Structure-based drug design (SBDD) is the process of using information (obtained by X-ray crystallography) about the three-dimensional shapes of proteins to model the interaction of protein molecules with small molecules (ligands) that are potential drugs.  In this way, structural biologists can design molecules that fit the ligand-binding site of a protein in competition with the natural ligand, thereby altering the protein’s function.  Specific inhibitors that fit the active site (the position where catalysis occurs) of a catalytic enzyme can be obtained. A helpful analogy is to think of using the inner shape of a lock to build a key that selectively fits just that lock.  By de novo design and optimization, an inhibitor molecule can be made to complement the ligand-binding cavity of a protein in terms of its shape and chemical properties.  Furthermore, an important advantage of this technology over trial-and-error methods is that cross-reactivity or side effects of drugs can be minimized.

Another approach is to conduct automated computer-based (virtual) screening of libraries containing millions of known compounds by docking them, one at a time, at a ligand-binding site and calculating a score for the goodness of fit.  Further optimization is carried out to obtain as snug a fit as possible.  The compounds with the highest scores can then be synthesized or purchased, assayed for inhibitory activity, and their effectiveness in animals determined.  Toxicity tests and other trials can proceed for promising drug candidates.

A technique for empirically relating the activities of potential drug candidates with their chemical structures is known as quantitative structure-activity relationships or QSAR.  When the activities of several compounds are known, the activities of related drug candidates can be predicted by QSAR analysis. Thus, virtual screening and QSAR provide two high-throughput SBDD tools for generating a large number of candidate drugs specific for a particular target protein. When the 3D structure of a target is available or can be obtained by using a similar protein as a model, these tools can be used to obtain lead compounds for initiating the rational drug development process. There are several success stories of SBDD to date, none perhaps more dramatic than the discovery (at Merck & Co.) of HIV protease inhibitors for the treatment of AIDS. Other examples are given at: www.nigms.nih.gov/Publications/structure_drugs.htm

At HWI, scientists are working on several proteins that are candidates for the SBDD approach. The figure above illustrates virtual screening and docking of compounds at the active site of human steroid sulfatase, a target for breast and prostate cancer drug design.