Modulating protein-protein interactions: from structural determinants of binding to druggability prediction to application.

During the last decades, a large amount of evidence has been gathered on the importance of protein-protein interactions in tuning and regulating most important biological processes. Since many of these pathways are deeply involved in diseases, extensive research efforts have been undertaken towards the modulation of protein-protein interactions. At the early stage of this challenge most of the attention was drawn to the drawbacks of such a therapeutic approach. Encouragingly, however, several recent studies provided a proof of concept that protein-protein interactions are actually valuable targets and that they do have a promising therapeutic potential. This review is divided into three sections. In the first section we summarize the general features of protein-protein interfaces, focusing on the characteristics that make them different from classical protein-ligand binding sites, as well as on problematic aspects that hamper the application of classical drug discovery approaches. In the second section, we present how some of the characteristics of protein-protein interactions can be exploited fruitfully in drug design, hence focusing on the druggability of protein-protein interfaces. Methods successfully applied to protein-protein interactions will be introduced, giving special attention to the computational ones. In the third section, three case studies are presented. First, we describe protein-protein interaction modulators targeting HDM2 and the computational methods applied to identify them. Next, we present the retrospective application of the discussed approaches on the well-examined target IL-2. We conclude with a prospective application to the NHR2 protein, a target just recently validated experimentally with the aid of computational methods.