Advanced Kidney Models In Vitro Using the Established Cell Line Renal Proximal Tubular Epithelial/Telomerase Reverse Transcriptase1 for Nephrotoxicity Assays

Abstract

Nephrotoxicity stands as one of the most limiting effects in the development and validation of new drugs. The kidney, among the organs evaluated in toxicity assessments, has a higher susceptibility, with nephrotoxic potential frequently evading detection until late in clinical trials. Traditional cell culture, which has been widely used for decades, does not recapitulate the structure and complexity of the native tissue, which can affect cell function, and the response to cytotoxins does not resemble what occurs in the kidney. In the current study, we aimed to address these challenges by creating in vitro kidney models that faithfully biomimic the dynamics of the renal proximal tubule, using the well-established RPTEC/TERT1 cell line. For doing so, two models were developed, one recreating tubule-like structures (2.5D model) and the other using microfluidic technology (kidney-on-a-chip). The 2.5D model allowed tubular structures to be generated in the absence of hydrogels, and the kidney-on-a-chip model allowed shear stress to be applied to the cell culture, which is a physiological stimulus in the renal tissue. After characterization of both models, different nephrotoxic compounds such as cisplatin, tacrolimus, and daunorubicin were used to study cell responses after treatment. The developed models in our study could be a valuable tool for pre-clinical nephrotoxic testing of drugs and new compounds.

Keywords: 3D structures; hydrogel; in vitro models; kidney; kidney-on-a-chip; nephrotoxicity.

Figure 1

(a) Formation of self-organized structures. RPTEC/TERT1 were seeded on top of hydrogels with different proportions of collagen I and MG by adding 5% MG to the cell suspension before cell seeding or without 5% MG. Tubule-like structures were formed mostly in the 90% Col I + 10% MG hydrogel when compared with the other conditions. (b) RPTEC/TERT1 with 5% MG seeded on top of different coatings formed self-organized structures in all conditions, even in the control without coating. Scale bar 50 µm.

Figure 2

RPTEC/TERT1 with 5% MG on collagen I + Matrigel hydrogel (90:10). Confocal microscopy images show tubular formations of RPTEC/TERT1 on hydrogel (a,b) and without hydrogel (c,d). Acetylated tubulin is located in the primary cilium. The yellow arrows point to the presence of primary cilia in both conditions. The location of the actin in the apical region of the cells defines the lumen of the tubule-like structures. ZO-1 shows the tight junction between the proximal tubule cells. Scale bar: 20 µm. (e) The expression of transporters and dedifferentiation markers was analyzed in the three models, both in the control and in the tubules with or without gel. No significant differences were found between models, but increased expression of OCT1, fibrin, vimentin, α-SMA, and AQP1 was observed in tubulogenesis models. Two-way ANOVA, n = 3.

Figure 3

(a) RPTEC/TERT1 in the kidney-on-a-chip model. After 72 h at low flow (0.003 dyne/cm²) and high flow (0.2 dyne/cm²), cells maintained the monolayer in the channels of the device. (b) In a different experiment, the presence of ZO-1 (green) and acetylated tubulin (red) was analyzed. Both in low-flow and high flow conditions, the presence of ZO-1 and acetylated tubulin was confirmed. The yellow line indicates the height at which the Z-stack was made. Scale bar: 20 µm.

Figure 4

Expression of different markers present in the kidney-on-a-chip model. (a) Significant differences were found in the expression of OAT1, AQP1, and SGLT2, increasing their expression in flow conditions when compared with control. (b) Expression of different markers related to cellular metabolism were analyzed and no significant differences were found in flow conditions when compared with control. Two-way ANOVA, * p < 0.1, ** p < 0.01, *** p < 0.001, **** p < 0.0001. n = 2.

Figure 5

(a,b) Effect of daunorubicin and CDDP treatment on the viability of tubule-like structures. RPTEC/TERT1 on hydrogel showed lower drug resistance to daunorubicin, and the opposite happened when cells were exposed to CDDP. Mean ± SD, n = 5. (c,d) EC50 values were calculated for daunorubicin and CDDP. Tubules in the 2.5D model were significantly more sensitive when they were exposed to daunorubicin when compared to control. Mean with ± SD One-way ANOVA * p < 0.05. n = 5.

Figure 6

Effect of daunorubicin and tacrolimus in the kidney-on-a-chip model. The percentage of cell viability was measured by PrestoBlue (a,e) and the percentage of surviving cells with crystal violet (b,f). Mean ± SD, n = 7. EC50 was also measured using data from PrestoBlue (c,g) and crystal violet (d,h). Significant differences were found between cells exposed to fluid flow when compared to the control in cells treated with daunorubicin and tacrolimus. Mean with ± SD. One-way ANOVA. * p < 0.05, ** p < 0.005. n = 7.