Engraftment of Bacteria after Fecal Microbiota Transplantation Is Dependent on Both Frequency of Dosing and Duration of Preparative Antibiotic Regimen

Abstract

The gut microbiota has emerged as a key mediator of human physiology, and germ-free mice have been essential in demonstrating a role for the microbiome in disease. Preclinical models using conventional mice offer the advantage of working with a mature immune system. However, optimal protocols for fecal microbiota transplant (FMT) engraftment in conventional mice are yet to be established. Conventional BALB/c mice were randomized to receive 3-day (3d) or 3-week (3w) antibiotic (ABX) regimen in their drinking water followed by 1 or 5-daily FMTs from a human donor. Fecal samples were collected longitudinally and characterized using 16S ribosomal RNA (rRNA) sequencing. Semi-targeted metabolomic profiling of fecal samples was also done with liquid chromatography-mass spectrometry (LC-MS). Lastly, we sought to confirm our findings in BKS mice. Recovery of baseline diversity scores were greatest in the 3d groups, driven by re-emergence of mouse commensal microbiota, whereas the most resemblance to donor microbiota was seen in the 3w + 5-FMT group. Amplicon sequence variants (ASVs) that were linked to the input material (human ASVs) engrafted to a significantly greater extent when compared to mouse ASVs in the 3-week groups but not the 3-day groups. Lastly, comparison of metabolomic profiles revealed distinct functional profiles by ABX regimen. These results indicate successful model optimization and emphasize the importance of ABX duration and frequency of FMT dosing; the most stable and reliable colonization by donor ASVs was seen in the 3wk + 5-FMT group.

Keywords: antibiotics; conventional mice; engraftment; microbiome.

Figure 1

Experimental schema for primary experiment in Balb/c mice. Classification and sampling scheme for mice (n = 5/group) receiving either the (A) 3-day or (B) 3-week antibiotic regimen. Representative data from two experiments shown.

Figure 2

Analysis of diversity. (A) Box plots of longitudinal changes in alpha diversity scores over time (n = 5 mice/group; p-value by Kruskal–Wallis test). PCoA plot of unweighted UniFrac beta-diversity distances in (B) all groups, (C) G3 mice and (D) G6 mice (n = 5/group; p-value by ANOSIM). Other groups in Figure S4.

Figure 3

Comparison of engraftment kinetics. Unsupervised hierarchical clustering heatmap of ASVs uniquely identified in the input FMT material in (A) G3 and (B) G6 mice (n = 5 mice/group). Columns represent ASVs. Rows are samples temporally arranged from T1 through T5. Line plot to compare changes in ‘human’ and ‘mouse’ ASV abundances over time in (C) G3 and (D) G6 mice (n = 5 mice/group). The mean and standard deviation of ASV abundances is represented at each time point. The p-value is calculated from a one-sided paired Wilcoxon signed-rank test and adjusted for false discovery rate [36]. Other groups in Figure S5. Additionally, see line plot for ASVs uniquely derived from input material in Figure S6.

Figure 4

Taxonomic landscape. Stacked bar plot showing engraftment of phyla uniquely derived from the input material in recipient mice in (A) G3 and (B) G6 mice, n = 5 mice/group. See other groups in Figure S7. Additionally, see analogous plot for families in Figure S8. Stacked bar plot comparing the overall taxonomic landscape at the phylum level of the input material and recipient mice over time in (C) G3 and (D) G6, n = 5 mice/group. Only the most abundant taxa are shown. See other groups in Figure S9. See analogous plot for families in Figures S10 and S11. (E) Principal component analysis plot of metabolite abundances. Data points represent individual samples (n = 5 mice/group) and arrows show an overlaid biplot of loadings for each metabolite.

Figure 5

Validation experiment in BKS mice using the G6 regimen. (A) Experimental schema. (B) PCOA plot of unweighted UniFrac beta-diversity distances with p-values calculated by ANOSIM. (C) Line plot to compare changes in ‘human’ and ‘mouse’ ASV abundances over time. The p-value is calculated from a one-sided paired Wilcoxon signed-rank test and adjusted for false discovery rate. (D) Stacked bar plot showing engraftment of phyla uniquely derived from the input material in recipient mice.