Life-long microbiome rejuvenation improves intestinal barrier function and inflammaging in mice

Abstract

Background: Alterations in the composition and function of the intestinal microbiota have been observed in organismal aging across a broad spectrum of animal phyla. Recent findings, which have been derived mostly in simple animal models, have even established a causal relationship between age-related microbial shifts and lifespan, suggesting microbiota-directed interventions as a potential tool to decelerate aging processes. To test whether a life-long microbiome rejuvenation strategy could delay or even prevent aging in non-ruminant mammals, we performed recurrent fecal microbial transfer (FMT) in mice throughout life. Transfer material was either derived from 8-week-old mice (young microbiome, yMB) or from animals of the same age as the recipients (isochronic microbiome, iMB) as control. Motor coordination and strength were analyzed by rotarod and grip strength tests, intestinal barrier function by serum LAL assay, transcriptional responses by single-cell RNA sequencing, and fecal microbial community properties by 16S rRNA gene profiling and metagenomics.

Results: Colonization with yMB improved coordination and intestinal permeability compared to iMB. yMB encoded fewer pro-inflammatory factors and altered metabolic pathways favoring oxidative phosphorylation. Ecological interactions among bacteria in yMB were more antagonistic than in iMB implying more stable microbiome communities. Single-cell RNA sequencing analysis of intestinal mucosa revealed a salient shift of cellular phenotypes in the yMB group with markedly increased ATP synthesis and mitochondrial pathways as well as a decrease of age-dependent mesenchymal hallmark transcripts in enterocytes and TA cells, but reduced inflammatory signaling in macrophages.

Conclusions: Taken together, we demonstrate that life-long and repeated transfer of microbiota material from young mice improved age-related processes including coordinative ability (rotarod), intestinal permeability, and both metabolic and inflammatory profiles mainly of macrophages but also of other immune cells. Video Abstract.

Keywords: Aging; Lifespan extension; Microbiome; Rejuvenation.

Fig. 1

Schematic representation of microbial rejuvenation experiment. Eight-week-old male C57BL6/J wildtype mice (n = 40) were aged until 120 weeks of age and every 8 weeks mice (n = 20 per group) received a microbial transfer from unrelated wildtype mice being either of the same age as the recipients (isochronic microbiome transfer—iMB) or being 8-week-old (young microbiome transfer—yMB) with the goal to rejuvenate the aging recipient mice with a young microbiome. Mice were monitored throughout the experiment and extensively phenotyped at final analysis by functional and behavioral tests (rotarod, grip strength, intestinal barrier function, glucose homeostasis). Furthermore, colonic and small intestinal tissue were subjected to bulk and single-cell transcriptome sequencing, and fecal samples collected throughout the experiment were subjected to 16S rRNA amplicon sequencing and metagenomics

Fig. 2

Microbial rejuvenation improves several host physiological traits. A Weight development. Grey areas denote phases of recurrent microbiome treatments. B Survival curve. C Blood glucose measurements during intraperitoneal glucose tolerance test (ipGTT) including area under the curve (AUC) summary. n = 5–6 per group. D yMB performed better in RotaRod performance test. *p < 0.05. n = 8–11 per group. E Grip strength as measured using a metal grid. n = 8–11 per group. F FITC dextran quantified in serum from vena cava 1 h after oral gavage. n = 5–6 per group. G Reduced LPS levels in vena cava serum of rejuvenated yMB mice. **p < 0.01. n = 7 per group

Fig. 3

Microbiome rejuvenation reverts various age-associated microbial metabolic functions. A Constrained β-diversity plot of yMB and iMB microbiomes over time as determined by 16S rRNA gene amplicon sequencing. B Bray–Curtis distances from the baseline microbiome configuration (week 8) are smaller in yMB compared to iMB. A greater Bray–Curtis distance indicates more dissimilar microbiome compositions. C Relative abundance of the genera Akkermansia and Lactobacillus. D The majority of metagenome-derived metabolic functions (Humann3) had an inverted linear dependence on host age in yMB and iMB. E Pathway enrichment of those metabolic functions revealed distinct aging patterns for yMB and iMB. The aging effect in yMB is reduced, i.e. slowed down, in all enriched processes or even inverted for 12 out of 19 metabolic pathways. F Increased stabilizing antagonistic predicted microbe-microbe interactions in aged yMB compared to iMB mice

Fig. 4

Transcriptional adaptations to the microbiome rejuvenation are specific to subtypes of intestinal epithelial cells. A–C Bulk RNA sequencing was performed on colon tissue of 40- and 120-week-old yMB and iMB mice. A Principal component plot. B Genes which expression changes contributed to the observed variance attributed to timepoint or treatment model. C Enriched functions among genes that were contributed to treatment explained variance. D–L Single-cell RNA sequencing of colonic samples of 120-week-old yMB and iMB mice. D Ten epithelial cell clusters were identified based on marker gene expression. E Treatment identity of all individual analyzed cells. F Number of up- and downregulated genes per epithelial cell type when comparing yMB vs iMB. G Shared and unique DEGs among epithelial cell types for up- and downregulated genes. H Transcriptional regulators enriched in upregulated DEGs. I Enriched functions for up- and downregulated genes (yMB vs iMB) in enterocytes, immature enterocytes, TA, and enteroendocrine cells. J Number and strength of predicted receptor-ligand interactions in yMB and iMB. K Outgoing and incoming ligand-receptor interaction strength of each epithelial cell type for iMB and yMB treatment. L Mesenchymal score reduced in yMB compared to iMB for multiple epithelial cell types indicating reduced EMT. M Reduced mesenchymal score in multiple colonic epithelial cells of 3- compared to 30-month-old mice of the tabula muris senis database [27, 28]

Fig. 5

Macrophages are the main responding immune cell type to microbial rejuvenation. Single-cell RNA sequencing of small intestinal lamina propria immune cells of 120-week-old yMB and iMB mice. A 16 immune cell clusters were identified based on marker gene expression. B Treatment identity of all individual analyzed cells. C Number of up- and downregulated genes per immune cell type. D Shared and unique DEGs among immune cell types for up- and downregulated genes (yMB vs iMB). E Enriched functions for up- and downregulated genes (yMB vs iMB) in macrophages, B, CD8 + T, and MAIT cells. F Number and strength of predicted receptor-ligand interactions in yMB and iMB. G Outgoing and incoming ligand-receptor interaction strength of each immune cell type for iMB and yMB treatment. H Reduced inflammation score in yMB compared to iMB in multiple immune cell-types indicating a lower inflammatory tone. I Reduced inflammation score in splenic immune cells of 3- compared to 30-month-old mice of the tabula muris senis database [27, 28] indicating an increased inflammation score as a sign of aging