Potential Roles of the Free Salivary Microbiome Dysbiosis in Periodontal Diseases

Abstract

Saliva is a vital mediator in the oral cavity. The dysbiosis of free bacteria in saliva might be related to the onset, development, prognosis, and recurrence of periodontal diseases, but this potential relationship is still unclear. The objective of this study was to investigate the potential roles of the free salivary microbiome in different periodontal statuses, their reaction to nonsurgical periodontal therapy, and differences between diseased individuals after treatment and healthy persons. We recruited 15 healthy individuals, 15 individuals with gingivitis, and 15 individuals with stage I/II generalized periodontitis. A total of 90 unstimulated whole saliva samples were collected and sequenced using full-length bacterial 16S rRNA gene sequencing. We found that as the severity of disease increased, from healthy to gingivitis and periodontitis, the degree of dysbiosis also increased. A higher abundance of Prevotella intermedia and Catonella morbi and a lower abundance of Porphyromonas pasteri, Prevotella nanceiensis, and Haemophilus parainfluenzae might be biomarkers of periodontitis, with an area under curve (AUC) reaching 0.9733. When patients received supragingival scaling, there were more pathogens related to recolonization in the saliva of periodontitis patients than in healthy persons. Even after effective nonsurgical periodontal therapy, individuals with periodontitis displayed a more dysbiotic and pathogenic microbial community in their saliva than healthy individuals. Therefore, the gradual transition in the entire salivary microbial community from healthy to diseased includes a gradual shift to dysbiosis. Free salivary pathogens might play an important role in the recolonization of bacteria as well as the prognosis and recurrence of periodontal diseases.

Keywords: dysbiosis; full-length 16S rRNA gene sequencing; generalized periodontitis (stage I/II); gingivitis; non-surgical therapy; salivary microbiome.

Figure 1

The flow chart of this study showing the enrollment and classification of participants to subsequent microbiota analyses.

Figure 2

Comparisons of alpha diversity and beta diversity among the three baseline groups, H, G0, and P0. (A) Microbial richness presented by Chao1 and microbial diversity presented by Shannon and Simpson indices. (B) PCoA based on Bray-Curtis distance exhibiting the variation of community structure in the three groups. (C) Intragroup dispersion presented by the corresponding interindividual weighted UniFrac distances for each group. **P < 0.01, ***P < 0.001 by Kruskal-Wallis test. Each color represents one group: green for H, yellow for G0, and red for P0.

Figure 3

Differential species and ROC analysis. (A) Comparison of the relative abundance of bacteria by species between the H and P0 groups at baseline. P < 0.05 by Welch’s t-test. (B) Receiver operating characteristic (ROC) analysis of distinguishing group H from group P0 using detected species assessed by the area under the curve (AUC). (C) ROC curve for the logistic regression model of combined diagnosis by the five detected species.

Figure 4

Random forest classification model for different periodontal statuses based on free salivary bacteria. The top 30 OTUs with the highest importance for classification are ranked on the left (see Appendix Table 3 for the complete list of taxa), and the score of importance is shown on the right. Different colors on the top identify the six sample groups. The heat map shows the abundance distribution of these taxa in each sample. Boxes 1, 2, and 5 were characterized by high levels of disease-associated taxa. Boxes 3 and 4 were characterized by health-associated taxa and other bacteria considered to have low pathogenicity.

Figure 5

MetagenomeSeq analyses of significantly different taxa (occurrence frequency of OTUs in the upregulated group was greater than 0.3) between different groups. The differential taxa are annotated at the genus level at the bottom and at the species level at the top. (A) Upregulated OTUs in group P2 compared to P0. (B) Downregulated OTUs in group P2 compared to P1. (C) Upregulated OTUs in group P1 compared to H. (D) Upregulated OTUs in group P2 compared to H.

Figure 6

Diagrams of the ecological plaque hypothesis in periodontal disease. (A) Chart showing changing trends in the degree of dysbiosis from healthy to diseased and after nonsurgical treatment of periodontitis. (B) Chart showing changing trends in the degree of dysbiosis resulting from the combined action of three factors: the amount of bacteria, the species composition, and their metabolic function. The length of the arrows represents their respective influence. The angle between any two of the arrows represents an interaction between the factors, while the interaction is not yet well characterized.