Biol. Pharm. Bull. 30(8) 1593—1595 (2007)

of drugs through changes in clearance and distribution volume. In general, for drugs that show a high degree of hepatic or renal extraction (flow-limited drugs) after intravenous administration, alterations of plasma protein binding cause changes in the unbound plasma drug concentration at steady state but little change in the total concentration. Recently, these theoretical changes have been confirmed clinically for propofol in vivo in terms of alterations in the pharmacokinetics. It is important to characterize the binding characteristics of such drugs in order to predict and understand alterations of drug disposition and also interaction between drugs. In particular, elucidation of drug binding site(s) on plasma protein is useful for predicting the pharmacokinetic alterations caused by drug–drug interaction. Human a1-acid glycoprotein (AGP) is separable isoelectrophoretically into several variants, known as the F1/S, and A variants. These variants are genetically determined and can be explained by the presence of two different AGP genes: the AGP-A gene encoding the variants ORM1 F1 and S, and the AGP-B/B gene encoding the variant ORM2 A. Herve et al. reported that disopyramide and imipramine, and warfarin, bind specifically to A variant and F1/S variant mixture, respectively, whereas chlorpromazine and lignocaine bind specifically to both variants. Accordingly, it has been suggested that human AGP has two main genetic variants for drug binding, and that each respective variant has different binding characteristics. Three dimensional quantitative structure–activity model (3D-QSARs) analysis using CoMFA have been reported some common structural features and they showed A variant have two rings present in high affinity tricyclic drugs such as imipramine and high affinity of N-methylated amines. On the other hand, the binding site of F1/S variant to be a large hydrophobic area with no obvious structural requests for binding and they pointed out that the insufficient number of high affinity ligands for 3D-QSARs analysis. Warfarin is a well known model compound for studies of binding to F1/S variants. However, warfarin is supplied as a racemate and its binding to AGP shows stereoselectivity. The affinity constant of S-warfarin for human AGP is higher than that of the R-isomer. Furthermore, warfarin shows wide inter-individual variation in its pharmacokinetics, and binds mainly to albumin. Therefore other suitable marker compounds are required for studying the characteristics of drug binding to the AGP F1/S variant in clinical situations. Tamsulosin is an a1-adrenoceptor antagonist used for the pharmacological treatment of benign prostatic hyperplasia, acting in part by relaxing prostatic smooth muscle (Fig. 1). Tamsulosin shows marked binding to AGP and that in human patients there is a significant correlation between the concentration ratio of bound and unbound (Cb/Cu) tamsulosin and the plasma AGP level, but not between Cb/Cu and the plasma albumin level. As the binding of tamsulosin was not significantly affected by the presence of amitriptyline, which binds specifically to the AGP A variant, tamsulosin may bind to the AGP F1/S variants with high specificity. In the present study, we attempted to identify and quantify the characteristics of tamsulosin binding to human AGP variants.