5-Aza, 6-aza, 7-aza and 8-aza-phthalazinone, and 5,8-diazaphthalazinone templates were synthesised by stereoselective routes starting from the appropriate pyridine/pyrazine dicarboxylic acids by activation with CDI, reaction with 4-chlorophenyl acetate ester enolate to give a β-ketoester, which was hydrolysed, and decarboxylated. The resulting ketone was condensed with hydrazine to form the azaphthalazinone core. The azaphthalazinone cores were alkylated with N-Boc-D-prolinol at N-2 by Mitsunobu reaction, de-protected, and then alkylated at the pyrrolidine nitrogen to provide the target H(1) receptor antagonists. All four mono-azaphthalazinone series had higher affinity (pK(i)) for the human H(1) receptor than azelastine, but were not as potent as the parent non-aza phthalazinone. The 5,8-diazaphthalazinone was equipotent with azelastine. The least potent series were the 7-azaphthalazinones, whereas the 5-azaphthalazinones were the most lipophilic. The more hydrophilic series were the 8-aza series. Replacement of the N-methyl substituent on the pyrrolidine with the n-butyl group caused an increase in potency (pA(2)) and a corresponding increase in lipophilicity. Introduction of a β-ether oxygen in the n-butyl analogues (2-methoxyethyl group) decreased the H(1) pA(2) slightly, and increased the selectivity against hERG. The duration of action in vitro was longer in the 6-azaphthalazinone series. The more potent and selective 6-azaphthalazinone core was used to append an H(3) receptor antagonist fragment, and to convert the series into the long acting single-ligand, dual H(1) H(3) receptor antagonist 44. The pharmacological profile of 44 was very similar to our intranasal clinical candidate 1.
Copyright © 2012 Elsevier Ltd. All rights reserved.