Age and sex-associated variation in the multi-site microbiome of an entire social group of free-ranging rhesus macaques

Abstract

Background: An individual's microbiome changes over the course of its lifetime, especially during infancy, and again in old age. Confounding factors such as diet and healthcare make it difficult to disentangle the interactions between age, health, and microbial changes in humans. Animal models present an excellent opportunity to study age- and sex-linked variation in the microbiome, but captivity is known to influence animal microbial abundance and composition, while studies of free-ranging animals are typically limited to studies of the fecal microbiome using samples collected non-invasively. Here, we analyze a large dataset of oral, rectal, and genital swabs collected from 105 free-ranging rhesus macaques (Macaca mulatta, aged 1 month-26 years), comprising one entire social group, from the island of Cayo Santiago, Puerto Rico. We sequenced 16S V4 rRNA amplicons for all samples.

Results: Infant gut microbial communities had significantly higher relative abundances of Bifidobacterium and Bacteroides and lower abundances of Ruminococcus, Fibrobacter, and Treponema compared to older age groups, consistent with a diet high in milk rather than solid foods. The genital microbiome varied widely between males and females in beta-diversity, taxonomic composition, and predicted functional profiles. Interestingly, only penile, but not vaginal, microbiomes exhibited distinct age-related changes in microbial beta-diversity, taxonomic composition, and predicted functions. Oral microbiome composition was associated with age, and was most distinctive between infants and other age classes.

Conclusions: Across all three body regions, with notable exceptions in the penile microbiome, while infants were distinctly different from other age groups, microbiomes of adults were relatively invariant, even in advanced age. While vaginal microbiomes were exceptionally stable, penile microbiomes were quite variable, especially at the onset of reproductive age. Relative invariance among adults, including elderly individuals, is contrary to findings in humans and mice. We discuss potential explanations for this observation, including that age-related microbiome variation seen in humans may be related to changes in diet and lifestyle. Video abstract.

Keywords: Aging; Genital microbiome; Gut microbiome; Non-human primates; Oral microbiome; Sex differences.

Fig. 1

Community diversity in the microbiome of rhesus macaque rectums, oral cavities, penises, and vaginas. Alpha-diversity (Shannon Index) varies between (a-d) infants and non-infants and (e) across sample sites. Beta-diversity varies (f) across sample sites and (g-j) between age groups, with differences driven by infants vs. non-infants in (f) rectal, (g) oral, and (i) penile samples, but (j) no age differences in vaginal samples

Fig. 2

Differences in relative abundances of the top 10 genera were driven by differences between infants and non-infants in (a) rectal and (b) oral microbiomes. Top 10 genera did not vary by age in (c) vaginal samples but did in (d) penile microbiomes. The youngest two age groups are pooled for vaginal samples because of small sample sizes for infant vaginas

Fig. 3

Pairwise comparisons of differentially abundant taxa between age groups in (a) rectal, (b) oral, and (c) penile microbiomes. Figures show bacterial taxa that are significantly differentially abundant between adjacent age groups with a log2fold change of at least +/− 2 (subheadings identify the denominator vs. numerator). NAs reflect bacteria for which order, family, and genus were unidentified. No taxa differed between oral samples after the first age comparison and no taxa differed between vaginal samples in any age group comparison

Fig. 4

Functional pathways predicted by PiCrust2 that differentiate age groups in (a) rectal, (b) oral, (c) vaginal, and (d) penile communities. Differentiating pathways identified by LefSe with LDA effect size ≥2 and alpha ≤ 0.01