Intralesional microbial community signatures across histopathologic categories of tongue lesions.

BACKGROUND

The tissue-resident microbiome of oral epithelial lesions remains incompletely characterized, with most prior studies relying on saliva or surface sampling. This study aimed to characterize intralesional microbial communities across histopathologic categories of tongue lesions using formalin-fixed paraffin-embedded (FFPE) tissue.

METHODS

This cross-sectional study characterized the intralesional microbiome of 63 formalin-fixed, paraffin-embedded tongue tissues, including fibroma (n = 15), low malignant potential dysplasia (LMP; n = 24), high malignant potential dysplasia (HMP; n = 24), and additional OSCC samples. Amplicon sequencing of the V3-V4 16S rRNA region was used to assess taxonomic composition, alpha and beta diversity, phylogenetic structure, predicted functional pathways (PICRUSt2), and machine-learning-based discrimination of lesion categories.

RESULTS

Microbial community profiles differed significantly across histopathologic groups. Non-dysplastic tissues exhibited higher richness and greater representation of commensal genera such as Streptococcus, Rothia, and Veillonella. Dysplastic tissues demonstrated reduced diversity and increased abundance of stress-adapted Proteobacteria including Bosea, Novosphingobium, Sphingomonas, and Pseudomonas. Beta diversity analyses revealed distinct community structures between fibroma, LMP, and HMP categories. Predicted functional profiles suggested differences in inferred metabolic potential, including pathways related to carbohydrate metabolism and xenobiotic degradation in dysplastic lesions. A supervised classifier demonstrated separation between groups (AUC 0.83-1.00), with several taxa contributing to classification; however, these findings should be interpreted cautiously given the sample size.

CONCLUSION

Intralesional microbial communities differ across fibroma, dysplasia, and OSCC of the tongue in both taxonomic composition and predicted functional profiles. These findings describe lesion-associated microbial signatures within tissue and provide a foundation for future studies incorporating longitudinal designs and multi-omics approaches to clarify their biological and clinical relevance.