The purinome, a complex mix of drug and toxicity targets

Much attention has focused on the development of protein kinases as drug targets to treat a variety of human diseases including diabetes, cancer, hypertension and arthritis. To date, Gleevec is one example of a drug targeting protein that has successfully treated human cancer. Several other protein kinase inhibitors are in clinical development. However, protein kinases are in fact part of a larger collection of some 2000 distinct proteins expressed by the genome that like the protein kinases also bind purines (the purinome), either to be utilized as substrates or as co-factors in the form of NAD, NADP and co-enzyme A. The solution structures of many representative gene family members within the purinome show these proteins bind purines in a similar orientations to that observed in all protein kinases. Several non-protein kinase purine utilizing proteins are established drug targets such as HMG CoA reductase, dihydrofolate reductase, phosphodiesterase and HSP90. Searches of OMIM identifies many purine utilizing enzymes that are associated with inborn errors in metabolism. Inhibition of any one of which by a drug could lead to an undesirable side effect. The purinome is therefore somewhat of a drug discovery mixed blessing. It is a rich source of therapeutic targets, but also contains a large collection of diverse proteins whose inhibition could result in an adverse outcome. Drug discovery within the purinome should therefore encompass strategies that enable broad assessment of selectivity across the entire purinome at the earliest stages of the discovery process. In this article we review the purinome within the context of drug discovery and discuss approaches for avoiding off target binding during the discovery/lead optimization process with particular emphasis on use of proteome mining technology.