Evaluation of intestinal biopsy tissue preservation methods to facilitate large-scale mucosal microbiota research

Abstract

Background: Large-scale multicentre studies are needed to understand complex relationships between the gut microbiota, health and disease. Interrogating the mucosal microbiota may identify important biology not captured by stool analysis. Gold standard tissue cryopreservation ('flash freezing') limits large-scale study feasibility. We aimed to compare gut microbiota in gold standard and pragmatic mucosal biopsy storage conditions.

Methods: We collected endoscopic recto-sigmoid biopsies from 20 adults with inflammatory bowel disease. Biopsies were preserved using three methods: (i) flash freezing (most proximal and distal biopsy sites); (ii) nucleic acid preservative reagents (QIAGEN Allprotect®, Invitrogen RNAlater™, and Zymo DNA/RNA Shield™); and (iii) formalin fixation with paraffin embedding (FFPE), which is used to preserve tissue for clinical histopathology within healthcare settings. Microbiota were sequenced on the MiSeq platform (V4 region, 16S rRNA gene).

Findings: Tissue microbiota were consistent between most proximal and distal tissue suggesting any within-patient variation observed reflected storage condition, not biopsy location. There was no significant difference in alpha-diversity or microbial community profiles of reagent-preserved versus gold standard tissue. FFPE community structure was significantly dissimilar to other tissue samples, driven by differential relative abundance of obligate gut anaerobes; Faecalibacterium, Anaerostipes and Lachnospiraceae. Despite these differences, tissue microbiota grouped by participant regardless of preservation and storage conditions.

Interpretation: Preservative reagents offer a convenient alternative to flash freezing tissue in prospective large-scale mucosal microbiota studies. Whilst less comparable, FFPE provides potential for retrospective microbiota studies using historical samples.

Funding: Medical Research Council (MR/T032162/1) and The Leona M. and Harry B. Helmsley Charitable Trust (G-2002-04255).

Keywords: Formalin fixed paraffin embedded; Gut microbiome; Inflammatory bowel disease; Precision medicine; Tissue microbiome; Tissue preservative reagents.

Fig. 1

Schematic summarising study sample collection and biopsy processing. (a) Participant samples collected during the study (2 stool and 8 tissue samples per participant); (b) Illustrative locations and storage conditions for tissue biopsy samples collected from participants at lower gastrointestinal endoscopy procedures; (c) Overview of sample processing and storage conditions. Samples were immediately placed into pre-defined conditions within the endoscopy room. Conditions 1 and 8: Samples were flash-frozen with dry ice and transferred to −80 °C storage within 4 h. Conditions 2–6: Samples were placed into a cryovial pre-filled with one of three storage buffers before placing into wet ice for a maximum of 4 h before transfer to 4 °C. Samples remained at 4 °C for 24 h (Condition 2), 72 h (Conditions 3, 5 and 6) or 1 month (Condition 4) before transfer to −20 °C. Condition 7: Tissue sample placed into formalin pre-filled pot and sent to the Cellular Pathology Department at the Royal Victoria Infirmary, Newcastle upon Tyne Hospitals NHS Foundation Trust, for processing into a formalin-fixed paraffin embedded research block using the standard local laboratory protocol. Created in BioRender. Wyatt, N. (2024) BioRender.com/p99f243.

Fig. 2

Comparison of flash frozen (FF) and reagent-preserved tissue. (a) Violin plots illustrate comparison of unrarefied library sizes (Kruskal–Wallis P = 0.77); (b) rarefied alpha-diversity (taxonomic richness) (Kruskal–Wallis P = 0.99); and (c) rarefied alpha-diversity (Shannon diversity index) (Kruskal–Wallis P = 0.99), where each point represents an individual sample. (d) Beta-diversity measures of bacterial community composition by preservation method (PERMANOVA P = 1, R² = 0.007), and (e) participant (PERMANOVA P = 0.001, R² = 0.55) are also shown. (Abbreviations: RNAl, RNAlater; AP, AllProtect).

Fig. 3

Comparison of flash frozen (FF), AllProtect (AP), and formalinfixation with paraffin embedding (FFPE) tissue, with stool preserved using the OMNIgene®·GUT kit. Violin plots (where each point represents an individual sample) showing comparison of: (a) taxonomic richness (Kruskal–Wallis P = 0.09); (b) Shannon diversity (Kruskal–Wallis P = 0.11); and (c) community similarity to flash frozen (distal) tissue (Kruskal–Wallis P < 0.01). Beta-diversity measures of bacterial community composition are shown by: (d) preservation method (PERMANOVA P = 0.001, R² = 0.22), and (e) participant (PERMANOVA P = 0.001, R² = 0.55). Differential taxa between FF, AP, and FFPE tissue, as well as stool are illustrated via a heatmap (f) where boxes highlight taxa identified as significantly differential between each sample type and flash frozen tissue (MaAsLin2 q < 0.05). Borderline differential taxa (q < 0.25) are annotated along the y axis by points according to sample type.