Antisense transcription reveals disease-associated adaptations in the human gut microbiome

Abstract

The gut microbiome is a dynamic ecosystem in which microorganisms constantly adjust their transcriptional programmes. Here we developed metastrand, a framework that integrates strand-aware metatranscriptomics and metagenomics to quantify mRNAs and antisense RNAs (asRNAs) in complex microbial communities at gene-level resolution. In inflammatory bowel disease (IBD), microbial asRNA programmes converged across patients during active disease, correlated with faecal metabolites and calprotectin levels and remained stable during persistent inflammation, highlighting their potential as biomarkers of inflammatory activity in the gut. These programmes involved antisense-to-sense transcriptional shifts at insertion sequence elements with functionally diverse passenger genes and preceded their detection at new genomic locations, linking asRNA dynamics to structural genome rearrangements and redistribution of adaptive functions under selective pressure. Similar dynamics were observed in a mouse model of colitis, oxidative stress in vitro and in patients with pathogen-confirmed gastroenteritis, establishing asRNAs as an important dimension of microbial adaptation in health and disease.