New Imide 5-HT1A Receptor Ligands – Modification of Terminal Fragment Geometry

Abstract

Two sets of new o-methoxyphenylpiperazine (MPP; series a) and 1,2,3,4-tetrahydroisoquinoline (THIQ; series b) derivatives, containing various imide moieties derived from NAN190, were synthesized and evaluated in vitro for their ability to bind to the serotonin 5-HT_1A and 5-HT_2A receptors. All new derivatives from series a demonstrated high 5-HT_1A affinities, whereas THIQ analogues were much less active. With respect to 5‑HT_2A receptors, three MPP derivatives presented moderate activity but the rest of the investigated compounds were practically inactive. The influence of changes in terminus geometry on 5-HT_1A receptor affinity was analyzed in regard to model compounds NAN190 and MM199.

Introduction

During the last decade a large number of structurally different compounds have been proposed as 5‑HT_1A receptor ligands. Among these, 4-substituted long chain 1-arylpiperazines (LCAPs) have been extensively studied. In particular, much effort has been devoted to the role of the terminal part in a ligand-receptor interaction, and in consequence, a great many different fragments were used [1]. The imide containing fragments constituted one of the most thoroughly investigated termini type and among others, compounds like buspirone (anxiolytic drug; 5-HT_1A partial agonist [2]) or NAN190 (postsynaptic antagonist [3,4]) were found (Figure 1).

Figure 1. Structure and binding constants (K_i) of the selected 5-HT_1A receptor ligands.

Figure 1. Structure and binding constants (K_i) of the selected 5-HT_1A receptor ligands.

Our systematic SAR studies of the long chain 5-HT_1A ligands resulted in identification of several new imide derivatives with interesting biological activity. Two NAN190 analogues: MM77 (postsynaptic antagonist [5]) and its constrained form MP349 (full antagonist [6,7]) exhibited anxiolytic-like activity in some animal models [7,8]. Replacement of 1-pyrimidinylpiperazine fragment of buspirone with 1,2,3,4-tetrahydroisoquinoline (THIQ) led to the compound MM199 [9] (Figure 1), for which anxiolytic-like and antidepressant-like effects in rats were described [10]. Continuing exploration of this type of agents we synthesized two sets of new o-methoxyphenylpiperazine (MPP) and THIQ derivatives and evaluated their affinity for 5-HT_1A and 5-HT_2A receptors.

Results and Discussion

The structures of NAN190 and its simplified analog MM77 were chosen as lead molecules for modification of imide terminus. All the new compounds were examined in vitro for their ability to displace [³H]-OH-DPAT and [³H]-ketanserine binding to rat 5-HT_1A and 5-HT_2A receptors respectively, as described in the Experimental Section. The results are presented in Table 1, where additionally, the affinities of parent compounds and previously reported derivatives 1b, 2a, 2b and 4b are also included [9,11].

Table 1. Structure and 5-HT_1A and 5-HT_2A binding affinities of compounds 1–7.

No.	Imide moiety	a:						b:
		Ki ± SEM (nM)						Ki ± SEM (nM)
		5-HT1A				5-HT2A		5-HT1A				5-HT2A
1		0.55±0.14a,b				451±65		355±28b				2770±130
2		4.0±0.2b				109±22		50±11b				880±12
3		1.7±0.5				185±42		157±17				1780±25
4		6.4±0.3c				1510±95		2920±90d				NT
5		6.8±0.5				690±40		453±15				4000±30
6		7.5±0.7				835±45		690±20				4400±40
7		46±3				86±2		620±18				1230±34

^a K_i value according to Glennon et al. was 0.58 nM [3]. ^bK_i value from ref 11. ^c ref 5. ^d ref 9.

Table 1. Structure and 5-HT_1A and 5-HT_2A binding affinities of compounds 1–7.

No.	Imide moiety	a:						b:
		Ki ± SEM (nM)						Ki ± SEM (nM)
		5-HT1A				5-HT2A		5-HT1A				5-HT2A
1		0.55±0.14a,b				451±65		355±28b				2770±130
2		4.0±0.2b				109±22		50±11b				880±12
3		1.7±0.5				185±42		157±17				1780±25
4		6.4±0.3c				1510±95		2920±90d				NT
5		6.8±0.5				690±40		453±15				4000±30
6		7.5±0.7				835±45		690±20				4400±40
7		46±3				86±2		620±18				1230±34

^a K_i value according to Glennon et al. was 0.58 nM [3]. ^bK_i value from ref 11. ^c ref 5. ^d ref 9.

All four newly synthesized MPP derivatives (3a and 5a–7a) showed a high affinity for 5-HT_1A receptors (K_i = 1.7–46 nM), however, lower than that of NAN190. This is thus another confirmation that the flat phthalimide terminus optimally binds to the respective receptor region. All modifications of this fragment, namely saturation (2a and 3a), removal of the aromatic ring (4a), changes in its position (5a and 6a) or enlargement (7a) decreased the observed 5-HT_1A receptor affinity. Compound 7a, the least active at the 5-HT_1A sites, was found to be the most potent 5-HT_2A ligand. All the other MPP derivatives displayed a moderate to low affinity for 5-HT_2A receptors (K_i = 109–1510 nM), and except for 2a, were at the same time at least 100-fold selective 5-HT_1A agents. This is not unexpected, since it is known that the MPP system prefers 5-HT_1A binding sites.

The THIQ moiety has often been used as a replacement for the arylpiperazine fragment in our studies on the role of individual pharmacophoric groups of LCAPs. Although THIQ itself does not show any affinity for the 5-HT_1A receptors (K_i > 50 000 nM) [12] its substitution with butyl-azaspiro[4.5]decane-7,9-dione (MM199, Figure 1) increased the affinity to nanomolar level [9]. The same was observed in the case of the 1-adamantoyloaminobutyl THIQ derivative, for which K_i = 0.95 nM was determined [13]. It was then concluded that bulky alicyclic termini should be especially well accommodated by the 5-HT_1A binding site. This finding is additionally confirmed by the results obtained for the second group of imide derivatives b. Indeed, the highest 5-HT_1A receptor affinity was observed for the alicyclic systems (2b and 3b), whereas aromatic groups were less favorable, although they were better than the small succinimide alone (4b, K_i = 2920 nM). Thus, the replacement of MPP by a THIQ fragment caused a decrease of 5-HT_1A affinity and we did not obtain any more active compounds than the previously examined 2b [11]. A more detailed comparison of terminal fragments of MM199, 2b and 3b reveals that their geometry may play a role, and that a bent configuration of the cycloalkyl part (Figure 2) is preferred. With regard to 5-HT_2A receptors, we did not detect a significant activity in the investigated THIQ derivatives.

Figure 2. The geometry of terminal fragments: azaspiro[4.5]decane-7,9-dione (a), cis- (b) and trans-1,2-cyclohexanedicarboxyimide (c).

Figure 2. The geometry of terminal fragments: azaspiro[4.5]decane-7,9-dione (a), cis- (b) and trans-1,2-cyclohexanedicarboxyimide (c).

As a result of present investigation three new highly active 5-HT_1A ligands were found (3a, 5a and 6a), which displayed a 100-fold selectivity over 5-HT_2A receptors. During the preparation of this manuscript, Hackling and coworkers published a paper in which compound 7a was investigated as a dopamine D₂ and D₃ receptor ligand [14]. In our hands this derivative was a dual 5-HT_1A/5-HT_2A ligand and based on the data obtained by Hackling, it is also a potent but unselective D₂/D₃ agent (K_i = 40 and 29 nM, respectively). Pharmacological work is currently in progress to evaluate the 5-HT_1A functional profile of these four compounds.

Experimental

General

Melting points were determined on a Boetius apparatus and are uncorrected. ¹H-NMR spectra were obtained on a Varian EM-360L (60 MHz) in the CDCl₃ solution with Me₄Si as an internal reference. Elemental analyses were performed in the Institute of Organic Chemistry PAS (Warsaw, Poland), and were within 0.5% of the theoretical values. The syntheses of 2-(4-aminobutyl)-1,2,3,4-tetrahydroisoquinoline [9] and 4-(4-aminobutyl)-1-(2-methoxyphenyl)piperazine [15] have been previously reported.

General method for the preparation of compounds 3a, 3b, 5a, 5b, 6a, 6b, 7a and 7b.

Equimolar amounts (1 mmol) of the respective, commercially available acid anhydride and 4-(4-aminobutyl)-1-(2-methoxyphenyl)piperazine or 2-(4-aminobutyl)-1,2,3,4-tetrahydroisoquinoline were refluxed in xylene (20 mL) for 3 h. After the mixture was allowed to cool to room temperature, the xylene was evaporated under reduced pressure and the products were isolated on a silica gel column. Free bases were converted into the hydrochloride salts in CHCl₃ or acetone solution by treatment with excess of Et₂O saturated with gaseous HCl.

4-{4-[2-(trans-1,2-Cyclohexanedicarboxyimido)]butyl}-1-(2-methoxyphenyl)piperazine (3a).

Oil, yield (base) 90%; M.p. (salt) 192-194°C (acetone); R_f = 0.45 (CHCl₃/MeOH 19:1); ¹H-NMR (base): δ (ppm) = 7.1-6.7 (m, 4H, arom), 3.9 (s, 3H, OCH₃), 3.7-3.3 (m, 2H), 3,3-2.9 (m, 4H), 2.9-2.1 (m, 8H), 2.1-1.2 (m, 12H); C₂₃H₃₃N₃O₃·2HCl (472.03), Calcd. C 58.52, H 7.90, N 8.90; Found C 58.37, H 7.52, N 8.90.

1-(2-Methoxyphenyl)-4-[4-(3-phenylsuccinimido)butyl]piperazine (5a).

Oil, yield (base) 76%; M.p. (salt) 205-207°C (acetone); R_f = 0.46 (CHCl₃/MeOH 19:1); ¹H-NMR (base): δ (ppm) = 7.5-7.1 (m, 5H, arom), 7.1-6.7 (m, 4H, arom), 4.2-3.8 (m, 1H), 3.8 (s, 3H, OCH₃), 3.8-3.3 (m, 2H), 3,3-2.2 (m, 12H), 2.9-1.3 (m, 4H); C₂₅H₃₁N₃O₃·2HCl (494.464), Calcd. C 60.72, H 6.72, N 8.49; Found C 60.24, H 6.90, N 8.43.

1-(2-Methoxyphenyl)-4-[4-(3-phenylmaleicimido)butyl]piperazine (6a).

Oil, yield (base) 56%; M.p. (salt) 209-210°C (acetone); R_f = 0.53 (CHCl₃/MeOH 19:1); ¹H-NMR (base): δ (ppm) = 7.5-7.0 (m, 5H, arom), 7.0-6.8 (m, 4H, arom), 4.1-3.6 (m, 3H), 3.8 (s, 3H, OCH₃), 3,3-2.9 (m, 4H), 2.9-2.1 (m, 10H), 1.9-1.3 (m, 4H); C₂₆H₃₃N₃O₃·2HCl (508.49), Calcd. C 61.41, H 6.93, N 8.26; Found C 61.00, H 6.99, N 8.08.

1-(2-Methoxyphenyl)-4-[4-(1,8-naphthalimido)butyl)]piperazine (7a).

Oil, yield (base) 84%; M.p. (salt) 280-281°C (acetone); R_f = 0.52 (CHCl₃/MeOH 19:1); ¹H-NMR (base): δ (ppm) = 8.6(d, J=8Hz, 2H, arom), 8.2 (d, J=8Hz, 2H, arom), 7.7 (t, J=8Hz, arom), 7.1-6.7 (m, 4H, arom.), 4.5-4.0 (m, 2H), 3.9 (s, 3H, OCH₃), 3,3-2.9 (m, 4H), 2.9-2.3 (m, 6H), 2.1-1.5 (m, 4H); C₂₇H₃₂ N₃O₃·2HCl (552.926), Calcd. C 58.65, H 5.83, N 7.60; Found C 58.83, H 6.36, N 7.63.

2-{4-[2-(trans-1,2-Cyclohexanedicarboxyimido)]butyl}-1,2,3,4-tetrahydroisoquinoline (3b).

Oil, yield (base) 79%; M.p. (salt) 168-170°C (acetone); R_f = 0.49 (CHCl₃/MeOH 19:1); ¹H-NMR (base): δ (ppm) = 7.3-6.9 (m, 4H, arom), 3.7-3.3 (m, 4H) 3.1-2.0 (m, 8H), 2.0-1.2 (m, 12H); C₂₁H₂₈N₂O₂·HCl·0.5H₂O (385.919), Calcd. C 65.35, H 7.83, N 7.26; Found C 64.99, H 7.89, N 7.08.

2-[4-(3-Phenylsuccinimido)butyl]-1,2,3,4-tetrahydroisoquinoline (5b).

Oil, yield (base) 81%; M.p. (salt) 127-129°C (acetone); R_f = 0.39 (CHCl₃/MeOH 19:1); ¹H-NMR (base): δ (ppm) = 7.6-7.2 (m, 5H, arom), 7.2-7.0 (m, 4H, arom), 4.1-3.3 (m, 5H), 3.3-2.3 (m, 8H), 2.1-1.3 (m, 4H); C₂₃H₂₆N₂O₂ · HCl · 0.25H₂O (403.439), Calcd. C 68.47, H 6.87, N 6.94; Found C 68.47, H 6.89, N 6.85.

2-[4-(3-phenylmaleicimido)butyl]-1,2,3,4-tetrahydroisoquinoline (6b).

Oil, yield (base) 47%; M.p. (salt) 121-123°C (acetone); R_f = 0.43 (CHCl₃/MeOH 19:1); ¹H-NMR (base): δ (ppm) = 7.6-7.2 (m, 5H, arom), 7.2-6.9 (m, 4H, arom), 4.2-3.5 (m, 5H), 3.3-2.0 (m, 10H), 1.9-1.4 (m, 4H); C₂₄H₂₈N₂O₂·HCl·1.25H₂O (435.482), Calcd. C 66.19, H 7.29, N 6.43; Found C 66.43, H 7.26, N 6.16.

2-[4-(1,8-naphthalimido)butyl]-1,2,3,4-tetrahydroisoquinoline (7b).

Oil, yield (base) 40%; M.p. (salt) 265-267°C (acetone); R_f = 0.43 (CHCl₃/MeOH 19:1); ¹H-NMR (base): δ (ppm) = 8.6 (d, J=8Hz, 2H, arom), 8.2 (d, J=8Hz, 2H, arom), 7.7 (t, J=8Hz, 2H arom), 7.3-6.9 (m, 4H, arom), 4.5-4.0 (m, 2H, CH₂), 3.8-3.5 (m, 2H), 3,1-2.4 (m, 6H), 2.1-1.5 (m, 4H); C₂₅H₂₄ N₂O₂· HCl·0.5H₂O (429.942), Calcd. C 69.84, H 6.09, N 6.51; Found C 70.17, H 5.63, N 6.17.

Radioligand binding studies

The in vitro affinity of the investigated compounds for 5-HT_1A and 5-HT_2A receptors was assessed on the basis of their ability to displace [³H]-8-OH-DPAT (222 Ci/mmol, Amersham, England) and [³H]-ketanserin (66.4 Ci/mmol, NEN Chemicals, USA), respectively. Radioligand binding experiments were carried out on rat brain using tissues from the hippocampus for 5-HT_1A receptors, and from the cortex for 5-HT_2A receptors, according to the previously published procedures [16].

K_i values were determined from at least three competitive binding experiments in which 10–14 sample concentrations, run in triplicate, were used. The Cheng and Prusoff [17] equation was used for K_i calculations.