Open AccessComparison of human gut microbial profiles of matched faecal and biopsy samples from healthy individuals using culture dependent and independent methods
Author(s) -
Indrani Mukhopadhya,
Jenny C. Martin,
Sophie Shaw,
Irini Lazou Ahrén,
Niklas Larsson,
Titti Martinsson Niskanen,
Silvia Gratz,
Karen P. Scott
Publication year - 2020
Publication title -
access microbiology
Language(s) - English
Resource type - Journals
ISSN - 2516-8290
DOI - 10.1099/acmi.ac2020.po0261
Subject(s) - unifrac , biology , feces , concordance , 16s ribosomal rna , microbiology and biotechnology , gut flora , bacteroidetes , metagenomics , ribosomal rna , microbiome , bacteria , immunology , genetics , gene
Faecal samples have often been used to characterise the gut microbiota in health and disease. There is significant debate whether faecal bacterial communities accurately reflect the mucosa associated bacterial populations, which are considered critical in the aetiopathogenesis of several gastrointestinal diseases. We simultaneously assessed faecal and mucosal microbiota from healthy volunteers to unravel the degree of concordance between the two profiles. Paired fresh rectal biopsies and faecal samples were obtained from ten healthy volunteers and processed under stringent anaerobic conditions. Composition and diversity of the microbiota were studied using next generation sequencing targeting the 16S ribosomal nucleic acid (rRNA) gene and culturomics. Bacterial richness and diversity were comparable between mucosal and faecal samples with no significant statistical differences. The relative abundance of Oxalobacteraceae, Propionibacteriaceae, Campylobacteraceae and Corynebacteriaceae were significantly increased (Corncob analysis; FDR=0.00027, 0.000046, 0.011 and 0.025 respectively) in biopsy compared to faecal samples at the family level. Conversely, there was increased abundance from the family Ruminococcaceae and Clostridiaceae (Corncob analysis; FDR=0.025 and 0.025 respectively) in faecal samples. Principal Coordinates Analysis of a Bray Curtis distance matrix generated from sequence variant tables did not show distinct clustering of biopsy and faecal samples (PERMANOVA; p=0.991). A total of 528 bacteria were isolated from a subset of 6 volunteer samples (biopsy and faeces) out of which there were 97 unique and 39 novel species identified. Our study showed good concordance between faecal and gut mucosal microbial profile, corroborating that faecal samples can act as a convenient surrogate to study gut microbiota.