Muscarinic acetylcholine receptor binding affinities of pethidine analogs

Abstract

A series of pethidine analogs were synthesized and their affinities for the [³H]N-methylscopolamine (NMS) binding site on muscarinic acetylcholine receptors (mAChRs) were determined using M₁, M₃ or M₅ human mAChRs expressed by Chinese hamsters ovary (CHO) cell membranes. Compound 6b showed the highest binding affinities at M₁, M₃ and M₅ mAChRs (K_i = 0.67, 0.37, and 0.38 μM, respectively).

Graphical abstract

Many drugs of abuse including cocaine, amphetamine, methamphetamine, and morphine increase extracellular dopamine (DA) in the nucleus accumbens (NAc) at doses that produce rewarding effects.¹ Indeed, bilateral microinjection of 6-hydroxydopamine which produces dopaminergic neuron damage in NAc inhibited initiation of amphetamine self-administration in rats when 6-hydroxydopamine was administered before self-administration training, and disrupted responding during maintenance of amphetamine self-administration.² DA containing ventral tegmental area (VTA) neurons project to NAc, and are important for drug seeking behavior. Electrical stimulation of acetylcholine-containing laterodorsal tegmental nucleus (LDT) neurons, which innervate VTA dopaminergic neurons, results in an increase in extracellular DA concentrations in NAc.^3,4 The five subtypes of mAChRs (M₁ to M₅) are separated into two groups depending on their G_α protein coupling functionality. M₁, M₃, M₅ mAChR subtypes preferentially activate G_αq/11 proteins, leading to an increase in cytosolic calcium ion concentration. M₂ and M₄ mAChR subtypes preferentially couple with G_αi/o protein, inhibiting the conversion of adenosine triphosphate to cyclic adenosine monophosphate in the cytosol.⁵ Important to the current study, M₅ mAChRs are highly expressed on the postsynaptic DA neurons in VTA.^6-9 In M₅ mAChRs knockout (KO) mice, LDT stimulation and morphine-induced DA release in NAc is reduced relative to wild-type mice.^10,11 M₅ KO mice also have decreased cocaine self-administration and cocaine- or morphine-induced conditioned place preference when compared to wild-type controls.^11,12 Microinfusion into VTA of scopolamine, a mAChR antagonist, robustly decreased cocaine-seeking behavior during withdrawal in rats.¹³ Together, these findings led us to hypothesize that selective antagonism of M₅ mAChRs represents a novel target for the treatment of drug abuse.

We recently reported on a class of M₅-preferring orthosteric antagonists based on the scaffold of 1,2,5,6-tetrahydropyridine-3-carboxylic acid.¹⁴ Compound 1 (Figure 1, Table 1) was identified as the most selective M₅ mAChRs antagonist in this series. Interestingly, removal of the meta-methoxy group in 1 (compound 2) significantly increased binding affinities at both M₁ and M₅ receptors, but resulted in a complete loss in selectivity for M₅ over M₁. To further explore the structure-activity relationship (SAR), we planned to reposition the carboxylate group in 1 and 2 from C-3 to C-4 of the piperidine ring. New analogs resulted from such rearrangement resembling pethidine (3, Figure 1), a once popular analgesic. Interestingly, pethidine has been identified as an antagonist at mAChRs in guinea-pig ileum assays.¹⁵ Thus, analogs based on the pethidine scaffold may afford interesting SAR at mAChRs. In addition to ester containing analogs (4 and 6), we also planned to evaluate amides (5 and 7), carbamates (8 and 9), and carbamides (10). Herein, we describe the synthesis of these novel analogs and the evaluation of their binding affinity at mAChRs.

Figure 1.

Structure of compounds 1 and 2, pethidine (3), and design of pethidine analogs 4-10 as novel mAChR ligands.

Table 1. Structures and binding affinity for analogs at M₁, M₃, and M₅ mAChRs^a.

		[3H]NMS binding Ki ± SEM (μM)
compd	R	M1	M3	M5
1b	-	25.3	>100	2.24
2b	-	0.02 ± 0.002	NDc	0.03 ± 0.005
4a		10.8 ± 0.67	5.26 ± 0.33	6.95 ± 0.47
4b		1.20 ± 0.11	0.64 ± 0.037	0.87 ± 0.053
5a		> 30	> 10	> 30
5b		> 10	3.29 ± 0.80	6.97 ± 0.77
6a		3.63 ± 0.22	4.60 ± 0.61	2.14 ± 0.22
6b		0.67 ± 0.078	0.37 ± 0.045	0.38 ± 0.011
7a		> 10	> 30	> 10
7b		3.44 ± 0.93	1.54 ± 0.19	1.81 ± 0.11
8		5.09 ± 0.29	4.03 ± 0.57	5.41 ± 0.59
9		2.91 ± 0.17	2.30 ± 0.22	2.05 ± 0.30
10a		> 10	> 10	>10
10b		5.40 ± 1.22	3.55 ± 0.11	4.30 ± 0.29

^aThree independent experiments, each experiment included duplicate samples, were performed to obtain K_i values (Mean ± SEM)

^bData from reference 13

^cNot determined

Compounds 4a, 4b, 5a, and 5b were synthesized by converting 1-methyl-4-phenylpiperidine-4-carboxylic acid (pethidinic acid, 14) to the corresponding carbonyl chloride in the presence of SOCl₂, followed by reacting with a phenyl ring substituted 2-phenylethanol or 2-phenylethanamine (Scheme 1). Compound 14 was synthesized by initial catalytic hydrogenolysis of compound 11 under Pd(OH)₂ to afford 12, followed by Nmethylation to form compound 13 and hydrolysis of the cyano group. Conversion of 14 to alcohol 16 was achieved by a standard two-step procedure. Esterification between phenyl ring substituted 3-phenylpropanoyl chloride and 16 afforded compounds 6a and 6b. Treatment of 16 with 4-nitrophenyl chloroformate followed by reaction with 2-(3,4-dimethoxyphenyl)ethanamine provided carbamate 8. Similarly, amides 7a and 7b, carbamate 9, and carbamides 10a and 10b were synthesized from amine intermediate 17. A three-step process of an initial LAH reduction of the cyano group in 12 to amino group, followed by Boc protection to reduce polarity for column purification and de-Boc afforded 17.

Scheme 1.

Reagents and conditions: (a) Pd(OH)₂, H₂, MeOH; (b) HCHO (37% aqueous), NaBH(OAc)₃, THF; (c) 6N HCl, reflux; (d) 1. SOCl₂, DCM, reflux; 2. alcohol or amine, TEA, DCM; (e) EtOH, SOCl₂, reflux; (f) LAH, THF, 0 °C-rt; (g) 3-(3,4-dimethoxy or 4-methoxy)phenylpropanoyl chloride, TEA, DCM; (h) 1. 4-nitrophenyl chloroformate, Na₂CO₃, THF; 2. 2-(3,4-dimethoxyphenyl)ethanamine Na₂CO₃, THF; (i) 1. LAH, THF, 0 °C-rt; 2. Boc₂O, Na₂CO₃, THF/H₂O; 3. TFA/DCM; (j) 1. 4-nitrophenyl chloroformate, Na₂CO₃, THF; 2. alcohol or amine, Na₂CO₃, THF.

Subtype selectivity for M₅ over M₁ and M₃ mAChRs is particularly difficult to achieve because M₅ exhibit the high amino acid sequence identity with M₃ and M₁ mAChRs (85%, 79%, 73%, and 68% with M₃, M₁, M₄, and M₂ mAChRs respectively).¹⁶ In addition, all three subtypes prefer to bind Gαq/11. Thus, in this study, we evaluated binding affinity of our analogs at M₁, M₃, and M₅ mAChR subtypes as a first approach, and then compared selectivity of analogs at M₅ over M₁ and M₃ mAChRs. Analog affinities for M₁, M₃ and M₅ mAChRs were determined by measuring inhibition of [³H]N-methylscopolamine (NMS) binding to Chinese hamster ovary (CHO) cell membranes expressing M₁, M₃, or M₅ recombinant human mAChRs. CHO cells stably expressing each of the human mAChRs were obtained from Dr. Tom Bonner of National Institute of Mental Health (NIMH). Detailed materials and methods for cell culture and cell membrane preparation were described previously.^14,17 IC₅₀ values were obtained and K_i values were calculated using the equation of Cheng and Prusoff.¹⁸ Results are summarized in Table 1.

Similar to the SAR generated for the parent compounds 1 and 2, mono-methoxy substituted analogs (4b, 5b, 6b, 7b, and 10b) consistently exhibited higher affinity (2 to 9-fold at M₁; 3 to 19-, fold at M₃; 2 to 8-fold at M₅) when compared to their corresponding di-methoxy substituted analogs (4a, 5a, 6a, 7a, and 10a, respectively). The corresponding carboxylate moiety repositioned in molecule 4a exhibited 2- and 19-fold higher affinity at M₁ and M₃ mAChRs, respectively, compared with compound 1. However, the affinity of 4a at M₅ mAChRs was decreased by 30%. Thus, compound 4a was not subtype selective. It is unclear why reposition of the carboxylate group in 1 resulted in a completed loss of subtype selectivity.

In addition, replacement of the ester link in 4a/b or 6a/b with an amide link (5a/b and 7a/b, respectively) resulted in a loss of affinity at all three mAChR subtypes (4a vs 5a, 2 to 4-fold; 4b vs 5b, 5 to 8-fold; 6a vs 7a, 3 to 7-fold; 6b vs 7b, 4 to 5-fold). In general, analogs with carbamate and carbamide linkers exhibited up to an 8-fold lower affinity compared to the esters. Furthermore, the reverse ester of 4a (i.e., 6a) exhibited a moderate 1 to 3-fold increase in affinity at all three mAChRs. A similar increase in affinity was observed for the other reverse ester/amide series, i.e., 4b vs 6b and 5b vs 7b. Analog 6b was identified as the most potent compound at M₅ in this series.

In summary, a series of pethidine analogs was synthesized and evaluated to determine binding affinity for the [³H]NMS binding site on M₁, M₃, and M₅ human mAChRs expressed by CHO cell membranes. Compound 6b showed the highest binding affinity at M₁, M₃ and M₅ mAChRs (K_i = 0.67, 0.37, and 0.38 μM, respectively). However, this series of new analogs did not exhibit selectivity for M₅ mAChRs over M₁ and M₃ subtypes. Further SAR and pharmacological evaluations are needed to identify potent and selective M₅ mAChR antagonists. Additionally, pethidine has been reported to have weak μ-opioid receptor agonist activity.¹⁹ Thus, in future studies, once analogs are demonstrated have high affinity and selectivity for M₅ mAChRs, they will be evaluated also for μ-opioid receptor affinity to assure selectivity at the M₅ mAChR target.