The development of a cellular model of nephrosis to evaluate nephrotoxic biomarkers

Abstract

Nephrotoxicity is one of the major causes for compound failure late in the drug development process. Pharmaceutical companies are interested in identifying biomarkers of nephrotoxicity which can be used to identify potential toxic compounds earlier in the development process and hence reduce the overall time and costs involved in bringing a drug to market. I developed a cellular model of nephrosis, in NRK cells, using the well characterized nephrotoxicant compound puromycin aminonuceloside (PAN). Using this cellular model I examined the expression of kidney specific genes. Two podocyte specific proteins, podoplanin and podocalyxin were found to be specifically down-regulated. Podoplanin showed an almost universal 65% reduction in the level of gene expression after PAN treatment. Podocalyxin showed a dose-dependent reduction in expression, which reached a peak of 85% reduced expression at the highest PAN dose tested. A cell aggregation assay was developed to quantify the effect of PAN induced nephrosis on the cell adhesion properties of the NRK cells. It was found that PAN nephrosis had a significant effect on the cells ability to aggregate and to remain adhesive. However cells which lost the ability to adhere were still found to be viable. Integrin a3 protein expression was found not be altered in response to PAN treatment as determined by immunofluorescence microscopy however Laminin β2 was found to form aggregates in response to PAN treatment. The actin cytoskeleton was also found to be severely disrupted as a result of PAN induced nephrosis. Based on these studies podocalyxin has been identified as a potential genetic biomarker of nephrosis, however further study of podocalyxin expression in other models of nephrosis is required before podocalyxin can be routinely used as a predictor of nephrotoxicity.