Bioengineered 3D human kidney tissue, a platform for the determination of nephrotoxicity

Abstract

The staggering cost of bringing a drug to market coupled with the extremely high failure rate of prospective compounds in early phase clinical trials due to unexpected human toxicity makes it imperative that more relevant human models be developed to better predict drug toxicity. Drug-induced nephrotoxicity remains especially difficult to predict in both pre-clinical and clinical settings and is often undetected until patient hospitalization. Current pre-clinical methods of determining renal toxicity include 2D cell cultures and animal models, both of which are incapable of fully recapitulating the in vivo human response to drugs, contributing to the high failure rate upon clinical trials. We have bioengineered a 3D kidney tissue model using immortalized human renal cortical epithelial cells with kidney functions similar to that found in vivo. These 3D tissues were compared to 2D cells in terms of both acute (3 days) and chronic (2 weeks) toxicity induced by Cisplatin, Gentamicin, and Doxorubicin using both traditional LDH secretion and the pre-clinical biomarkers Kim-1 and NGAL as assessments of toxicity. The 3D tissues were more sensitive to drug-induced toxicity and, unlike the 2D cells, were capable of being used to monitor chronic toxicity due to repeat dosing. The inclusion of this tissue model in drug testing prior to the initiation of phase I clinical trials would allow for better prediction of the nephrotoxic effects of new drugs.

Figure 1. Human renal proximal tubule cells maintain epithelial markers and function with immortalization and 2D cell culture.

(A) Expression of epithelial cell markers, E-cadherin (red) and cytokeratin 8/18/19 (green), the proximal tubule cell marker GGT1 (green), and the organic anion transporter OAT4 (red) in NKi-2 cells. Nuclei are stained with dapi (blue). Scale bars  =  100 µm. (B) Adenylate cyclase activity in NKi-2 cells upon treatment with ADH or PTH. Treatment with forskolin acted as a control for production of cAMP. *p<0.05, ***p<0.0001, n  =  8. (C) GGT and LAP hydrolase activity in NKi-2 cells. Fold increase is calculated against human fibroblast activity. n  =  48. (D) Na⁺ dependent glucose uptake in NKi-2 cells. Cells were exposed to 2-DG with or without Na⁺ in the form of NaCl. *p<0.05, n  =  5.

Figure 2. NKi-2 cells formed tubular structures when grown in 3D tissues.

(A) Schematic of 3D tissue formation. An acellular layer of 1:1 Matrigel:rat tail collagen I (1 mg/mL) is layered onto a tran-well membrane. Following polymerization, a layer of the same ECM mixture containing NKi-2 cells is added. The tissues are maintained in growth media and after approximately 2 weeks, the cells organize into branching structures. (B) Morphology of NKi-2 cells after extended growth in 3D tissues. Tissue sections, 10 µm, were stained with H&E and whole tissue sections were stained with carmine for whole mounts. Arrows indicate areas of branching tubular-like structures. Scale bars  =  100 µm. (C) Expression of epithelial cell markers, e-cadherin (red) and cytokeratin 8/18/19 (green), kidney proximal tubule cell marker GGT1 (green) and organic anion transporters OAT1 (green) and OAT4 (red) within the 3D tissues after 4 weeks of growth. Scale bars  =  25 µm.

Figure 3. NKi-2 cells maintain kidney function when grown in 3D tissues.

(A) Adenylate cyclase activity in NKi-2 cells within the bioengineered 3D tissue upon treatment with ADH or PTH. Treatment with forskolin acted as a control for production of cAMP. n  =  4. (B) Na⁺ dependent glucose uptake in NKi-2 cells within the bioengineered 3D tissues. Tissues were exposed to 2-DG with or without Na⁺ in the form of NaCl. *p<0.05, n  =  5. (C) RT-PCR of drug tranporters found in kidney epithelial cells using RNA from both NKi-2 cells and 4 week 3D tissues. (C) RT-PCR of different transport proteins found in human kidney cells in both NKi-2 cells (2D) and 4 week old 3D tissues.

Figure 4. Cytotoxicity assay development for bioengineered 3D tissues. (A)

DNA content of bioengineered kidney tissues over time with treatment of a range of cisplatin concentrations. Tissues were grown for 2 weeks and treated with a 5 concentrations of cisplatin from 0.01 µM to 100 µM. A subset of tissues was lysed at 0, 3, 7, 10, and 14 days, for each treatment and DNA content was measured using Picogreen fluorescence. n  =  4. (B) LDH secretion over time with treatment of various concentrations of cisplatin. Tissues were grown for 2 weeks and treated with 5 concentrations of cisplatin from 0.01 µM to 100 µM. Supernatant was taken at 0, 3, 7, 10, and 14 days and assayed for LDH activity. n  =  4. (C) NKi-2 cell and 3D tissue secretion of LDH over time. NKi-2 cells were seeded at a density of 100,000 cells per well of a 6-well plate and the media supernatant was assayed for LDH secretion 2, 5, 9, 12, and 16 days after seeding. 3D tissues were grown for 2 weeks and the supernatant was assayed for LDH secretion 14, 17, 21, 24, and 28 days after formation of the tissues. ***p<0.001, n  =  4.

Figure 5. Drug treatment range and strategy.

(A) NKi-2 cells were subjected to a range of drug concentrations for 48 hours and cell proliferation was determined by reduction of MTT. n  =  6. (B) Schematic of the process of drug treatment followed for both the NKi-2 cells and the 3D tissues.

Figure 6. Morphology of 3D tissues after treatment with cisplatin, gentamicin, and doxorubicin.

Tissues were fixed after 2 weeks of drug treatment. (A) 8 µm sections were stained with hematoxylin and eosin to visualize changes in tissue structure. (B) Whole fixed tissues were stained with carmine for whole mount visualization of tissue structure. Low dose refers to the lowest does tested and high dose refers to the highest dose tested. Scale bars  =  50 µm.

Figure 7. LDH measured cytotoxicity in NKi-2 cells and tissues upon treatment with cisplatin, gentamicin, or doxorubicin.

LDH secretion was measured in the supernatant of both NKi-2 2D cell cultures and 3D tissues at 0, 3, 7, 10, and 14 days following treatment with (A) 0.01 µM to 100 µM cisplatin, (B) 0.2 mM to 21.6 mM gentamicin, or (C) 0.001 µM to 20 µM doxorubicin. n  =  7.

Figure 8. Kim-1 and NGAL secretion in untreated NKi-2 cells and 3D tissues.

Untreated NKi-2 cells were grown for 24 hours followed by media sampling at 0, 3, 7, 10, and 14 days. The media was assayed for the presence of (A) Kim-1 or (B) NGAL by ELISA. NKi-2 3D tissues were grown for 2 weeks followed by media sampling at 0, 3, 7, 10, and 14 days and assayed for the presence of (A) Kim-1 or (B) NGAL by ELISA. **p<0.01, ***p<0.001, n  =  7.

Figure 9. Kim-1 measured cytotoxicity in NKi-2 cells and tissues upon treatment with cisplatin, gentamicin, or doxorubicin.

Kim-1 secretion was measured by ELISA in the supernatant of both NKi-2 2D cell cultures and 3D tissues at 0, 3, 7, 10, and 14 days following treatment with (A) 0.01 µM to 1 µM cisplatin, (B) 0.2 mM to 2.2 mM gentamicin, or (C) 0.001 µM to 0.2 µM doxorubicin. n  =  7.

Figure 10. NGAL measured cytotoxicity in NKi-2 cells and tissues upon treatment with cisplatin, gentamicin, or doxorubicin.

NGAL secretion was measured by ELISA in the supernatant of both NKi-2 2D cell cultures and 3D tissues at 0, 3, 7, 10, and 14 days following treatment with (A) 0.01 µM to 1 µM cisplatin, (B) 0.2 mM to 2.2 mM gentamicin, or (C) 0.001 µM to 0.2 µM doxorubicin. n  =  7.