With the rise of antibiotics in the 1930s, phage therapy (i.e. the use of viruses called bacteriophages to fight bacterial infections) was abandoned. Today, with the rise of antibiotic resistance making the treatment of bacterial infections increasingly difficult, phage therapy is once again attracting the interest of doctors and researchers, despite the complexity of its application due to the great diversity and specificity of bacteriophages. Scientists from the Institut Pasteur, Inserm, AP-HP and Université Paris Cité have developed a new tool to help select the best possible cocktail of bacteriophages for a given patient, simply and effectively. To achieve this, they have developed and trained an artificial intelligence-based model capable of selecting tailor-made bacteriophages based solely on the genome of the targeted bacteria. The results of this work were published on October 31, 2024 in the journal Nature Microbiology , and pave the way for personalized phagotherapies to combat antibiotic-resistant bacterial infections.
Some bacteria, such as Escherichia coli, are proving increasingly resistant to conventional antibiotics, becoming what are known as "superbugs". To circumvent this resistance, which represents a major public health problem, research teams are exploring phage therapy. The principle is to use viruses, known as phages or bacteriophages, which infect only bacteria, to eliminate those that are pathogenic to humans.
" Phage therapy was invented by Pasteurian researcher Félix d’Hérelle in the 1920s, then abandoned with the rise of antibiotics in the late 1930s, which are much simpler and cheaper to manufacture and use. Today, only a few Eastern European countries, such as Georgia, still use phage therapy, while in Western countries, "broad-spectrum" phages are used on an ad hoc, compassionate basis to treat chronic infections that are multi-resistant to antibiotics(1), when no authorized drug is effective any more.Baptiste Gaborieau, co-first author of the article, a resuscitation physician at Hôpital Louis Mourier (AP-HP) and researcher in the IAME laboratory (Université Paris Cité-Inserm). Over the last twenty years or so, thanks to its promotion by the WHO(2)and more recently the setting up of clinical trials, notably in Europe, phagotherapy has once again aroused interest. "
One of the challenges is to know which bacteriophage will be effective against a given infection, bearing in mind that each phage can only infect certain strains(3) of bacteria. In soil or water, where phages are naturally present, they circulate until they find the right target. Scientists from Institut Pasteur, Inserm, AP-HP and Université Paris-Cité therefore decided to take a closer look at bacterial-phage interactions, to see whether it was possible to predict the effectiveness of a bacteriophage on a given bacterial strain. The first step was to create a high-quality database containing 403 strains of Escherichia coli bacteria and 96 bacteriophages. This work took over two years to complete.
" We brought phages into contact with bacteria in culture and observed which bacteria were killed. We studied 350,000 interactions and succeeded in identifying, at the level of the bacterial genome, the characteristics likely to predict phage efficacy", summarizes Aude Bernheim, principal author of the study and head of the Molecular Diversity of Microbes laboratory at the Institut Pasteur.
" Contrary to what was initially thought, it is the receptors on the surface of bacteria, and not their defense mechanisms, that primarily determine the ability of bacteriophages to infect bacteria, and thus predict their effectiveness ", continues Florian Tesson, co-first author of the article and doctoral student in the Molecular Diversity of Microbes laboratory at the Institut Pasteur and the IAME laboratory at the Université Paris Cité-Inserm.
Thanks to this precise and comprehensive analysis of the interaction mechanisms between bacteria and phages, the team’s bioinformaticians have been able to design an optimized and efficient artificial intelligence program. The program is based on the analysis of the bacterial genome, and more specifically on the analysis of regions involved in the coding of bacterial membrane receptors, the entry point for phages.
" We’re not dealing here with a ’ black box’, and that’s the strength of our AI-based model. We know exactly how it works, which helps us to improve its performance ", emphasizes Hugo Vaysset, co-first author of the article and a doctoral student in the Molecular Diversity of Microbes laboratory at the Institut Pasteur.
After more than two years of design and training, the AI was able to correctly predict the efficacy of the bacteriophages against the E. coli bacteria in the database in 85% of cases, simply by analyzing the bacteria’s DNA.
" It’s a result that exceeds our expectations," admits Aude Bernheim.
To take this a step further, the researchers tested their model on a new collection of d-E. coli bacterial strains responsible for pneumonia, and selected a tailor-made "cocktail" of three bacteriophages for each strain. In 90% of cases, the customized bacteriophages selected by the AI succeeded in their mission, destroying the bacteria present. This method, which can easily be used in hospital biology laboratories, paves the way for rapid, personalized selection of bacteriophage treatments for the diagnosis of bacterial infections with Escherichia coli that are highly resistant to antibiotics.
" We still have to test the behavior of phages in different environments, but the proof of concept is there. We hope to be able to extend it to other pathogenic bacteria, as our AI has been designed to adapt easily to other scenarios, and offer personalized phage therapy treatments in the future," concludes Aude Bernheim.
- In France, phages can be used within the framework of a Temporary Authorization for Use (ATU).
- A group of bacteria with common characteristics within a given species.
https://www.who.int/europe/fr/news/item/25-06-2024-building-evidence-for-the-use-of-bacteriophages-against-antimicrobial-resistance
