Inhibition of ferroptosis protects House Ear Institute‐Organ of Corti 1 cells and cochlear hair cells from cisplatin‐induced ototoxicity

Abstract Ferroptosis is a recently recognized form of non‐apoptotic cell death caused by an iron‐dependent accumulation of lipid hydroperoxides, which plays important roles in a wide spectrum of pathological conditions. The present study was aimed to investigate the impact of ferroptosis on cisplatin‐induced sensory hair cell damage. Cell viability was determined by Cell Counting Kit‐8 and lactase dehydrogenase assays. The reactive oxygen species (ROS) levels were evaluated by 2,7‐Dichlorodi‐hydrofluorescein diacetate (DCFH‐DA) and MitoSox‐Red staining. Mitochondrial membrane potential (MMP) was measured by tetramethylrhodamine methyl ester (TMRM) staining. Lipid peroxidation, intracellular and mitochondrial iron were detected by Liperfluo, C11‐BODIPY581/591, FerroOrange and Mito‐FerroGreen, respectively. We found that cisplatin treatment not only markedly augmented ROS accumulation, decreased the MMP, but increased lipid peroxidation and iron accumulation in House Ear Institute‐Organ of Corti 1 (HEI‐OC1) cells. Of note, treatment with the specific ferroptosis inhibitor ferrostatin‐1 could effectively abrogate the cisplatin‐induced toxicity and subsequent cell death. Specifically, the improvement of mitochondrial functions is important mechanisms for protective action of ferroptosis inhibitor against cisplatin‐induced damages in HEI‐OC1 cells. Moreover, inhibition of ferroptosis significantly protected murine cochlear hair cells against cisplatin damage. In addition, treatment murine cochlear hair cells with ferroptosis inducer, RSL3, significantly exacerbated cisplatin‐induced damage, which could be alleviated by ROS inhibitor N‐acetyl‐L‐cysteine. Collectively, our study indicated that ferroptosis inhibition could alleviate the cisplatin‐induced ototoxicity via inactivation of lipid peroxide radical and improvement of mitochondrial function in hair cells.


| INTRODUC TI ON
Hearing loss is a very common sensory disorder in humans and is mainly attributable to hair cell damage caused by ototoxic pharmaceutical agents, excessive noise, ageing and genetic disorders.
Cisplatin is one of the largest classes of valuable clinical chemotherapeutic agents against various malnancies. 1,2 However, the ototoxicity induced by cisplatin is one of the most serious adverse effects of cisplatin administration, which results in irreversible, progressive, bilateral and accumulative hearing impairment, thereby limiting the clinical use of cisplatin. 3,4 Numerous studies have shown that cochlear hair cells and spiral ganglion neurons are major targets, and exposure of these cells to cisplatin induces a significant cell death. [5][6][7] Though it has been shown that the cisplatin-induced ototoxicity is associated with the reactive oxygen (ROS)/nitrogen species accumulation, mitochondrial dysfunction and caspase activation in auditory organs, 8,9 exact mechanisms involved in cisplatin damage have still not been fully clear, thereby better understanding the underlying mechanisms of cisplatin-induced ototoxicity are crucial for seeking more effective therapies for preventing hearing loss.
Ferroptosis is a recently recognized type of non-apoptotic cell death that is characterized by the overload of free ferrous ion and the accumulation of lipid-based ROS; it is biochemically and morphologically distinct from apoptosis, necroptosis and autophagy. 10 The accumulation of lipid peroxides is mainly caused by the reduced activity of glutathione peroxidase 4 (GPX4), which is a unique intracellular antioxidant enzyme that suppresses lipid peroxidation generation in the cell membrane. 11 Another key regulator in ferroptosis is the cystine/glutamate antiporter system x − c (xCT), which exchanges extracellular cystine for intracellular glutamate. 12 There are increasing studies showing that ferroptosis inducers, such as RSL3, inhibiting the function of GPX4, 13 and erastin, inhibiting xCT, 14,15 have been confirmed to enhance sensitivity of drug-resistant cancer cells to chemotherapeutic drugs such as cisplatin and temozolomide thereby exhibiting anticancer effects. Several inhibitors of ferroptosis have recently been identified, including liproxstatin-1, 16 ferrostatin-1 (FER-1) 17 and the iron chelator deferoxamine (DFO). Inhibition of accumulation of lipid peroxidation that inhibits ferroptosis could present highly promising way to treat pathological conditions by protecting from the cell loss in the brain, liver, kidney and other tissues. 16,18,19 In vivo studies with ferroptosis inhibitors highlighted the importance of inhibition of this death pathway in mitigating cell damage. 16,18 To date, there has been no study with regard to ferroptosis involvement in cisplatin-induced ototoxicity.
In this study, we investigated the involvement of ferroptosis in cisplatin-induced hair cell damage, and the potential protective effect of ferroptosis inhibition in alleviating the impairment of hair cells induced by cisplatin administration in both auditory House Ear Institute-Organ of Corti 1 (HEI-OC1) cells and murine cochleae. Our results showed that inhibition of ferroptosis with FER-1 significantly attenuated cisplatin-induced hair cell damage by preserving mitochondrial function, suggesting that inhibition of ferroptosis might be a novel therapeutic target for future hearing loss treatment.

| Postnatal cochlear explants culture
Postnatal day (P) 2 C57BL/6 mice were sacrificed and soaked in 75% alcohol, and the cochleae tissues were carefully dissected using scissors and placed in cooled PBS. The cochlea was then stuck to a glass of coverslip coated with Cell-Tak (BD Biosciences, Franklin Lakes, NJ, USA). Finally, DMEM/F12 medium supplemented with N2/B27 (Invitrogen) and ampicillin was added, and the cochleae tissues were cultured in a 5% CO 2 /95% air atmosphere at 37°C overnight prior to each treatment. All experimental procedures on animals in this study were conducted in accordance with the laboratory animals care guidelines and approved by the Institutional Animal Care and Use Committee of Fudan University.

| Cell viability quantification
Cell Counting Kit-8 (CCK-8) (Sigma, Saint Louis, USA) reagent was used to examine cell viability according to the manufacturer's instructions. In brief, the cultured HEI-OC1 cells were seeded at the density of 5000 cells/well in 96-well plates in three replicates and allowed to attach overnight. At the end of different treatments, 10 μL per well CCK-8 reagent was added to each well in incubation for 2 hours. Absorbance at 450 nm was detected by a plate reader (BioRad, Hercules, CA, USA).

| Lactase dehydrogenase release assay
The extent of cellular injury was determined by lactase dehydrogenase (LDH) leakage using LDH cytotoxicity detection kit (Dojindo Laboratory, Kumamoto, Japan). Briefly, HEI-OC1 cells were seeded at the density of 5000 cells/well in 96-well plates. According to the manufacturer's instruction, at the end of different treatments, 100 μL of fresh reaction mixture was added to each well and incubated for 30 minutes, the absorbance at 490 nm was determined using a microplate reader (BioRad).

| ROS measurement
The level of intracellular and mitochondrial ROS was respec-

| Lipid peroxides measurement
To visualize the lipid ROS, cells were seeded in 8-well ibidi plates and treated with the designate conditions. After different treatments, cells were stained with 5 μmol/L Liperfluo (Dojindo) or 2 μmol/L C11-BODIPY 581/591 probe (Molecular Probes Inc., Eugene, Oregon, USA) in accordance with the manufacturer's instructions.

| Mitochondrial membrane potential measurement
Mitochondrial membrane potential (MMP) was estimated by TMRM

| Fe 2+ detection
To detect intracellular and mitochondrial Fe 2+

| Measurement of GSH/GSSG ratio
The intracellular total glutathione (GSH) and glutathione disulphide (GSSG) levels were detected by GSH Kit (Dojindo) according to the manufacturers' protocols. The relative levels were analysed on the microplate reader (BioRad).

| Propidium iodide staining
Cell death was determined using Hoechst 33342/propidium iodide (PI) staining (Dojindo). After HEI-OC1 cells were seeded in 8-well ibidi plates and treated with different conditions, the cells were stained with PI (5 μg/mL) and Hoechst 33 342 (5 μg/mL). After 10 minutes in the dark, the cells were then examined using a Leica SP8 confocal fluorescence microscope (Leica Microsystems) equipped with a 60 × 1.40 NA oil immersion objective. Slices were collected every 1.5 μm to generate z-stacks, and each image represented the maximum projection of all slices along the z-stack. For evaluation, HEI-OC1 cells were seeded in 24-well plates with 50 000 cells/well. After different treatment, cells were stained with PI (5 μg/mL) for 10 minutes in the dark. After collecting and washing with PBS, cells were re-suspended in PBS and red fluorescence was detected by FACS analysis (FACScan; BD Biosciences, San Jose, CA, USA). Data were collected from at least 20 000 cells.

| Immunofluorescence
The cochleae fixed in 4% paraformaldehyde were rinsed three times with PBS and permeabilized for 30 minutes with 1% Triton X-100 in PBS (PBST), blocked with 10% donkey serum in PBST

| Western blot analysis
Cochleae were lysed with cold radio-immunoprecipitation assay Protein bands were detected using a chemiluminescence solution, electrochemiluminescence (ECL) kit (Millipore, Massachusetts, MA, USA). Each experiment was repeated three times, and all protein expression was normalized to that of GAPDH.

| Cochlear hair cell counts
For quantitative assessment of hair cells, the intact myosin 7a-positive hair cells were separately counted along the apical, middle and basal turns in each cochlea. The average number of hair cells per F I G U R E 1 Effects of ferrostatin-1 (FER-1) and RSL3 on cell viability in cisplatin-damaged House Ear Institute-Organ of Corti 1 cells. A and B, Cells treated with varying concentrations of RSL3 for 24 h were analysed by Cell Counting Kit-8 (CCK-8) and lactase dehydrogenase (LDH) assays. (C-D) Cells were pre-treated with varying concentrations of FER-1 for 2 h, followed by addition of 3 μM RSL3 for 24 h and analysed by CCK-8 and LDH assays. E and F, Cells treated with varying concentrations of cisplatin for 24 h were analysed by CCK-8 and LDH assays. G and H, Cells were pre-treated with varying concentrations of RSL3 for 2 h, followed by addition of 30 μmol/L cisplatin for 24 h and analysed by CCK-8 and LDH assays. I and J, Cells were pre-treated with varying concentrations of FER-1 for 2 h, followed by addition of 30 μmol/L cisplatin for 24 h, and analysed by CCK-8 and LDH assays. All the data represent the mean ± SEM. of three independent experiments. ***P < 0.001, ****P < 0.0001 vs the control group; # P < 0.05, ## P < 0.01, ### P < 0.001, #### P < 0.0001 vs the group treated with RSL3 (C, D) or cisplatin (G-J) alone 200 µm at different regions in each cochlear explant was calculated from experimental groups. Each experiment was repeated three times, and exact numbers of cochlear explants (n) are indicated in the legends.

| Statistical analyses
Statistical analyses were performed using the GraphPad Prism statistical software (version 6; GraphPad Software, Inc, San Diego, CA).
Data were shown as mean ± SEM and analysed by one-way ANOVA. P < 0.05 was considered statistically significant.

| Inhibition of ferroptosis protected against cisplatin-induced loss of cell viability
To assess whether RSL3 exposure could induce ferroptosis in hair cells, HEI-OC1 cells, a widely used auditory hair cell line derived from murine organ of Corti, 20-24 were exposed to RSL3 at different concentrations for analysis of cell viability and LDH release by CCK-8 and LDH assays. As seen in Figure 1A

| Inhibition of ferroptosis prevented cisplatininduced iron overload in HEI-OC1 cells
Given that ferroptosis is mainly dependent on intracellular iron accumulation and lipid peroxidation, we first measured the level of However, these changes could be remarkably ameliorated by FER-1 pre-treatment or enhanced by RSL3 challenge (Figures S2 and S3).

| Inhibition of ferroptosis ameliorated cisplatininduced ROS accumulation in HEI-OC1 cells
As ROS accumulation is closely related to ferroptosis, 10,25 we thus determined the production of total ROS by using DCFH-DA reagent.   (Figure 4B,C). Additionally, when we used another C11-BODIPY 581/591 probe as a lipid peroxide indicator, 10 FER-1 pre-treatment significantly decreased the lipid peroxide level that was increased by cisplatin ( Figure S4). Treatment with FER-1 only was unable to produce an increase of both Liperfluo and C11-BODIPY 581/591 markers. Given that GSH plays key roles in exerting cellular antioxidant functions, we next investigated the mechanism underlying the otoprotective effect of FER-1 by testing GSH/GSSG ratio. As shown in Figure 4D, after treatment with cisplatin, the GSH/GSSG ratio in HEI-OC1 cells markedly decreased in contrast to the control group. However, pre-treatment with FER-1 significantly increased the GSH/GSSG ratio compared with the cisplatin group ( Figure 4D). Taken together, these results hinted that cisplatin disrupted cellular antioxidant capacity and induced ferroptosis in HEI-OC1 cells.

| Inhibition of ferroptosis prevented cisplatininduced breakdown of MMP in HEI-OC1 cells
To investigate the effect of ferroptosis on MMP in cisplatin-damaged HEI-OC1 cells, MMP was measured by TMRM staining. As shown in Figure 5, the TMRM fluorescence reduced significantly in the cisplatin group, compared with the control group, indicating that cisplatin challenge provoked a loss of membrane potential; and exacerbated vanishing membrane potential was observed after cisplatin and RSL3 treatment ( Figure 5). Conversely, FER-1 pre-treatment protected the cells against mitochondrial damage after exposure of cisplatin, as evidenced by enhanced fluorescence intensity ( Figure 5).

F I G U R E 4
Effects of ferrostatin-1 (FER-1) on lipid reactive oxygen species (ROS) production and ratio of glutathione (GSH) and glutathione disulphide (GSSG) in cisplatin-damaged House Ear Institute-Organ of Corti 1 (HEI-OC1) cells. A, The experimental workflow. The HEI-OC1 cells were pre-treated with 3 μmol/L RSL3 or 30 μmol/L FER-1 for 2 h and then treated with or without 30 μmol/L cisplatin for another 24 h, or treated with 30 μmol/L cisplatin alone for 24 h, and then lipid ROS was detected by Liperfluo. B, Representative images of Liperfluo staining. Scale bar, 20 µm. C, The fluorescence intensity was quantified by ImageJ software. The data are shown as mean ± SEM. of three independent experiments. *P < 0.05, **P < 0.01, ****P < 0.0001 and n.s. no significant vs the control group; ## P < 0.01 vs the cisplatin group. D, GSH/GSSG assay. The data is shown as mean ± SEM. of three independent experiments. ***P < 0.001, ****P < 0.0001 and n.s. no significant vs the control group; # P < 0.05 vs the cisplatin group

| Inhibition of ferroptosis protected cochlear hair cells against cisplatin-induced damage
Moreover, we strengthened the findings from HEI-OC1 cells in organotypically cultured cochlear explants from P2 C57BL/6 mice in vitro. 26 Cochlear explants were treated with culture media, 30 μmol/L cisplatin alone, or pre-treatment with

| D ISCUSS I ON
As the roles of ferroptosis in the pathological situations have been recently investigated, the importance of ferroptosis in ototoxic pharmaceutical agents-mediated hearing loss is increasing. We first demonstrated that ferroptosis plays a pivotal role in the development and progression of cisplatin-induced hair cell injury, and inhibition of ferroptosis potentially protected hair cells from cisplatin-mediated F I G U R E 6 Effect of ferrostatin-1 (FER-1) on iron accumulation in cisplatindamaged cochlear hair cells. A, The experimental workflow. Cochlear explants were treated with 30 μmol/L cisplatin alone (Cis) for 24 h, or pre-treatment with 30 μmol/L FER-1 for 2 h and addition of 30 μmol/L cisplatin for 24 h followed by 3 d recovery. B, Representative images of myosin 7a (green) and FerroOrange (red) staining of middle cochlear turns from different groups. Scale bar, 20 µm. C, Representative images of myosin 7a (red) and Mito-FerroGreen (green) staining of middle cochlear turns from different groups. Scale bar, 20 µm. D and E, Relative fluorescence intensity of FerroOrange and Mito-FerroGreen. The data are shown as mean ± SEM. **P < 0.01, ***P < 0.001, ****P < 0.0001 vs the control group; #### P < 0.0001 vs the cisplatin group, n = 12-18 cochlear explants from three independent experiments damage through mechanisms involving preservation of mitochondrial parameters, such as maintaining MMP and inhibiting ROS production.
Although the production of excessive ROS is considered to be the major cause of cisplatin-induced ototoxicity; increasing ROS production leads to downstream cytokine deprivation, mitochondrial The data are shown as mean ± SEM. ***P < 0.001, ****P < 0.0001 and n.s. no significant vs the control group; ### P < 0.001, #### P < 0.0001 vs the cisplatin group, n = 6 cochlear explants from three independent experiments dysfunction and eventually leading to cell death, 33-36 the exact molecular mechanism of cisplatin ototoxicity is still not fully understood. Ferroptosis is a type of newly discovered programmed cell death dependent on iron 37 ; it is distinct from other cell death forms such as apoptosis, autophagy or necrosis. 10 Ferroptosis has been implicated in multiple pathological situations, such as cancers 38 and neurodegenerative diseases. 17 Recent studies have indicated that triggering ferroptosis in cancer cells is a promising approach to increase the sensitivity to cisplatin. 39 Liperfluo-OX; it responds to membrane lipid hydroperoxides and reports intracellular sites of lipid hydroperoxide accumulation. 44,45 Moreover, due to its high solubility in various organic solvents, such as ethanol and dimethylsulphoxide, this probe is extremely useful for the imaging of lipid hydroperoxides in living cells. 45 In the current study, we exposed cultured HEI-OC1 cells to cisplatin and estimated LPO production based on the Liperfluo signal. Liperfluo signal significantly increased in the presence of cisplatin; this illustrated that cisplatin-induced damage in cultured HEI-OC1 cells is directly due to iron-mediated lipid peroxidation. In contrast, the increment of lipid LPO in response to cisplatin injury was remarkably ameliorated by FER-1 treatment. Additionally, we further validated this finding using another lipid peroxidation-sensitive dye (C11-BODIPY 581/591 ) confirming the data obtained by Liperfluo. C11-BODIPY 581/591 is a fatty acid analogue and binds to hydroxyl and superoxide radicals; it is a fluorescent radio-probe for indexing lipid peroxidation and antioxidant efficacy in model membrane systems. C11-BODIPY 581/591 has a high quantum yield of fluorescence emission, assuring a good signal measurement, and it has a good photo-stability and its fluorescence emission is virtually insensitive to environmental changes such as solvent polarity or pH. Oxidation of the C11-BODIPY 581/591 induces a shift of the fluorescence emission maximum from red (non-oxidized) to green (oxidized) fluorescence. 46 Furthermore, C11-BODIPY 581/591 was reported to be very sensitive to free radicals formed by the decomposition of hydroperoxides, but not to hydroperoxides themselves. 16,[46][47][48] Use of both the Liperfluo and C11-BODIPY 581/591 assays allow for detection of LPO, but in contrast to C11-BODIPY 581/591 , which does not react with a lipid hydroperoxide, Liperfluo is the only compound that can specifically detect lipid peroxides; thus, the Liperfluo assay may be the most closely related to the mechanisms of ferroptosis. Moreover, previous studies have provided evidence that GSH serves a major role in protecting cells against damage, while GSH depletion could induce an iron-dependent accumulation of lipid peroxidation, eventually triggering ferroptotic cell death. 16 Here, we discovered that cisplatin significantly inhibited the GSH/GSSG ratio in HEI-OC1 cells compared to the undamaged controls; the decrease, however, was reversed by FER-1 treatment. Together, these findings illustrated that cisplatin led to ferroptosis through enhancing iron overburden and lipid peroxidation, reducing GSH in HEI-OC1 cells.
Accumulating studies showed that cisplatin ototoxicity was closely related to the enhanced ROS production in the inner ear. 49,50 Cisplatin triggers overproduction of ROS in the cochlea resulting in sensory epithelial cell death. 33  Mitochondria is a key regulator of cellular processes, and mitochondrion dysfunction plays an important role in cell death.
Increasing evidence suggested that the mitochondria was important sources of ROS generation that led to cisplatin-induced ototoxicity. 51 In this research, to investigate whether ferroptosis inhibition has a role in the modulation of mitochondrial dysfunction induced by cis- tin-damaged cochlear hair cells with RSL3 both without and in combination with radical scavenger NAC for 24 hours, respectively, we found significant protection by application of NAC; these findings were in consonance with those studies showed that NAC protected cochlea against cisplatin ototoxicity. 53,54 Glutathione peroxidase 4, a GSH-dependent enzyme, is one of the most important antioxidant enzymes that selectively catalyses lipid hydroperoxides, reducing lipid hydroperoxides to lipid alcohols. [55][56][57] Previous studies have reported that inhibition of GPX4 activity leads to rapid lipid peroxide accumulation, triggering an iron-dependent ferroptotic cell death, 38,58 and deletion of GPX4 in mice is embryonic lethal. 16 In the present study, our data suggested that cisplatin toxicity was not only a result of increase in cleaved-PARP protein expression but, was also attributed to a remarkable reduction of protein levels of GPX4 within 24 hours of cisplatin exposure, while addition of FER-1 reversed GPX4 expression, but had no effect on cleaved-PARP. These observations, in combination with our observations that iron overburden, ROS/LPO production, mitochondrial dysfunction secondary to cisplatin exposure were largely ameliorated by FER-1 pre-treatment, imply an important role for ferroptosis in cisplatin-induced ototoxicity, and the administration of ferroptosis inhibitor, significantly prevented the ferroptotic cell death secondary to cisplatin exposure.
It is important to note that although inhibition of ferroptosis in vitro has provided highly promising protective effects, further exact mechanism and in vivo mouse model studies are needed.
Although we showed a critical function and underlying mechanisms of ferroptosis in cisplatin-induced ototoxicity in vitro and demonstrated that its inhibition may be a potential therapeutic approach for treating cisplatin-induced hearing loss, the effects of ferroptosis inhibitor on some key proteins related to cell survival or other mitochondrial parameters after cisplatin is currently unclear, which would require further examination.
In conclusion, the present study reveals that ferroptosis plays a vital role in cisplatin-induced hair cell damage. Inhibition of ferroptosis antagonizes cisplatin-induced ototoxicity through inhibiting oxidative toxicity and mitochondrial dysfunction in both cultured auditory HEI-OC1 cells and neonatal mouse cochlear explants. Our findings suggest that ferroptosis inhibitor might serve as potential otoprotectant for the treatment of cisplatin-induced hearing loss.

ACK N OWLED G EM ENTS
The authors would like to thank Xu Wang and Shaoyang Sun for their technical assistance and Yalin Huang for help with the confocal microscope. This work was supported by grants from the National Natural Science Foundation of China (nos. 81870728, 81900931, 81800912), and Sponsored by Shanghai Rising-Star Program (19QA1401800).

CO N FLI C T O F I NTE R E S T S
The authors declare no competing financial interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.