The experimental design presents a timeline for the developmental toxicity and behavioral neurotoxicity experiments from spawning through assessments. Procedures for the two studies were very similar, differing primarily in the type of plate (mesh or square-well), distribution of the chemicals on the plate (by column, or random), chemical renewal process due to plate type, and behavioral assessments included in the behavioral neurotoxicity experiment. Typical developmental toxicity plates included 2 test chemicals, represented by the blue and orange circles (color gradient represents the 8 concentrations for each chemical), light-grey circles represent DMSO vehicle control, and the green circles represent the positive control fluoxetine (40 µM). Only one test chemical was tested on each behavioral neurotoxicity plate, and concentrations were randomly distributed using the ECHO system. The color gradient represents the different chemical concentrations, and light-grey squares represent DMSO vehicle control. The different shades of blue represent an example of how the chemical concentrations might be distributed across the plate by the ECHO, with each shade of blue representing a concentration of the one chemical on that plate.

Fluoxetine results following the developmental toxicity and behavioral neurotoxicity experiments. The bar graph in Panel (A) presents data from the developmental toxicity observational assessments indicating the status (percent normal, not hatched, abnormal, severely abnormal, and dead); the vertical axis indicates percent (%) affected while the horizontal axis shows fluoxetine concentrations ≤ 40 µM (n = 28 for 40 µM and n = 8 for all lower concentrations). Beneath the graph are pictures of common malformations previously published by our laboratory [45]. The photos illustrate various malformations and normal-looking images. Panel (B) combines the four fluoxetine (0–4 µM) behavioral neurotoxicity experiments together and shows the average distance moved every two minutes in a line graph on the left, and a bar graph on the right with the average distance moved in the light and dark periods separately, with the average for each individual larva represented by the circles. Panel (C) presents the behavioral neurotoxicity results from a 13-endpoint analysis in a heatmap, identifying 9 of the 13 endpoints that displayed significant concentration-related changes.

Example of a benchmark concentration (BMC) graph for light habituation 1 following fluoxetine exposure, which depicts the response of each animal (small black dots), the average for each concentration (large blue dots), the benchmark response (dashed black line) with the cutoff range (blue shading), the best-fit model (solid black line), and BMC (µM) (dashed red line) with a confidence interval (red shading).

Bar graph showing the developmental toxicity results for ibuprofen and omeprazole at concentrations ≤ 100 µM (n = 8 for all concentrations). Observational assessment data is presented for the following percentages: normal, not hatched, abnormal, severely abnormal, and dead. The vertical axis indicates percent affected, while the horizontal axis shows concentration (µM).

Line graphs for the nine candidate negative control chemicals show the average distance moved every 2 min (vertical axis) during the light–dark assay. Time (min) is shown on the horizontal axis with the first half of the session in the light (3500 lux) and the second half in the dark (12 lux). Vehicle control (DMSO) is represented by the navy-blue line and circle, while each chemical concentration is a different shape and line color.

Bar graphs for the nine candidate negative control chemicals show the average distance moved every 2 min (vertical axis) during the light–dark assay. Concentration (≤100 µM) is shown on the horizontal axis, with the light (3500 lux) and dark (12 lux) conditions shown separately, and the average for each individual embryo represented by the colored circles.

Behavioral neurotoxicity results from a 13-endpoint analysis presented in a heatmap identifying only one endpoint (light average speed for selegiline hydrochloride) that revealed a concentration-related change based on the concentration–response modeling analysis, shaded dark blue. The endpoints that did not show any changes are shaded light blue. The 13 endpoints are listed on the vertical axis and chemicals on the horizontal axis.