Dynamics of tongue microbial communities with single-nucleotide resolution using oligotyping. Academic Article uri icon


  • The human mouth is an excellent system to study the dynamics of microbial communities and their interactions with their host. We employed oligotyping to analyze, with single-nucleotide resolution, oral microbial 16S ribosomal RNA (rRNA) gene sequence data from a time course sampled from the tongue of two individuals, and we interpret our results in the context of oligotypes that we previously identified in the oral data from the Human Microbiome Project. Our previous work established that many of these oligotypes had dramatically different distributions between individuals and across oral habitats, suggesting that they represented functionally different organisms. Here we demonstrate the presence of a consistent tongue microbiome but with rapidly fluctuating proportions of the characteristic taxa. In some cases closely related oligotypes representing strains or variants within a single species displayed fluctuating relative abundances over time, while in other cases an initially dominant oligotype was replaced by another oligotype of the same species. We use this high temporal and taxonomic level of resolution to detect correlated changes in oligotype abundance that could indicate which taxa likely interact synergistically or occupy similar habitats, and which likely interact antagonistically or prefer distinct habitats. For example, we found a strong correlation in abundance over time between two oligotypes from different families of Gamma Proteobacteria, suggesting a close functional or ecological relationship between them. In summary, the tongue is colonized by a microbial community of moderate complexity whose proportional abundance fluctuates widely on time scales of days. The drivers and functional consequences of these community dynamics are not known, but we expect they will prove tractable to future, targeted studies employing taxonomically resolved analysis of high-throughput sequencing data sampled at appropriate temporal intervals and spatial scales.

publication date

  • 2014