Eight-week-old female BALB/c mice were purchased from the Experimental Animal Center of Anhui (Anhui, China). The TCE transcutaneous sensitization mouse model mimicking the state of OMDT patients was established as we previously described (11). All mice were divided into the blank control group (Blank), vehicle control group (Vehicle), TCE treatment group (TCE), and TCE+CD59 cotreatment group (CD59) 24 hr before treatment, and the dorsal hair was shaved to expose the dorsal skin. Details of the TCE-sensitized mouse model are shown in Figure 1 . All animal experiments were conducted in accordance with the “Administrative Regulations on Laboratory Animals” issued by the Ministry of Science and Technology of the People’s Republic of China and approved by Anhui Medical University (Animal Ethics Committee No.: 20210897).

Mouse renal tubular epithelial cells were isolated and cultured as we previously described (13). Briefly, fresh renal cortex was diced into pieces of approximately 1 mm³ after the mice were sacrificed. Tissue fragments were digested in 5 ml of DMEN/F12 serum-free medium containing 1 mg/ml collagenase and 1% penicillin-streptomycin solution at 37°C for 30 min. After digesting the tissue fragments, the cell suspension was poured through a nylon sieve with 70 μm pore. The suspension was centrifuged at 1000 RPM for 8 min, and the supernatant was discarded. Then, the cell clump was washed with 4 mL PBS and cultured in DMEM/F-12 medium supplemented with 10% mouse’s own serum at 37°C and 5% CO₂.

Human renal proximal tubular epithelial cells (HK-2 cells) purchased from Fuheng Biotechnology (Shanghai, China) were cultured in DMEM/F12 medium (Gibco, Cat. 11320033) supplemented with 10% fetal bovine serum (FBS, Biological Industries, Cat. 04-001-1A) and 1% penicillin-streptomycin solution (Beyotime, Cat. C0222) at 37°C in a humidified atmosphere containing 5% CO2. Complement C5b6 (Complement Technology, Cat. A122) and normal human serum (NHS) were used to assemble C5b-9 on the cytomembrane according to a previous report (24). To explore the underlying mechanism whereby C5b-9 induces ferroptosis, HK-2 cells were treated with IP3R siRNA, 2 μM BAPTA-AM (Selleck, Cat. S7534) or 10 nM cyclosporin A (CsA, Selleck, Cat. S2286).

HK-2 cells were seeded into 96-well cell plates, and the cell viability was detected using a CCK8 kit (Beyotime, Cat. C0037) as the suppliers’ manual.

Human IP3R-specific small interfering RNA (siR, GenePharma) and lipofectamine 3000 (ThermoFisher Scientific, Cat. L3000015) were mixed in serum-free DMEM/F-12 medium for 15 min. Then, the mixture was added to the culture medium to transfect HK-2 cells for 6 h. After this, the cells continued to be cultured in fresh medium for 48 h and treated with the test compounds. After washing, the cells were harvested and subjected to specific experiments. The sequences of IP3R siR were 5’-GCAGAUCUUCAAGUUGUUATT-3′ (forward) and 5’-UAACAACUUGAAGAUCUGCTT-3′ (reverse). The sequences of the negative control siR were 5’-UUCUCCGAACGUGUCACGUTT-3′ (forward) and 5’-ACGUGACACGUUCGGAGAATT-3′ (reverse).

Lipid peroxidation was detected by BODIPY™ 581/591 C11 (ThermoFisher Scientific, Cat. D3861). Briefly, HK-2 cells were incubated in serum-free DMEN/F-12 medium containing 10 μM BODIPY lipid probe for 30 min at 37°C and 5% CO₂ after the test compounds were treated and subsequently counterstained with Hoechst 33342 (Beyotime, Cat. C1028). Then, the medium was removed, and the cells were washed 3 times with PBS protected from light. The oxidized lipid (green), non-oxidized lipid (red) and nuclei (blue) were imaged using a fluorescence microscope.

Mitochondrial membrane potential was detected using an enhanced mitochondrial membrane potential assay kit with JC-1 (Beyotime, Cat. C2003S). Mitochondrial accumulation of JC-1 is membrane potential dependent and leads to the formation of red fluorescent J-aggregates. The mouse primary tubular epithelial cells or HK-2 cells were incubated with 10 μg/ml JC-1 probe for 20 min at 37°C and 5% CO₂ after the test compounds were administered and subsequently counterstained with Hoechst 33342. The J-monomers (green), J-aggregates (red) and nuclei (blue) were imaged using a fluorescence microscope. Moreover, flow cytometry was also used to detect the mitochondrial membrane potential. Briefly, HK-2 cells were cultured in a 24-well plate treated with the test compounds and harvested by trypsinization. Then, cell clumps were resuspended in 500 μl serum-free DMEN/F-12 medium containing 10 μg/ml JC-1 probe and cultured at 37°C and 5% CO₂ for 20 min. JC-1 monomers and aggregates were detected from the FITC and PE channels, respectively, using a flow cytometer.

mPTP was detected using a mitochondrial permeability transition pore assay kit (Abcam, Cat. 239704). Mouse primary tubular epithelial cells or HK-2 cells were cultured in a 12-well plate treated with the test compounds and harvested by trypsinization. The cells were incubated in 1,000 μl pre-warmed mPTP wash buffer containing 2 μg/ml mPTP staining dye and 5 μl CoCl2 solution at 37°C and 5% CO2 for 15 min protected from light. Cells were centrifuged at 4°C and 1000 × g for 5 min after incubation. Then, the cells were resuspended in 1 mL of mPTP wash buffer to remove excess staining and quenching reagents. Fluorescence intensity was detected from the FITC channel using a flow cytometer.

Renal cortex and HK-2 cells were homogenized in RIPA buffer (Beyotime, Cat, P0013B) and centrifuged at 15,000 × g and 4°C for 15 min to remove cell debris. A BCA protein assay kit (ThermoFisher Scientific, Cat. 23225) was used to detect the protein concentration. The total kidney protein was quantified as 3 μg/μl and the HK-2 cell protein was quantified as 2 μg/μl. The proteins were separated by SDS-PAGE electrophoresis and subsequently transferred to a polyvinylidene fluoride membrane. Membrane was incubated overnight at 4°C using the specific primary antibody after blocking 2 hr with 5% skimmed milk at room temperature. The specific primary antibodies include: anti-ACSL4 antibody (Santa Cruz, Cat. sc-365230, dilution 1:1,000), anti-COX2 antibody (Santa Cruz, Cat. sc-19999, dilution 1:1,000), anti-GPX4 antibody (Abcam, Cat. ab125066, dilution 1:2,000), anti-IP3R antibody (Santa Cruz, Cat. sc-398434, dilution 1:1,000), anti-MCU antibody (Santa Cruz, Cat. sc-515930, dilution 1:1,000), anti-OXPHOS antibody (Abcam, Cat. ab110413, dilution 1:5,000) and anti-GAPDH antibody (Abcam, Cat. ab8245, dilution 1:10,000). Then, the membrane was incubated with secondary antibodies for 2 hr at room temperature after washing with TBST buffer 4 times for 10 min each time. The WesternLumaxLight Sirius HRP substrate kit (Zeta Life, Cat. 310231) was used for color reaction. GAPDH was used as a loading control for total proteins. ImageJ software was used to quantify the protein band intensities (National Institutes of Health, Bethesda, MD, USA).

Total RNA was extracted from the renal cortex or HK-2 cells using TRIzol reagent (ThermoFisher Scientific, Cat. 15596026) and 1-bromine-3-chloropane (1-BCP, Sigma-Aldrich, Cat. B9673). The isolated RNA was reverse-transcribed to cDNA using the reverse transcription system (Promega Corporation, Cat. A3500) following the instruction manual. Real-time PCR was carried out using LightCycler^® 480 SYBR^® Green I Master (Roche, Cat. 04887352001). Gapdh served as a loading control, and the gene-specific primers are listed in Table S1 .

The GSH levels in the renal cortex and HK-2 cells were detected using a GSH detection kit (Nanjing Jiancheng Bioengineering Institute, Cat. A006-2). The MDA levels in the renal cortex and HK-2 cells were detected using an MDA detection kit (Nanjing Jiancheng Bioengineering Institute, Cat. A003-1). GSH and MDA levels were normalized to per gram of protein.

Intracellular Fe²⁺ was detected using the Fe²⁺ detection probe FerroOrange (DOJINDO, Cat. F374). Briefly, HK-2 cells were incubated with Hank’s balanced salt solution (HBSS) containing 1 μM FerroOrange dye at 37°C and 5% CO₂ for 30 min after the test compounds were applied. Cells were imaged immediately using a fluorescence microscope.

Cytosolic Ca²⁺ was detected using a Fluo-8 calcium flux assay kit (Abcam, Cat. ab112129). HK-2 cells and mouse primary tubular epithelial cells were cultured with 100 μl Fluo-8 dye-loading solution at 37°C for 30 min and at room temperature for another 30 min in a black wall and clear bottom 96-well plate. The relative cytosolic Ca²⁺ level was detected at Ex/Em = 490/525 nm using a multimode plate reader (PerkinElmer).

Mitochondrial Ca²⁺ was detected with the mitochondrial Ca²⁺ fluorescent probe Rhod-2/AM probes (Yeasen, Cat. 40776ES72). Briefly, HK-2 cells were incubated with HBSS (without Ca²⁺ and Mg²⁺) containing 4 μM Rhod-2/AM dye at 37°C and 5% CO₂ for 30 min after the test compounds were applied. Cells are washed 3 times with HBSS to adequately remove residual Rhod-2/AM working solution. Then, cells were incubated with in a 37°C incubator for another 30 min to ensure complete de-esterification of AM bodies within the cells. The MitoTracker Green probes were used to localize the mitochondrial. Cells were imaged immediately using a fluorescence microscope.

All statistical analyses were performed using SPSS 23.0 software (SPSS Inc., Chicago, USA). The mean values between different groups were compared using two-way ANOVA with LSD’s multiple comparison post-hoc test. All quantitative data are expressed as the means ± S.E.M. Differences were considered statistically significant at P<0.05.

TCE- and CD59-group mice were divided into a sensitization-positive group (erythema or edema) and a sensitization-negative group (no reaction) according to the dorsal skin reaction ( Figure S1 ). As shown in Table 1 , the sensitization rate of the TCE group was 40.00% (8/20), and the sensitization rate of the TCE+CD59 treatment group was 36.84% (7/19). No obvious skin reaction was observed in Blank- and Vehicle-group mice.

Our previous results show that C5b-9 increases the cytosolic Ca²⁺ level in TCE-sensitization-positive mouse primary tubular epithelial cells and C5b-9-attacked HK-2 cells, but the underlying mechanism remains unclear. Here, we detected the upregulation of IP3R, an endoplasmic reticulum calcium release channel protein, using immunoblotting and RT-PCR methods. As expected, the IP3R protein and mRNA level was significantly higher in the TCE+ group than in the Blank, Vehicle and TCE- groups ( Figures S2A, B ), while pretreatment with CD59 reversed the abnormal increase of the IP3R protein and mRNA in the renal tubules of TCE sensitization-positive mice ( Figures 2A–C ). To explore the role of C5b-9 in the upregulation of IP3R, we assembled C5b-9 on the cytomembrane of HK-2 cells as described in our previous study (13). Immunofluorescence results showed that C5b-9 was assembled after C5b6 and NHS co-treatment ( Figure S3 ). Interestingly, IP3R was clearly upregulated by C5b-9, as IP3R protein and mRNA levels in HK-2 cells were markedly increased after C5b-9 was assembled ( Figures 2D–F ). The consistent changes in mRNA levels support the IP3R gene expression underlying the increased IP3R protein levels. These results suggest that the cytosolic Ca²⁺ overload caused by TCE sensitization may be due to an upregulation of IP3R.

To explore the potential mechanism whereby C5b-9 induces renal tubule ferroptosis, we examined Ca²⁺ levels and mitochondrial damage in tubular epithelial cells from TCE-sensitized mouse kidneys and C5b-9-attacked HK-2 cells. Mouse primary tubular epithelial cells were extracted and cultured with 10% of their own sera, and immunofluorescence was used to identify tubular epithelial cells. CK-18 and Nephrin were used to mark tubular epithelial cells and podocytes, respectively ( Figure S4A ). Figure 3A shows that the Ca²⁺ concentration was significantly higher in primary tubular epithelial cells of TCE+ group mice than in Blank, Vehicle and TCE- group mice. C5b-9 inhibition markedly reduced the increased cytosolic Ca²⁺ level by TCE sensitization ( Figure S5A ). To explore the role of C5b-9 in cytosolic Ca²⁺ overload in HK-2 cells, we used NHS and 0.2 μg/ml C5b6 to assemble C5b-9 on the surface of HK-2 cells ( Figure S3 ). As expect, the relative cytosolic Ca²⁺ levels of C5b-9-attacked HK-2 cells were higher than those of normal cells ( Figure 3B ). The mitochondrial calcium uniporter (MCU) protein and mRNA was increased in the kidneys of TCE+ group mice, and CD59 alleviated the abnormal upregulation ( Figures S2A, C ; 3C–E ). Consistently, the mitochondrial membrane potential of primary tubular epithelial cells of TCE+ group mice was decreased, as seen by the increased JC-1 monomers fluorescence intensity ( Figure S4B ). Meanwhile, immunoblotting analysis showed that the essential components of oxidative phosphorylation system (OXPHOS) was significant lower in TCE sensitization-positive mice as compared to Vehicle and TCE sensitization negative mice ( Figures 3G–L ). Interestingly, mitochondrial membrane potential depolarization and the decrease in the essential components of OXPHOS induced by TCE sensitization were markedly ameliorated by CD59 pretreatment ( Figures 3F–L ). These data provide experimental evidence that TCE sensitization induced C5b-9 deposition causes IP3R upregulation, cytosolic Ca²⁺ overload, mitochondrial calcium influx and dissipation of mitochondrial membrane potential and consequent loss of the mitochondrial function - oxidative phosphorylation.

The key role of IP3R was also verified in HK-2 cells. To further investigate the role of IP3R in C5b-9-induced Ca²⁺ overload and mitochondrial dysfunction, IP3R siR was used in HK-2 cells before C5b-9 assembled. RT-PCR was used to detect the effectiveness of IP3R RNA interference ( Figure S5B ). As expected, IP3R siR pretreatment reduced C5b-9-induced increase in cytosolic Ca²⁺ levels in HK-2 cells ( Figure 3B ). Meanwhile, IP3R siR pretreatment normalized C5b-9-augmented IP3R and MCU protein and mRNA levels ( Figures 3M–Q ). In addition, to study the alteration of mitochondrial Ca²⁺ levels, we detected mitochondrial Ca²⁺ levels with the mitochondrial Ca²⁺ fluorescent probe Rhod-2 AM and the mitochondrial fluorescent probe MitoTracker Green. As shown in Figure S5C , C5b-9 significantly cause an increase in mitochondrial Ca²⁺ levels, while pretreatment with IP3R siR alleviated the abnormal changes in mitochondrial Ca²⁺ caused by C5b-9. Mitochondrial membrane potential was analyzed by fluorescence microscopy and flow cytometry after staining with JC-1 dye. As shown in Figures 3R, S and Figure S6A , IP3R siR pretreatment markedly counteracted C5b-9-induced mitochondrial membrane potential depolarization. In addition, IP3R siR effectively restored C5b-9-downregulated expression of the essential components of OXPHOS ( Figures 3T–Y ).

In summary, these results suggest that IP3R upregulation plays an important role in C5b-9-induced Ca²⁺ overload and mitochondrial damage in the renal tubules of TCE sensitization-positive mice and HK-2 cells.

To explore whether the ferroptosis of HK-2 cells induced by C5b-9 was caused by IP3R upregulation, we detected the expression of ferroptosis-related indicators after IP3R siR pretreatment. As Figure 4A shows, IP3R siR abolished the cell death induced by C5b-9 in HK-2 cells. CoA synthetase long-chain family member 4 (ACSL4), an enzyme that converts fatty acids to fatty acyl-CoA esters, regulates lipid biosynthesis and contributes to ferroptosis. Additionally, cyclooxygenase-2 (COX2) and glutathione peroxidases 4 (GPX4) have been shown to participate in ferroptosis by regulating the cellular redox status. Our previous study showed that C5b-9 caused markedly abnormal changes in ACSL4, COX2 and GPX4 protein and mRNA levels in HK-2 cells. Interestingly, IP3R siR effectively reduced the C5b-9-increased ACSL4 and COX2 protein levels ( Figures 4B–D ). Furthermore, the C5b-9-suppressed GPX4 protein level was also clearly reversed by IP3R siR treatment ( Figures 4B, E ). Consistently, the C5b-9-augmented mRNA levels of Acsl4 and Ptgs2, the encoding gene of COX2, were also antagonized by IP3R siR, while the decreased Gpx4 mRNA level was blocked by IP3R siR ( Figures 4F–H ). GSH and MDA play an important role in maintaining the balance of oxidation and reduction, which was disrupted in C5b-9-attacked HK-2 cells. Further investigation showed that IP3R siR pretreatment attenuated C5b-9-induced aberrant changes in GSH and MDA levels ( Figures 4I, J ). The accumulation of lipid peroxide and intracellular iron metabolism disorders are other characteristics of ferroptosis and were detected by the BODIPY™ 581/591 C11 probe and FerroOrange dye, respectively. As shown in Figure 4K , the increase in lipid peroxide induced by C5b-9 was antagonized by IP3R siR. Finally, IP3R siR alleviated the upregulated Fe²⁺ level induced by C5b-9 ( Figure 4L ). In conclusion, these data show that IP3R activation causes lipid peroxidation and iron deposition and hence ferroptosis and suggest that the ferroptosis induced by C5b-9 likely occurs through activation of IP3R.

To investigate the role of cytosolic Ca²⁺ overload in mitochondrial damage, we treated C5b-9-attacked HK-2 cells with BAPTA-AM, an intracellular Ca²⁺ chelator. The mPTP is a non-specific channel formed by components of the inner and outer mitochondrial membranes. Continuous mPTP opening causes subsequent loss of mitochondrial membrane potential. mPTP opening in primary tubular epithelial cells and HK-2 cells was detected using mPTP staining dye. As shown in Figure 5A and S6B , the mPTP opening was observed in the TCE+ group, while CD59 pretreatment alleviated this abnormal activity. Consistently, the mPTP was also opened in C5b-9-attacked HK-2 cells, and BAPTA-AM effectively rescued the cells from mPTP opening ( Figures 5B and S6C ). Immunoblotting analysis showed that BAPTA-AM reduced IP3R activation-augmented MCU protein and mRNA levels ( Figures 5C–E ). In addition, fluorescence imaging and flow cytometry also demonstrated that BAPTA-AM pretreatment antagonized the loss of mitochondrial membrane potential of HK-2 cells caused by IP3R activation( Figures 5F, G ; S6D ). Furthermore, the IP3R-mediated decrease in the essential components of OXPHOS was alleviated by BAPTA-AM ( Figures 5H–M ). Collectively, these results suggest that IP3R causes mPTP opening and mitochondrial dysfunction in TCE-sensitized mice and HK-2 cells and that cytosolic Ca²⁺ overload may play a key role.

To investigate the role of mPTP activation in IP3R-induced mitochondrial dysfunction, CsA, a mPTP inhibitor, was used in HK-2 cells. The effectiveness of CsA was detected. As expected, CsA pretreatment clearly inhibited IP3R activation-induced mPTP opening ( Figure 6A ; S6E ). Mitochondrial membrane potential was analyzed by fluorescence microscopy and flow cytometry after staining with JC-1 dye. As shown in Figures 6B, C ; S6F , mPTP inhibition markedly counteracted IP3R activation-induced mitochondrial membrane potential depolarization. Next, the effect of CsA inhibition on the essential components of OXPHOS was analyzed. Interestingly, CsA inhibition effectively restored IP3R activation-downregulated expression of the essential components of OXPHOS ( Figures 6D–I ). In summary, these results suggest that mPTP opening plays an important role in IP3R activation-induced mitochondrial damage in HK-2 cells.

The counter effect of mPTP inhibition by CsA on ferroptosis in a C5b-9-attacked HK-2 cell model was examined. Treatment with CsA significantly mitigated the C5b-9-induced decrease in cell viability ( Figure 7A ). Immunoblotting analysis showed that the defective expression of ferroptosis-associated proteins, such as ACSL4, COX2 and GPX4, was clearly rescued by pretreatment with CsA ( Figures 7B–E ). Consistently, the abnormal changes in Acsl4, Ptgs2 and Gpx4 mRNA levels were also counteracted after CsA treatment ( Figures 7F–H ). In addition, treatment with CsA markedly mitigated the IP3R activation-induced decrease in GSH levels and increase in MDA levels ( Figures 7I, J ). Furthermore, the BODIPY™ 581/591 C11 probe was used to detect the lipid peroxidation level. Fluorescence imaging showed that the oxidized lipid level was significantly decreased after CsA treatment ( Figure 7K ). To explore the intracellular free iron level, Fe²⁺ was measured using FerroOrange dye. As shown in Figure 7L , mPTP inhibition markedly alleviated intracellular free Fe²⁺ levels. Taken together, these results suggest that the activation of mPTP plays a key role in IP3R activation-induced ferroptosis of HK-2 cells.