Effects of the Antivitamin Roseoflavin on the Structure of the RNA-Derived Bacterial Community in Pooled Human Saliva: An Exploratory In Vitro Study

Background/Objectives: The antivitamin roseoflavin is a structural analogue of riboflavin and may affect oral bacterial communities through competitive antagonism. Methods: In this exploratory in vitro study, saliva from 14 volunteers was pooled before allocation to the experimental conditions and incubated for 24 h with roseoflavin or ampicillin. RNA was isolated and RNA-derived bacterial community profiles were analyzed by 16S rRNA sequencing using QIIME 2 and ANCOM-BC2. Because all experimental replicates originated from a single pooled biological source, the findings do not support donor-level or population-level inference. Results: Total RNA levels and alpha diversity showed treatment-associated tendencies, with the strongest reduction after ampicillin treatment; however, these differences were not statistically significant. Beta diversity indicated modest treatment-associated variation within the pooled experimental system. Differential abundance analysis indicated moderate shifts in the relative representation of taxa within the RNA-derived oral bacterial community. Valid ANCOM-BC2-derived log2 fold change estimates for both vehicle-matched treatment comparisons were obtained for 45 taxonomic units. Most taxa showed concordant directional responses to roseoflavin and ampicillin, whereas only a small subset showed treatment-specific descriptive patterns. Ampicillin tended to induce larger absolute log2 fold changes than roseoflavin. Conclusions: Within this pooled exploratory model, roseoflavin treatment was associated with moderate shifts in the RNA-derived oral bacterial community rather than broad community-wide changes. Whether the comparatively limited response reflects selective activity, insufficient antimicrobial potency, community-level buffering, or a combination of these mechanisms remains unresolved. The findings require validation using independent donors, controlled dose–response experiments, and more physiologically relevant oral models.