A series of 1-(4-fluorophenyl)-1H-indoles substituted at the 3-position with 1-piperazinyl, 1,2,3,6-tetrahydro-4-pyridinyl, and 4-piperidinyl was synthesized. Within all three subseries potent dopamine D-2 and serotonin 5-HT2 receptor affinity was found in ligand binding studies. Quipazine-induced head twitches in rats were inhibited by most derivatives as a measure of central 5-HT2 receptor antagonism. Piperazinyl and tetrahydropyridyl indoles were cataleptogenic, while piperidyl substituted indoles surprisingly were found to be noncataleptogenic or only weakly cataleptogenic. Noncataleptogenic piperidyl derivatives also failed to block dopaminergic-mediated stereotypies, that is methyl phenidate-induced gnawing behavior in mice. These profiles resemble that of the atypical neuroleptic clozapine. 1-Ethyl-2-imidazolidinone was found to be the optimal substituent of the basic nitrogen atom in order to avoid catalepsy. The atypical neuroleptic 1-[2-[4-[5-chloro-1-(4-fluorophenyl)-1H-indol-3-yl]-1-piperidinyl] ethyl]-2-imidazolidinone (sertindole, compound 14c) was selected for further development as a result of these structure/activity studies.

A series of 3-(4-fluorophenyl)-1H-indoles substituted in the 1-position with 4-piperidinyl, 1,2,3,6-tetrahydro-4-pyridinyl, and 4-piperazinyl was synthesized. By variation of the substituents in the benzene part of the indole nucleus in 1-[2-[4-[3-4-fluorophenyl)-1H-indol-1-yl]-1-piperidinyl]-ethyl]-2- imidazolidinones, the highest 5-HT2 receptor affinity and selectivity with respect to dopamine D2 receptors and alpha 1 adrenoceptors were obtained by 5-methyl substitution. Different substituents were introduced in the 1-position of the piperidine ring in the 5-methyl-substituted derivative. Thus replacement of the 2-(2-imidazolidinon-1-yl)ethyl side chain with a 2-(1,3-dimethyl-1-ureido)ethyl or methyl substituent resulted in unchanged affinity and selectivity for 5-HT2 receptors. Replacement with a 2-[3-(2-propyl)-2-imidazolidinon-1-yl]ethyl side chain reduced binding to alpha 1 adrenoceptors with a factor of four, while affinities for 5-HT2 and D2 receptors were retained, compared to the 3-unsubstituted imidazolidinone. Indoles substituted in the 1-position with 4-piperazinyl had generally weaker affinity for both 5-HT2 and D2 receptors compared to corresponding 4-piperidinyl- and 1,2,3,6-tetrahydro-4-pyridinyl-substituted indoles. Introduction of a methyl group in the 2-position of the 5-methyl-substituted indole resulted in further increase of selectivity for the 5-HT2 receptor. Compounds with potent receptor binding also potently inhibited the quipazine-induced head twitch syndrome in rats. The compounds were equally active after oral and subcutaneous administration and showed a long duration of action (> 24 h). In general, the derivatives were found to be considerably more potent at 24 h than at 2 h after the administration. The compounds within this series were prepared as analogues of the previously described 1-(4-fluorophenyl)-3-(4-piperidyl)-1H-indoles by interchange of the C-3 carbon atom and the nitrogen atom in the indole nucleus. The pharmacological results indicate that this isosteric replacement results in higher selectivity for 5-HT2 receptors compared to the former series. The 1-[2-[4-[2,5-dimethyl-3-(4-fluorophenyl)-1H-indol-1-yl]-1- piperidinyl]ethyl]-2-imidazolidinone has high affinity for 5-HT2 receptors (IC50 = 3.4 nM) and extremely low affinity for both dopamine D2 receptors (IC50 = 6900 nM) and alpha 1 adrenoceptors (IC50 = 2300 nM).

A series of trans-1-piperazino-3-phenylindans were synthesized with the goal of replacing their established neuroleptic profile with that of peripheral 5-hydroxytryptamine (5-HT2) antagonism. Compounds with an unsubstituted or fluoro-substituted 6-position in the indan ring, and which had a five- or six-membered heterocyclic ring attached by an ethylene chain to the piperazine ring, satisfied this objective. Some of the compounds had potent antihypertensive activity in conscious, spontaneously hypertensive rats (SHR). In pithed rats they antagonized the pressor effect induced by 5-HT in doses 100-1000 times lower than doses needed to antagonize the pressor effect of phenylephrine. The effect was stereoselective and associated with enantiomers with 1R,3S absolute configuration. 1S,3R enantiomers inhibited the uptake of dopamine and norepinephrine in vitro. The compound with the best antihypertensive activity was (+)-(1R,3S)-1-[2-[4-[3-(4-fluorophenyl)-1-indanyl]-1- piperazinyl]ethyl]-2-imidazolidinone (Lu 21-098, irindalone). Its pharmacological profile resembled that of the standard compound ketanserin. There was a close structural correspondence between ketanserin and irindalone in a conformation that we recently identified as a D-2 receptor-relevant configuration of its neuroleptic "parent" tefludazine. This suggests that the dopaminergic (D-2) and the serotonergic (5-HT2) pharmacophores are structurally closely related.

Octoclothepin (1) was resolved into its R and S enantiomers via the diastereomeric tartaric acid salts. The enantiomers were shown to be of high optical purity by 1H NMR with use of the chiral shift reagent (R)-(-)-2,2,2-trifluoro-1-(9-anthryl)ethanol. Pharmacological and biochemical testing confirmed that (S)-1 is the more potent dopamine (DA) D-2 antagonist both in vitro and in vivo, although the R enantiomer still has significant D-2 antagonistic activity. In contrast, both enantiomers were equally active in test models detecting activity at D-1 receptors, serotonin-2 (5-HT2) receptors and alpha 1 adrenoceptors. Contrary to a previous prediction, it was found that norepinephrine (NE) uptake inhibition was confined solely to the S enantiomer. Overall, (S)-1 has a "classical" neuroleptic profile, while the R enantiomer has a more "atypical" profile. These pharmacological profiles seem to be in agreement with the reported clinical profiles of the two enantiomers. A previous conformational study was revised in light of the biochemical test results with enantiomers of known optical purity. Their relative D-2 receptor affinity corresponded well with the calculated conformational energy difference between their "active conformations" deduced from a previously reported new D-2 receptor model. Also the high enantioselectivity of (S)-1 at the NE uptake site could be explained after a detailed conformational analysis showing strict requirements for the orientation of the piperazine lone-pair direction at the NE uptake site.

Eight indolizidine and quinolizidine derivatives of 3-PPP were synthesized and tested for possible dopamine (DA) autoreceptor activity. The equatorial indolizidine derivative 19e had the profile of a selective autoreceptor agonist and was half as active as 3-PPP. However, resolution of the compound revealed that the 8R enantiomer was an unselective DA agonist with a profile similar to (+)-3-PPP, while the 8S enantiomer was a weak DA antagonist without any DA agonist activity. The unsaturated quinolizidine derivative 21 also had the profile of a DA antagonist while the axial quinolizidine derivative 18a had an amphetamine-like profile in 6-OHDA-lesioned rats. All other derivatives were inactive. The observed structure-activity relationships were in agreement with existing DA receptor models, although these models are not apparently detailed enough to explain why the 8S enantiomer of 19e is inactive as a DA agonist.