Discovery of antimicrobials by massively parallelized growth assays (Mex)

Philipp Koch, Steven Schmitt, Mathias Cardner, Niko Beerenwinkel, Sven Panke, Martin Held

TL;DR

Researchers have developed a high-throughput screening method called "Mex" that leverages E. coli self-screening and next-generation sequencing to discover novel antimicrobial peptides from a library of over 10,000 naturally occurring sequences. This approach has identified over 1000 previously unknown antimicrobial peptides, with a significant portion showing potent activity at low micromolar or nanomolar concentrations, offering a promising new avenue for antibiotic discovery.

What Problem Does This Paper Address?

The pipeline for discovering new antibiotics has been drying up for decades, leading to a critical shortage of effective treatments against increasingly resistant bacteria. While natural compounds, particularly peptides, hold immense promise due to their structural diversity and biological relevance, traditional methods for screening large peptide libraries are often slow, costly, and inefficient. This paper tackles the challenge of rapidly and effectively identifying novel, potent antimicrobial peptides from vast natural sources.

Methods

The core innovation is the Mex (Massively parallelized growth assay) method. Here's a breakdown:

  • Library Generation: They started with known antimicrobial peptides (PARENTS) from the Antimicrobial Peptide Database (APD) and used sequence similarity searches (tblastn) to identify thousands of related sequences (SIMILARS) from genomic databases. This created a diverse library of over 10,000 peptides.
  • DNA Synthesis and Cloning: These peptide sequences were reverse-translated into E. coli-optimized DNA sequences and synthesized as oligonucleotides. These were then cloned into a plasmid under the control of an inducible promoter (PBAD).
  • E. coli Self-Screening: The plasmid library was transformed into E. coli TOP10 cells. When the peptide expression was induced, any antimicrobial peptide produced intracellularly would inhibit the growth of the host cell.
  • Next-Generation Sequencing (NGS): The abundance of each peptide-encoding DNA sequence was tracked over time using NGS. A decrease in the relative abundance of a specific DNA sequence indicated that the corresponding peptide was inhibiting the growth of its host cell.
  • Growth Curve Analysis: By analyzing the NGS read counts at different time points, the researchers could construct approximate growth curves for each peptide-expressing strain. This allowed them to identify "Mex-active" peptides that significantly reduced growth.
  • Mode of Action Clues: The kinetics of growth inhibition (how quickly it occurred) provided initial insights into the peptide's mechanism of action (e.g., targeting intracellular components vs. membrane damage).
  • Validation: Promising Mex-active peptides were chemically synthesized and tested in standard MIC (Minimum Inhibitory Concentration) assays, as well as assays for membrane damage and hemolysis, to confirm their activity and assess toxicity.

What's novel here is the combination of massively parallelized growth assays with NGS for direct readout, enabling the screening of a huge number of peptides simultaneously and providing kinetic data for initial mechanism-of-action insights.

Datasets

  • Antimicrobial Peptide Database (APD): Used as the source for initial "PARENTS" antimicrobial peptides.
  • NCBI non-redundant nucleotide collection (nr/nt) and Protein Database (PDB): Used for identifying "SIMILAR" peptide sequences through tblastn searches.
  • E. coli TOP10: The model organism used for the self-screening assay.

No new benchmark datasets were introduced, but the study effectively created and screened a large, novel library of naturally-derived peptide candidates.

Key Insights and Ideas

  • Nature's Richness: The study confirms that natural genomic databases are an incredibly rich source of undiscovered antimicrobial peptides, far exceeding the number of previously characterized ones.
  • Intracellular Activity is Key: A significant portion of the identified active peptides likely exert their antimicrobial effect by targeting intracellular components, a mechanism less explored than membrane disruption. This opens up a new frontier for antibiotic design.
  • Kinetics as a Predictor: The speed at which a peptide inhibits growth in the Mex assay can offer valuable clues about its mechanism of action, helping to prioritize candidates and understand their potential.
  • Phylogenetic Clues: The distribution of active peptides across different host organisms suggests that insects, in particular, may harbor a vast, untapped reservoir of antimicrobial peptides.
  • Mex is Robust: The high validation rate of Mex-active peptides in subsequent MIC assays demonstrates the reliability and effectiveness of the Mex screening platform.

Key Results

  • Library Size: The study screened a library of over 10,000 naturally-encoded peptides.
  • Hit Rate: 1240 out of 10,663 peptide-expressing strains (11.6%) were identified as "Mex-actives" (significantly inhibiting growth).
  • Novelty: Over 1000 of these Mex-actives were previously unknown antimicrobial peptides.
  • Potency: Ten out of 15 chemically synthesized and tested peptides efficiently eradicated bacteria at minimal inhibitory concentrations in the lower µM or upper nM range.
  • Mechanism Insights: Approximately one-third of the active peptides were estimated to interact with intracellular targets, rather than solely relying on membrane damage.
  • Low Cytotoxicity: Most of the identified active peptides showed low toxicity towards human erythrocytes at their MIC.

Limitations and Future Work

  • Host Specificity: The Mex screen was performed in the laboratory strain E. coli TOP10. The activity of identified peptides might differ in other bacterial species, especially Gram-positive bacteria or clinical isolates.
  • Intracellular vs. Extracellular Activity: Some peptides active intracellularly in the Mex assay showed reduced or no activity when tested extracellularly in MIC assays. This suggests that uptake mechanisms are crucial for some candidates, and strategies like linking them to cell-penetrating peptides might be necessary.
  • Chemical Modifications: The E. coli system may not be able to perform certain post-translational modifications found in natural antimicrobial peptides, potentially limiting the discovery of some active peptide classes.

The authors suggest that Mex can be extended for de novo design and optimization of natural peptides through directed evolution. They also propose its application for screening against drug-resistant strains (e.g., Pseudomonas aeruginosa) and for discovering peptides that do not rely on membrane damage, which could lead to next-generation antibiotics with improved specificity and reduced resistance development.

Summary updated

Abstract

The number of newly approved antimicrobial compounds has been steadily decreasing over the past 50 years emphasizing the need for novel antimicrobial substances. Here we present Me<sup>x</sup>, a method for the high-throughput discovery of novel antimicrobials, that relies on E. coli self-screening to determine the bioactivity of more than ten thousand naturally occurring peptides. Analysis of thousands of E. coli growth curves using next-generation sequencing enables the identification of more than 1000 previously unknown antimicrobial peptides. Additionally, by incorporating the kinetics of growth inhibition, a first indication of the mode of action is obtained, which has implications for the ultimate usefulness of the peptides in question. The most promising peptides of the screen are chemically synthesized and their activity is determined in standardized susceptibility assays. Ten out of 15 investigated peptides efficiently eradicate bacteria at a minimal inhibitory concentration in the lower µM or upper nM range. This work represents a step-change in the high-throughput discovery of functionally diverse antimicrobials.
BibTeX
@inproceedings{koch2022discovery,
  author = {Philipp Koch and Steven Schmitt and Mathias Cardner and Niko Beerenwinkel and Sven Panke and Martin Held},
  title = {{Discovery of antimicrobials by massively parallelized growth assays (Mex)}},
  booktitle = {Scientific},
  year = {2022},
  doi = {10.1038/s41598-022-07755-7},
}
Figure