Oral Microbiome Shifts From Caries-Free to Caries-Affected Status in 3-Year-Old Chinese Children: A Longitudinal Study.

As one of the most prevalent human infectious diseases, dental caries results from dysbiosis of the oral microbiota driven by multiple factors. However, most of caries studies were cross-sectional and mainly focused on the differences in the oral microbiota between caries-free (CF) and caries-affected (CA) populations, while little is known about the dynamic shift in microbial composition, and particularly the change in species association pattern during disease transition. Here, we reported a longitudinal study of a 12-month follow-up of a cohort of 3-year-old children. Oral examinations and supragingival plaque collections were carried out at the beginning and every subsequent 6 months, for a total of three time points. All the children were CF at enrollment. Children who developed caries at 6-month follow-up but had not received any dental treatment until the end of the study were incorporated into the CA group. Children who remained CF at the end of the study were incorporated into the CF group. Using Illumina Miseq Sequencing of the 16S rRNA gene, we monitored the shift of supragingival microbiome during caries initiation and progression in children who developed caries over the 12-month study period. Intriguingly, principle coordinates analyses revealed two major shifting patterns in microbial structures during caries initiation and progression in CA group, but not in CF group. Dynamic co-occurring OTU network study showed that compared to CF group, there was significant increase in both number and intensity of correlations between microbial taxa, as well as the formation of tight clusters of specific bacteria in CA group. Furthermore, there were enhanced correlations, positive ones between CA-enriched taxa, and negative ones between CF-enriched and CA-enriched species within CA group. Our data suggested coordinated microbial interactions could be essential to caries pathogenesis. Most importantly, our study indicated that significant microbial shifts occur not only during caries development, but even in the sub-clinical state. Using supragingival microbiome profiles, we were able to construct a caries-onset prediction model with a prediction accuracy of 93.1%. Our study indicated that the microbial shifts prior to the onset of caries might potentially be used for the early diagnosis and prediction of caries.