Structure, function and diversity of the healthy human microbiome

Abstract

Studies of the human microbiome have revealed that even healthy individuals differ remarkably in the microbes that occupy habitats such as the gut, skin and vagina. Much of this diversity remains unexplained, although diet, environment, host genetics and early microbial exposure have all been implicated. Accordingly, to characterize the ecology of human-associated microbial communities, the Human Microbiome Project has analysed the largest cohort and set of distinct, clinically relevant body habitats so far. We found the diversity and abundance of each habitat's signature microbes to vary widely even among healthy subjects, with strong niche specialization both within and among individuals. The project encountered an estimated 81-99% of the genera, enzyme families and community configurations occupied by the healthy Western microbiome. Metagenomic carriage of metabolic pathways was stable among individuals despite variation in community structure, and ethnic/racial background proved to be one of the strongest associations of both pathways and microbes with clinical metadata. These results thus delineate the range of structural and functional configurations normal in the microbial communities of a healthy population, enabling future characterization of the epidemiology, ecology and translational applications of the human microbiome.

Figure 1. Diversity of the human microbiome is concordant among measures, unique to each individual, and strongly determined by microbial habitat

A) Alpha diversity within subjects by body habitat, as measured using the relative inverse Simpson index of 16S rRNA gene OTUs (red), genus-level phylotypes (blue), shotgun metagenomic reads matched to reference genomes (green), functional modules (yellow), and enzyme families (white). The mouth generally shows high within-subject diversity and the vagina low diversity, with other habitats intermediate; variation among individuals often exceeds variation among body habitats. B) Bray-Curtis beta diversity among subjects by body habitat, colors as for A. Skin differs most between subjects, with oral habitats and vaginal genera more stable. Although alpha- and beta-diversity are not directly comparable, changes in structure among communities (A) occupy a wider dynamic range than do changes within communities among individuals (B). C) Principal coordinates plot showing variation among samples demonstrates that primary clustering is by body area, with the oral, gastrointestinal, skin, and urogenital habitats separate; the nares bridge oral and skin habitats. D) Repeated samples from the same subject (red) are more similar than microbiomes from different subjects (yellow). Technical replicates (green) are in turn more similar; these patterns are consistent for all body habitats and for both phylogenetic and metabolic community composition. See previously described sample counts for all comparisons.

Figure 2. Carriage of microbial taxa varies while metabolic pathways remain stable within a healthy population

Vertical bars represent microbiome samples by body habitat in the seven locations with both shotgun and 16S data; bars indicate relative abundances colored by A) microbial phyla from binned OTUs and B) metabolic modules. Legend indicates most abundant phyla/pathways by average within one or more body habitats; RC = retroauricular crease. A plurality of most communities’ memberships consists of a single dominant phylum (and often genus; see Supp. Fig. 2), but this is universal neither to all body habitats nor to all individuals. Conversely, most metabolic pathways are evenly distributed and prevalent across both individuals and body habitats.

Figure 3. Abundant taxa in the human microbiome, which has been metagenomically and taxonomically well-defined in the HMP population

A–C) Prevalence (intensity, color denoting phylum/class) and abundance when present (size) of clades in the healthy microbiome. The most abundant A) metagenomically-identified species, B) 16S-identified genera, and C) PATRIC “pathogens” (metagenomic). The population size and sequencing depths of the HMP have well-defined the microbiome at all assayed body sites, as assessed by saturation of added community D) metabolic configurations (rarefaction of minimum Bray-Curtis β-diversity of metagenomic EC abundances to nearest neighbor, inter-quartile range over 100 samples) and E) phylogenetic configurations (min. 16S OTU weighted UniFrac distance to nearest neighbor).

Figure 4. Microbial carriage varies between subjects down to the species and strain level

Metagenomic reads from 127 tongue samples spanning 90 subjects were processed with MetaPhlAn to determine relative abundances for each species. A) Relative abundances of 11 distinct Streptococcus spp. In addition to variation in broader clades (see Fig. 2), individual species within a single habitat demonstrate a wide range of compositional variation. Inset illustrates average tongue sample composition. B) Metabolic modules present/absent (grey/white) in KEGG reference genomes of tongue streptococci denote selected areas of strain-specific functional differentiation. C) Comparative genomic coverage for the single Streptococcus mitis B6 strain. Grey dots are median Reads Per Kilobase per Million reads (RPKM) for 1kb windows, gray bars are the 25th to 75th percentiles across all samples, red line the lowess smoothed average. Red bars at the bottom highlight predicted genomic islands. Large, discrete, and highly variable islands are commonly under-represented. D) Two islands highlighted, V = V-type H+ ATPase subunits I,K,E,C,F,A & B, and CH = Choline binding proteins cbp6 and cbp12, indicating functional cohesion of strain-specific gene loss within individual human hosts.

Figure 5. Microbial community membership and function correlates with host phenotype and sample metadata

The pathway and clade abundances most significantly associated (all FDR q<0.2) using a multivariate linear model with A) subject race or ethnicity, B) vaginal posterior fornix pH, C) subject age, and D) BMI. Samples’ scatter plots are shown with lines indicating best simple linear fit. Race/ethnicity and vaginal pH are particularly strong associations; age and BMI are more representative of typically modest phenotypic associations (Sup. Table 3), suggesting that variation in the healthy microbiota may correspond to other host or environmental factors.