Metatranscriptomic analysis reveals toxin-antitoxin system shifts in caries-associated oral microbiomes

BackgroundToxin-antitoxin (TA) systems are bacterial regulatory elements involved in persistence, dormancy, and biofilm stability under stress conditions such as limited nutrients, acid exposure, and antimicrobial treatment. Although their roles are well-characterized in isolated pathogens, TA systems remain largely unexamined in the context of the oral microbiome. Given their potential involvement in microbial adaptation during oral diseases and treatment, we aimed to identify transcriptionally active TA systems and evaluate their condition-specific expression patterns across oral health states using metatranscriptomic data. ResultsWe re-analyzed two publicly available supragingival metatranscriptomic datasets: a longitudinal dataset by Carda-Dieguez et.al. (Dieguez) capturing treatment transitions (fluoride (Fl), fluoride-arginine (Fl-Ar)) in caries-active and caries-free individuals, and a cross-sectional dataset by Ev et.al., (Ev) profiling caries-active lesions (CA), caries-inactive (CI), and caries-free (CF) states. After quality control, host sequences removal, rRNA removal, and HUMAnN3-based functional profiling, we identified 1,189 unique UniRef90 gene clusters with known TA associations. Differential analysis using ANCOM-BC revealed 197 TA-related gene clusters as significantly modulated in at least one condition. Of these, 22 gene clusters were differentially expressed in Dieguez, 38 in Ev, with 77 shared across both datasets. Distinct TA expression signatures were observed across healthy (CF, CI, baseline), lesion types (CAa, CAi, CAs), and treatment stages (Fl, Fl-Ar), suggesting condition-specific regulatory activity. Functional annotation of differentially expressed TA genes using the Toxin-antitoxin system database, eggNOG-mapper, and InterProScan, revealed 18 genes with strong evidence for their involvement in key bacterial stress-response pathways. These included members of the ParD-ParE, RelE/StbE-RelB/StbD, FitA, and YafQ families, which are annotated with mRNA catabolism, transcriptional repression, and prokaryotic defense-associated pathways. ConclusionThis is the first metatranscriptomic analysis profiling functional shifts in toxin-antitoxin systems across oral healthy, caries, and treatment states. TA systems showed dynamic and condition-specific expression, with pathway annotations suggesting their roles in microbial persistence, stress adaptation, and ecological remodeling during caries progression. These findings open new avenues for targeting microbial stress modules in precision microbiome therapeutics.