The search for new antibiotics has turned to an unexpected source: brain proteins. Past work suggested that fragments from certain proteins could combat microbes, but no large-scale effort had examined prion and prion-like proteins for hidden antimicrobial peptides. Earlier studies had hinted at this link by reporting that fragments from amyloid-beta, which is involved in neurodegenerative diseases like Alzheimer’s disease, and the cellular prion protein, could fight microbes. A research group at the University of Pennsylvania developed a deep-learning tool called APEX 1.1 to analyze 19.3 million short peptide fragments from 2,897 of these proteins. The system predicted which sequences might act as antibiotics, identifying 1,179 candidates. The researchers named this new class “prionins,” and the findings were published in Nature Microbiology.
The group chose 75 peptides for lab testing, prioritizing those the platform ranked highest against 11 bacterial pathogens, including drug-resistant strains. Testing revealed that 59 of them blocked at least one pathogen, while 42 worked effectively even at low doses. These peptides destroy bacteria by breaking down their membranes, a method many antimicrobial peptides share. Signs of toxicity were limited, and 16 active peptides showed no measurable harm to red blood cells or human cells at the highest concentrations tested. This safety profile is particularly significant for potential therapeutic applications, as previous experiments with the most promising candidates did not result in treatment-related weight loss.
Next, the scientists moved from lab dishes to living organisms. To verify these findings, researchers tested two of the most promising peptides—derived from a fungus and a roundworm—in mice. They found that the approach reduced bacteria levels in a standard skin infection model caused by Acinetobacter baumannii, a difficult-to-treat pathogen. The effects were comparable to polymyxin B, a last-resort antibiotic, and researchers saw no adverse effects on the animals’ weight.
Related: Tiny particles beat stubborn cancer cells
“This work changes where we think antibiotics might be hiding,” said César de la Fuente, PhD, FRSB, Presidential Associate Professor and director of the Machine Biology Group at the University of Pennsylvania Perelman School of Medicine and senior author of the study. The work builds on the de la Fuente Lab’s broader effort to mine the biological world for “encrypted peptides”—short, hidden sequences inside larger proteins that can have biological functions when isolated.
Marcelo D. T. Torres, co-first author of the study, noted that the AI search gave a short list, but the validation demonstrated that many molecules worked in lab and animal models. “That is what makes this a discovery platform, not just a prediction exercise.” The team plans further tests to refine the peptides and explore their effects in larger animals. Machi
