Electronic Cigarette Use Promotes a Unique Periodontal Microbiome

Abstract

Electronic cigarettes (e-cigs) have become prevalent as an alternative to conventional cigarette smoking, particularly in youth. E-cig aerosols contain unique chemicals which alter the oral microbiome and promote dysbiosis in ways we are just beginning to investigate. We conducted a 6-month longitudinal study involving 84 subjects who were either e-cig users, conventional smokers, or nonsmokers. Periodontal condition, cytokine levels, and subgingival microbial community composition were assessed, with periodontal, clinical, and cytokine measures reflecting cohort habit and positively correlating with pathogenic taxa (e.g., Treponema, Saccharibacteria, and Porphyromonas). α-Diversity increased similarly across cohorts longitudinally, yet each cohort maintained a unique microbiome. The e-cig microbiome shared many characteristics with the microbiome of conventional smokers and some with nonsmokers, yet it maintained a unique subgingival microbial community enriched in Fusobacterium and Bacteroidales (G-2). Our data suggest that e-cig use promotes a unique periodontal microbiome, existing as a stable heterogeneous state between those of conventional smokers and nonsmokers and presenting unique oral health challenges. IMPORTANCE Electronic cigarette (e-cig) use is gaining in popularity and is often perceived as a healthier alternative to conventional smoking. Yet there is little evidence of the effects of long-term use of e-cigs on oral health. Conventional cigarette smoking is a prominent risk factor for the development of periodontitis, an oral disease affecting nearly half of adults over 30 years of age in the United States. Periodontitis is initiated through a disturbance in the microbial biofilm communities inhabiting the unique space between teeth and gingival tissues. This disturbance instigates host inflammatory and immune responses and, if left untreated, leads to tooth and bone loss and systemic diseases. We found that the e-cig user's periodontal microbiome is unique, eliciting unique host responses. Yet some similarities to the microbiomes of both conventional smokers and nonsmokers exist, with strikingly more in common with that of cigarette smokers, suggesting that there is a unique periodontal risk associated with e-cig use.

Keywords: dysbiosis; electronic cigarette; microbiome; periodontitis; subgingival plaque.

FIG 1

Clinical measures validate patient inclusion and demonstrate disease status and progression in specific cohorts. (A) Patient breath carbon monoxide levels in parts per million, saliva cotinine concentration, and the average distance from the free gingival margin to the depth of the pocket (pocket depth); sample number is given below cohort designation. Kruskal-Wallis H with post hoc Dunn’s test was performed, with multiplicity-adjusted P values reported. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. On each visit, patients were evaluated for the degree of periodontitis, as described by Xu et al. (7) (B). CS, conventional cigarette smokers; ES, e-cigarette users; NS, nonsmokers. Number of subjects is given below each cohort.

FIG 2

Periodontal microbiome α-diversity remains similar among cohorts, yet community structure is unique. α- and β-diversities of periodontal microbial communities for conventional cigarette smokers, e-cigarette users, and nonsmokers are shown. Sample numbers are provided in Materials and Methods. α-Diversity between visits tended to increase within a cohort (A) (paired-sample Wilcoxon). The degree of change in α-diversity measures between visits (v2 – v1) for a given cohort was not significantly different between the cohorts (B) (Mann-Whitney U test). When visits were merged within a cohort, no significant differences in α-diversity were observed (C). Mean and SEM are shown for panels A to C; box indicates the interquartile range of the data. β-Diversity of periodontal microbial communities was significantly different between cohorts (PERMANOVA) (D). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. A Venn diagram depicts shared and unique amplicon sequence variants (ASVs) among the cohorts; percentages of total ASVs are in parentheses (E).

FIG 3

Cohorts displayed distinct patterns in taxa mean relative abundance. Periodontal microbial community composition at the class level and genera-based differential relative abundance are shown for CS, ES, and NS. (A) Per-visit bar plots for relatively abundant classes in the different cohorts. (B) Statistically significant differentially relatively abundant genera grouped based on abundance patterns in the three cohorts (visits merged within a cohort) (Mann-Whitney U test; *, P < 0.05; **, P < 0.01; ***, P < 0.001). Rel. Abd., relative abundance.

FIG 4

Mean-relative-abundance patterns for genera distinguish nonsmokers from e-cig users and conventional smokers. A hierarchical-clustering relative-abundance heat map of the 20 most relatively abundant genera in the three cohorts with row z-score is displayed.

FIG 5

The e-cig user periodontal microbiome resembles those of both conventional smokers and nonsmokers. Supervised learning sample classification can accurately predict sample inclusion in the CS and NS but struggles with the ES cohort (A). The model accuracy was tested on 34 samples that were excluded from the training data set (n = 134). All cohorts had areas under the curve (AUC) well above what would be expected by chance (B). A hierarchical-clustering mean-relative-abundance heat map of the 50 most important features (ASVs) shows important clusters with distinct abundance patterns and pathological relevance (C).

FIG 6

Cytokine abundance patterns differ among cohorts, and cytokines positively correlate with known pathogens. (A) Cytokine concentrations (picograms per milliliter) for the three cohorts; sample numbers are displayed below cohort designation. Kruskal-Wallis H with post hoc Dunn’s test was performed, with multiplicity-adjusted P values reported. *, P < 0.05; **, P < 0.01; ***, P < 0.001. (B) Hierarchical-clustering correlation heat map with cytokines and the 20 most relatively abundant genera. *, P < 0.05.

FIG 7

Pathogens and commensals correlate with clinical measures and cytokines. Positive and negative correlation cord diagrams demonstrate correlations between matched samples for the 20 most abundant genera, clinical measures of periodontal disease, and cytokines. BoP, bleeding on probing; CO, carbon monoxide.