Semi-synthesis of nisin-based peptide antibiotics

Abstract

There is a growing need for novel antibiotics since there are more and more cases of infections caused by resistant bacteria. Possible novel antibiotics are antimicrobial peptides, especially the lantibiotic nisin. Lantibiotics are ribosomally synthesized cationic peptides that contain several unnatural amino acids like dehydroalanine (Dha), dehydrobutyrine (Dhb) and have multiple cyclic structures derived by thio ether bonds or lanthionines, which are very important for its antimicrobial activity. Nisin can inhibit the cell-wall synthesis by binding to lipid II (an important lipid in the bacterial cell-wall) with its N-terminal A and B rings. After binding, nisin is able to form a pore-complex by inserting its C-terminus (C, D/E rings) into the bacterial cell membrane resulting in a collapse of vital ion gradients. Although nisin holds a promising template for novel peptide-based antibiotics, synthesis of nisin-like structures is not trivial and especially the introduction of the lanthionine rings is challenging. Moreover, the lanthionine bridges are oxidation sensitive and the approach as described in this research to replace the lanthionines by dicarba bonds via ring-closing metathesis (RCM) could be used to access stabilized nisin-like structures via (chemical) synthesis. In the first part of this thesis, the design, synthesis and evaluation of nisin fragments is described. Native nisin fragments were prepared by using optimized enzymatic and chemical cleavage methods, and two previously unknown cleavage sizes were discovered. Furthermore, through solution and solid phase peptide synthesis, mimics of nisin fragments were successfully prepared replacing the lanthionines by dicarba bonds using RCM. Native and synthetic nisin fragments were assembled via copper-I catalyzed click chemistry to obtain full-length nisin hybrids and their biological activities were evaluated. This orthogonal ligation strategy via click chemistry was also successfully applied on full-length native nisin to ligate several conjugation partners, to obtain biologically active fluorescent nisin derivatives, which are important tools to further study nisin’s mode of action. Finally, a lipophilic amino acid was designed, synthesized and incorporated into the short antimicrobial peptide anoplin as a model peptide. This strategy demonstrated an approach to improve the antimicrobial activity by increasing the lipophilicity of anoplin, while retaining its selectivity for bacterial membranes, and could potentially be applied as a general strategy to improve the activity of membrane-acting peptides. The synthesized nisin dicarba AB analogs proved to be biologically active and the importance of the native backbone structure was shown. The nisin hybrids were biologically active, however, they lacked pore-formation activity possibly due to a non-optimal C-terminus. The results of this thesis provide valuable information about the importance of each individual ring structure, which gives more insight into the potential of nisin to improve its metabolic stability and thereby its antibiotic properties.