Cell-based screening has been used in drug development, if the receptors are either not known, isolable, or functional upon isolation;or if cell-level responses are sought. Unresolved contributions of the drug disposition and receptor binding to the measured effects hamper lead structure optimization. Valuable ligands pass unnoticed through cellular assays, if slow transport and/or interactions with non-receptor cell constituents prevent the attainment of the effective drug concentration in the receptor surroundings. To account for the factors complicating the interpretation of cellular data, we will develop a structure-based computational tool, called the disposition function (DF), estimating the kinetics of intracellular drug disposition. To make the DF as general and practical as possible, conformation-averaged interactions during passive transport will be described using drug properties measured in surrogate systems, and conformation-specific binding to macromolecules will be expressed via 3-dimensional quantitative structure-activity relationships (3D-QSAR). Accumulation in phospholipid bilayers, composed of more than 50% phosphatidylcholine (PC) in mammals, comprises two distinct types of solvation - in the core and in the headgroup region. While hexadecane (C16) is a good surrogate solvent for the core, drug solvation in the headgroup regions, occupying about a third of the bilayer volume, is not understood properly. A novel surrogate solvent, diacetyl-PC (DAcPC, i.e. the PC headgroup with the truncated fatty acid chains), hydrated to the extent typical for a fluid bilayer, will be used to measure the headgroup-like solvation energies of hundreds of drugs. For all common drug fragments, the DAcPC and C16 solvation energies will be deconvoluted into fragment contributions, and adjusted to express the experimental headgroup and core quantities using conceptual correlations with the bilayer data. We hypothesize that for the maximum trans-bilayer transport rates, drugs must exhibit intermediate interaction affinities for the headgroups, core, and the interface between them. Based on this hypothesis, experimental transport kinetics data will be conceptually correlated with the headgroup and core solvation energies, to obtain a baseline form of the DF that estimates drug disposition for cellular systems, using just the lipid and protein contents. The baseline DF will be refined for representative G+ and G- bacteria, and two human cell lines using the uptake data and conceptual 3D-QSAR for conformation-specific binding to inert proteins. The DFs will be combined with current ligand-based and receptor-based QSAR techniques, developed for isolated receptor data, to provide cell-QSAR models suitable for processing cell-level bioactivities. The calibrated cell-QSAR models can be utilized by others as software or a web service to extract receptor affinities from cell-assay data, optimize drug structures individually for disposition and receptor binding, convert good binders to promising drug candidates, describe drug disposition in systems biology models, and annotate drug structures in PubChem and other databases by the estimates of accumulation in bilayer regions and intracellular disposition.
For many drugs, transport into the cells is one of the key processes for the therapeutic effect. The planned research will find out, how the rate and extent of the cellular entry depend on molecular structure and physicochemical properties of drug candidates. This information will be used do create software and a web service that will help other scientists to develop better and safer drugs faster.
Showing the most recent 10 out of 12 publications
