Computationally Designed Peptides Self-Assemble Into Pores That Kill Drug-Resistant Bacteria
Researchers report in Nature Chemical Biology a computational de novo design framework that generates short α-helical peptides engineered to insert into microbial membranes and assemble into transmembrane barrel-stave pores. Rather than screening natural antimicrobial peptides, the team designed the pore-forming behavior from first principles, then tuned a lead candidate for selectivity and potency.
The optimized peptide selectively killed drug-resistant ESKAPEE pathogens, including Acinetobacter baumannii, while sparing human cells, and it showed anti-infective efficacy in preclinical mouse infection models. The authors frame the same design logic as extensible to nanopore sensing and drug delivery.
Why it matters: antimicrobial resistance is one of the most pressing threats in medicine, and membrane-disrupting peptides are attractive because bacteria struggle to evolve resistance against physical membrane damage. Demonstrating that these pores can be designed computationally — rather than discovered by trial and error — points toward a programmable pipeline for next-generation antibiotics.
Suggested PeptideWiki angle: a short explainer on how de novo-designed pore-forming peptides work as antibiotics, contrasting the "designed from scratch" approach with classic natural AMPs like the defensins, and why selectivity for bacterial over human membranes is the central challenge.