acase@tulane.edu / Development of antimicrobial peptides (AMPs) as next generation clinical antibiotics has been a pursuit of the scientific community for several decades. AMPs are attractive drug candidates because of their potent antibacterial activity and a low propensity for eliciting antibiotic resistant bacterial phenotypes. However, despite substantial efforts and myriad development approaches, AMPs have yet to make inroads in the clinic due to toxicity concerns and activity loss in vivo. We hypothesized that eukaryotic cytotoxicity and antibacterial activity loss are intricately related in that peptide-induced tissue or host cell damage corresponds to depletion of free peptide available to target bacterial cells. Using human red blood cells (RBCs) as a model eukaryotic cell, we demonstrate that a cross-section of AMPs lose appreciable antibacterial activity when preincubated with concentrated eukaryotic cells (1x109 red blood cells/mL) and that this behavior can be explained by plasma membrane binding. To approach this problem in a unique manner, we synthesized a combinatorial peptide library based on the potent AMP, ARVA, and screened the library for activity in the presence of concentrated RBCs. We isolated nine unique, but similar sequences from the screen. During the screening program, we discovered that RBC-peptide interactions lead to peptide degradation through the release of cytosolic proteases. We used this knowledge to design a consensus sequence based on the nine peptides isolated from the library screen and synthesized it using only D-isomer amino acids. The novel peptide displays excellent antimicrobial activity against several human pathogens in the presence and absence of concentrated RBCs, has reduced toxicity towards eukaryotic cells, and is not susceptible to cleavage by cellular proteases. We attempted to use this peptide, D-NOGCON, to combat P. aeruginosa in a mouse model of acute pneumonia, but were unable to ameliorate the negative outcomes associated with infection. We ultimately suggest alternative models of bacterial infection in which the peptide may be more effective and future approaches to further refining the sequence of D-NOGCON. / 1 / Charles Starr