Janssen M J, Ensing K, de Zeeuw R A
Groningen Institute for Drug Studies (GIDS), University Centre for Pharmacy, Department of Analytical Chemistry and Toxicology, The Netherlands.
Pharmazie. 2000 Feb;55(2):102-6.
In this article, the binding affinities of the fluorescent-labeled benzodiazepines described in Part 1 are compared to assess the influence of the labeling position and the choice of fluorophore on the binding affinity. This comparison was extended by taking into account the data of other fluorescent-labeled benzodiazepines in the literature. The differences in the binding affinities observed could partly be explained by structure-activity relationships (SAR). On the basis of this comparison, fluorescent-labeled desethylflumazenil (Ro15-3890, 19) derivatives were selected as the most suitable labeled ligands in fluorescent receptor assays. A methyl-methoxycoumarin derivative (Mmc-O-CO-(CH2)3-Ro15-3890) (20b) had a Ki-value of 6.5 nM, and a 7-nitrobenz-2-oxa-1,3diazole derivative (NBD-NH-(CH2)3-Ro15-3890, 21) had a Ki-value of 5.7 nM. In order to yield sufficient sensitivity in the final receptor assay, a suitable fluorescent labeled ligand should have a Ki < 10 nM. A further advantage of the above two ligands is that the benzodiazepine moiety has no receptor affinity of its own. Thus, if some hydrolysis of the labeled ligand were to occur, the resulting Ro15-3890 (18) would hardly affect the outcome of the assay. In the second part of this paper the prerequisites of the fluorophore are being examined. In this regard, 20b is preferred, because the coumarin derivative has higher fluorescence intensities in aqueous media than the NBD-derivative. Therefore, 20b was selected as a fluorescent-labeled ligand in the development of a non-radioactive receptor assay for benzodiazepines.