Compounds that bind to an enzyme generally contain a pharmacophore that is specific for the enzyme. The geometric criteria of a pharmacophore, deduced from the structure of known substrates, has been used to design and discover alternative substrates, which can behave as competitive inhibitors of the enzyme. Such inhibitors are lead drugs which can be developed into drug candidates by chemical modification, to modify solubility or toxicity for example. Methods have been developed to conduct 3D searches through structural databases for compounds with specific geometries. Such searches routinely produce compounds which behave as enzyme inhibitors. In an alternative approach, computer programs which can design chemical structures having predetermined geometric characteristics are being used to design enzyme inhibitors. Some 30 inhibitors or protein kinase C have been developed in this way. These inhibitors have an optimum lipophilicity and the nature of their binding to the enzyme has been characterized structurally. Typically between 5 and 12 hydrogen bonds, corresponding to binding energies of 5-25 kcal/mol are involved. Non-peptidic, competitive inhibitors of HIV protease have also been developed by these methods. These occupy the active site of the enzyme, forming hydrogen bonds to Asp25, Asp25', Ile50 and Ile50' in the enzyme dimer. Likewise, numerous inhibitors of HIV integrase have been identified. Such inhibitors are all active at the nM-mu M level in enzyme assays: a small proportion of the enzymatically active compounds are also active in cellular assays and these are being pursued as lead drug candidates.
