CORM‑2 reduces cisplatin accumulation in the mouse inner ear and protects against cisplatin-induced ototoxicity

Graphical abstract

Highlights

• •CORM-2 attenuates cisplatin-induced hearing loss in young adult mice.
• •CORM-2 co-treatment decreases platinum accumulation in the inner ear.
• •CORM-2 protects against cisplatin-induced toxic cellular responses including necroptosis, plasma membrane disruption, and apoptotic cell death.
• •CORM-2 co-treatment reverses the persistent inflammatory environment created by cisplatin.
• •CORM-2 co-treatment reduces cochlear capillary leakage (hyperpermeability) and maintains the integrity of blood-labyrinth barrier.

Abstract

Introduction

Cisplatin is a life-saving anticancer compound used to treat multiple solid malignant tumors, while it causes permanent hearing loss. There is no known cure, and the FDA has not approved any preventative treatment for cisplatin-based ototoxicity.

Objectives

This study investigated whether the carbon monoxide (CO)-releasing tricarbonyldichlororuthenium (II) dimer, CORM-2, reverses cisplatin-induced hearing impairment and reduces cisplatin accumulation in the mouse inner ear.

Methods

Male 6-week-old BALB/c mice were randomly assigned to one of the following groups: control (saline-treated, i.p.), CORM-2 only (30 mg/kg, i.p., four doses), cisplatin only (20 mg/kg, i.p., one dose), and CORM-2 + cisplatin, to determine whether cisplatin-based hearing impairment was alleviated by CORM-2 treatment.

Results

Our results revealed CORM-2 significantly attenuated cisplatin-induced hearing loss in young adult mice. CORM-2 co-treatment significantly decreased platinum accumulation in the inner ear and activated the plasma membrane repair system of the stria vascularis. Moreover, CORM-2 co-treatment significantly decreased cisplatin-induced inflammation, apoptosis, and cochlear necroptosis. Because the stria vascularis is the likely cochlear entry point of cisplatin, we next focused on the microvasculature. Cisplatin induced increased extravasation of a chromatic tracer (fluorescein isothiocyanate [FITC]-dextran, MW 75 kDa) around the cochlear microvessels at 4 days post-treatment; this extravasation was completely inhibited by CORM-2 co-therapy. CORM-2 co-treatment effectively maintained the integrity of stria vascularis components including endothelial cells, pericytes, and perivascular-resident macrophage-type melanocytes.

Conclusion

CORM-2 co-therapy substantially protects against cisplatin-induced ototoxicity by reducing platinum accumulation and toxic cellular stress responses. These data indicate that CORM-2 co-treatment may be translated into clinical strategy to reduce cisplatin-induced hearing loss.

Introduction

Cisplatin, a platinum-based compound, effectively treats a wide range of solid tumors including head and neck, testicular, ovarian, bladder, cervical, breast, esophageal, lung, and brain cancers [1]. Cisplatin was first licensed for medical use in 1978 [2], and the drug is on the List of Essential Medicines of the World Health Organization for both adults and children; drugs on this list are often the safest and most effective medications [3]. However, cisplatin causes permanent hearing loss in 20–80 % of patients [4], [5], greatly compromising their quality of life [4], [5]. Thus, there is an unmet clinical need for therapies that can prevent cisplatin-based ototoxicity.

There are two possible mechanisms of cisplatin-induced hearing loss: long-term cisplatin retention in the cochlea after treatment, and induction of toxic stress responses [6], [7], [8]. After systemic administration, cisplatin persisted for an extended duration in the inner ears of both humans and mice [6], [9], [10]. Cisplatin was present in various cochlear sub-regions including the stria vascularis (SV), organ of Corti, and spiral ganglion neurons; all of these structures are essential for sensitive hearing [11], [12], [13], [14]. Toxic cisplatin-related responses include the induction of heat shock proteins (HSPs), activation of Toll-like receptor 4 (TLR4) downstream pathways, and upregulation of cell death pathways (apoptosis and necroptosis). HSP induction protects hair cells from major stressors [15], [16], [17]. HSP32, an otoprotective HSP, is also known as heme oxygenase-1 (HO-1) because it degrades heme to free iron, CO, and biliverdin; subsequently, biliverdin is converted to bilirubin. Baker et al. reported that HO-1/HSP32 induction reduced cisplatin-induced hair cell death in cultures of utricles from adult mice [15]. Thus, pharmacological regulators of HO-1/HSP32 may prevent cisplatin-induced hearing impairment. One novel stress response involves the TLR4 pathway. A recent study showed that cisplatin activated TLR4, a receptor that recognizes bacterial lipopolysaccharide, triggering hair cell damage in vivo; it also triggered hair cell oxidative, inflammatory, and cell death responses in vitro, indicating that the targeting of TLR4 may be otoprotective [18].

The activation of cell death pathways (necroptosis and apoptosis) greatly damages hair cells exposed to major stressors. Necroptosis is a form of programmed cell death mediated by activation of the mixed lineage kinase-like (MLKL) pseudokinase, which is phosphorylated by protein kinase RIPK3. The phosphorylated MLKL enzyme induces plasma membrane rupture, leading to the release of intracellular components [19], [20], [21], [22], [23]. Ruhl et al. recently showed that both necroptosis and apoptosis contribute to cisplatin-induced ototoxicity in vivo using a caspase-8/Ripk3 double-knockout mouse line and pharmacological inhibitors of RIPK1 (necrostatin-1s, 0.1 mg/mL) and all caspases (zVAD-FMK). However, only apoptosis contributed to cisplatin-based ototoxicity ex vivo; necroptosis did not [24]. Thus, pharmacological inhibition of necroptosis and/or apoptosis may mitigate cisplatin-induced ototoxicity.

CO is a gaseous mediator (i.e., “gasotransmitter”) endogenously produced by humans, similar to the gasotransmitters nitric oxide (NO) and hydrogen sulfide (H₂S) [25]. CO is generated by two isoforms of heme oxygenases (HOs): constitutive (HO-2 and HO-3) and inducible (HO-1). CO exerts anti-tumor, anti-inflammatory, and anti-apoptotic effects [25]. The carbon monoxide (CO)-releasing tricarbonyldichlororuthenium (II) dimer, CORM-2 treatment [30 mg/kg, intraperitoneally (i.p.)] before and after administration of doxorubicin protected against doxorubicin-induced cardiotoxicity by reducing oxidative stress and apoptosis [26]; it also partially inhibited neomycin-induced hair cell death in whole-organ cultures of utricles from adult mice, confirming that CORM-2 exerted both anti-tumor activity and protected utricle hair cells against neomycin [27]. The present study explored whether CORM-2 prevented cisplatin-induced ototoxicity. CORM-2 and CO gas therapy exert anticancer activities; thus, if CORM-2 alleviates cisplatin-based ototoxicity, CORM-2 co-therapy could be rapidly translated to the clinic to protect hearing in cancer patients undergoing cisplatin treatment. The current study demonstrates that CORM-2 co-treatment reduces cisplatin accumulation in the inner ear and protects against cisplatin-based ototoxicity by reducing toxic cellular stress responses (apoptosis, necroptosis, and inflammation) and maintaining vascular integrity.

Materials and methods

Animals and drug administration

Male BALB/c mice, aged 5 weeks, were purchased (total of 60 mice) from Orient Bio (Seoul, South Korea). After one week of acclamation, animals within each group (control, CORM-2, cisplatin, cisplatin + CORM-2) were randomly assigned to experiments. A single animal was considered as an experimental unit except for molecular tests (single cochlea/unit). All mice were maintained in an identical environment (temperature 22 °C, humidity 45–55 %) with a 12/12-h dark-light cycle (0700 to 1900 h) and fed pelleted food (2018 Teklad global 18 % protein rodent diets, Envigo) and water ad libitum.

Mice in cisplatin-treated groups (Cis only, Cis + CORM-2) were intraperitoneally (i.p.) injected with one dose of 20 mg/kg cisplatin (Ildong pharmaceutical Co. Seoul, South Korea). Mice in non-cisplatin-treated groups (Saline control) received comparable volumes of sterile saline (0.9 % NaCl, i.p.) on drug injection days. All CORM-2-treated mice (CORM-2, Cis + CORM-2) received once daily doses of CORM-2 (30 mg/kg, i.p., Sigma-Aldrich, St. Louis, MO, USA) beginning two hours prior to the start of cisplatin administration in addition to daily CORM-2 administration for three days (total of 4 injections). Mice in cisplatin-treated groups received twice daily nutrition and hydration support to maintain body weight and overall health. Samples were collected for biochemical and molecular analysis at 4 days post treatment (Fig. 1A).

Fig. 1.

Cisplatin-induced ototoxicity is reduced by CORM-2 treatments in BALB/c mice. (A) Hearing sensitivity was evaluated by ABR in all mice at baseline (−1 day) and 4 days following cisplatin (20 mg/kg, i.p., one dose) treatment. CORM-2 (30 mg/kg, i.p., four doses) was pretreated intraperitoneally 2 h prior to cisplatin administration and at 1, 2, and 3d following cisplatin injection. Threshold shifts were recorded as the difference in ABR threshold between baseline and post-treatment measurements. Animals were sacrificed at 4 days after cisplatin treatment. (B) Cisplatin-treated mice (orange line) demonstrated significantly increased threshold shifts compared to saline-treated mice (dashed line). CORM-2 treatment (green line) did not impact auditory threshold shifts. Mice pre-treated with CORM-2 and cisplatin had significantly smaller threshold shifts relative to cisplatin alone. n = 8, saline; n = 20, cisplatin; n = 16, cisplatin + CORM2; n = 8, CORM2. * p < 0.05 cisplatin vs. other groups; within each frequency, compared treatments (simple effects within frequency); a, main effect of treatment, p < 0.05. Two-way ANOVA, Tukey's multiple comparisons test.

Ethics statement

All procedures were approved by the IACUC of the Chungnam National University (IACUC committee’s reference number #CNU-00937).

Auditory brainstem response

ABR thresholds were obtained separately from both ears using the TDT System-3 (Tucker Davis Technologies, Gainesville, FL, USA) hardware and software as described previously [28]. The waveforms were analyzed using a custom program (BioSig RP, ver. 4.4.1; Tucker Davis Technologies) with the researcher blinded to the treatment group. Threshold was defined as the lowest stimulus intensity to evoke a wave III response > 0.2 μV.

Histology

Cochlear samples from mice were collected at 4 days post treatments. Tissues were obtained and fixed in 4 % paraformaldehyde in phosphate-buffered saline (PBS) for 30 min at room temperature. The separated individual cochlear turns were processed for immunostaining as described previously [29].

Inductively coupled plasma mass spectrometry

Platinum measurements were performed using a triple quadrupole inductively coupled plasma mass spectrometer (ThermoScientific, Germany) at Korea Basic Science Institute (KBSI). Samples were labeled only by a unique identifying number, sent to the KBSI, so the investigators performing the analyses were blinded to the treatment group. All whole organs were weighed prior to analysis. Organs were digested with trace metal nitric acid (Fisher Chemical) for 2 h at 100 °C in 15 ml Teflon vial (Savillex, USA) with a cap on a heat block in the chemical fume hood. An equal volume of hydrogen peroxide (Optima Grade, Fisher Chemical) was added and evaporated. Samples were diluted 1:100 with 2 % nitric acid prior to running. A quantitative analysis method in standard mode measuring the platinum 195 isotope was used to determine platinum concentrations. A platinum standard (AccuStandard, Inc., USA) curve was generated with 0.1, 1.0, and 10.0 ppb standards. Thallium isotope 205 was used as the internal standard (1.0 ppb stock concentration). Each sample was run in duplicate and platinum concentrations were normalized to organ weight.

Protein extraction and quantification

Whole cochlear tissue samples were collected and homogenized as described previously [29]. Protein concentrations were determined using the BCA protein assay kit (Thermo Scientific, #23225) following manufacturer's recommendations and the plates were read at 562 nm in a microplate reader (Tecan US Inc., Durham, NC, USA).

Western blotting. Samples prepared in protein sample buffer were run on 4–15 % precast Tris-HCl SDS-polyacrylamide gels (BioRad, Hercules, CA, USA) and transferred to PVDF membranes (Millipore, Burlington, MA, USA). Blots were successively probed with primary and secondary antibodies (Table 1) in TBS containing 3 % protease free BSA (Sigma, St. Louis, MO, USA) and were visualized using Immobilon Western Chemiluminescent HRP Substrate (Millipore, Burlington, MA, USA) and the images were acquired and quantitated using Azure 300 Chemiluminescent Western Blot Imaging System (Azure Biosystems, Dublin, CA, USA).

Table 1.

Antibody list.

Antibody	Company	Catalog #	Dilution	Source (Host)
β-actin	Santa Cruz	sc-47778-HRP	1:3000	–
Bax	cell signaling	3498	1:1000	Rabbit
Bcl-2	cell signaling	2772	1:1000	Rabbit
Caspase3	Cell signaling	9661S	1:1000	Rabbit
Caspase7	cell signaling	9494	1:1000	Mouse
CHMP4B	Invitrogen	PA5-100092	1:200	Rabbit
DAPI	Thermo Scientific	62248	1:4000	–
HIF-1a	cell signaling	3716	1:1000	Rabbit
Hoechst 33342	Invitrogen	H3570	1:1000	–
Iba1	Abcam	ab234437	1:200	Rabbit
IL-1b	Abcam	ab234437	1:1000	Rabbit
MyoVIIa	Proteus	25-6790	1:200	Rabbit
NG2	Santa Cruz	sc-33666	1:200	Mouse
PECAM	EMP Millipore	MAB1398Z	1:200	hamster
p-Ser358-MLKL	Invitrogen™	PA5105678	1:200	Rabbit
RIPK1	Cell signaling	3493S	1:200	Rabbit
RIPK3	Cell signaling	9661S	1:200	Rabbit
Alexa Fluor 488 Phallodin (F-actin probe)	Invitrogen	A12379	1:500	Conjugated
Alexa Fluor 594, Goat anti-Rabbit IgG (H + L) Cross-Adsorbed ReadyProbes™ Secondary Antibody	Invitrogen	R37117	1:500	Goat
Alexa Fluor 594, Goat anti-Rabbit IgG (H + L) Highly Cross-Adsorbed Secondary Antibody	Invitrogen	A11037	1:500	Goat
Alexa Fluor 488, Goat anti-Rabbit IgG (H + L) Highly Cross-Adsorbed Secondary Antibody	Invitrogen	A11034	1:500	Goat
Alexa Fluor 488, Goat anti-Mouse IgG (H + L) Cross-Adsorbed Secondary Antibody	Invitrogen	A11001	1:500	Goat
Alexa Fluor 488, Goat anti-Chicken IgY (H + L) Secondary Antibody	Invitrogen	A11039	1:500	Goat

Assessment of capillary permeability

Animals were intravenously administered FITC-dextran (molecular weight 75 kDa, 40 mg/ml, Sigma, USA) at 4 days post treatment via tail vain. Five minutes later, cochlear stria vascularis was immediately harvested and fixed by 4 % paraformaldehyde for 30 min at room temperature. Mounted specimens were observed under a confocal microscope (ZEISS, LSM 900, Germany). Adobe Photoshop CS6 was used for adjustment of image contrast, superimposition of images, and colorization of monochrome fluorescence images.

Statistics

One- or two-way ANOVA was used, followed by planned comparisons. Data normality was tested by the D’Agostino–Pearson omnibus normality test when sample size was sufficient to do so. All measurements were taken from distinct samples and sample sizes were determined without any expectation of the effect size. All tests were performed using GraphPad Prism 10 (GraphPad Software, San Diego, CA, USA).

Results

CORM-2 reduces cisplatin-induced ototoxicity in BALB/c mice

Hearing in male 6-week-old BALB/c mice was evaluated prior to cisplatin and CORM-2 administration, using the auditory brainstem response (ABR) to determine baseline hearing sensitivity. Mice were randomly assigned to one of the following groups: control (saline-treated, i.p.), CORM-2 only (30 mg/kg, i.p., four doses as shown in Fig. 1A), cisplatin only (20 mg/kg, i.p., one dose), and CORM-2 + cisplatin, to determine whether cisplatin-based hearing impairment was alleviated by the CO donor CORM-2. ABR measurements were performed 4 days later. ABR threshold shifts were the differences between baseline (–1 day) and 4 days. Cisplatin treatment significantly modulated the threshold shift (main effect of treatment, F_3,240 = 77.38, p < 0.0001) but not the frequency (Fig. 1B, F4, 240 = 1.707, p = 0.1491). Control and CORM-2-treated mice showed no significant change in hearing sensitivity at any frequency (4, 8, 16, and 32 kHz). In contrast, cisplatin-treated mice (n = 20) exhibited significant increases in ABR thresholds across all frequencies (asterisks indicate significant differences between the cisplatin group and all other groups, *p < 0.05). Notably, mice that received both cisplatin and CORM-2 (Fig. 1B, purple line) exhibited significant decreases in the ABR threshold shifts at all frequencies, compared with cisplatin-treated mice (Fig. 1B, orange line). Thus, CORM-2 co-treatment significantly protected hearing sensitivity.

CORM-2 co-treatment reduces cisplatin-induced cochlear outer hair cell loss

After the completion of ABR measurements, the cochleae were removed and whole mounts were prepared. Tissues from cisplatin- (Fig. 2A) and cisplatin + CORM-2- (Fig. 2B) treated mice were subjected to immunofluorescent staining for myosin-VIIa (red) and phalloidin (green) to visualize sensory hair cells and β-actin, respectively. Inner hair cell (IHC) and outer hair cell (OHC) survival rates were measured. As shown in Fig. 2C, OHC survival was significantly affected by CORM-2 treatment (main effect of treatment, F_1,12 = 72.04, p < 0.0001); the frequency was also significantly affected by CORM-2 treatment (main effect of cochlear turns, F_2,12 = 70.64, p < 0.0001). Cisplatin-related OHC loss was not significant in the apical cochlear turn, but it was significant in both the middle (# p = 0.0003, apex:cisplatin vs. middle:cisplatin) and basal (# p < 0.0001, apex:cisplatin vs. base:cisplatin) turns. Mice treated with both cisplatin and CORM-2 exhibited better OHC survival in the middle (*p < 0.0001, middle:cisplatin vs. middle:cisplatin + CORM-2) and basal (*p = 0.0003, base:cisplatin vs. base:cisplatin + CORM-2) cochlear turns. However, IHC survival (Fig. 2D) was not affected by cisplatin or cisplatin + CORM-2 treatment (p > 0.05). Overall, consistent with the auditory data (Fig. 1), these findings indicated that CORM-2 was protective against cisplatin-induced hearing impairment and OHC loss.

Fig. 2.

Cisplatin-induced OHC loss is inhibited by CORM-2 co-treatment. (A and B) Whole-mount preparations of the auditory epithelium collected at 4 days post cisplatin- (A1–A3) or cisplatin + CORM2- (A4–A6) treated mice. Tissues were immunofluorescently stained for myosin-VIIa (red) to visualize the hair cells. Representative images are shown from the cochlear apex (A1 and A4), middle (A2 and A5) and basal (A3 and A6) turns. The three rows of outer hair cells (OHCs) and a single row of inner hair cells (IHCs) are indicated in each picture. Photographed using epifluorescence microscope. (C and D) Quantitative analysis of hair cell survival on OHCs (B) and IHCs (C). (B) Cisplatin-treated mice (orange) showed significant OHC death at middle and basal turns of cochlea (>∼16 kHz). CORM2 treatments (purple) were protective against cisplatin-induced OHC loss in the middle and basal turns. Scale bar = 50 μm. All graphs represent mean ± S.E.M. Two-way ANOVA, Tukey's multiple comparisons test. a, p < 0.05, main effect of treatment; b, p < 0.05, main effect of frequency; c, p < 0.05, main effect of interaction. n = 3. *p < 0.01, differs from matching cisplatin; # p < 0.01, differs from apex cell count of matching treatment. (C) IHCs remained intact in all cochlear turns of cisplatin- or cisplatin + CORM2-treated mice.

CORM-2 co-therapy reduces cochlear platinum accumulation

One mechanism of cisplatin-induced ototoxicity involves long-term cochlear retention of the drug [6]. To determine whether CORM-2 reduced cochlear accumulation of cisplatin, platinum levels were measured by inductively coupled plasma mass spectrometry in whole cochleae that had been collected 4 days after treatment. As expected, platinum levels were significantly elevated in cisplatin-treated mice (Fig. 3, orange, 3,230 ± 221.0 ng/g) compared with vehicle- or CORM-2-treated controls. CORM-2 co-therapy (Cis + CORM-2, purple, 2,486 ± 118.7 ng/g) significantly reduced platinum levels (by 23 %) in inner ear tissue (one-way analysis of variance with Tukey multiple comparisons test; F_3,57 = 92.89; p = 0.0033) at 4 days after cisplatin treatment. Thus, the protective effect of CORM-2 in terms of cisplatin-induced hearing impairment may be partly explained by reduced cochlear accumulation of cisplatin.

Fig. 3.

CORM-2 co-therapy reduces cisplatin accumulation in the mouse inner ear. Platinum levels were measured by ICP-MS in whole cochleae collected at 4 days post treatments. Platinum levels were elevated in cisplatin-treated mice (orange) relative to vehicle or CORM2-treated controls. CORM-2 co-therapy significantly reduced platinum levels in the inner ear. Mean ± SEM, n = 10, vehicle; n = 11, CORM2; n = 20, cisplatin; n = 20, cis + CORM2. One-way ANOVA with Tukey's multiple comparisons test. F_{(3, 57)} = 92.89; **p = 0.0033, ****p < 0.0001.

Cisplatin-induced SV hyperpermeability is inhibited by CORM-2 co-treatment

Vascular hyperpermeability is a feature of various pathological conditions including chronic inflammatory diseases, diabetes, neurological disorders, and retinal diseases [30]. Because the SV is the probable cochlear entry point of cisplatin [6], [31], we investigated whether cisplatin increased the endothelial cell barrier and whether CORM-2 co-therapy inhibited this effect. Mice were administered fluorescein isothiocyanate (FITC)-dextran (MW 75 kDa) at 4 days post-treatment via the tail vein. Five minutes later, the cochlear SV was removed and observed under a confocal microscope. Substantially increased extravasation of FITC-dextran fluorescence was evident in the cisplatin group (Fig. 4C) compared with the control (Fig. 4A) and CORM-2 (Fig. 4B) groups, indicating that cisplatin caused the SV to become “leaky.” Strikingly, CORM-2 co-treatment almost completely inhibited the cisplatin-induced hyperpermeability (Fig. 4D). Thus, CORM-2 co-treatment effectively protected the site of damage (i.e., the SV) by decreasing cisplatin-induced capillary leakage. CORM-2 co-therapy maintained the integrity of the endothelial barrier, which may help to prevent cochlear influx of cisplatin and subsequent accumulation.

Fig. 4.

Cisplatin-induced SV hyperpermeability is inhibited by CORM-2 co-therapy. Animals were administered FITC-dextran (MW 75 kDa) at 4 days post treatment via tail vain. Five minutes later, cochlear stria vascularis was dissected out and observed under confocal microscopy. Optical images showed markedly increased extravasation of FITC-dextran fluorescence at 4 days post cisplatin treatment (C) in comparison to a clear structure of microvessels in control (A) or CORM-2-alone (B). CORM-2 co-treatment reversed the cisplatin-based hyperpermeability (D). Scale bar, 50 μm.

CORM-2 co-treatment maintains SV component integrity

To determine how cisplatin and CORM-2 affected blood labyrinth barrier/vascular permeability, we targeted specific markers of endothelial cells (platelet endothelial cell adhesion molecule-1 [PECAM]), pericytes (NG-2), and perivascular-resident macrophage-type melanocytes (PVM/Ms; Iba-1). PECAM-1/CD31 immunofluorescence was used to visualize endothelial cells in the SV (Fig. 5A). Cisplatin induced a substantial increase in vessel thickness (Fig. 5, A3) compared with control or CORM-2 treated mice, but this increase was inhibited by CORM-2 co-treatment (Fig. 5, A4). PVM/Ms are SV-specific cells that are essential for the maintenance of blood-lymph barrier integrity via communication with endothelial cells and pericytes [32], [33]. Pericytes have key roles in the regulation of angiogenesis, blood flow, vascular integrity, and tissue fibrogenesis [34]. Immunofluorescent NG-2 (green) and Iba-1 (red) were observed on SV tissues collected 4 days after treatment (Fig. 5B). The untreated control exhibited long dense pericytes and PVM/Ms (Fig. 5, B1), whereas the cisplatin-treated cochlea displayed shorter and more numerous protruding pericytes and PVM/Ms (Fig. 5, B3). Evenly distributed rich populations of NG-2- and Iba-1-expressing cells were apparent in the CORM-2 co-treatment group (Fig. 5, B4). Thus, cisplatin significantly damaged endothelial cell, pericyte, and PVM/M structure and morphology; CORM-2 co-therapy prevented these effects.

Fig. 5.

Cisplatin exposure changes endothelial cells, pericytes and perivascular-resident macrophage-like melanocytes. (A) Endothelial cells of stria vascularis were stained with anti-PECAM antibody and imaged with a confocal microscope. Scale bar represents 50 µm. A marked increase in the vessel thickness was observed in the cisplatin group (A3), which was reversed by CORM-2 co-treatment (A4). (B) NG-2 (pericyte marker, green) and Iba-1 (PVM/Ms marker, red) was observed on the stria vascularis collected at 4 days post treatments. Untreated control (B1) and CORM-2 (B2) treatment contained lengthy and densely populated pericytes and PVM/Ms while cisplatin-treated (B3) cochlea displayed shorter and protrude pericytes and PVM/Ms. CORM-2 co-treatment group presents evenly distributed and a rich population of NG-2 and Iba-1 (B4). Scale bar represents 25 µm.

CORM-2 co-therapy reduces cochlear cisplatin-induced necroptosis

Because necroptosis contributes to cisplatin-based ototoxicity [24], [35], we next explored whether cisplatin activated necroptotic markers, along with the possible effects of CORM-2 on such activation. Whole cochlear lysates were collected at 4 days post-treatment; the expression levels of the necroptosis markers receptor-interacting serine/threonine-protein kinase (RIPK) 1 and 3 were examined (Fig. 6A). Cisplatin significantly increased the levels of RIPK1 and RIPK3 expression compared with control- or CORM-2-treated cochleae (Fig. 6A). Importantly, CORM-2 co-therapy significantly decreased the levels of RIPK1 and RIPK3, compared with the cisplatin-alone group. Quantitative data (Fig. 6, B–C) confirmed this finding.

Fig. 6.

CORM2 co-treatments reduce cisplatin-induced necroptosis in cochlea. (A–C) Necroptosis markers, RIPK1, RIPK3 and pSer358-MLKL, in the whole cochleae were analyzed by Western blot and quantified by ImageJ. Cisplatin-treated mice demonstrated significantly increased RIPK1 and RIPK3 expression compared to saline-treated mice (con). Cisplatin-induced RIPK1 and RIPK3 were reduced in cisplatin mice treated with CORM2. All graphs represent mean ± S.E.M. One-way ANOVA, Tukey's multiple comparisons test. n = 4. *p < 0.05. (D and E) Cochlear sections were double-stained with fluorescence-tagged antibodies against necroptosome, RIPK3 (D) and p-ser358-MLKL (E). CORM-2 co-treatments reduced cisplatin-induced RIPK3 and p-MLKL increase in the cochlea. (F) Whole mounts were double-stained with fluorescence-tagged antibodies against RIPK3 (green) and phalloidin (red) from apex, middle, and basal turns of cochlea. No significant RIPK3 expression was examined in saline- or CORM2-treated animals. Cisplatin-treated mice demonstrated significant RIPK3 expression in OHCs and IHCs at all turns of cochlea. CORM2 co-treatments reduced cisplatin-induced RIPK3 expression. Scale bar represents 50 μm. (G–J) Quantification of RIPK3+ cells/phalloidin+ cells were graphed. Cisplatin-treated mice (orange bar) showed a significant increase in RIPK3 expression in OHCs (G) and IHCs (H) at all turns of cochlea. CORM2 co-treatment (purple) significantly reduced cisplatin-induced RIPK3 expression in OHCs from apex, middle, and basal turns and of IHCs of basal turn. The number of RIPK3+ IHCs from middle and basal turns did not show significant differences The intensity of green fluorescence (RIPK3+) analysis recapitulated that CORM2 treatments alleviate cisplatin-induced RIPK3 expression in middle (I) and basal (J) turns of cochlea. No significant RIPK3 expression was examined in saline- or CORM2-treated animals. All graphs represent mean ± S.E.M. Two-way ANOVA (G and H), One-way ANOVA (I and J), Tukey's multiple comparisons test. a, p < 0.05, main effect of treatment; b, p < 0.05, main effect of frequency; c, p < 0.05, main effect of interaction. n = 3. *p < 0.05.

Immunohistochemical analysis of RIPK3 and the MLKL pseudokinase that acts downstream of RIPK3 confirmed that CORM-2 co-therapy reduced the cisplatin-related increases in necroptotic markers within sectioned cochleae (Fig. 6D and E). We also examined whole-mount samples. RIPK3 expression (green) was significantly increased by cisplatin; this effect was inhibited by CORM-2 co-treatment of cochlear hair cells (Fig. 6F). The ratios (cell counts) of RIPK3-positive cells (green) to phalloidin (beta-actin)-positive cells (red) were analyzed in OHCs (Fig. 6G) and IHCs (Fig. 6H). Saline- and CORM-2-treated mice did not express RIPK3 in the OHCs of all cochlear turns, or the apical and middle turns of IHCs. Cisplatin-treated mice exhibited significantly increased RIPK3 levels in the OHCs and IHCs of all cochlear turns. Notably, CORM-2 co-treatment significantly lowered cisplatin-induced RIPK3 expression in the OHCs of all cochlear turns and the basal turns of IHCs.

Fig. 6G and H shows only the numbers of cells positive for RIPK3 (green) and phalloidin (red), rather than the protein expression levels. Thus, we performed immunohistochemistry, then measured the levels of each protein in the IHCs of the middle and basal cochlear turns using ImageJ software (Fig. 6I and J). The intensity of green fluorescence (RIPK3+) confirmed that CORM-2 co-treatment alleviated cisplatin-induced RIPK3 expression in the middle (Fig. 6I) and basal (Fig. 6J) turns. Thus, cisplatin activates cochlear markers of necroptosis; CORM-2 co-treatment negates this effect.

CORM-2 co-treatment activates the plasma membrane repair system

Activation of the MLKL protein by RIPK3 is presumed to trigger plasma membrane disruption and induce necroptotic cell death [36]. However, recent studies have shown that the phosphorylated MLKL pseudokinase also enhances cell survival by acting as a “brake” at the final stage of “necroptosis execution” via facilitation of endosomal trafficking; it also acts upstream of the endosomal sorting complex required for transport (ESCRT)-III [21], [36], [37]. ESCRT machinery plays key roles in plasma membrane repair, membrane budding, and multivesicular body biogenesis [38], [39], [40]. Thus, high levels of ESCRT-III components impart resistance to cell death [41]. To explore whether CORM-2 co-therapy impacted plasma membrane integrity, we used fluorescence analysis to visualize the ESCRT III-associated protein known as charged multivesicular body protein 4B (CHMP4B). Fig. 7 shows that cisplatin significantly decreased CHMP4B levels in mouse cochlear sections; this effect was completely inhibited by CORM-2 co-treatment (Fig. 7A–F). These data suggest that high levels of major ESCRT-III components impart resistance to plasma membrane disruption; thus, CORM-2 co-treatment may increase cellular survival. Notably, although most CHMP4B in cisplatin-treated SVs was intracellular (Fig. 7D), most CHMP4B in the cisplatin + CORM-2 group was located around the plasma membrane. This finding suggests that CORM-2 co-treatment protects against cisplatin-induced ototoxicity partly by activating/stabilizing ESCRT-III components including CHMP4B; partly by repairing the cell membrane; and partly by shedding membrane regions damaged through toxic cisplatin-induced cellular stress responses.

Fig. 7.

Cisplatin treatment significantly decreased CHMP4B in mouse cochlea, which was reversed by CORM-2 co-treatment. ESCRT-III associating protein, charged multivesicular body protein 4B (CHMP4B) was immunostained using anti-CHMP4B antibody and DAPI. Cisplatin-treated mice demonstrated significantly decreased CHMP4B expression in the cochlea (A), organ of Corti (B), stria vascularis (D), SGN (E), and ligament (F), as compared to saline-treated mice (con). Cisplatin-induced CHMP4B reduction were recovered by cisplatin mice treated with CORM-2. CHMP4B expression in cochlea was quantified and graphed (C). All graphs represent mean ± S.E.M. One-way ANOVA, Tukey's multiple comparisons test. n = 4. *p < 0.05.

CORM-2 co-treatment reduces cisplatin-induced inflammation and apoptosis in the inner ear

To explore whether CORM-2 co-therapy alleviated other toxic cisplatin-induced cellular stress responses, multiple components of the inflammatory and apoptotic pathways were quantified by Western blotting in whole cochlear lysates that had been collected 4 days after cisplatin treatment. The levels of major regulators of inflammation (i.e., interleukin-1β [IL-1β] and hypoxia-inducible factor-1α [HIF-1α] [42]) were increased by cisplatin; CORM-2 co-treatment significantly reduced these levels (Fig. 8A–C), indicating that CORM-2 countered the inflammatory action of cisplatin in the cochlea.

Fig. 8.

CORM-2 co-treatment reduces cisplatin-induced inflammation and apoptosis in the whole cochlea lysate. (A) Cisplatin treatment increased pro-inflammatory cytokine, interleukin-1β (IL-1β) and hypoxia–inducible factor 1α (HIF-1α), which were stabilized by CORM-2 co-treatment by Western blot (A) and quantified by ImageJ (B and C). (D) Apoptosis markers, Bax, Bcl-2, caspse 7 and caspse 3,were analyzed by Western blot (D) and quantified by ImageJ (E–I). Cisplatin-treated mice showed significantly increased apoptosis marker expression relative to saline- or CORM2-treated mice. CORM-2 co-treatment (Cis + CORM2) reduced cisplatin-induced apoptosis. All graphs represent mean ± S.E.M. One-way ANOVA, Tukey's multiple comparisons test. n = 4. *p < 0.05.

To assess the roles of apoptosis pathways in the CORM-2-mediated effects, protein markers of the intrinsic and extrinsic apoptotic pathways were subjected to blotting and quantification. Cisplatin-treated cochleae exhibited increased apoptotic signals, which comprised decreases in B-cell lymphoma 2 (Bcl2), procaspase-7, and procaspase-3, as well as increases in Bcl-2-associated × protein (Bax) and cleaved caspase-7. Importantly, CORM-2 co-treatment completely inhibited these changes (Fig. 8D–I). CORM-2 exerted anti-inflammatory and anti-apoptotic effects on cisplatin-treated cochleae, partly explaining the observed CORM-2-induced protection against cisplatin-based ototoxicity.

Discussion

Over the past decade, interventions and strategies that treat and prevent cisplatin-induced ototoxicity have been extensively explored. Fernandez et al. reported that lovastatin reduced cisplatin-induced hearing impairment without decreasing cochlear accumulation of cisplatin in the cochlea [7]. Other antioxidants and anti-inflammatory drugs have demonstrated otoprotective effects in various preclinical models [43]. Multiple laboratory-based otoprotective approaches have been described; these include randomized controlled trials of intratympanic dexamethasone and N-acetylcysteine, as well as a multicenter randomized phase 3 clinical trial of sodium thiosulfate. The hearing losses were less than in cisplatin-only groups (reviewed and summarized in [43]). However, no intervention has been approved by the Food and Drug Administration. Considering the number of patients, both children and adults, negatively affected by the drug, there is an urgent need to prevent cisplatin-based ototoxicity. We explored whether CORM-2 could facilitate this prevention.

Cisplatin-induced ototoxicity is explained by long-term drug retention and toxic cellular stress responses [6], [7], [8], [15], [44], [45], [46]. Our data strongly suggest that CORM-2 addresses both problems. First, it decreases cochlear cisplatin levels (Fig. 3). Second, it protects against cisplatin-induced toxic cellular responses including high capillary leakage (Fig. 4), damage to pericytes and PVMs (Fig. 5), necroptosis (Fig. 6), plasma membrane disruption (Fig. 7), and apoptotic cell death (Fig. 8D). Third, it reverses the persistent inflammatory environment created by cisplatin (Fig. 8A). To our knowledge, this is the first report of a “cisplatin-reducing” agent active in the inner ear. Importantly, CORM-2 co-treatment effectively protected the SV by decreasing cisplatin-induced capillary leakage (Fig. 4), maintaining the integrity of SV components (Fig. 5), and preventing the cochlear influx of cisplatin (a toxic substance) and subsequent accumulation (Fig. 3).

Similar to the gasotransmitters NO and H₂S, CO is generated by heme oxygenases (HOs). CO elicits anti-tumor, anti-inflammatory, and anti-apoptotic responses [25]. Since CO treatment is already used in clinical settings as a chemo reagent [25], cisplatin and CO may synergistically suppress tumor progression and CO significantly decreases cisplatin-based ototoxicity, thereby enhancing quality of life. However, the results of preclinical studies regarding gasotransmitters differed in terms of the reported efficacies of tumor prevention [25]. The various effects of CO are attributable to the biphasic (bimodal or bell-shaped) pharmacological profile of CO [25]; accordingly, both the CO concentration and the model system must be carefully chosen. In vivo studies using CD1 athymic mice revealed that CORM-2 (35 mg/kg/day, i.p.) could attenuate the growth rate and the peritumor angiogenic response of pancreatic cancer cells (CAPAN-2/PaTu-8902 cells); the effects of injected CORM-2 were recapitulated by exposure to safe doses of CO (500 ppm 1 h/day) [47]. CORM-2 (30 mg/kg, i.p.) administered both before and after doxorubicin protected against doxorubicin-induced cardiotoxicity by reducing oxidative stress and apoptosis [26]. CORM-2 treatment partially inhibited neomycin-induced hair cell death in whole-organ cultures of utricles from adult mice. Thus, our CORM-2 dose, 30 mg/kg, appears to be safe in mice, and does not seem to reduce the therapeutic efficacy of cisplatin.

CORM-2 enhances barrier function and mitochondrial biogenesis. Joshi et al. reported that solid lipid nanoparticles containing CORM-2 were able to protect the blood-spinal cord barrier after spinal cord injury [48]. Niu et al. showed that CORM-2 maintained intestinal mucosal barrier integrity by reducing epithelial tight junction damage after cardiopulmonary resuscitation [49]. Notably, CORM-2 exerted cytoprotective effects on hepatocytes and liver tissues by activating nuclear translocation of the transcription factor-endoplasmic reticulum protein, which induces lysosomal and mitochondrial biogenesis [50]. Defective endothelial barrier function is apparent in many pathological conditions, including neurological disorders and chronic inflammatory diseases [30]. Because uncontrolled vascular hyperpermeability contributes to the progression of cisplatin-induced hearing impairment [51], our findings are particularly important in that CORM-2 is protective against cisplatin ototoxicity via maintenance of endothelial barrier function.

The MLKL protein has been reported to regulate necroptotic cell death by disrupting plasma membrane integrity, however, recent studies show that the protein functions as a “brake” at the final stage of “necroptosis execution” by acting upstream of ESCRT-III [20], [21], [22]. The ESCRT machinery plays key roles in plasma membrane repair, membrane budding, and multivesicular body biogenesis [38], [39], [40]. Activation of the ESCRT-III machinery is followed by membrane repair via removal of damaged parts of cell membranes, extracellular vesicle generation, inhibition of death, and extension of survival [20]. Our data show that CHMP4B (an ESCRT-III associating protein) expression was very low in the cisplatin-treated group but high in the cisplatin + CORM-2 group (Fig. 7, A-C). Furthermore, while most CHMP4B in cisplatin-treated SVs was intracellular (Fig. 7D), that in the cisplatin + CORM-2 group was located around the plasma membrane. Thus, CORM-2 co-treatment protects against cisplatin-induced ototoxicity by acting on the phospho-MLKL-regulated ESCRT system, repairing the cell membrane, and postponing the necroptosis induction.

CORM-2 co-treatment reduced the cisplatin level in the inner ear (Fig. 3). However, the underlying cellular and molecular mechanisms remain unknown. The CORM-2-induced decrease in platinum concentration may reflect decreased cochlear entry or increased removal/excretion. Cisplatin cellular uptake is presumed to involve both passive diffusion and membrane transporters. However, the existence of transporter/protein-mediated uptake has been questioned because of differences in the inner ear accumulations of cisplatin, oxaliplatin, and carboplatin [12], [45]; irreversible binding of platinum to the cysteine-rich copper transporter CTR1 protein [52]; and the observation that platinum uptake is unsaturable [52]. Our findings suggest that the cellular and molecular mechanisms underlying CORM-2-mediated protection against cisplatin ototoxicity are multifactorial in nature; CORM-2 acts through the pathways responsible for cellular cisplatin hypersensitivity; it also inhibits hyper-accumulation of cisplatin within the cochlea. Overall, our data demonstrate that CORM-2 co-treatment could be translated into clinical strategy to reduce cisplatin-induced hearing loss.

Our study had some limitations. First, we administered single high-dose cisplatin injections with a study interval of 4 days; previous studies used multiple low doses of cisplatin over long periods as similar to clinical practice [45], [53]. Second, we used only male mice. Others reported that the extent of cisplatin-induced ototoxicity did not differ between sexes, however, a difference was observed in the effectiveness of treatments [7]. Further studies involving female mice are required.

Conclusion

Cisplatin is an effective and affordable standard therapy for several solid tumors, however, it causes irreversible hearing loss in cancer survivors. There are no known cures or preventative treatments approved by FDA for cisplatin-based ototoxicity. Herein, we showed the potential for CORM-2 to reverse cisplatin-induced hearing impairment and to reduce cisplatin accumulation in the inner ear. To our knowledge, this was the first report of a “cisplatin-reducing” agent active in the inner ear. Importantly, CORM-2 co-therapy effectively protected against cisplatin-induced toxic cellular responses including high capillary leakage (Fig. 4), damage to pericytes and PVMs (Fig. 5), inflammation (Fig. 8A), necroptosis (Fig. 6), plasma membrane disruption (Fig. 7), and apoptotic cell death (Fig. 8D). These findings suggest that CORM-2 co-treatment provides a promising therapeutic strategy against cisplatin-induced hearing impairment.

Consent for publication

Not applicable.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Author statement

AL, hypothesized, designed, conducted experiments, analyzed data, and prepared figures; SJK, designed and conducted experiments, analyzed data, prepared graphical abstract; MJP, conceptualized, hypothesized, designed research studies, refined data analysis and interpretation, prepared figures, and wrote manuscript; Y-H.P., conceptualized and refined experimental design and interpretation, and edited manuscript.

Funding

This research was supported by National Research Foundation (NRF) of Korea grants to A-R. L. (2022R1C1C2007705); Y.-H.P. (NRF-2021R1A2B5B02001612); and M.J.P. (NRF-2020R1I1A3052557).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.