BACKGROUND: Faecal samples are often used to characterise gut microbiota, since they are easily collected. However, whether or not the faecal microbiota differ from the mucosa-associated microbiota remains largely unknown. This may be specifically relevant in conditions that are characterised by complex mucosal microbe-host interactions, such as Crohn's disease. We aimed to determine the degree of agreement between faecal and mucosal microbiota profiles in healthy individuals, using two commonly used collection procedures.
MATERIAL AND METHODS: The gut microbiota composition of faecal samples (sent at ambient temperature before storage at -70°C) and of colonic biopsies (obtained at endoscopy and immediately stored at -70°C) was determined by sequencing the 16S rRNA gene. Thirty-one randomly selected healthy individuals from the population-based colonoscopy (Popcol) study were included.
RESULTS: Faecal samples were characterised by a reduced degree of richness (p<0.0001) and diversity (p=0.016), and also differences in several phyla, including a lower relative abundance of Proteobacteria (p<0.0001) and Verrucomicrobia (p=0.008) than in biopsies. Only 3 of 30 individuals had a similar faecal and mucosal microbiota profile, based on weighted UniFrac analysis. A difference in Crohn's disease dysbiosis-associated bacteria was observed, including a lower relative abundance of Faecalibacterium (p=0.004) and a higher relative abundance of Ruminococcus (p=0.001) in faeces than in biopsies.
CONCLUSIONS: Analysis of faecal samples that have been transported at ambient temperature does not adequately reflect the colonic mucosa-associated microbiota in healthy individuals. These findings have implications for the interpretation of the previous literature, and may be specifically relevant to studies on Crohn's disease.