Exploration of zebrafish larvae as an alternative whole-animal model for nephrotoxicity testing

Abstract Due to an increasing demand for testing of new and existing chemicals and legal restrictions for the use of animals, there is a strong need for alternative approaches to assess systemic toxicity. Embryonic and larval zebrafish (Danio rerio) are increasingly recognized as a promising alternative whole-animal model that may be able to overcome limitations of cell-based in vitro assays and bridge the gap between high-throughput in vitro screening and low-throughput in vivo tests in animals. Despite the relatively simple anatomical structure of the zebrafish larval kidney (pronephros) – composed of only two nephrons - the pronephros shares major functions and cell types with mammalian nephrons. Glomerular filtration begins at 48 h post fertilization. The aim of the present study was to investigate if early zebrafish larvae might be a suitable model for nephrotoxicity testing. On day 3 post fertilization, larval zebrafish were treated with selected nephrotoxins (aristolochic acid, cadmium chloride, potassium bromate, ochratoxin A, gentamicin) for 48 h. Histological evaluation of zebrafish larvae exposed to model nephrotoxins revealed tubule injury as evidenced by dilated tubules with loss of the brush border, tubule cell necrosis and disorganization of the tubular epithelium. These changes were most severe after treatment with gentamicin, which also impaired pronephros function as evidenced by reduced clearance of FITC-dextran. Whole-mount in situ hybridization showing loss of cdh17 expression revealed site-specific injury to the proximal tubule segment. Analysis of genes previously identified as novel biomarkers of kidney injury in mammals showed upregulation of the kidney injury marker genes heme oxygenase 1 (hmox1), clusterin (clu), secreted phosphoprotein/osteopontin (spp1), connective tissue growth factor (ctgf) and kim-1 (havcr-1) in response to nephrotoxin treatment, although the response of individual genes varied across compounds. Consistent with the severity of lesions and impaired kidney function, the most prominent gene expression changes occurred in larvae exposed to gentamicin. Overall, our results suggest that larval zebrafish may be a suitable alternative model organism for nephrotoxicity screening, yet further improvements and integration with quantitative in vitro to in vivo extrapolation will be needed to predict human toxicity.