Skip to main content
Premium Trial:

Request an Annual Quote

Microbiome Metabolite Screening Method Uncovers Gut Microbe-Drug Interactions, Variability

NEW YORK – The metabolites produced by microbes in the human gut may impact the activity of drugs that move through the small or large intestine, according to new research from a Princeton University team.

"Everyone's microbiome is unique, and we were able to see this in our study," co-first author Bahar Javdan, a graduate student in molecular biology at Princeton, said in a statement.

For a paper appearing in Cell on Wednesday, Javdan and colleagues relied on a personalized culture-, sequencing-, quantitative metabolomics-, and genomics-based method dubbed microbiome-derived metabolism screening (MDM-Screen) to systematically search for interactions between drugs and representatives from a gut microbial community.

The approach was built on an ex vivo mixed culture system developed to be as close to the uncultured gut microbiome of one "pilot" individual as possible — a metric measured with the help of 16S ribosomal RNA gene region V4 sequencing.

"To identify the medium and culturing period that can support the growth of a batch culture whose composition is maximally similar to the original [pilot donor] microbiome, freshly collected and glycerol-stocked human feces from PD were cultured in 14 different media and sampled daily for four days," the authors explained. "We then extracted DNA from all samples, amplified the V4 region of the bacterial 16S rRNA gene, and deeply sequenced the amplicons."

After narrowing in on a culturing method that appeared to most closely represent the gut microbiome composition with the amplicon sequence variant data, the team attempted to screen for interactions between the microbes and drugs of interest with mixed culture incubations followed by high-performance liquid chromatography with mass spectrometry.

From a set of 575 drugs, the researchers were able to screen more than 400 compounds after weeding out drugs with stability issues or other factors that interfered with the MDM-Screen. Of those, the investigators identified 57 drugs that were significantly altered or depleted through interactions with gut microbial metabolites.

From there, the researchers ramped up the MDM-Screen to assess ex vivo interactions between 23 oral drugs and the gut microbial community collections from 20 more individuals, identifying personalized drug-microbiome interactions that sparked deeper analyses of the drug-metabolizing enzymes that can turn up in the gut microbiome.

"This inter-person variability underscores why studying a single bacterial species makes it impossible to compare the microbiome's metabolism of drugs between individuals," senior and corresponding author Mohamed Donia, a molecular biology, chemical, and biological engineering researcher at Princeton, said in a statement. "We need to study the entire intestinal microbial community."

Based on results from the initial screens, combined with follow-up metagenomic and mouse model experiments, the authors speculated that gut microbial community clues may prove useful for everything from drug development to the more personalized use of existing treatment compounds.

"We observed three main categories — drugs that were consistently metabolized by all the microbiomes in our study; drugs that were metabolized by some and not by others; and drugs that were not subject to any microbiome-derived metabolism," Javdan said.

Those involved in the study noted that results from even larger MDM-Screen experiments should provide a clearer understanding between-individual variability in drug-gut microbiome interactions, along with the interplay between compounds produced by microbes in the gut and those found across the medications that make their way through the intestine.

"A simultaneous expansion into hundreds of drugs and hundreds of donor samples is necessary to reveal the complete biochemical potential of [microbiome-derived metabolism (MDM)]: it is very likely that the types of MDM transformations observed here are an underestimation of all possible ones," the authors wrote.