Zebrafish larvae as a model system for high-throughput, image-based screens in diabetes and insulin resistance

JSON

Emmanouilidou A, Ranefall P, Bandaru M, Allalou A, Ingelsson E, den Hoed M

ASHG 2017 Annual Meeting - The American Society of Human Genetics - (-) - [2017-10-17; online 2017-10-17]

Background Genome-wide association studies identified >100 loci that are robustly associated with the risk of insulin resistance and diabetes. The causal genes and mechanisms remain unidentified for most loci. We aim to examine if zebrafish larvae can serve as a model system for image-based genetic screens in insulin resistance and diabetes. Methods In a dietary screen, ~500 larvae with transgenically expressed fluorescent labels on β-cells and/or hepatocytes were overfed on a normal or cholesterol-supplemented diet in the presence or absence of 3% glucose in the medium from 5 to 10 days post-fertilization. In a separate effort, ~800 larvae were metabolically challenged with or without concomitant treatment with rosiglitazone or metformin. At 10dpf, larvae were soaked in a lipid staining dye, followed by optical sectioning of the pancreatic islet, liver and subcutaneous lipid stores using an automated positioning system and fluorescence microscope. ß-cell volume and subcutaneous and liver fat accumulation were quantified objectively using custom-written scripts. Quantitative outcomes were inverse normally transformed and data were analyzed using linear (beta±SE) or logistic (OR [95% CI]) regression, adjusting for time of imaging and batch. Results Continuous exposure to 3% glucose showed a positive main effect (0.38±0.17 SD) and a positive interaction with cholesterol supplementation (0.47±0.23 SD) for β-cell volume. Glucose exposure also increased the odds of subcutaneous (2.24 [1.28-3.90]) and hepatic fat accumulation( 2.44 [1.20-4.96]). Dietary cholesterol supplementation increased the odds of hepatic fat accumulation (2.15 [1.10-4.22]) but reduced the odds of subcutaneous fat accumulation (0.40 [0.23-0.71]). Treatment with rosiglitazone reduced hepatic fat accumulation by -0.82±0.10 SD and increased the odds of subcutaneous fat accumulation (2.16 [1.07-4.37]). Treatment with metformin had the opposite effect (0.33±0.09 SD; 0.38 [0.16-0.92]). Neither rosiglitazone nor metformin affected ß-cell volume (P>0.3). Conclusion Challenging zebrafish larvae metabolically for five days induces an insulin resistant state that can to some extent be prevented by treatment with rosiglitazone. Hence, zebrafish larvae are a promising model system for genetic screens in insulin resistance and diabetes. Multiplex CRISPR-Cas9 mutant models for proof-of-concept genes are currently being screened to further validate the model system.

BioImage Informatics [Collaborative]