Antimicrobial peptides (AMPs) are ubiquitous amongst living organisms and are part of the innate immune system with the ability to kill pathogens directly or indirectly by modulating the immune system. AMPs have potential as a novel therapeutic against bacteria due to their quick-acting mechanism of action that prevents bacteria from developing resistance. Additionally, there is a dire need for therapeutics with activity specifically against Gram-negative bacterial infections that are intrinsically difficult to treat, with or without acquired drug resistance. Development of new antibiotics has slowed in recent years and novel therapeutics (like AMPs) with a focus against Gram-negative bacteria are needed. We designed eight novel AMPs, termed PHNX peptides, using computational design (database filtering technology combined with the novel positional analysis on APD3 dataset of AMPs with activity against Gram-negative bacteria) and assessed their theoretical function using published machine learning algorithms, and finally, validated their activity in our laboratory. These AMPs were tested to establish their minimum inhibitory concentration (MIC) and half-maximal effective concentration (EC) under CLSI methodology against antibiotic resistant and antibiotic susceptible and . Laboratory-based experimental results were compared to computationally predicted activities for each of the peptides to ascertain the accuracy of the computational tools used. PHNX-1 demonstrated antibacterial activity (under high and low-salt conditions) against antibiotic resistant and susceptible strains of Gram-positive and Gram-negative bacteria and PHNX-4 to -8 demonstrated low-salt antibacterial activity only. The AMPs were then evaluated for cytotoxicity using hemolysis against human red blood cells and demonstrated some hemolysis which needs to be further evaluated. In this study, we successfully developed a design methodology to create synthetic AMPs with a narrow spectrum of activity where the PHNX AMPs demonstrated higher antibacterial activity against Gram-negative bacteria compared to Gram-positive bacteria. Thus, these peptides present novel synthetic peptides with a potential for therapeutic use. Based on our findings, we propose upfront selection of the peptide dataset for analysis, an additional step of positional analysis to add to the database filtering technology (DFT) method, and we present laboratory data on the novel, synthetically designed AMPs to validate the results of the computational approach. We aim to conduct future studies which could establish these AMPs for clinical use.
