Performance evaluation and methods comparison of transcriptomic-based approaches for the characterization of wastewater treatment effluent
Wastewater treatment effluents (WWTE) present complex risks to aquatic ecosystems that are difficult to characterize using traditional methods. This study systematically evaluated the consistency and performance of transcriptomic-based approaches over time with repeated sampling and with differing experimental approaches (selection of reference condition, grab vs. composite sampling, deployed vs. laboratory exposed). RNA-seq was performed on larval fathead minnow (FHM) exposed in the laboratory to moderately hard reconstituted water (MRHW) or individual grab samples collected from an upstream site and a WWTE in the morning and afternoon over two successive days, as well as FHM deployed concurrently with grab sampling at the same sites. Composite transcriptional profiles were generated by pooling count data from grab sample exposures. The choice of comparator significantly affected results. The use of the upstream site as the reference consistently yielded fewer differentially expressed genes (DEGs) and minimal overlap compared to DEG sets from the other comparisons. Using MRHW as a comparator, DEG sets showed high consistency across grab samples, with morning samples demonstrating larger, highly consistent gene expression sets (96 % overlap) compared to afternoon samples, revealing clear and consistent within-day expression patterns. With the MHRW comparator, DEG sets from grab sample composites and deployments also overlapped substantially, indicating that transcriptional profiles accurately reflect WWTE composition regardless of exposure method. Comparisons with non-targeted (NTA) and targeted analytical datasets confirmed that gene expression interpretations aligned with effluent composition while highlighting limitations of relying solely on targeted analyte sets for connecting expression to specific chemicals. Though highly dependent on experimental design, these results demonstrate that transcriptomic-based methods provide significant utility for characterizing the bioactivity of complex environmental mixtures, and when paired with NTA datasets, have the potential to deliver a comprehensive assessment of their environmental risk.