Podocytes derived from human induced pluripotent stem cells: characterization, comparison, and modeling of diabetic kidney disease.

BACKGROUND

In diabetic kidney disease, high glucose damages specialized cells called podocytes that filter blood in the glomerulus. In vitro culture of podocytes is crucial for modeling of diabetic nephropathy and genetic podocytopathies and to complement animal studies. Recently, several methods have been published to derive podocytes from human-induced pluripotent stem cells (iPSCs) by directed differentiation. However, these methods have major variations in media composition and have not been compared.

METHODS

We characterized our accelerated protocol by guiding the cells through differentiation with four different medias into MIXL1+ primitive streak cells with Activin A and CHIR for Wnt activation, intermediate mesoderm PAX8+ cells via increasing the CHIR concentration, nephron progenitors with FGF9 and Heparin for stabilization, and finally into differentiated podocytes with Activin A, BMP-7, VEGF, reduced CHIR, and retinoic acid. The podocyte morphology was characterized by scanning and transmission electron microscopy and by flow cytometry analysis for podocyte markers. To confirm cellular identity and niche localization, we performed cell recombination assays combining iPSC-podocytes with dissociated mouse embryonic kidney cells. Finally, to test iPSC-derived podocytes for the modeling of diabetic kidney disease, human podocytes were exposed to high glucose.

RESULTS

Podocyte markers were expressed at similar or higher levels for our accelerated protocol as compared to previously published protocols that require longer periods of tissue culture. We confirmed that the human podocytes derived from induced pluripotent stem cells in twelve days integrated into murine glomerular structures formed following seven days of culture of cellular recombinations. We found that the high glucose-treated human podocytes displayed actin rearrangement, increased cytotoxicity, and decreased viability.

CONCLUSIONS

We found that our accelerated 12-day method for the differentiation of podocytes from human-induced pluripotent stem cells yields podocytes with comparable marker expression to longer podocytes. We also demonstrated that podocytes created with this protocol have typical morphology by electron microscopy. The podocytes have utility for diabetes modeling as evidenced by lower viability and increased cytotoxicity when treated with high glucose. We found that multiple, diverse methods may be utilized to create iPSC-podocytes, but closely mimicking developmental cues shortened the time frame required for differentiation.