Massively-parallelized discovery and optimization of antimicrobials

Abstract

The number of newly approved antimicrobial compounds has been steadily decreasing over the past 50 years emphasizing the need for novel antimicrobial substances. Besides, current antimicrobial therapies become increasingly ineffective due to the rapid spread of drug-resistant pathogens, and even the most potent small-molecule antibiotics may eventually fail to cure infections with highly resistant bacteria. Peptides are a promising class of potential drug substances, as they are still underexplored, offer high structural complexity, and at least partly cover the chemical space left open between biologics and small molecule drugs. Ribosomally produced antimicrobial peptides are of particular interest, as they already act as effective antimicrobials in the defense against invading pathogens of most organisms. Consequently, they are currently being investigated for potential use to fight infectious diseases in humans. The fact that antimicrobial peptides are produced using transcription and translation makes them particularly attractive as they can be produced recombinantly and their modification can be simply achieved by straightforward manipulation of the synthesis template at the DNA level. In this thesis, we developed a high-throughput method for the discovery of antimicrobial peptides which is based on the principle of self-screening: Different peptides are expressed in a recombinant Escherichia coli strain, and their effect on growth rate is recorded. In our case, this is done by next-generation sequencing of expression plasmids in the bacterial culture, enabling the recording of massive numbers of growth curves for single strains in a single flask. We termed this technology massively parallelized growth assays (Mex). We applied this method to discover novel candidates for antimicrobial peptides by screening a library of ~12’000 naturally occurring peptides with a length between 5 and 42 amino acids and diverse properties. Analysis of thousands of growth curves allowed us to identify more than 1,000 previously unknown antimicrobials. Additionally, by incorporating the kinetics of growth inhibition, we were able to obtain a first indication of the mode of action, with important implications for the ultimate usefulness of the peptide in question. We chemically synthesized the most promising peptides of the screen and determined their activity when applied externally. Notably, the results indicated that 10 out of 15 investigated peptides efficiently eradicated bacteria at a minimal inhibitory concentration in the upper nM / lower µM range. We think that this work represents a step-change in the high-throughput discovery of functionally diverse antimicrobial peptides. Next, we applied a simplified version of Mex to optimize a single antimicrobial peptide in high-throughput. As a model peptide, we optimized the 23 amino acid version of the well-researched and highly active antimicrobial peptide Bac71-23 using deep mutational scanning, consisting of a first random and then a semi-rational approach. The random library of ~600,000 different Bac71-23 variants allowed us to derive a fitness landscape of the peptide and to identify residues that are essential for growth inhibition and residues with potential for activity optimization. A smaller semi-rational library of ~160,000 Bac71 23 variants enabled us to extract the most beneficial amino acid combinations, thereby generating an antimicrobial peptide that, if synthesized chemically, is non-toxic and superior to Bac71-23 against a large panel of bacterial pathogens. We thus created a new-to-nature peptide lead with a great potential to be further developed in pre-clinical stages. To our best estimation, these novel methods exceed competing approaches in terms of throughput, hit-rate and sensitivity, while also offering an opportunity for the direct functional characterization of large libraries. These methods will accelerate the discovery and optimization of antimicrobials drastically, and may thus provide a path forward to master one of today’s most urgent challenges, the antimicrobial resistance crisis.