Zebrafish embryo/larval toxicity testing with PFAS to address uncertainties in transcriptomic point of departure estimates across species.

Background: Mounting evidence suggests that chemical concentrations that do not elicit concerted molecular responses over relatively short exposure durations (e.g., 24 h to 5 d) generally do not elicit adverse effects, even over much longer exposure durations. This has led to proposals to implement an omics-based regulatory testing paradigm which would use transcriptomic points of departure (tPOD; a benchmark dose/concentration -based treatment level below which a concerted gene expression response is not observed) as a health protective exposure level for risk assessment (Johnson et al. 2022). While initial research on the application of tPODs has focused on human health protection, more recently our research team and others have started to explore application of this concept for ecosystem protection as well. To evaluate the scientific underpinnings of the approach, two critical questions as it pertains to regulatory application are: 1) How variable are tPODs as a function of common experimental design variables? 2) How do tPODs for the same chemical compare across species? To examine these questions, zebrafish (Danio rerio) embryos were exposed to three different per- or polyfluoroalkyl substances (PFAS), perfluorooctanesulfonic acid (PFOS; DTXSID3031864); perfluorooctanoic acid (PFOA; DTXSID8031865); and perfluorohexane sulfonate (PFHxS; DTXSID90892476) in concentration-response, and then whole-body gene expression was determined using RNA sequencing (RNAseq). Embryos were exposed to each of the three PFAS, as well as a chlorpyrifos positive control, over seven distinct exposure periods (6 hours post fertilization [hpf]-120 hpf; 6 hpf-48 hpf; 6 hpf-24 hpf; 24-48 hpf; 24 hpf – 120 hpf; 48 hpf-120 hpf; and 96 hpf-120 hpf; Figure 1) with the goal of determining how much the tPOD varies as a function of the developmental window over which the organisms were exposed. In a separate study (CSS.1.7.5), tPODs for PFOS, PFOA, and PFHxS were generated for several other species including fathead minnow (Pimephales promelas), Daphnia magna (a crustacean), Chironomous dilutus (an insect) and Raphidocelis subcapitata (an algae). Thus, the tPODs generated for zebrafish embryos can also be compared against these data. Dataset: Processed RNAseq data are provided as normalized count matrices (counts per million reads; filtered to remove genes with <15 reads across all samples; log2 transformed with a pseudo count of 1 added to avoid negative value) with individual genes as rows and samples as columns. Data sets are organized into 21 count matrices, one for each exposure window evaluated for each chemical (T1-T7; Table 1). A meta-data file that that defines the exposure window, test chemical, treatment concentration, and replicate number for each sample is also provided. RNAseq was repeated for a sub-set of samples. Repeated samples are indicated in the metadata file.