Računalniško načrtovanje neribosomalne peptidne sintetaze za vključevanje 4-azido-L-alanina

Non-ribosomal peptidyl synthases (NRPS) are huge multimodular megaenzymes. Each of their modules incorporates one monomer, and that is how they synthesize a very diverse range of products in an assembly line fashion. The adenylation (A) domain was recognized very early on as a 'specificity guard' of the mechanism due to its role in the specific binding of monomers before their incorporation. Therefore, A domains have been of interest in the context of protein engineering since the discovery of NRPSs. Recent work by the host group has shown that with a single point mutation in the active site of an A domain, they can successfully incorporate 4-azido alanine into a peptide. However, since worse specificity and much lower production was observed than in the wild type, the aim of this work was to try to improve these properties by introducing multipoint mutations in the active site. In order to make this otherwise demanding task manageable, a new automated protocol was developed. In the first step, diverse and relatively stable candidates are generated using the FuncLib tool. In the next step, complexes with the monomer and cofactors are prepared for all the candidates and are evaluated according to the interaction energies. The most promising complexes are then simulated using molecular dynamics (MD) simulations, and a more detailed affinity analysis is then performed on the trajectory using the MMPBSA method. In this specific case, the vast majority of candidates showed a stable conformational dynamics, looking at the trajectories. Unfortunately, time ran out for a more detailed analysis with MMPBSA, and we are still waiting for the results of experimental testing.