"Nonribosomal peptides are a highly important class of natural products and display a broad spectrum of bioactivity, ranging from antimicrobial and immunosuppressant to anticancer. They are synthesized by non-ribosomal peptide synthetases (NRPSs), a diverse group of large multifunctional enzymes that operate as molecular assembly lines. They consist of modules that are responsible for coupling, processing and modification of an extending chain of amino acid monomers, heterocyclic rings and N-methylated residues. Re-programming of these unique biosynthetic routes should lead to a range of novel products of pharmaceutical importance. Previous studies which have focused on module swapping led to so called chimeric assembly-line enzymes. However most of those chimeras are heavily impaired due to the drastic changes applied, which could partially be compensated by site directed mutagenesis (SDM). SDM was also used to directly alter the specificity of several domains (e.g. the A-domain of TycA). Still no high throughput in vitro assay to evolve NRPS systems is available. We will therefore adapt the Megavalent Bead Surface Display (BeSD) system that has recently been developed for the directed evolution of protein binders to monitor peptide forming reactions by C-domains. In contrast to widely-used in vivo systems (such as bacterial, phage and yeast display) it is able to deal with larger library sizes and is not limited by host compatibility. BeSD will be the first in vitro system that is able to rank enzymes according to their bond forming efficiency since it can be coupled to fluorescence-activated cell sorting (FACS). Finally variants with altered specificities will be incorporated in modular assembly lines. This work will open the road to analyze larger, more diverse libraries of potential drug candidates. It is expected to be a crucial step towards the generation of novel NRPSs and is going to enrich our knowledge on how these multi-enzyme complexes work."