, L.; Kruljac-

̈

Letunic, A.; Krop, F.; Blaukat, A.; Fang, Y.; Tran, E.; Ulven, T.;

Drewke, C.; Whistler, J.; Pardo, L.; Gomeza, J.; Kostenis, E. The C-

terminal tail of CRTH2 is a key molecular determinant that constrains

Galphai and downstream signaling cascade activation. J. Biol. Chem.

2009, 284, 1324−1336.

Q

dx.doi.org/10.1021/jm500860w | J. Med. Chem. XXXX, XXX, XXX−XXX

Article Doi

Rational design of partial agonists for the muscarinic M1 acetylcholine receptor

DOI: 10.1021/jm500860w

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm500860w

Full text of DOI:10.1021/jm500860w

Products guided by the article

R&D Labs maybe for 179942-68-8

Relevant to this article

Hot Product