Better Drugs by Evolution

This thesis describes four years of work toward the selection of natural product-like macrocyclic peptides (MPs) by bacteriophage display. It all began with a fascination for certain potent MPs that can be found in nature. With their backbones constrained into (bi-)cyclic structures including various non-peptidic modifications, these bio-chemical hybrid compounds possess improved pharmacological properties compared to their linear and purely peptidic counterparts. Notably, they display higher binding affinity, selectivity, stability, and membrane permeability, and can furthermore bind to protein interfaces that are generally considered to be 'undruggable'. As such, hybrid MPs are a promising class of compounds for the development of new pharmaceuticals and/or chemical probes. During my PhD we worked on strategies to speed up the discovery process for hybrid MP lead compounds. In nature, the structures of these compounds are fine-tuned over eons of time by evolution. We reasoned that for drug discovery efforts, it would also be highly efficient to apply an evolutionary approach and enable the selection of natural product-like MP binders for various targets of interest from large genetically-encoded phage-displayed peptide libraries in the laboratory. Toward this end, we developed highly selective and mild phage-compatible chemistry to be able to cyclize peptides displayed on whole bacteriophages. Using a wide range of synthetic scaffolds for cyclization, diverse libraries of hybrid MPs can be obtained. Furthermore, the incorporation of specific synthetic scaffolds that are known to weakly interact with certain protein targets is expected to tailor pharmacological properties and form a better starting point for drug discovery.