Uncovering new antibiotics inside the human gut

The average human gut is home to roughly 100 trillion microbes, many of which are constantly competing for limited resources. “It’s a very hostile environment,” says Cesar de la Fuente, dean’s assistant professor of bioengineering and chemical and biomolecular engineering in the School of Engineering and Applied Science, psychiatry and microbiology in the Perelman School of Medicine, and chemistry in the School of Humanities and Sciences.

“Bacteria coexist but also fight each other. This environment could foster innovation.”

In this conflict, de la Fuente’s lab sees the potential for new antibiotics that might one day contribute to humanity’s own defense against drug-resistant bacteria. After all, if the bacteria in our gut must develop new tools to fight each other to survive, why not use the bacteria’s own weapons against them?

in paper in cellDe la Fuente and his colleagues, including Ami S. Bhatt, a professor of medicine (hematology) and genetics at Stanford University, have studied the gut microbiomes of nearly 2,000 people and discovered dozens of potential new antibiotics.

“We think of biology as a source of information,” de la Fuente says, “it’s all just code. If we can develop algorithms that can organize that code, we can dramatically speed up antibiotic discovery.”

In recent years, de la Fuente’s lab has attracted attention for uncovering antibiotic candidates from the genetic code of extinct organisms, all over the place. Neanderthals From mammoths to bacteria in abundance, the institute has used artificial intelligence to analyze the genetic material.

“One of our main goals is to collect biological information from around the world as a source of antibiotics and other useful molecules,” de la Fuente says.

“Instead of relying on the traditional, laborious method of taking soil or water samples and purifying active compounds, we are leveraging the vast amounts of biological data contained in genomes, metagenomics and proteomes. This allows us to discover new antibiotics at digital speed.”

Because bacteria evolve rapidly, de la Fuente and his co-authors hypothesized that a competitive environment like the human gut could harbor many undiscovered antibacterial compounds. “When resources are scarce, that’s when biology really comes up with innovative solutions,” de la Fuente says.

The research team focused on peptides, short chains of amino acids that have previously been shown to have potential as new antibiotics.

“We computationally mine more than 400,000 proteins,” de la Fuente said, referring to the process in which the AI ​​reads the letters of genetic code and predicts which gene sequences are likely to have antibacterial properties after being trained on a set of known antibiotics.

“Interestingly, these molecules have a different composition than those traditionally thought to be antibacterial,” said Marcelo DT Torres, a researcher in the de la Fuente lab and lead author of the paper. “The compounds we discovered constitute a new class, and their unique properties will help us understand and expand the sequence space of antibacterial agents.”

Of course, these predictions must be verified experimentally: after finding hundreds of antibiotic candidates, the researchers selected 78 to test against real bacteria.

After synthesizing the peptides, the researchers exposed bacterial cultures to each peptide and waited 20 hours to see which peptides effectively inhibited bacterial growth. The team then tested the antibiotic candidates in animal models.

More than half of the peptides tested were effective, i.e. inhibiting the growth of both beneficial and pathogenic bacteria, and the lead candidate, prevotelin-2, showed anti-infective capabilities similar to those of polymyxin B, an FDA-approved antibiotic currently used to treat multidrug-resistant infections, suggesting that the human gut microbiome may contain antibiotics with future clinical applications.

“It was a great surprise to me to identify prevotelin 2, which has activity comparable to polymyxin B, one of the antibiotics of last resort,” Bhatt says.

“This suggests that searching for new and exciting classes of antimicrobial peptides from the human microbiome is a promising future avenue for researchers, physicians, and especially patients.”