Interdisciplinary approaches to drug discovery An academic approach

Abstract

Drug discovery has traditionally been based on a process of molecular roulette in which large numbers of compounds are tested for biological activity in animals. This is slow and expensive, and does not have very good odds for success. Typically, introduction of a new drug costs $100m–200m in research and development, and only about 40 new chemical entities are introduced each year despite an estimated $20 000m annual expenditure on pharmaceutical research worldwide. Increasingly, research on drug discovery involves highly integrated interdisciplinary teams, and the use of modern technology. Three themes can be discerned in drug discovery research: computer aided drug design, development of new therapeutic targets, and exploitation of new sources of lead compounds. Most benefits will result from synergistic interactions between these three themes. Computer based drug design can focus either on postulated targets or on postulated lead compounds. Although gene cloning provides a vast database of sequences of potential target proteins, modelling of structure from ah initio calculations has yet to be convincingly achieved. There have been advances based on crystal structures, but these are limited to enzymes (for example, thymidylate synthase and purine nucleoside phosphorylase) and DNA. Other approaches concentrate on lead compounds. When these are also proteins, a wide range of evolutionary analyses can be used to predict active site structures, leading to the design of simpler analogues. Gene cloning also provides the means to a molecular dissection of disease, revealing new targets for therapeutic intervention. Transgenic animals may offer human diseases in laboratory mice. Receptors, enzymes, and second messenger systems can be obtained in quantities sufficient to form the basis of new assays for testing compounds. The use of robotics and microelectronic biosensors enables large numbers of compounds to be tested routinely. However, such molecular assays are absolutely dependent on the correct choice of target, and the screening methods lose the subtlety of integrated biological systems. New compounds can come from biotechnology, from randomly generated pep tides, or from exploiting natural biodiversity. The therapeutic protein ‘revolution’ has still to arrive, but new approaches to natural products may be more promising. Combination of high throughput micro-screening technology with computer aided structural refinement should generate novel lead compounds.