Further Advances in Optimizing (2-Phenylcyclopropyl)methylamines as Novel Serotonin 2C Agonists: Effects on Hyperlocomotion, Prepulse Inhibition, and Cognition Models

Abstract

A series of novel compounds with two halogen substituents have been designed and synthesized to further optimize the 2-phenylcyclopropylmethylamine scaffold in the quest for drug-like 5-HT2C agonists. Compound (+)-22a was identified as a potent 5-HT2C receptor agonist, with good selectivity against the 5-HT2B and the 5-HT2A receptors. ADMET assays showed that compound (+)-22a possessed desirable properties in terms of its microsomal stability, and CYP and hERG inhibition, along with an excellent brain penetration profile. Evaluation of (+)-22a in animal models of schizophrenia-related behaviors revealed that it had a desirable activity profile, as it reduced d-amphetamine-stimulated hyperlocomotion in the open field test, it restored d-amphetamine-disrupted prepulse inhibition, it induced cognitive improvements in the novel object recognition memory test in NR1-KD animals, and it produced very little catalepsy relative to haloperidol. These data support the further development of (+)-22a as a drug candidate for the treatment of schizophrenia.

Figure 1.

Structures of lorcaserin (1), vabicaserin (2) and compounds 3–6.

Figure 2.

Representative drugs having a multi-halogen substitution pattern and the general structure of new 2-phenylcyclopropylmethylamines.

Figure 3.

Effects of compound (+)-22a on d-amphetamine (AMPH)-stimulated hyperlocomotion in the open field. (A) Locomotor activity in the open field in 5-min blocks reflecting baseline responses (0–15 min) and AMPH stimulated activity with its reduction by (+)-22a (15–120 min). (B) Cumulative baseline locomotor activities (0–15 min). (C) Cumulative AMPH-stimulated activities and reductions in activity with (+)-22a (16–90 min). N = 7–10 mice/treatment; *, p < 0.05, compared to the Veh + Veh group; #, p < 0.05, compared to the Veh + AMPH group.

Figure 4.

Compound (+)-22a restores d-amphetamine (AMPH)-disrupted PPI in C57BL/6 mice. (A) Mice treated with vehicle-vehicle (Veh + Veh) showed prepulse dependency in the PPI response. Four mg/kg (+)-22a alone did not affect PPI relative to the Veh + Veh controls. Mice administered Veh + AMPH had a marked reduction in PPI. Treatment with 0.5, 1, or 2 mg/kg (+)-22a restored overall AMPH-disrupted PPI to the levels of the Veh + Veh group; with 1 mg/kg fully rescuing prepulse dependency of the response. Mice given the 4 mg/kg dose + AMPH showed a reduction in PPI similar to that found in Veh + AMPH animals. N = 8–11 mice/treatment; *, p < 0.05, compared to the Veh + Veh control; #, p < 0.05, compared to the Veh + AMPH group; ‡, p < 0.05, within treatment 8 or 12 dB compared to the 4 dB response; §, p < 0.05, within treatment 12 dB compared to the 8 dB response. (B) Pulse-only responses during PPI testing were not affected by treatment in the (+)-22a+Veh or the Veh + AMPH groups but when (+)-22a was given before AMPH the startle-only response was attenuated in dose-dependent manner. N = 8–11 mice/treatment; *, p < 0.05, compared to the Veh + Veh control; #, p < 0.05, compared to the Veh + AMPH group; +, p < 0.05 compared to 4 mg/kg (+)-22a + Veh treatment; †, p < 0.05, compared to the 4 mg/kg (+)-22a + AMPH treatment.

Figure 5.

Compound (+)-22a rescues novel object recognition memory in NR1-KD mice. Mice were given vehicle (Veh) or 0.5 or 1 mg/kg (+)-22a 30 min before training or at testing for long-term memory (LTM). (A) At testing for LTM, Veh-treated wild-type (WT) controls displayed a preference for the novel object, whereas NR1-KD mice showed no preference for either object. In contrast, 1 mg/kg (+)-22a rescued this deficit in the mutants while producing some impairment in the WT animals. (B–C) As a control, the duration of object exploration was examined during (B) training and (C) LTM testing. At training and at testing Veh-treated NR1-KD mice spent more time exploring the objects than WT mice, while (+)-22a reduced this exploration time to levels that were not significantly different from the WT controls. N = 9–10 mice/genotype/treatment; *, p < 0.05, (+)-22a compared to the Veh within genotype; ^x, p < 0.05, NR1-KD compared to the WT vehicle.

Figure 6.

Comparison of catalepsy induced by haloperidol and (+)-22a in C57BL/6 mice with the horizontal bar test. Separate groups of mice were injected with vehicle (Veh), or 0.01, 0.1, 1, or 10 mg/kg haloperidol (Hal), or 0.01, 0.1, 1, 10, or 30 mg/kg compound (+)-22a and evaluated for catalepsy 60 min later. N = 8–10 mice/treatment; *, p < 0.05, compared to the Veh control; ^V, p < 0.05, compared to all other groups.

Scheme 1.

Synthesis of target compounds 20a–g, 21a–g, and 22a–g.^{a a} Reagents and conditions: (a) Ph₃P=CHC(O)N(OMe)Me, CH₂Cl₂, rt, overnight; (b) Me₃S⁺(O)I⁻, NaH, DMSO, rt, overnight; (c) DIBAL-H, THF, −78 °C, 2 h; then NaBH₄, MeOH, 0 °C to rt, 0.5 h; (d) phthalimide, PPh₃, DEAD, THF, rt, overnight; (e) N₂H₄−H₂O, EtOH, reflux, 3h; then Boc₂O, Et₃N, CH₂Cl₂, rt, 0.5 h; (f) BBr₃, CH₂Cl₂, −78 °C to rt; then Boc₂O, Et₃N, CH₂Cl₂; (g) 2-fluoroethanol, Ph₃P, DEAD, THF, 0 °C to 60 °C, 1h; (h) allyl bromide, Cs₂CO₃, DMF; (i) (1) chiral preparative-HPLC; (2) 2M HCl in Et₂O, rt, 24 h. 11–22a: 4, 5-diF; 11–22b: 5, 6-diF; 11–22c: 4-F, 5-Cl; 11–22d: 5-Cl, 6-F; 11–22e: 4, 5-diCl; 11–22f: 5, 6-diCl; 11–22g: 5-F, 6-Cl.