The human oral cavity microbiota composition during acute tonsillitis: a cross-sectional survey

Abstract

Background: Microbial culture-based investigations of inflamed tonsil tissues have previously indicated enrichment of several microorganisms such as Streptococcus, Staphylococcus and Prevotella. These taxa were also largely reflected in DNA sequencing studies performed using tissue material. In comparison, less is known about the response of the overall oral cavity microbiota to acute tonsillitis despite their role in human health and evidence showing that their compositions are correlated with diseases such as oral cancers. In addition, the influence of subject-specific circumstances including consumption of prescription antibiotics and smoking habits on the microbiology of acute tonsillitis is unknown.

Methods: We collected oral rinse samples from 43 individuals admitted into hospital for acute tonsillitis and 165 non-disease volunteers recruited from the public, and compared their microbial community compositions using 16S rRNA gene sequencing. We assessed the impact of tonsillitis, whether subjects were prescribed antibiotics, the presence of oral abscesses and their smoking habits on community composition, and identified specific microbial taxa associated with tonsillitis and smoking.

Results: Oral rinse community composition was primarily associated with disease state (tonsillitis vs non-tonsillitis) although its effect was subtle, followed by smoking habit. Multiple Prevotella taxa were enriched in tonsillitis subjects compared to the non-tonsillitis cohort, whereas the non-tonsillitis cohort primarily showed associations with several Neisseria sequence variants. The presence of oral abscesses did not significantly influence community composition. Antibiotics were prescribed to a subset of individuals in the tonsillitis cohort but we did not observe differences in community composition associated with antibiotics consumption. In both tonsillitis and non-tonsillitis cohorts, smoking habit was associated with enrichment of several Fusobacterium variants.

Conclusions: These findings show that the oral cavity microbial community is altered during acute tonsillitis, with a consistent enrichment of Prevotella during tonsillitis raising the possibility of targeted interventions. It also supports the possible link between smoking, Fusobacteria and oral cancers.

Keywords: 16S ribosomal RNA gene; Fusobacteria; Microbial community; Mouth rinse; Prevotella; Smoking.

Fig. 1

Ordination of oral rinse microbial communities in healthy individuals and patients with tonsillitis. a Principal component analysis (PCA) of centered log ratio-transformed exact sequence variant (ESV) counts. Each circle represents community composition of one sample; the closer two dots are the more compositionally similar their microbial communities. b Three dimensional representation of PCA in panel A, with spokes connecting samples to their respective centroids according to cohort (healthy or tonsillitis). c Principal coordinate analysis (PCoA) of weighted UniFrac distances calculated in QIIME 2. Green circles represent samples from non-tonsillitis individuals, orange circles represent samples from patients with tonsillitis. Community composition of samples from tonsillitis patients and healthy individuals significantly differed irrespective of metric used (transformed ESV counts and weighted UniFrac) (p < 0.001, PERMANOVA)

Fig. 2

Top 10 exact sequence variants (ESVs) associated with the tonsillitis and healthy cohorts ranked according to linear discriminant analysis (LDA) effect size. ESVs were identified using the LDA effect size (LEfSe) algorithm with default parameters. Taxonomy labels indicate genus level classification (or lowest rank available) with reference to the SILVA database. Green bars represent ESVs associated with non-tonsillitis individuals, orange bars represent ESVs associated with patients with tonsillitis

Fig. 3

Exact sequence variants (ESVs) associated with smokers and non-smokers in the tonsillitis cohorts ranked according to linear discriminant analysis (LDA) effect size. ESVs were identified using the linear discriminant analysis effect size (LEfSe) algorithm. Only ESVs with an LDA score of at least 2 in their respective groups are shown. Genus level classification (or lowest rank available) with reference to the SILVA database is provided. Red bars represent ESVs associated with smokers, whereas blue bars represent ESVs associated with non-smokers