3-Aryl-3-arylmethoxyazetidines. A new class of high affinity ligands for monoamine transporters

Abstract

A series of 3-aryl-3-arylmethoxy-azetidines were synthesized and evaluated for binding affinities at dopamine and serotonin transporters. The 3-aryl-3–arylmethoxyazetidines were generally SERT selective with the dichloro substituted congener 7c (K_i = 1.0 nM) and the tetrachloro substituted derivative 7i (K_i = 1.3 nM) possessing low nanomolar affinity for the SERT. The 3-(3,4-dichlorophenyl-3-phenylmethoxyazetidine (7g) exhibited moderate affinity at both DAT and SERT transporters and suggests that substitution of the aryl rings can be used to tune the mononamine transporter affinity.

To date there is no pharmacological therapies for psychostimulant-dependence or the adverse side effects associated with craving and withdrawal. A variety of medications have been investigated as potential treatment strategies, however, none have been identified as having significant promise.¹ The failure of these drugs to be effective may be due to the limited scope of action, targeting single monoaminergic systems. To date there is mounting evidence that in addition to dopaminergic systems, brain serotonergic systems also modulate responses in psychostimulant-induced behaviors.^2–4 A single dopaminergic or serotonergic agent may not adequately attenuate the behavioral effects associated with psychostimulant abuse. It has been suggested by Rothman and colleagues that the development of an appropriately calibrated dual acting DAT/SERT agent may be more effective as a medication than an agent selective for a single transporter.⁵ This model of psychostimulant addiction suggests that drug-induced dopamine and serotonin dysfunction contribute to the withdrawal symptoms, drug craving, and relapse. The model further postulates that decreased levels of synaptic dopamine during stimulant withdrawal are the source of anhedonia and psychomotor retardation. Moreover, decreased levels of synaptic serotonin results in depression, obsession, and lack of impulse control. Based upon this rationale, it should be possible to treat stimulant addicts that exhibit depleted synaptic levels of dopamine and serotonin with medications capable of restoring dopaminergic and serotonergic tone to disrupted neuronal systems. In this vein, both releaser-type drugs or uptake inhibitors could be developed.^5,6 To date, dopamine/norepinephrine uptake inhibitors (e.g. buproprion, methylphenidate), SSRIs (e.g. fluoxetine, paraoxetine) and SNRIs (e.g. duloxetine) and NRIs (e.g. reboxetine) have been widely prescribed for depression, ADHD and obesity and have a good safety record.^7–11 This would suggest that a dual uptake inhibitor could be a promising pharmacological target for the treatment of psychostimulant addiction.

Our efforts to develop novel molecular scaffolds targeting monoamine transporter systems has led to the discovery of the 3α-arylmethoxy-3β-aryltropanes (1) as a unique class of monoamine transporter ligands that possess tunable affinity for dopamine and serotonin transporters.¹² This has prompted a broader examination of the structure-activity relationships of this molecular scaffold in search of compounds with dual affinity for dopamine and serotonin transporters. It was of interest to explore condensed ring systems that would accommodate the pharmacophore requirements while reducing the overall molecular weight and lipophilicity inherent to the tropane derivatives. To this end, we identified the 3-arylmethoxy-3-arylazetidines (2) as viable targets for synthesis and biological evaluation at monoamine transporters.

The azetidine ring system has recently become an attractive molecular scaffold for the development of CNS active compounds.¹³ Preliminary computational studies revealed that the replacement of the tropane ring system with an azetidine scaffold would lead to significant decrease in molecular weight and lipophilicity (cLogP values, Table 1).¹⁴ In addition, superposition of the of predicted favorable solvated conformers of 1 and 2 (Figure 1)¹⁵ suggested that the azetidine scaffold would lead to a favorable alignment of the major structural elements of the two compounds.

Table 1.

Monoamine Transporter Affinity and Selectivity

Cmpda	Code	R1	R2	R3	cLogPb	DAT (Ki, nM)c	SERT (Ki, nM)c	DAT/SERT
1ad	HK2-151	H	H	H	3.67	117 ± 19	247 ± 27	0.47
1bd	HK3-77	H	H	4-Cl	4.27	22 ± 8.0	6.1 ± 0.50	3.6
1cd	HK3-45	H	H	3,4-Cl2	4.87	16 ± 1.0	0.061 ± 0.024	258
6a	ANT-I-35	H	H	H	3.02	4,860 ± 450	1,230 ± 70	4.0
6be	ANT-I-74	H	H	4-Cl	3.63	3,610 ± 25	2,590 ± 20	1.4
6ce	ANT-I-57	H	H	3,4-Cl2	4.23	2,820 ± 110	7.3 ± 0.70	390
6d	ANT-I-129	H	4-Cl	H	3.63	3,180 ± 360	825 ± 8.0	3.9
6e	ANT-I-87	H	4-Cl	4-Cl	4.23	3,830 ± 42	3.5 ± 0.20	1,100
6f	ANT-I-92	H	4-Cl	3,4-Cl2	4.83	3,770 ± 320	8.1 ± 3.2	470
6g	ANT-I-110	H	3,4-Cl2	H	4.23	1,300 ± 86	208 ± 8.0	6.3
6h	ANT-I-126	H	3,4-Cl2	4-Cl	4.83	3,020 ± 340	2.9 ± 0.30	1,000
6i	ANT-I-108	H	3,4-Cl2	3,4-Cl2	5.44	3,670 ± 30	4.2 ± 0.30	870
7a	ANT-I-47	CH3	H	H	3.40	1,730 ± 320	73 ± 9.4	24
7b	ANT-I-73	CH3	H	4-Cl	4.01	3,910 ± 280	4.0 ± 0.3	980
7c	ANT-I-72	CH3	H	3,4-Cl2	4.61	2,820 ± 110	1.0 ± 0.20	2,800
7d	ANT-I-85	CH3	4-Cl	H	4.01	1,410 ± 107	39 ± 7.0	36
7e	ANT-I-84	CH3	4-Cl	4-Cl	4.61	2,030 ± 600	7.8 ± 3.3	490
7f	ANT-I-125	CH3	4-Cl	3,4-Cl2	5.22	976 ± 61	3.0 ± 1.1	330
7g	ANT-I-133	CH3	3,4-Cl2	H	4.61	620 ± 140	23 ± 1.6	27
7h	ANT-I-124	CH3	3,4-Cl2	4-Cl	5.22	2,290 ± 513	4.8 ± 2.2	630
7i	ANT-I-106	CH3	3,4-Cl2	3,4-Cl2	5.82	436 ± 66	1.3 ± 0.0	340

^aCompounds were tested as the hydrochloride salts unless otherwise noted.

^bSee reference 13.

^cAll values are the mean ± SEM of three experiments preformed in triplicate.

^dData taken from reference 12.

^eCompound was tested as the oxalate salt.

Figure 1.

Superimposed predicted favorable solvated conformers of 1 (green) and 2 (cyan).

As illustrated in Scheme 1, the syntheses of a series of chlorinated target compounds were achieved from commercially available N-Boc-3-azetidinone (3). The chlorinated derivatives were selected as initial targets since the corresponding tropane analogues (1a–1c, Table 1) had demonstrated a broad range of DAT/SERT selectivity. Introduction of the substituted 3-aryl group was achieved by addition of a preformed aryl lithium reagent to the ketone moiety of 3.¹⁶ To minimize degradation the azetidine ring the work-up was performed under weakly acidic conditions [10 % NH₄Cl (aq.)] at cold temperatures (5–10 °C). This afforded the alcohols 4 in high yields (86–94%).

Scheme 1.

Reagents and conditions: (i). R²-C₆H₄Br, n-BuLi, THF, −78 °C. (ii)R³-C₆H₄CH₂Br, Bu₄NBr, 4N NaOH/CH₂Cl₂, reflux. (iii) 2M HCl/EtOAc, rt. (iv) LiAlH₄, THF, 65 °C.

To complete the synthesis of the initial series of target compounds, the alcohols 4 were alkylated with a variety of substituted benzyl bromides using phase-transfer conditions to furnish the N-Boc-3-arylmethoxy-3-arylazetidines (5). Typically the azetidines 5 were not isolated and purified but rather converted into the 3-arylmethoxy-3-arylazetidine hydrochloride salts 6 or the N-methyl-3-arylmethoxy-3-arylazetidine salts 7. Treatment of 5 with 2M HCl in EtOAc at room temperature afforded the corresponding hydrochloride salts 6 in 21–64% overall yield. Alternatively, the N-methyl derivatives 7 were prepared in 18–55% overall yield by heating 5 with 1M lithium aluminum hydride in THF at 65 °C. After work-up with Glauber’s salt the N-methyl derivatives were treated with 2M HCl in EtOAc at room temperature to give the hydrochloride salts. In general, we found that direct conversion of the azetidine bases to the hydrochloride salts gave stable solid compounds. In the few exceptions that solid hydrochloride salts were not formed, the acidic oils were treated with sodium carbonate solution and the resulting free based azetidine was converted into the oxalate salt. Many of the salts gave crystalline compounds which allowed the structure to be confirmed unequivocally by X-ray crystallography (Figure 2, 6a).¹⁷

Figure 2.

X-ray crystal structure of 6a

Binding affinities for the dopamine and serotonin transporters were determined by the ability of the drug to displace the radiolabeled ligands [³H]WIN 35,428,and [³H]citalopram, respectively, from the monoamine transporters in rat brain tissue using previously reported assays.¹⁸ The monoamine transporter binding affinities of the 3-arylmethoxy-3-aryl-azetidines (6) were generally selective for the SERT over the DAT exhibiting nanomolar affinity for the SERT and micromolar affinity for the DAT. The dichloro substituted congeners 6e (K_i = 3.5 nM) and 6h (K_i = 2.9 nM) were the most potent of the azetidines at SERT. The binding affinities of N-methyl 3-arylmethoxy-3-aryl-azetidines 7 also exhibited high affinity for the SERT. The N-methyl analogues typically exhibited higher affinity at the SERT than the corresponding unsubstituted analogues 6. The dichloro substituted congener 7c (K_i = 1.0 nM) and the tetrachloro substituted derivative 7i (K_i = 1.3 nM) were the most potent analogues of the entire series. However, the NH analogues 6e and 6h exhibited similar SERT affinity to the corresponding N-methyl analogues 7e and 7h, while the DAT affinity of 6e and 6h was not similarly affected. Overall, N-methyl, 3,4-dichloro analogue 7c exhibited the greatest selectivity for SERT (DAT/SERT = 2800) of all the azetidine derivatives.

Inspection of the SAR of the azetidines 6 and 7, revealed that for compounds with the 3,4-dichlorophenyl group attached as the 3-aryl substituent (e.g. 6g and 7g) the relative DAT affinity was improved and the DAT/SERT selectivity was decreased. A similar effect had been observed for a series of meperidine derivatives.¹⁹ However, if the 3-arylmethoxy moiety possessed a chloro substituent (e.g. 6h and 7h) then the DAT affinity decreased and the SERT selectivity dominated. Alternatively, if the 3-arylmethoxy moiety possessed the 3,4-dichloro substituent pattern (e.g. 6c and 7c) then the ligand exhibited high SERT affinity. However, when both aryl groups possessed the 3,4-dichlorophenyl moiety there was little improvement in affinity or selectivity. The effect of the 3-(3,4-chlorophenyl-methoxy) moiety on SERT affinity and selectivity is consistent with the SAR observed for the 3α-arylmethoxy-3β-aryltropanes.¹² Among the tropane congeners, the 3β-(3,4-chlorophenylmethoxy) derivatives (e.g. 1c) typically exhibited some of the highest SERT affinity and selectivity for the class of ligands.

Of the azetidine derivatives, 7g exhibited a binding profile which most closely approached that of a dual DAT/SERT uptake inhibitor (DAT/SERT = 27) while maintaining nanomolar affinity at both transporters. Although SERT selective, the moderate affinity for the DAT observed for 7g (K_i = 620 nM) was encouraging. While the SAR of the azetidines did not mirror the SAR of the tropanes, it was clear that substitution of the aryl rings can be used to tune the monoamine transporter affinity and suggests that the proper substitution of the aryl rings might lead to a potent DAT/SERT dual inhibitor.

In conclusion, we have synthesized a novel class of azetidine-based monoamine transporter ligands. In general, the 3-aryl-3-arylmethoxyazetidines 6 and 7 exhibited high affinity for the SERT and were generally selective for the SERT over the DAT. However, it is clear that DAT affinity can be improved with the proper substitution of the 3-aryl ring. Based upon these preliminary studies, the azetidine scaffold seems to be suitable replacement for the tropane scaffold and the 3-aryl-3-arylmethoxy-azetidines are well suited for the development of new compounds that display a broad spectrum of multi-targeted monoamine transporter affinity and selectivity.