F to 120 hpf (before the feeding stage), several lethal or sublethal endpoints based on the DarT protocol [3], including survival rates, hatching rate, edema, tail detachment, somite formation, spontaneous movement, heart beat, pigmentation and touch response were observed and recorded as indicators for chemical toxicity.Imaging and data analysisGFP fluorescence was observed under a fluorescent microscope (ZEISS Axiovert 200M) with a GFP filter and photographed with a digital camera (ZEISS AxiocCam HRC). For direct comparison in the same set of experiment, images were taken for the same exposure time at a fixed aperture. At least 5 embryos/larvae wereTransgenic Zebrafish for Neurotoxin TestFigure 1. Survival rates of Tg(nkx2.2a:mEGFP) fry in the presence of different concentrations of testing chemicals. Survival rates at 8, 24, 
48 and 96 hpf were plotted against different concentrations of the chemicals. Chemical names are indicated above each panel. doi:10.1371/journal.pone.0055474.gT-test was used; significant difference (P = 0.01?.05) and highly significant difference (P,0.01) from the control groups are indicated in Figure 2 and Table S1. Highly significant differences for all or most of these endpoint measurements were observed only from high dosage groups of five of these chemicals and their starting concentrations were: 10 mg/L acetaminophen, 4 mg/L atrazine, 0.5 ethanol, 5 mg/L Lindane and 10 mg/L mefenamic acid. For atenolol, most endpoints did not show significant difference but hatching 1516647 edema.Transgenic Zebrafish for Neurotoxin TestFigure 2. Summary of selected DarT endpoint measurements. (A) Hatching (96 hpf), (B) Heartbeat (48 hpf), (C, D) Tail detachment (24 hpf, 48 hpf), (E, F) Normal somite (24 hpf, 48 hpf). Names and concentrations of chemicals are indicated at the bottom of Panel F. 0.01 DMSO was used as control except that egg water was used as control for ethanol test. Hearbeat is shown as numbers per 15-second. Statistical significance: **P,0.01; *P,0.05. doi:10.1371/journal.pone.0055474.gAxon length provides a more sensitive and measurable marker 
for evaluation of neurotoxixityIn order to demonstrate that GFP fluorescence may provide more sensitive markers for phenotypical changes induced by thesechemicals, GFP fluorescence was observed and photographed for each treatment group. As reported previously [16,17], GFP fluorescence was observed in the developing neural tube and brain from 1 dpf. By 3 dpf, obvious GFP-labeled axons were observedTransgenic Zebrafish for Neurotoxin TestFigure 3. E.F to 120 hpf (before the feeding stage), several lethal or sublethal endpoints based on the DarT protocol [3], including survival rates, hatching rate, edema, tail detachment, somite formation, spontaneous movement, heart beat, pigmentation and touch response were observed and recorded as indicators for chemical toxicity.Imaging and data analysisGFP fluorescence was observed under a fluorescent microscope (ZEISS Axiovert 200M) with a GFP filter and photographed with a digital camera (ZEISS AxiocCam HRC). For direct comparison in the same set of experiment, images were taken for the same exposure time at a fixed aperture. At least 5 embryos/larvae wereTransgenic Zebrafish for Neurotoxin TestFigure 1. Survival rates of Tg(nkx2.2a:mEGFP) fry in the presence of different concentrations of testing chemicals. Survival rates at 8, 24, 48 and 96 hpf were plotted against different concentrations of the chemicals. Chemical names are indicated above each panel. doi:10.1371/journal.pone.0055474.gT-test was used; significant difference (P = 0.01?.05) and highly significant difference (P,0.01) from the control groups are indicated in Figure 2 and Table S1. Highly significant differences for all or most of these endpoint measurements were observed only from high dosage groups of five of these chemicals and their starting concentrations were: 10 mg/L acetaminophen, 4 mg/L atrazine, 0.5 ethanol, 5 mg/L Lindane and 10 mg/L mefenamic acid. For atenolol, most endpoints did not show significant difference but hatching 1662274 and edema appeared to be quite sensitive indicators with the highly significant difference (p,0.01) at the concentration of 5 and 7.5 mg/L respectively while most other traits did not show highly significant difference even at the highest dosage (10 mg/L) used (Table S1).In addition, we also observed some specific effects for these tested chemicals. Atrazine had a dosage-dependent increase of heartbeat rate (but with a smaller magnitude of heart contraction) while all other five chemicals caused a dosage-dependent decrease of heartbeat (Figure 2B and Table S1). High dose ethanol led to, obvious edema with shorter body length in a high percentage of treated fry. High dose lindane generally resulted in coiled body and shorter body length; when these treated were touched, they had spiral-locally swimming pattern. For mefenamic acid, high dose groups of fry had light or no pigmentation (Figure 3F), in addition to high percentage of 1516647 edema.Transgenic Zebrafish for Neurotoxin TestFigure 2. Summary of selected DarT endpoint measurements. (A) Hatching (96 hpf), (B) Heartbeat (48 hpf), (C, D) Tail detachment (24 hpf, 48 hpf), (E, F) Normal somite (24 hpf, 48 hpf). Names and concentrations of chemicals are indicated at the bottom of Panel F. 0.01 DMSO was used as control except that egg water was used as control for ethanol test. Hearbeat is shown as numbers per 15-second. Statistical significance: **P,0.01; *P,0.05. doi:10.1371/journal.pone.0055474.gAxon length provides a more sensitive and measurable marker for evaluation of neurotoxixityIn order to demonstrate that GFP fluorescence may provide more sensitive markers for phenotypical changes induced by thesechemicals, GFP fluorescence was observed and photographed for each treatment group. As reported previously [16,17], GFP fluorescence was observed in the developing neural tube and brain from 1 dpf. By 3 dpf, obvious GFP-labeled axons were observedTransgenic Zebrafish for Neurotoxin TestFigure 3. E.