Structural and biochemical studies on tailoring enzymes involved in ribosomal peptides biosynthesis

Human history is also a history of the ceaseless fight against various diseases, especially those caused by pathogenic infections. Both scientists and pharmaceutical companies have been dedicated in developing antimicrobial compounds to combat the micro-enemies. Despite the efforts, the resistant strains keep emerging with the introduction of new therapeutic compounds, which implies an urgent need for compounds with modified or novel scaffolds to overcome the resistance. One good source of antimicrobial compound discovery is natural products. And research on their biosynthesis is of great importance in understanding and also utilizing nature’s tools to create bioactive compounds and variants. One major class of bioactive natural products is ribosomally synthesized and post-translationally modified peptides (RiPPs). Precursor peptides and tailoring enzymes for post-translational modifications (PTMs) are encoded in the same gene cluster, which is efficient in producing peptide variants providing different precursor sequences. The peptides are first synthesized on ribosomes and then undergo extensive modifications by the tailoring enzymes to generate active, rigid and stable products. Elucidation of the biosynthetic process, especially the enzymatic mechanisms in PTMs provide the guidelines of engineering either the peptides or the tailoring enzymes to produce more effective and stable RiPP variants or derivatives. Lanthipeptides are RiPPs bearing characteristic lanthionine (Lan) and/or methyllanthionine (MeLan) bridges. Many of them display potent antimicrobial activities, such as the class I lanthipeptides nisin and microbisporicin. However, the mechanism of one key step in their biosynthesis – dehydration – was still mysterious, until recently researchers in Wilfred van der Donk group found class I lanthipeptide dehydratase catalyzes the reaction in a tRNA-dependent manner. To understand the structural basis of dehydration reaction, I determined the crystal structures of two class I lantibiotic dehydratases involved in the biosynthesis of nisin and microbisporicin, respectively. The structures supported the proposed two-step mechanism of dehydration reaction, illustrated the decisive factors for peptide substrate recognition, and provided clues on the tRNA-binding site based on a docking model. Plantazolicin is a linear azol(ine)-containing peptide (LAP) that specifically inhibits the growth of anthrax causing pathogen Bacillus anthracis. It contains two patches of thiazole and (methyl)oxazole/methloxazoline rings that endow its rigidity. It has an Nα, Nα-dimethyl-arginine at the N-terminus, which is critical for the antibacterial activity. The N-methyltransferase responsible for this modification in the gene cluster is very selective to thiazole/(methyl)oxazole substrates with an N-terminal arginine. By studying its crystal structures in complex with truncated substrate analogs as well as steady-state kinetics, I demonstrated the determinants of the substrate specificity and key residues involved in interacting with substrate. Cyanobactins are a group of small cyclic RiPPs produced by symbiotic or asymbiotic cyanobacteria. They contain rigid thiazoles and/or (methyl)oxazoles similar to LAP and additional head-to-tail cyclization on backbone to decrease its flexibility and susceptibility to proteases. Many cyanobactins are subsequently decorated with isoprenoid groups on Tyr, Ser or Thr residues for improved hydrophobicity. To understand the substrate requirements for cyanobactin prenyltransferases, I performed structural study of selected cyanobactin prenyltransferases with isoprenoid donor and various acceptors to illustrate the substrate selectivity. Examination of the substrate spectrum further supported the proposed substrate requirements based on structural information