Abstract 2110 The effect of sequence on structure and microbial properties of insect antimicrobial peptides

The prevalence of antibiotic-resistant bacteria is rising as the number of newly approved antibiotics decreases, creating a global crisis of infections unable to be treated by antibiotics. Antimicrobial peptides (AMPs) are short proteins that act as a part of the innate immune system and naturally occur in a wide range of organisms. It is believed that the development of bacterial resistance to AMPs will be more challenging due to their mode of action, and are therefore under development to discover new classes of antimicrobial therapeutics. This work focuses on cecropins, a family of AMPs found in many insect species which have low toxicity to mammalian cells and diverse amino acid sequences. Cecropins typically contain a "hinge" region connecting the two ɑ-helical regions of the peptide. It has previously been shown that amino acid changes in this hinge region impact both the flexibility of the overall structure and the antimicrobial properties of the peptide. In this work, we analyzed 250 peptide sequences based on their amino acid properties, specifically categorizing peptides based on their hydrophobicity and amphiphilicity. A subset of peptides with divergent structure predictions were then synthesized in order to analyze their secondary structure and antibacterial properties. To gauge the efficacy of the representative peptides in preventing bacterial growth, using a variety of bacteria species, we performed minimum inhibitory concentration (MIC) assays, while the secondary structure and stability are analyzed by circular dichroism spectroscopy (CD). We discovered that even single amino changes in the peptide sequence can result in large differences in an observed MIC. Further, we have confirmed that these peptides are unfolded in solution, but obtain a primarily ɑ-helical structure when presented with solution conditions that mimic lipid membranes. Our long term goals are to identify relationships between peptide structure and antimicrobial activity, enabling the design of more effective AMPs. We are also using machine learning to discern underlying patterns in the sequences of divergent AMPs, ultimately enabling the prediction of novel peptides from genomic data that exhibit AMP characteristics. Thank you to Bucknell Presidential Fellowship Program and Cell Biology and Biochemistry Undergraduate Research Fund for funding this research.