Inter-personal diversity and temporal dynamics of dental, tongue, and salivary microbiota in the healthy oral cavity

Abstract

Oral microbes form a complex and dynamic biofilm community, which is subjected to daily host and environmental challenges. Dysbiosis of the oral biofilm is correlated with local and distal infections and postulating a baseline for the healthy core oral microbiota provides an opportunity to examine such shifts during the onset and recurrence of disease. Here we quantified the daily, weekly, and monthly variability of the oral microbiome by sequencing the largest oral microbiota time-series to date, covering multiple oral sites in ten healthy individuals. Temporal dynamics of salivary, dental, and tongue consortia were examined by high-throughput 16S rRNA gene sequencing over 90 days, with four individuals sampled additionally 1 year later. Distinct communities were observed between dental, tongue, and salivary samples, with high levels of similarity observed between the tongue and salivary communities. Twenty-six core OTUs that classified within Streptococcus, Fusobacterium, Haemophilus, Neisseria, Prevotella, and Rothia genera were present in ≥95% samples and accounted for ~65% of the total sequence data. Phylogenetic diversity varied from person to person, but remained relatively stable within individuals over time compared to inter-individual variation. In contrast, the composition of rare microorganisms was highly variable over time, within most individuals. Using machine learning, an individual's oral microbial assemblage could be correctly assigned to them with 88-97% accuracy, depending on the sample site; 83% of samples taken a year after initial sampling could be confidently traced back to the source subject.

Fig. 1

NMDS ordinations of samples showing sample dissimilarity patterns under the Bray–Curtis measure. a All samples colored by oral site (stress value: 0.16). b Tongue and saliva samples colored by oral site (stress value: 0.20). c All samples colored by subject number (stress value: 0.16). d Supragingival plaque samples colored by subject number (stress value: 0.16). e Tongue and saliva samples colored by subject number (stress value: 0.20)

Fig. 2

Boxplots of phylogenetic diversity measures. a PD values grouped by sample site (*** = significant at p < 0.001 by pairwise Wilcox test with false discovery rate correction). b PD values grouped by sample site and subject

Fig. 3

Venn diagram showing the taxonomic classifications of some of the core OTUs present within ≥95% of the samples in the indicated subsets of oral sites. For example, Bulleidia moorei was found in ≥95% of tongue plaque and saliva samples, but <95% of supragingival plaque samples, while Rothia dentocariosa was found in ≥95% of the samples from all three sites. This list is not exhaustive; see Supplementary File 1 for a complete list

Fig. 4

Lag plots of beta-diversity measures between sample pairs against the time elapsed between sample dates. Yellow lines are the linear least squares regression fits with 95% confidence bands and blue lines are the Loess locally weighted regression curves with 95% confidence bands. a Bray–Curtis distances between saliva samples from subject 7. b Sorensen distances between saliva samples from subject 7. c Sorensen distances between saliva samples from subject 1. d Sorensen distances between saliva samples from subject 10