Human iPSCs were obtained from CAS key laboratory of regenerative biology. The cells were cultured on growth factor-reduced Matrigel (354277, Corning Life Sciences, Kennebunk, ME, United States)-coated 6-well plates (0.013 mg/cm²) in mTeSR medium (85875, Stemcell Technologies, Vancouver, BC, Canada) and cells were subcultured every 3–4 days at a ratio of 1:6.

The culture of iPSC-derived kidney organoids in this study was initiated according to the method proposed by Przepiorski et al. (2018). Before differentiation, iPSCs were maintained on a 10-cm Matrigel-coated cell culture dish to approximately 75% confluence. On day 0, embryoid bodies were generated by detaching colonies into single cells with1 mg/mL dispase (17105041, GIBCO, Grand Island, NY, United States) and then cells were collected at 200 g for 5 min by centrifugation at room temperature. The cells were resuspended in BPEL (supplemented with 8 mM CHIR99021 (S2924, Selleck Chemicals, Houston, TX, United States), 3.3 mM Y27632 (S1049, Selleck), and 1 mM β-mercaptoethanol (M3148, Sigma-Aldrich, St. Louis, MO, United States) and seeded into a 6-well ultra-low attachment plate (3,471, Corning). After 48 h, half of the medium was replaced with BPEL and CHIR99021 (8 mM). From day 3 onwards, embryoid bodies were resuspended in Stage II culture medium until day 14 and agitated daily to prevent excessive fusion. BPEL and Stage II medium (10–100–455, Aimingmed, Hangzhou, Zhejiang, China) were prepared as described by Przepiorski et al. (2018).

Cisplatin (p4394, Sigma) and celastrol (C0869-10 MG, Sigma) were used for the nephrotoxicity testing. Cisplatin was reconstituted with a 0.9% sodium chloride solution to achieve a stock concentration of 0.5 mg/mL, and celastrol was dissolved in DMSO to reach a stock concentration of 10 mg/mL. The organoids were exposed to 50 μM cisplatin, as well as 100 nM, 200 nM, 500 nM, 1 μM, 2 μM, 5 μM, 10 μM, or 50 μM celastrol for a duration of 2 days starting from day 11 in vitro. In order to assess the kidney protective effect of celastrol, either 1 μM or 2 μM celastrol was co-administered with the cisplatin at a concentration of 50 μM. Organoids were harvested for immunostaining after fixation with 4% paraformaldehyde (P0099, Beyotime, Shanghai, China), or for qPCR and RNA-seq using TRIzol (15596018, Invitrogen, Carlsbad, CA, United States).

Total RNAs was extracted using TRIzol, and 1 μg of total RNA was used for cDNA synthesis using the GoScript reverse transcription mix (A2790, Promega, Madison, WI, United States). Real-time PCR was performed using the SsoAdvanced SYBR Green Supermix (1725274, Bio-Rad, Hercules, CA, United States) in an real-time PCR machine (StepOnePlus, Thermo Fisher Scientific, Waltham, MA, United States). For the miRNAs, 2.5 μg of total RNA was reverse transcribed using All-in-One MiRNA Q-PCR Detection Kit (GeneCopoeia, Rockville, MD, United States). The primers used for the reaction are listed in Table 1.

To characterize the differential expression of RNA transcripts between the control, cisplatin-induced nephrotoxicity, and celastrol kidney protection groups, whole genome transcript sequencing was performed by (Aimingmed). Up-sequencing using the library construction method with rRNA removal allowed for simultaneous detection of mRNA expression levels. Gene expression was analyzed to assess the correlation between gene expression characteristics and differentially expressed genes within and between groups. Then, using pheatmap package and the hierarchical clustering was performed. The results were visualized using a heatmap. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment and Gene Ontology (GO) enrichment analyses were performed.

The organoids were washed with PBS and fixed with 4% paraformaldehyde (PFA) for 1 h at room temperature (RT). Afterwards, they were incubated in sucrose (30% w/v) in PBS overnight at 4°C. Then, the organoids were embedded in optimal cutting temperature (OCT) compound (4,583, Tissue-Tek, Torrance, CA, United States) and cryosectioned into 10-µm sections. The primary antibodies and dilutions used for immunofluorescence analysis are listed in Table 2. After washing thrice with PBS, the sections were incubated with the corresponding secondary antibodies (Table 2) and diluted in PBS for 1 h at RT. The nuclei were counterstained with DAPI for 10 min. Images were obtained using a confocal microscope (LSM710, Zeiss, Oberkochen, Germany).

We used the GraphPad Prism (Version 7; GraphPad Software La Jolla, CA, United States) to conduct statistical analysis. Data were expressed as the mean ± SD. For comparison of more than three groups, Shapiro-Wilk test was used to check if a continuous variable follows a normal distribution. If P > 0.05, one-way ANOVA was applied; If p < 0.05, Kriskall-Wallis test was applied. Results were considered statistically significant with p values: ***p < 0.001, **p < 0.01.

The organoids were washed with cold PBS and dissociated into single cells using Accumax (07921, STEMCELL). The cells were stained with propidium iodide (PI; Invitrogen) in PBS for 15 min at 4°C. The cells were washed with PBS and strained through a 100 μm mesh. Flow cytometry measurements were performed using a BD FACSCalesta cytometer (BD Diagnostics, Franklin Lakes, NJ, United States).

The organoids were collected in a 2 mL tube and fixed with an electron microscopy fixation buffer consisting of glutaraldehyde (2.5%), paraformaldehyde (2%), and phosphate buffer (PB, 0.1 M, pH 7.4) overnight at 4°C. Post-fixation was incubated with 1% OsO4 in PB (0.1 M) for 1 h at 4°C. Kidney organoids were fixed in 1% Dehydrated in a graded series of ethanol solutions, and embedded in epoxy resin. Ultrathin sections (70 nm) were cut and stained with uranyl acetate and lead citrate. Slides were then imaged using an G2 Spirit transmission electron microscope (FEI Tecnai, Hillsboro, OR, United States).

The process of kidney organoids self-organization is illustrated in Figure 1A. At 14 days of culture, the 3D structures presented tubular (red arrows) and glomerular (green arrows) formation (Figure 1B). These constructs expressed NPHS1, a critical protein primarily found in podocytes of the glomerulus, and CD31, a marker highly expressed in endothelial cells, indicating the development and/or maturation of the kidney vasculature (Figure 1C). Additionally, the organoids presented LTL (Green), a highly specific marker of the kidney’s proximal tubules, and SLC12A1 (Red) staining, a marker for identifying the presence of thick ascending limb segments of the kidney (Figure 1D). Organoids contained multiple segmented nephron structures that resemble glomeruli and tubules marked by PODXL + podocytes (Figure 1E), HNF1B + tubules/collecting duct and CDH1+ distal tubules (Figure 1F), and MEIS1/2/3+ interstitial cells (Figure 1G). Typical images demonstrating the morphology of the kidney organoids exposed to celastrol or cisplatin under various conditions are shown in (Figure 1H). After the exposure of the kidney organoids to cisplatin, the organoids structure appeared partly collapsed. In addition, the collapsed structures were significantly reduced at celastrol was added.

We used 3D kidney organoid models to investigate the effects of cisplatin and celastrol on drug-induced kidney toxicity. The constructs were exposed to varying concentrations of the compounds to assess their effects on cell viability. Figure 2A shows a significant increase in cell death within organoids treated with 50 µM cisplatin, as evidenced by the increase in PI fluorescence, indicative of membrane damage and cell death. Moreover, the nephrotoxicity of celastrol at 10 µM (Figure 2B), 1 µM (Figure 2C) and 100 nM (Figure 2D) was significantly less pronounced than that observed in constructs treated with cisplatin-50 µM. Furthermore, the number of PI-stained cells in the 100 nM cL group (Figure 2D) was unnoticeable, similar to that found in Control (Figure 2E), and DMSO groups (Figure 2F).

Kidney organoids were exposed to varying concentrations of celastrol (0.1, 0.2, 0.5, 1, 2, 5, and 10 µM) to assess dose-related toxic effects (Figure 3). The 3D constructs were stained with propidium iodide (PI) to determine cell viability. At 10 µM celastrol, the proportion of PI-positive cells reached 33.8%, compared to 7.7% in the control group, while the number of PI-stained cells in the groups treated with Celastrol >1uM remained comparable to that of the basal levels (Ctrl). The Flow cytometer assay findings demonstrated that celastrol induced cytotoxicity in a dose-dependent manner, as evidenced by the increased number of dead cells, with concentrations to produce low cytotoxicity and yet retain the beneficial effects at an optimal range of 1–2 µM.

The dose-dependent response of celastrol-induced cytotoxicity was compared with that of cisplatin treatment. Flow cytometry analysis revealed that cisplatin at 50 µM resulted in only 19.7% of cells maintaining a normal structure, in stark contrast to the 55.8% observed in the control (Ctrl) and DMSO groups. Interestingly, Celastrol at 0.2 µM and 1 µM concentrations demonstrated normal structure cell percentages comparable to those in the DMSO and control groups (56.3%). However, at higher concentrations of 5, 10, and 50 μM, celastrol induced a dose-dependent decrease in the percentage of cells with normal structure, underscoring its nephrotoxicity at elevated doses (Figure 4).

To assess the impact of renal injury induced by individual treatments with cisplatin or celastrol, as well as their combined application, we adopted a holistic measurement strategy that encompasses the simultaneous quantification of several key renal mRNA markers. Specifically, we have focused on the mRNA levels of Interleukin-8 (IL-8), Interleukin-1 beta (IL-1β), Kidney Injury Molecule-1 (KIM-1), and Monocyte Chemoattractant Protein-1 (MCP-1), as these biomarkers are indicative of renal stress, inflammation, and injury (Figure 5). IL-1β and IL-8 are central to the kidney’s inflammatory response, with elevated mRNA levels indicating activation of inflammatory pathways typical of nephrotoxic agents, such as cisplatin. Remarkably, the mRNA levels of IL-1β, and IL-8 in the groups treated with a combination of 50 µM cisplatin and 1 (1+cp), and particularly 2 (2+cp) µM celastrol, were similar to those observed in the control groups (Ctrl, DMSO). Moreover, the mRNA expression levels in the 1+cp and 2+cp groups were significantly lower than those seen in the 10uM Celastrol or Cisplatin 50uM the transcriptional expression of KIM-1, a specific marker for kidney proximal tubule injury, showed similar expression patterns and were downregulated to basal levels when organoids were subjected to 50 µM cisplatin combined with 1 and 2 µM celastrol. However, MCP-1, known to recruit monocytes and macrophages to inflammation sites that can exacerbate tissue damage, also presented similar expression patterns to those observed in the previously described biomarkers. However, 1+cp did not bring MCP-1 mRNA levels to control levels. Taken together, by measuring these biomarkers, we can gain insights into the mechanisms of celastrol-induced nephrotoxicity, predict the extent of kidney damage, and potentially guide dosage adjustments to mitigate adverse effects, thus providing a holistic view of the nephrotoxic potential of the compound.

Figure 6A shows representative images of kidney injury markers (KIM-1 is a specific and sensitive biomarker of proximal tubules of kidney injury, γH2AX indicates DNA damage) in LTL-stained kidney organoids following the administration of celastrol at concentrations of 1 μM and 2μM, cisplatin at 50μM, or a combination of both drugs, for a duration of 2 days. The fluorescence intensity was showed in (Figure 6B), determined using the ImageJ software, revealed the cytotoxic effects of cisplatin at 50μM, as indicated by the elevated expression of both renal damage markers (Supplementary Figure 1). Notably, the co-administration of cisplatin 50 μM with 1μM and 2 μM celastrol not only mitigated drug-induced kidney injury, but also demonstrated protective actions. These findings suggest that 1 and 2 µM celastrol could potentially offer a novel therapeutic strategy that balances the potent anticancer action of cisplatin with the cytoprotective benefits of celastrol, thereby enhancing the overall safety and efficacy of cancer treatment.

In this study, human pluripotent stem cells derived kidney organoids comprised of not only complex interacting component cell types, but also some distinct segmenting nephrons, including distal tubule, proximal tubule, foot processes and podocytes of the glomerulus (Figure 1). With such various of kidney cell types, the advantages for use of these nephrons for nephrotoxicity screening have significantly promoted (Figure 7). Cisplatin induced-kidney injury was characterized by TEM, in cisplatin group, clear damages in tubular cells were identified, such as loss of cells shape, necrosis of renal tubular epithelium and vacuoles formation. 1 μM celastrol + cisplatin group resulted the significant injury recovery in tubular cells comparing with others.

For the RNA-seq result, compared 2 μM celastrol + cisplatin vs. Cisplatin group, GO analysis shown the most differentially expressed functional pathways are the renal system, renal tubular secretion and excretion function based on the GO database, indicating celastrol regulate renal tubular secretion and excretion function with protective effect on glomerulus. Meanwhile, Figure 8 shows the significant upregulation of SNORD3A in kidney organoids treated with cisplatin (Cis), with high expression levels correlating with nephrotoxicity. SNORD3A was primarily enriched in tubular epithelial cells in response to acute kidney injury (AKI) in tubular epithelial cells. We also observed a dramatic decline in SNORD3A expression in the celastrol + cisplatin group (2Cel-Cis) (Figure 8C), indicating that celastrol played a pivotal role in the inhibition of cisplatin-induced nephrotoxicity. Previous studies have demonstrated that miRNAs are involved in the pathophysiology of AKI, and aberrant miRNA expression levels serve as biomarkers for diagnosing AKI miRNA signatures. Our RNA-seq and RT-qPCR results showed that miR-3615 was positively associated with cisplatin-induced kidney injury. Moreover, miR-3615 was significantly downregulated in the 2Cel-Cis, further indicating its positive relationship with cisplatin-induced nephrotoxicity (Figure 8C). Furthermore, RPPH1, which is upregulated in diabetic nephropathy via an interaction with Gal-3, and the HIST1H3A gene, both presented expression patterns consistent with what we’ve observed for the previous analyzed genes, also indicating a possible relationship with the protective actions of Cis/Cel cotreatment (Figure 8C).