Time-resolved profiling reveals nonlinear dose-time trajectories of gut microbiota in PM_2.5 and PFOS exposure models

The gut microbiota is a dynamic ecological system, yet environmental microbiome studies often provide only limited temporal resolution. We characterized gut microbial responses across multiple exposure levels and sampling stages in two independent but complementary mouse models of PM_2.5 inhalation and perfluorooctane sulfonate (PFOS) drinking-water exposure. Time-resolved 16S rRNA gene sequencing, ecological analyses, PICRUSt2-based functional inference, and generalized additive models (GAMs) were integrated to examine temporal succession and nonlinear dose-time relationships. Under unexposed conditions, temporal microbial succession was evident in both models, highlighting the importance of accounting for background community dynamics. Within the PM_2.5 model, community restructuring was most apparent during early-to-intermediate exposure stages, with prominent nonlinear trajectories in Lactobacillus, Prevotella, and the Firmicutes/Bacteroidetes ratio. In the PFOS model, microbial restructuring extended across a longer observation period and involved taxa including Akkermansia, Allobaculum, and UBA7173. PICRUSt2 suggested temporal and dose-associated variation in predicted pathways related to nitrogen metabolism, carbohydrate utilization, and fermentation, although these functions were not directly measured. GAMs identified stage-specific nonlinear response windows in both models. These findings highlight the importance of jointly considering dose and time in environmental microbiome studies and demonstrate the value of time-resolved modeling for detecting microbial response patterns that may be missed by endpoint analyses.