A Kidney-Cell Assay Panel Separates Mitochondrial Stress from Later Platinum-Drug Injury
Human proximal-tubule cells showed distinct responses to cisplatin, oxaliplatin and carboplatin across five assay endpoints. Mitochondrial depolarization was generally more sensitive than overt injury, while oxidative-stress responses varied and lysosomal-enzyme release largely tracked later membrane damage.
Testing a Pathway Framework in Human Kidney Cells
Platinum chemotherapy can injure the kidney, but a single viability measurement gives limited information about the processes leading to that injury. Utrecht University researchers assembled a panel of functional assays around an adverse outcome pathway, or AOP, framework. Their NAM Journal study compares mitochondrial dysfunction, oxidative stress and downstream injury in the same human proximal-tubule cell model.
The experiments used conditionally immortalized proximal-tubule epithelial cells, abbreviated ciPTECs. These cells were expanded and then differentiated into monolayers before exposure to cisplatin, oxaliplatin or carboplatin. The study used at least three independent biological repeats with three technical replicates per condition; the model was a single immortalized cell line.
Initial concentration–response experiments measured metabolic reducing activity with PrestoBlue, a resazurin-based viability indicator. Each drug then received its own concentration range, spanning relatively mild effects through substantial cytotoxicity. This avoided treating very different nominal drug potencies as equivalent exposure conditions. Most injury measurements were collected after 24, 48 and 72 hours, with oxidative-stress measurements extending across a separate 2–48-hour time course.
The Drugs Differ in Potency and Stress Pattern
All three agents reduced metabolic viability in a concentration- and time-dependent manner. At 72 hours, the estimated half-maximal inhibitory concentrations were 10.5 μM for cisplatin, 45.5 μM for oxaliplatin and 97.0 μM for carboplatin. Their reported 95% confidence intervals were 9.8–11.1, 40.0–51.7 and 89.9–104.6 μM, respectively. Cisplatin was the most potent agent in this assay, while carboplatin required higher concentrations and longer exposure.
Mitochondrial membrane potential was assessed with JC-10, a fluorescent indicator sensitive to depolarization. This endpoint generally changed at lower concentrations than membrane leakage or loss of viability. Cisplatin produced the strongest and most rapid effects; oxaliplatin produced an early response that did not progress in the same way, while carboplatin’s effects were weaker or delayed.
Reactive oxygen species did not follow a uniform pattern across the drug class. Cisplatin produced strong and sustained increases. Oxaliplatin caused a more modest, transient response at higher concentrations, and carboplatin produced little or inconsistent oxidative-stress signal under the tested conditions. The weak carboplatin result matters because the panel was intended to distinguish mechanisms and response patterns, rather than assuming that every platinum drug activates every endpoint equally.
Membrane injury was measured through lactate dehydrogenase release. All three compounds showed concentration effects, but an additional time-dependent effect was apparent only for carboplatin. This readout complemented the metabolic assay by measuring leakage rather than the cells’ reducing activity.
A Lysosomal Enzyme Is an Injury Marker, Not a Complete Mechanism
The fifth endpoint was extracellular β-N-acetylglucosaminidase activity, abbreviated NAG. This lysosomal enzyme is abundant in proximal-tubule cells. Its measured release increased with injury, with cisplatin producing a rapid, sustained response and carboplatin requiring greater concentrations or longer exposure.
However, NAG activity largely paralleled lactate dehydrogenase release at more damaging exposures. The authors therefore interpret it as a supportive tubular-injury measurement that may reflect secondary enzyme leakage. They did not include a separate lysosome-specific positive control or directly measure lysosomal membrane destabilization. The NAG result cannot by itself identify an independent early lysosomal step.
From Ordered Measurements to Quantitative Causal Links
The panel produced a consistent sensitivity pattern: mitochondrial perturbation generally appeared at lower concentrations than overt membrane damage and viability loss, with drug-dependent oxidative responses between those measurements. That pattern is compatible with an AOP-based injury sequence. The experiment was not designed to establish a time-resolved causal chain, and it did not fit quantitative relationships showing that one measured event predicts the next.
Several upstream and kidney-specific processes also remain outside the panel. Platinum–DNA adducts and genomic DNA damage were not measured, so the experiments cannot apportion their contribution relative to mitochondrial injury. Transporter function, reabsorption, immune interactions and specific modes of cell death were also unmeasured. The static monolayer and exposures of up to 72 hours leave repeated treatment cycles and tubular flow outside the tested setting.
The study supplies concentration–response profiles for mechanistic screening within one drug class. Extending that use requires broader chemical testing, repeated-exposure experiments and pharmacokinetic translation of culture concentrations. Its present value lies in distinguishing stress and injury measurements—and their drug-specific gaps—before using them to build a quantitative kidney-toxicity prediction.