Age Associated Microbiome and Microbial Metabolites Modulation and Its Association With Systemic Inflammation in a Rhesus Macaque Model

Abstract

Aging is associated with declining immunity and inflammation as well as alterations in the gut microbiome with a decrease of beneficial microbes and increase in pathogenic ones. The aim of this study was to investigate the age associated gut microbiome in relation to immunologic and metabolic profile in a non-human primate (NHP) model. 12 geriatric (age 19-24 years) and 4 young adult (age 3-4 years) Rhesus macaques were included in this study. Immune cell subsets were characterized in peripheral blood mononuclear cells (PBMC) by flow cytometry and plasma cytokines levels were determined by bead based multiplex cytokine analysis. Stool samples were collected by ileal loop and investigated for microbiome analysis by shotgun metagenomics. Serum, gut microbial lysate, and microbe-free fecal extract were subjected to metabolomic analysis by mass-spectrometry. Our results showed that the gut microbiome in geriatric animals had higher abundance of Archaeal and Proteobacterial species and lower Firmicutes than the young adults. Highly abundant microbes in the geriatric animals showed a direct association with plasma biomarkers of inflammation and immune activation such as neopterin, CRP, TNF, IL-2, IL-6, IL-8 and IFN-γ. Significant enrichment of metabolites that contribute to inflammatory and cytotoxic pathways was observed in serum and feces of geriatric animals compared to the young adults. We conclude that aging NHP undergo immunosenescence and age associated alterations in the gut microbiome that has a distinct metabolic profile. Aging NHP can serve as a model for investigating the relationship of the gut microbiome to particular age-associated comorbidities and for strategies aimed at modulating the microbiome.

Keywords: age and immunity; age and metabolites; age and microbes; immunity and microbiome; microbiome and metabolites.

Figure 1

Plasma inflammatory cytokines in young adult and geriatric Rhesus Macaques: ELISA kits were used to quantify levels of CRP (A), and neopterin (B) in young adult and geriatric animals. Magpix multiplex analysis was used to quantify levels of TNF (C), IL-10 (D), IFN-γ (E), IL-8 (F), IL-6 (G), and IL-2 (H) in young adult and geriatric animals. Statistical analysis performed by Mann-Whitney U test with Benjamini-Hochberg correction for multiple comparisons to calculate the adjusted p-values (q value). Line and whiskers indicate the mean ± standard deviation. Q values are exact and a q < 0.05 was considered significant.

Figure 2

Gut microbial diversity indices of geriatric and young adult animals: (A), Qualitative differences between the two age groups (β-diversity), 2D rendering of the Bray-Curtis dissimilarity matrix showing distinct clustering between the two. (B–D) α-diversity indices, namely Shannon (B) Chao1 (C) and Simpson (D) do not exhibit any significant differences between geriatric (blue dots) and young adult (red dots) animals. Color gradients of the dots indicate the age range with lighter color indicate youngest and darker color indicates the oldest animals of respective group. Test of significance (B–D) by Mann-Whitney U test, using GraphPad Prism. P-values shown within the graphs are exact.

Figure 3

Phylum and Class-level differences in microbiota compositions with age: Stacked bar-plot representation of the percent distribution of microbiota compositions for each animals with taxonomic features at the level of phyla (A), and Class (B). The phyla, and class with low relative abundance are grouped as other. At the phylum level, Rhesus gut microbiome is Proteobacteria and Firmicutes dominated, followed by phylum Bacteroidetes. At the class level, classes belonging to Proteobacteria (Gammaproteobacteria) and Firmicutes (Bacilli, negetivicutes) dominates the landscape.

Figure 4

Heatmap showing the relative abundance values for the bacterial species-level differences in the old and young geriatric and young adult animals. All species have significant differences between groups. Species significantly higher in old geriatric animals are marked with ‘**’. The commits in the heatmap are the relative abundance values. Test of significance was pairwise Mann-Whitney U test and p < 0.05 was considered significant.

Figure 5

Correlation analysis of prevalent microbes in young adult or geriatric Rhesus Macaques and plasma inflammatory cytokines: Microbes were selected based off increased prevalence in young adult (A) or geriatric (B) rhesus macaques and then correlated with plasma cytokines. Positive correlations are in red. Negative correlations are in blue. Statistical differences are identified by *. Geriatric animals showed significantly more correlations with plasma cytokines, all of which were positive correlations. The list of bacteria and cytokines were derived from a Benjamini-Hochberg FDR corrected list. Test of significance for this analysis was a simple 2-tailed t-test for the correlation.

Figure 6

Bacterial association networks and differences with age: Bacterial association networks and differences with age: Markov Clustering Algorithm (MCL) evaluates correlations of vector abundance of bacterial species within each group, thereby generating a cluster map by using correlation scores as distance. Figure shows that microbiome of young adult animals group into a single large super-cluster at microbial homeostasis (A). However, with age, major constraints have been introduced into the network structure with the emergence of satellite clusters and scattering of the super cluster (B). Cluster identities are listed in Supplementary Figure 4 . Different colors signify different MCL minor clusters within the microbiome. The data used is species-level identified list, and the rendering for MCL includes only those with Pearson’s r > 0.85 and unidentifiable feature-level sequences were not included for this analysis.

Figure 7

Crosstalk between the bacterial and host metabolic pathways: Mass-spectrometry based Isotope Ratio Outlier Analysis (IROA) showed a few metabolites that were significantly different between the two groups (A). Bar graph showing -log(p) enrichment of the biosynthetic and stress pathways in the geriatric animals upon mapping these metabolites on pathways (B). Enrichment -log(p) of the all degradation pathways with the exception of carbohydrate and secondary metabolite catabolism in geriatric animals (C). LC-MS/MS analysis of the pure microbial fraction and pathway enrichment analysis showed that glycerolipid, amino acid and secondary metabolite metabolism predominated the metabolic landscape of both groups- Red color signifies significantly enriched pathways in geriatric macaques (D). Upon mapping the significantly different metabolites between geriatric and young adult macaques (E) to metabolic pathways, a synchrony was observed between microbial and host biosynthetic and degradation pathways. All metabolites used in the analysis were significantly different between young adult and geriatric animals (p < 0.05 in Mann-Whitney U test) with FDR of 0.1 (Benjamini-Hochberg correction).

Figure 8

Microbial metabolites and host disease pathways: In the microbe-free fecal compartment, after adjusting for diet-derived metabolites, several metabolites were significantly altered, that were mapped to several pathways in the aged animals. When all significantly changing metabolites in geriatric and young adult animals were mapped for human pathway impact (MetaboAnalyst 4.0),secondary metabolite synthesis, nucleotide synthesis and amino acid synthesis and degradation were still at synchrony with bacterial and host metabolic status for metabolites either absent (A) or down-regulated (B) in the microbial complement in geriatric animals (also see Supplementary Table 1 ). Metabolite-disease interaction network mapping identified several human diseases, mostly centered around alteration in Phosphate (C), L-Argenine/Oleic acid (D) and L-Methionine/L-Phenylalanine (E) metabolism. All metabolites used in the analysis were significantly different between young adult and geriatric animals with FDR of 0.1 (Benjamini-Hochberg correction).