“`html
With the swift advancement of antibiotics in the 1930s, phage therapy—which utilizes viruses called bacteriophages or phages to combat bacterial infections—was largely forgotten. However, with the current surge in antibiotic resistance complicating the treatment of bacterial infections, phage therapy is regaining attention from healthcare providers and researchers. Despite this renewed interest, implementing phage therapy is still challenging due to the vast variety and specificity of phages.
With the swift advancement of antibiotics in the 1930s, phage therapy—which utilizes viruses called bacteriophages or phages to combat bacterial infections—was largely forgotten. However, with the current surge in antibiotic resistance complicating the treatment of bacterial infections, phage therapy is regaining attention from healthcare providers and researchers. Despite this renewed interest, implementing phage therapy is still challenging due to the vast variety and specificity of phages. In light of this situation, scientists from Institut Pasteur, Inserm, the Paris Public Hospital Network (AP-HP), and Université Paris Cité have created a straightforward and effective tool to recommend the ideal phage cocktail for individual patients. This was achieved by developing an artificial intelligence model that can select phages tailored specifically to the genome of the bacteria being targeted. Findings from this research, published on October 31, 2024, in the journal Nature Microbiology, pave the way for customized phage therapies to combat antibiotic-resistant bacterial infections.
Some bacteria, such as Escherichia coli, are becoming increasingly resistant to standard antibiotics, evolving into what are termed “superbugs.” To address the issue of resistance, which poses a significant public health threat, research teams are investigating the potential of phage therapy. The concept involves using viruses known as phages or bacteriophages, which specifically target and eliminate pathogenic bacteria affecting humans. “Phage therapy was pioneered by the Institut Pasteur scientist Félix d’Hérelle in the 1920s, but it was gradually set aside from the late 1930s onward due to the emergence of antibiotics, which were easier and cheaper to produce and administer. Currently, only a few countries in Eastern Europe, like Georgia, still actively use phage therapy, whereas in Western nations, ‘broad host range’ phages might be used under compassionate care for patients with chronic multidrug-resistant infections,” explains Baptiste Gaborieau, co-first author of the paper, an intensive care physician at Louis Mourier Hospital (AP-HP), and a researcher at the IAME laboratory (Université Paris Cité-Inserm). “In the last two decades, following support from WHO and the initiation of clinical trials in Europe, interest in phage therapy has once again grown.”
A major challenge lies in determining which phages will effectively treat a specific infection since each phage can only infect certain bacterial strains. In natural settings like soil or water where phages exist, they move around until they locate their specific targets. Researchers from Institut Pasteur, Inserm, the Paris Public Hospital Network (AP-HP), and Université Paris Cité aimed to delve deeper into bacteria-phage interactions to predict the effectiveness of phages against particular bacterial strains. The initial step involved creating a comprehensive dataset of interactions between 403 varied strains of Escherichia coli and 96 phages. This extensive research spanned more than two years. “We exposed the phages to bacterial cultures and monitored which bacteria were eliminated. We analyzed 350,000 interactions and successfully identified characteristics in the bacterial genome that could indicate phage effectiveness,” summarizes Aude Bernheim, the last author of the study and Head of the Molecular Diversity of Microbes laboratory at Institut Pasteur. “Interestingly, contrary to our initial assumptions, it turns out that a phage’s ability to infect bacteria—an indicator of its efficacy—is determined by receptors on the bacterial surface, rather than the bacterial defenses against viral attack,” adds Florian Tesson, a co-first author of the paper and a PhD candidate from the Molecular Diversity of Microbes laboratory at Institut Pasteur and the IAME laboratory at Université Paris Cité-Inserm.
This detailed analysis of how bacteria and phages interact allowed the team’s bioinformaticians to develop a powerful AI program. The program evaluates the bacterial genome, especially focusing on the regions that code for bacterial membrane receptors, which act as gateways for phages. “Our AI model isn’t just a ‘black box’—its operations are transparent, which is what enhances its effectiveness. We thoroughly understand how it works, enabling us to refine its performance,” states Hugo Vaysset, co-first author of the paper and a PhD researcher at the Molecular Diversity of Microbes laboratory at Institut Pasteur. After more than two years of development and testing, the AI model successfully predicted the effectiveness of phages against the examined E. coli bacteria in 85% of cases by analyzing the bacterial DNA. “This outcome exceeded our expectations,” remarks Aude Bernheim. Looking ahead, the researchers applied their AI model to a separate collection of E. coli strains linked to pneumonia, selecting a customized “cocktail” of three phages for each strain. In 90% of instances, the AI-selected phages were successful in eliminating the bacteria. This approach, which can be easily adapted to hospital labs, sets the stage for future strategies that can quickly provide personalized phage treatments when highly antibiotic-resistant Escherichia coli infections are diagnosed. “While further testing of phage effects in different environments is necessary, we’ve established proof of concept. We aim to expand this work to other pathogenic bacteria, as our AI model is designed to easily adapt to various scenarios for future personalized phage therapy treatments,” concludes Aude Bernheim.
- In France, phages can be utilized if a temporary authorization for use (ATU) is granted for an individual patient.
- https://www.who.int/europe/news/item/25-06-2024-building-evidence-for-the-use-of-bacteriophages-against-antimicrobial-resistance
- A group of bacteria sharing similar traits within a specific species.
“`