Antimicrobial peptides are an interesting class on ribosomally and non-ribosomally synthesised peptides found in all kingdoms of life. Among these is the underexplored class of bacteriocins: Short ribosomally synthesized and post-translationally modified peptides that often show high potency, and low toxicity. One subclass requires a glycosylation to show antimicrobial activity and thus have been termed glycoactive bacteriocins, Glycocins. The glycosylation of glycocins is installed by a specific subfamily of glycosyltransferases, part of the CaZy transferase family 2. Despite being in the same family, these glycosyltransferases show a remarkable diversity in terms of donor specificity, peptide specificity, acceptor selectivity and number of glycosylations that are carried out. I identified several putative Glycocins and their cognate transferases using bioinformatics. The recombinant production and purification of several glycosyltransferases was established. These were characterised in regard of the metal ion dependency and the carbonucleotide specificity. Using SPPS several glycocin-peptides were synthesized and their selective glycosylation by the cognate transferase could be proven. A facile and scalable method to obtain the mature, glycosylated, and folded glycocin by heterologous expression of the peptide and its transferase in E. coli was established. To investigate the molecular determinants for the observed differences in specificity in terms of preferred sugars and their acceptor selectivity, X-ray crystallography was used as method of choice. For four glycosyltransferases with noticeably different specificities, crystals could be obtained, and the structure could be solved. For the transferase of Enterocin 96 it was possible to obtain snapshots of the enzyme-substrate complex during its reaction.
