Accurate determination of the in vitro kinetic parameters Km (Michaelis constant) and Ki (inhibition constant) is critical for the quantitative prediction of in vivo drug clearance and the magnitude of inhibitory drug interactions. A cause of inaccuracy in vitro arises from the assumption that all drug added to an incubation mixture is available for metabolism or inhibition. Many drugs bind non-specifically to the membrane of the in vitro enzyme source.
The aims of this thesis were to: 1) investigate the comparative importance of lipophilicity (as log P), and pKa as determinants of the non-specific binding of drugs to human liver microsomes; 2) develop and validate an ANS fluorescence technique for measuring the non-specific binding of drugs to human liver microsomes; 3) characterise the non-specific binding of a large dataset of physicochemically diverse drugs using the ANS fluorescence procedure; 4) evaluate relationships between selected physicochemical characteristics and the extent of non-specific binding of drugs to human liver microsomes and; 5) computationally model the non-specific binding of drugs to discriminate between high binding (fu(mic) less than 0.5) and low binding (fu(mic) greater than 0.5) drugs.
The comparative binding of the basic drugs atenolol (log P = 0.1; fu(mic) = 1.00), of propranolol (log P = 3.1; fu(mic) = 0.36 - 0.84), and imipramine (log P = 4.8; fu(mic) = 0.42 - 0.82) suggested that lipophilicity is a major determinant of non-specific binding. In contrast, the comparative binding of diazepam (pKa = 3.3; fu(mic) = 0.69 - 0.80), a neutral compound; and the bases propranolol (pKa = 9.5; fu(mic) = 0.36 - 0.84) and lignocaine (pKa = 9.5; fu(mic) = 0.98), indicated that pKa was not a determinant of the extent of non-specific binding. The non-binding of lignocaine, a relatively lipophilic base, was unexpected and confirmed by the non-binding of the structurally related compounds bupivacaine and ropivacaine. These results implicated physicochemical characteristics other than lipophilicity and charge as important for the non-specific binding of drugs to human liver microsomes.
An assay based on 1-anilinonaphthalene-8-sulfonate (ANS) fluorescence was developed using the seven drugs employed in the initial study. Non-specific binding data from equilibrium dialysis and the ANS fluorescence methods were compared and a linear correlation (r2 = 0.92, p less than 0.01) was observed at drug concentrations of 100 and 200 micromolar. The approach was further validated by characterising the microsomal binding of nine compounds (bupropion, chloroquine, chlorpromazine, diflunisal, flufenamic acid, meclofenamic acid, mianserine, triflupromazine, and verapamil) using both binding methods (i.e. equilibrium dialysis and ANS fluorescence). A significant logarithmic relationship (r2 greater than or equal to 0.90) was demonstrated between fu(mic) and the modulus of ANS fluorescence for all drugs and for basic drugs alone at concentrations of 100 and 200 micromolar, while the acidic/neutral drugs showed a significant linear relationship (r2 greater than or equal to 0.84) at these two concentrations (p less than 0.01). The non binding of bupropion provided further evidence that physicochemical properties other than log P and charge were important for non-specific binding of drugs to human liver microsomes.
The ANS fluorescence technique was then used to characterise the non-specific binding of 88 physicochemically diverse compounds. In general, acids and neutrals bound to a low extent (fu(mic) greater than 0.5) whereas bases bound the full fu(mic) range (0.0001 to 1). Statistically significant relationships were observed between the non-specific binding of bases and log P, the number of hydrogen bond donors and hydrogen bond acceptors per molecule, and molecular mass.
Preliminary in silico modeling of the dataset generated by the ANS fluorescence technique, using the program ROCS, provided discrimination of all but one (itraconazole) of the high binding bases. However, there were 14 false positives, resulting in low overall prediction accuracy.
Taken together, the studies conducted in this thesis provide important insights into the physicochemical factors that determine the non-specific binding of drugs to human liver microsomes.
| Identifer | oai:union.ndltd.org:ADTP/202469 |
| Date | January 2008 |
| Creators | McLure, James Alexander, james.mclure@flinders.edu.au |
| Publisher | Flinders University. Medicine |
| Source Sets | Australiasian Digital Theses Program |
| Language | English |
| Detected Language | English |
| Rights | http://www.flinders.edu.au/disclaimer/), Copyright James Alexander McLure |
Page generated in 0.002 seconds