The screening workflow and automated image analysis to detect zebrafish embryos, and count haematopoetic stem cells in the tails. (A) Schematic of screening platform workflow including plate preparation, image acquisition and image analysis of zebrafish embryos with approximate timings. (B,C) Brightfield image (B) and fluorescent image (C) of a Tg(itga2b:GFP) zebrafish at 3 dpf acquired from the Hermes showing some of the regions identified by Athena: tail (red), trunk (dark blue), head (white), eye (pink), yolk (yellow), tail fin (light blue) and fluorescent granules (green). (D) Cartoon showing the location of haematopoietic stem and progenitor cells (HSPCs) in the caudal haematopoietic tissue (CHT) in the tail at 3 dpf. (E) Correlation of HSPC counts (between manual and granule) in 93 individual images of Tg(itga2b:GFP) embryos at 3 dpf using the Athena Zebrafish Application, analysed using simple linear regression with r=0.844. Scale bars: 500 µm. n refers to the number of embryos analysed.

Validation of HSPC counts using the Athena Zebrafish Application to detect differences in stem cell populations due to age or genetics of zebrafish embryos. (A) Brightfield and fluorescent images of Tg(itga2b:GFP) embryos at 2-4 dpf and analysed for HSPC count (green) in the tail region (red). (B) HSPC counts at 2 dpf (n=48), 3 dpf (n=54) and 4 dpf (n=47), showing the increase in HSPC number over time. (C) Fluorescent images of the tail of rps14^+/+ and rps14^+/− Tg(itga2b:GFP) at 3 dpf following phz haemolytic stress for 24 h at 24 hpf alongside unstressed controls analysed for HSPC count (green) in the tail region (red). (D) HSPC count of the different conditions, showing no difference in HSPC count between unstressed rps14^+/− (n=27) and unstressed rps14^+/+ (n=22) embryos, but an increase in HSPC count in the phz-stressed rps14^+/+ embryos (n=29) compared with wild type, which does not occur in the phz-stressed rps14^+/− embryos (n=29), from triplicate experiments (Peña et al., 2020 preprint). Statistical analysis using unpaired t-tests. Error bars show mean±s.d. ns, P>0.05; **P<0.01. Scale bars: 500 µm. n refers to the number of embryos analysed for each condition. Cell counts give the Athena cell count for the example image shown.

Automated analysis of the effects of X-ray irradiation on HSPC count and myeloid count in dual fluorophore images, as well as apoptosis in zebrafish embryos. (A) Fluorescent images of the tail of Tg(itga2b:GFP) embryos at 3 dpf, irradiated at 2 dpf with 0 Gy, 40 Gy or 100 Gy X-ray and analysed for HSPC count (green) in the tail region (red). (B) HSPC counts showing the reduction in HSPC number with 40 Gy (n=46) and 100 Gy (n=48) compared with 0 Gy (n=48). (C) Fluorescent images of the tail of Tg(itga2b:GFP)(lyzC:mcherry) dual fluorophore embryos at 3 dpf, irradiated at 2 dpf with 0 Gy or 40 Gy X-ray and analysed for HSPC (green) and myeloid (red) in the tail (white). (D,E) HSPC count of GFP-positive cells (D) and myeloid cell count of mCherry-positive cells (E) in the tail, showing the decrease in both cell types with irradiation (n=43) compared with unirradiated (n=42). (F) Fluorescent images of Acridine Orange-stained embryos at 3 dpf, irradiated at 2 dpf with 0 Gy or 40 Gy X-ray and analysed for granule count (green) in the whole fish. (G) Granule counts in the whole fish showing an increase in apoptosis following irradiation (n=40) compared with unirradiated (n=42). Irradiation experiments were performed in triplicate. Statistical analysis using unpaired t-tests. Error bars show mean±s.d. ***P<0.001, ****P<0.0001. Scale bars: 500 µm. n refers to the number of embryos analysed for each condition. Cell counts give the Athena cell count for the example image shown, for each cell type in dual fluorophore images.

Extended applications of the analysis platform including automated analysis of hair cell loss, angiogenesis and eye size in zebrafish embryos. (A) Fluorescent images of YO-PRO-1-stained embryos at 4 dpf, following treatment for 1 h with DMSO, 100 µM pentamidine isethionate (PI) or 100 µM propantheline bromide (PB), analysed for fluorescent granules (red) in the whole fish. (B,C) Total fluorescence intensity in the whole fish (B) and average area of the fluorescent granules (C), showing decrease with drug treatment as previously described (Chiu et al., 2008). (D) Fluorescent images of Tg(kdrl:mCherry) embryos at 2 dpf, treated for 24 h from 24 hpf with DMSO, AG1478 or SU 4312 and analysed for mCherry fluorescence in the whole fish. (E,F) Total fluorescent area in the whole fish following treatment with AG1478 (E) and total fluorescent area in the whole fish following treatment with SU4312 (F), showing reduced angiogenesis in treated fish compared with the control, as previously described (Tran et al., 2007). (G) Brightfield images of mab21l2^+/+, mab21l2^+/− and mab21l2^−/− mutant embryos at 4 dpf analysed for eye size (pink). (H) Eye area, showing the reduced eye size in mab21l2^−/− mutants. Results consistent with previous publications were achieved in a single experiment for each of these assays. Statistical analysis using unpaired t-tests. Error bars show mean±s.d. ns, P>0.05; *P<0.05, ***P<0.001, ****P<0.0001. Scale bars: 500 µm. n refers to the number of embryos analysed for each condition.