Elucidating the Mechanism of Action of Unnatural Amino Acid Containing Antimicrobial Peptides in Membrane Environments

Organism resistance continues to develop to the currently available antimicrobial compounds necessitating the development of innovative new therapeutic compounds with different specificities and mechanisms of action that provide acceptable therapeutic indices. Unnatural amino acid containing antimicrobial peptides could provide a novel avenue for the development of therapies with improved efficacy and pharmacokinetics over natural amino acid containing peptides which are prone to protease degradation.  Molecular dynamics (MD) simulations of antimicrobial peptides containing unnatural amino acids have been performed using explicit water and multiple model membrane types in all-atom simulations. The structural properties of peptides were investigated using both the canonical and isothermal-isobaric ensembles to further understand the mechanism through which the collections of AMPs exert their in vitro activity.   Simulations with micelle membrane models were conducted at 300 K to correlate with experimental circular dichroism (CD) data showing the secondary structure the peptides adopt in the presence of an electrostatic membrane model. Analysis of the stabilized MD trajectory reflects peptide structural consistency with experimental data.  Simulations of the peptides with bilayer model membranes were conducted at the physiologically relevant 310 K to correlate with experimental cellular activity data which demonstrated the antimicrobial activity of the peptides without providing insight into the mechanism through which the activity was achieved. Long time scale simulations have noted distinct differences between bilayers in the presence of AMPs as compared to those without the peptide. Mixed bilayers with an anionic charge modeled bacterial membranes while a confluent zwitterionic bilayer modeled the mammalian membrane.  This research has demonstrated that force field parameters for unnatural amino acids can be derived from QM calculations. FF parameters derived from structures identified from a DFT approach have also been used to expand the AMBER ff03 force field. The FF parameters were able to able to model the interaction of the peptides which contain unnatural amino acids. The data is consistent with NMR data and further supported with CD spectroscopy.  Ph.D