AI-Guided Discovery of Enterololin Offers New Hope for IBD Treatment
核心洞察
Researchers at McMaster University and MIT have discovered enterololin (搜索), a narrow-spectrum antibiotic that specifically targets disease-causing bacteria in inflammatory bowel diseases while preserving beneficial gut microbiome.
The breakthrough utilized AI to predict the drug's mechanism of action in just six months for $60,000, compared to traditional methods that typically require two years and $2 million.
Enterololin (搜索) targets the Enterobacteriaceae (搜索) family including E. coli (搜索) strains that drive Crohn's disease (搜索), offering a promising treatment option for millions of IBD patients who currently have no cure.
Researchers at McMaster University and the Massachusetts Institute of Technology have achieved a dual breakthrough in drug discovery, unveiling a new antibiotic called enterololin (搜索) that specifically targets inflammatory bowel diseases while demonstrating the first successful use of AI to predict a drug's mechanism of action before experimental validation.
The discovery, detailed in Nature Microbiology on October 3, 2025, represents a significant advancement for the millions of people affected by Crohn's disease (搜索) and other IBD-related conditions. Unlike traditional broad-spectrum antibiotics that eliminate both harmful and beneficial bacteria, enterololin (搜索) operates as a precision therapeutic, targeting only specific disease-causing pathogens.
Precision Targeting for IBD
"This new drug is a really promising treatment candidate for the millions of patients living with IBD," says Jon Stokes, an assistant professor in McMaster's Department of Biochemistry and Biomedical Sciences and principal investigator on the study. "We currently have no cure for these conditions, so developing something that might meaningfully alleviate symptoms could help people experience a much higher quality of life."
Enterololin (搜索) specifically attacks the Enterobacteriaceae (搜索) family of bacteria, which includes the type of E. coli (搜索) that drives Crohn's disease (搜索). This narrow-spectrum approach preserves the gut microbiome while eliminating pathogenic bacteria, reducing the opportunity for drug-resistant strains to colonize the intestines.
"Most antibiotics used in clinics today are 'broad-spectrum' drugs, meaning they wipe out good bacteria in addition to those that cause disease — they're nukes," Stokes explains. "This can create opportunities for invasive and drug-resistant species of bacteria, like E. coli (搜索), to move in and colonize the intestines, which can exacerbate conditions like Crohn's."
AI Revolutionizes Drug Development Timeline
The research team achieved an unprecedented milestone by using MIT's DiffDock AI model to predict enterololin (搜索)'s mechanism of action in just 100 seconds. The AI predicted that the drug attacks a microscopic protein complex called LolCDE (搜索), essential for bacterial survival.
Traditional mechanism of action studies typically require up to two years and cost around $2 million. Using AI guidance, the McMaster team completed their analysis in six months for just $60,000, representing a 75% reduction in time and a 97% reduction in cost.
"A lot of AI use in drug discovery has been about searching chemical space, identifying new molecules that might be active," says Regina Barzilay, a professor in MIT's School of Engineering and developer of the DiffDock model. "What we're showing here is that AI can also provide mechanistic explanations, which are critical for moving a molecule through the development pipeline."
Experimental Validation Confirms AI Predictions
Despite the AI's rapid prediction, the research team, led by McMaster graduate student Denise Catacutan, conducted traditional laboratory validation studies. Within months, experimental results confirmed the AI's accuracy.
"We did all of our standard MOA workup to validate the prediction — to see if the experiments would back-up the AI, and they did," says Catacutan, a PhD candidate in the Stokes Lab. "Doing it this way shaved a year-and-a-half off of our normal timeline."
The successful validation represents what researchers believe to be a global first for AI in predicting drug mechanisms of action, opening new possibilities for accelerated drug development.
Path to Clinical Application
Stokes' spin-out company, Stoked Bio (搜索), has already licensed enterololin (搜索) from McMaster and is currently optimizing it for human use. The company is also testing modified versions against other drug-resistant bacteria, including Klebsiella, with early results showing promise.
"The identification of enterololin (搜索) underscores the remarkable science emerging at McMaster," says Jeff Skinner, CEO at Stoked Bio (搜索). "We are proud to partner with the university on translating this breakthrough into real therapies for patients."
If development proceeds as planned, the new drug could be ready for human trials within three years. The research addresses a critical unmet medical need, as IBD affects thousands of people across Canada alone, with no cure currently available.
Broader Implications for Drug Discovery
The success demonstrates AI's potential to address major bottlenecks in pharmaceutical development beyond initial drug discovery. "AI has expedited the rate at which we can explore chemical space for new drug candidates, but, until now, it has done little to alleviate a major bottleneck in drug development, which is understanding what these new drug candidates actually do," Stokes explains.
The collaboration between McMaster and MIT showcases how interdisciplinary partnerships can accelerate the translation of laboratory discoveries into potential therapies. As antimicrobial resistance continues to pose global health challenges, such innovations in drug development methodology could prove crucial for developing effective treatments.
"This work shows that we're still just scratching the surface as far as AI-guided drug discovery goes," Stokes notes. "The development of our new drug, which is designed to target IBD, has been fast-tracked thanks to the collaboration between humans and generative AI."