ABSTRACT:Acidic phospholipid binding plays an important role in determining the tissue distribution of basic drugs. This article describes the use of surface plasmon resonance to measure binding affinity (K D ) of three basic drugs to phosphatidylserine, a major tissue acidic phospholipid. The data are incorporated into mechanistic tissue composition equations to allow prediction of the steady-state volume of distribution (V ss ). The prediction accuracy of V ss using this approach is compared with the original methodology described by Rodgers et al. (J Pharm Sci 94:1259-1276), in which the binding to acidic phospholipids is calculated from the blood/plasma concentration ratio (BPR). The compounds used in this study [amlodipine, propranolol, and 3-dimethylaminomethyl-4-(4-methylsulfanylphenoxy)-benzenesulfonamide (UK-390957)] showed higher affinity binding to phosphatidylserine than to phosphatidylcholine. When the binding affinity to phosphatidylserine was incorporated into mechanistic tissue composition equations, the V ss was more accurately predicted for all three compounds by using the surface plasmon resonance measurement than by using the BPR to estimate acidic phospholipid binding affinity. The difference was particularly marked for UK-390957, a sulfonamide that has a high BPR due to binding to carbonic anhydrase. The novel approach described in this article allows the binding affinity of drugs to an acidic phospholipid (phosphatidylserine) to be measured directly and demonstrates the utility of the binding data in the prediction of V ss .
IntroductionAlong with clearance, the volume of distribution at steady state (V ss ) is a key pharmacokinetic parameter that determines how long a drug remains in the body. Several approaches have been used to prospectively predict human V ss , including scaling in vivo data from preclinical species, scaling in vitro data, and various in silico calculation methods (Obach, 2007).An approach that has gained popularity in recent years is to use in vitro and physicochemical data for a particular compound to estimate its tissue/plasma partitioning and by accounting for the volumes of different tissues and their composition. This information is used to predict V ss of a compound in an integrated physiologically based manner (eq. 1).where V is physical tissue volume, T is tissue, Kp is tissue to plasma partition coefficient, and p is plasma.Prediction of Kp on the basis of tissue composition and the physicochemical properties of a compound was first described by Poulin and Theil (2000) for neutral compounds, with emphasis on lipophilicity and binding to neutral lipids. This approach has subsequently been used by various authors to predict V ss for different compound datasets (Jones et al., 2006;De Buck et al., 2007). An extension to this mechanistic PBPK model that incorporates pH and binding of the ionized base to acidic phospholipids as an additional component of tissue binding was proposed by Rodgers et al. (2005a), initially for ionized bases. The affinity of a compound t...