Predict the gut microbiota resistome

The intestinal microbiota is considered to be a major reservoir of antibiotic resistance determinants (ARDs) that could potentially be transferred to bacterial pathogens via mobile genetic elements. A collaboration between several research institute including INRA, AP-HP and Institut Pasteur succeed to predict more than 6,000 ARDs from the gut microbiota using protein structural model.

A study¹ published in Nature Microbiology in November 2018 showed that a new developped method based on 3D structural modeling was able to predict 10-fold more ARDS genes than previous studies. To confirm those prediction, several distant hit were tested using gene synthesis confirming their ability to confer antibiotic resistance. In addition, authors were able to cluster individuals into six resistotypes very closely related to previous enterotypes.

Main findings

• First, authors developed and validated an annotation method on the basis of a three-dimensional structure, leading to the prediction of 6,095 ARDs in a catalogue of 3.9 million proteins from the human intestinal microbiota
• Secondly, authors found little evidence supporting their transfer between species.
• Third, according to the composition of their resistome, authors were able to cluster subjects from the MetaHIT cohort into six resistotypes that were connected to the previously described enterotypes..
• Finally, authors found that the relative abundance of pdARDs was positively associated with gene richness, but not when subjects were exposed to antibiotics.

Behind the scene

The first author, Etienne “ARDs hunter” Ruppe, explained how ideas behind the paper came in a blog post published in Nature community.