General caspase inhibitors are highly efficient anti-apoptotic small molecules in cisplatin-treated HEI-OC1 cells

Since the intrinsic apoptotic pathway is activated in cisplatin-treated cochlear cells, we screened small molecules that specifically interfere with this pathway for their suitability to serve as drugs for preventing or ameliorating cisplatin-induced ototoxicity (Fig. 1A) [14, 15]. In particular, we tested the caspase inhibitors Emricasan, Q-VD-OPh, and Ac-DEVD-CHO, as well as the BCL-XL-activating compound Muristerone A, in addition to the reference small molecules Z-DEVD-FMK and PFT-α, which were previously demonstrated to prevent cisplatin-induced apoptosis. To determine if the selected anti-apoptotic small molecules can prevent otic cells from cisplatin-induced cytotoxicity, we evaluated their effectiveness in vitro using the established HEI-OC1 cell line. HEI-OC1 cells are derived from the organ of Corti of the transgenic Immortomouse™ and express markers of supporting and HCs, suggesting them as a potential progenitor of both cell types [29]. Moreover, HEI-OC1 cells are sensitive to ototoxic drugs, making them a suitable HC-like in vitro model for investigating strategies to prevent ototoxicity [30].

Fig. 1

General caspase inhibitors Emricasan and Q-VD-OPh strongly alleviate cisplatin-mediated cytotoxicity in HEI-OC1 cells. A Schematic overview of the intrinsic apoptotic pathway, highlighting the targets of the anti-apoptotic small molecules employed in this study. The pathway is initiated by intrinsic lethal stimuli, which activate BH3-only proteins that activate the pro-apoptotic proteins BAK and BAX. BCL-XL, an anti-apoptotic protein, can prevent the continuation of the apoptotic pathway at this step. If p53 is involved in the apoptotic pathway, it activates pro-apoptotic proteins, leading to pore formation in the outer mitochondrial membrane, release of cytochrome c, and generation of the apoptosome with APAF1 (Apoptotic protease activating factor 1) and caspase-9. This initiates a cascade of caspase activation, followed by the final initiation of apoptosis. Muristerone A increases BCL-XL expression, while PFT-α inhibits p53. The general caspase inhibitors Emricasan and Q-VD-OPh can inhibit caspase-9 as well as the effector caspases caspase-3 and caspase-7, whereas Ac-DEVD-CHO and Z-DEVD-FMK specifically inhibit caspase-3 and caspase-7. B Annexin V and DAPI staining were used to analyze cell death at 72 h in cisplatin-treated (5 µM) and untreated (UT) HEI-OC1 cells. The performance of no inhibitor (NI) control cells was compared to cells to which the anti-apoptotic small molecules had been co-administered with the cisplatin. N = 3 independent experiments. The data in Fig. 1B are presented as mean ± standard deviation (SD) (P ≤ 0.05 (*), and P ≤ 0.001 (***), determined using one-way ANOVA together with Dunnett’s post-hoc test)

Application of the selected drugs to HEI-OC1 cells at pre-defined doses did not evoke any signs of cytotoxicity, as indicated by comparable cell death rates between untreated (UT) and small molecule-treated cells (Supplementary Fig. 1A). Therefore, these concentrations were used for all further experiments. To analyze the functionality of the caspase inhibitors, we induced apoptosis through cisplatin treatment and measured caspase-3/-7 activity to determine if the small molecules can reduce this activity in cisplatin-treated HEI-OC1 cells (Supplementary Fig. 1B). All tested caspase inhibitors could significantly reduce caspase-3/-7 activity and thus apoptosis, which was also reflected in reduced cell death compared to the cisplatin-treated no inhibitor (NI) control (Fig. 1B). For the cell death assay, cells were stained with Annexin V and DAPI to determine early apoptotic (Annexin V+), late apoptotic/dead (Annexin V+/DAPI+) and dead (DAPI+) HEI-OC1 cells. For Muristerone A, the reported feature of increased BCL-XL expression was confirmed by flow cytometry (Supplementary Fig. 1C), which contributed to the therapeutic effect of lower percentages of HEI-OC1 cells affected by cisplatin-induced cell death (Fig. 1B). PFT-α could reduce p53 phosphorylation seen in NI cells (Supplementary Fig. 1D and E) and could also lower cell death rates upon cisplatin treatment (Fig. 1B).

In summary, all tested anti-apoptotic small molecules showed molecular activity, and although they targeted different apoptosis mediators, they all achieved a therapeutic effect by reducing cisplatin-induced cytotoxicity in HEI-OC1 cells. However, since the general caspase inhibitors demonstrated the tendency of a slightly stronger reduction of cell death, we next focused on this group of inhibitors.

Emricasan alone is sufficient to prevent cisplatin-induced cytotoxicity in HEI-OC1 cells

The previous experiments showed that general caspase inhibitors may be promising compounds for reducing cisplatin-induced toxicity in otic cells. Since Emricasan has already been tested in clinical trials to treat certain liver diseases and was considered safe and well-tolerated, we focused on this general caspase inhibitor in the following experiments [31,32,33].

First, we investigated whether combining Emricasan with other anti-apoptotic small molecules targeting different proteins in the intrinsic apoptotic pathway could lead to additive effects. Therefore, we evaluated the efficacy in reducing cytotoxicity in cisplatin-treated HEI-OC1 cell cultures of Emricasan in combination with Muristerone A or with PFT-α, as well as of all three anti-apoptotic small molecules co-applied (Fig. 2). Annexin V/DAPI co-staining revealed reduced cell death in cisplatin-treated cultures treated with different small molecule combinations compared to NI, even though this effect was not significant (Fig. 2A). Emricasan alone was as efficient at reducing apoptosis in cisplatin-treated cultures as were the different drug combinations, demonstrating no additive effect of the tested molecules. We additionally measured dead-cell protease (Fig. 2B) and caspase-3/-7 (Fig. 2C) activity in these cultures. These parameters were significantly elevated in cisplatin-treated NI cells compared to UT cells. All small molecule combinations achieved significantly reduced levels of dead-cell protease and caspase-3/-7 activity compared to NI cells. However, the combinations were not superior to the Emricasan treatment alone. Strikingly, Emricasan treatment not only achieved a highly significant effect, but resulted in similar (dead-cell protease activity) or even lower (caspase-3/-7 activity) levels of cytotoxicity markers than seen in UT cultures.

Fig. 2

The combination of anti-apoptotic small molecules does not result in an additive effect to reduce cisplatin-induced cytotoxicity in HEI-OC1 cells. HEI-OC1 cells were left untreated (UT) or treated with cisplatin in the absence (No inhibitor, NI) or presence of the depicted small molecule combinations. Emricasan was combined with Muristerone A, PFT-α, or both small molecules and compared to the performance of NI control cells and cisplatin-treated Emricasan-only cells. A Percentage of early apoptotic (Annexin V+), late apoptotic/dead (Annexin V+/DAPI+), and dead (DAPI+) cells in the different HEI-OC1 cultures. B Analysis of the dead-cell protease activity using the ApoTox-Glo™ Triplex Assay. C Caspase-3/-7 activity determined using the ApoTox-Glo™ Triplex Assay. N = 3 independent experiments. The data is depicted as mean ± standard deviation (SD) (P ≤ 0.05 (*), P ≤ 0.01 (**), and P ≤ 0.001 (***), ns = non-significant, determined using one-way ANOVA together with Dunnett's post-hoc test). Cells were treated with 5 µM cisplatin for 72 h. The small molecules were added in the following concentrations: Emricasan = 10 µM, Muristerone A = 3 µM, PFT-α = 0.5 µM

In conclusion, no additive effect was observed when combining Emricasan with different anti-apoptotic small molecules to prevent cisplatin-induced cytotoxicity. Therefore, we next characterized Emricasan's effects on apoptotic processes in greater detail by analyzing additional hallmarks of apoptosis beyond caspase-3/-7 activation and Annexin V positivity [34, 35]. As one stringent hallmark, fragmented DNA was detected in cisplatin-treated HEI-OC1 cells by TUNEL staining (Fig. 3A). Strikingly, the addition of Emricasan strongly reduced the extent of DNA fragmentation. As additional evidence that Emricasan effectively inhibits cisplatin-induced apoptosis in HEI-OC1 cells, its addition significantly reduced PARP cleavage by caspase-3 compared to NI cells (Fig. 3B and C). We also assessed the confluency and morphology of cisplatin-treated HEI-OC1 cultures in the presence and absence of Emricasan as an indicator of cell viability. For this, live-cell imaging was performed. After 72 h of incubation, a confluent cell layer was observed in the UT and Emricasan-only treated cells. Notably, cisplatin treatment resulted in more dead cells in the supernatant and less confluent cells in the NI cultures. However, adding Emricasan to cisplatin-treated HEI-OC1 cells resulted in a more confluent cell layer than in the NI control cells, indicating a partial prevention of the cisplatin-induced decrease in confluency (Fig. 3D). The quantification of live-cell imaging data revealed increased confluency over time, with approximately 80% confluency at 72 h in both UT and Emricasan-only cells (Fig. 3E). NI control cultures reached approximately 20% confluency after 72 h, while the corresponding Emricasan co-treated cells showed approximately 40% confluency.

Fig. 3

Emricasan increases the viability of cisplatin-treated HEI-OC1 cells as evidenced by reduced apoptosis and increased confluency. The effects of Emricasan treatment were characterized by comparing untreated (UT) control, cisplatin-treated without Emricasan (No inhibitor, NI), and Emricasan plus cisplatin co-treated HEI-OC1 cultures. A Representative pictures from TUNEL assay to investigate DNA fragmentation, indicated by TUNEL-positive (pink) cells. All samples were counterstained with DAPI (blue). Scale bar = 20 µm. B Western blot analysis to investigate the expression of full-length (FL) PARP (116 kDa) and cleaved PARP (89 kDa) induced by caspase-3. Endogenous GAPDH (37 kDa) levels served as loading control. C Quantification of western blot signals showing the ratio of cleaved PARP to full-length PARP. D Representative microscopy pictures from the different cultures. Scale bar = 100 µm. E Live-cell imaging results using the CellCyte X™ to determine cell confluency over 72 h upon the administration of cisplatin in the presence or absence of Emricasan compared to UT cells. N = 3 independent experiments. The data is depicted as mean ± standard deviation (SD) (P ≤ 0.001 (***), determined using one-way ANOVA together with Dunnett's post-hoc test). Cisplatin (5 µM) and Emricasan (10 µM) co-treatment was performed over 72 h

In summary, Emricasan counteracted apoptotic hallmarks upon cisplatin administration in HEI-OC1 cells and increased cell confluency compared to NI cells.

One-time administration of Emricasan achieves prolonged beneficial effects in cisplatin-treated HEI-OC1 cells

Small molecules, including Emricasan (50 min), often have a short half-life in blood plasma [36]. Thus, for potential future clinical application, it is crucial to determine if the effects of a single administration continue even after metabolization of the drug or if the drug needs to be reapplied. To address this question in vitro, we analyzed the effects of a single Emricasan administration to reduce cisplatin-induced cytotoxicity in HEI-OC1 cells at multiple time points (Fig. 4). For this, HEI-OC1 cells were seeded on day 1, and cisplatin and Emricasan were co-administered on day 2 for 72 h (Fig. 4A). Annexin V and DAPI staining was performed on days 5, 8, and 12 to determine apoptosis and cell death. The analysis on day 5 confirmed that Emricasan treatment reduced cell death related to cisplatin treatment in HEI-OC1 cells compared to NI cells (Fig. 4B). Importantly, the analyses on day 8 (Fig. 4C) and day 12 (Fig. 4D) likewise demonstrated lower percentages of Annexin V+/DAPI+ cells in Emricasan-treated as compared to NI cultures, indicating that, despite its short serum half-life, a single Emricasan application is sufficient to achieve a prolonged alleviation of cisplatin-induced toxicity in vitro.

Fig. 4

The anti-apoptotic effect of Emricasan in cisplatin-treated HEI-OC1 cells is prolonged. A Schematic overview of the experimental layout. The experiment started on day 1 with cell seeding, followed by Emricasan and cisplatin co-administration on day 2, with the drugs left on the cells until day 5. Cell death analyses were performed on days 5, 8, and 12. Untreated (UT) cells and cisplatin-only (No inhibitor, NI) treated cells served as controls. B–D Analysis of HEI-OC1 cell death by determination of the percentage of Annexin V+ (early apoptotic), Annexin V+/DAPI+ (late apoptotic/dead), and DAPI+ (dead) cells in untreated (UT), cisplatin only- (NI), and Emricasan co-treated cultures. N = 3 independent experiments. The data is depicted as mean ± standard deviation (SD) (P ≤ 0.05 (*), ns = non-significant, determined using one-way ANOVA together with Dunnett's post-hoc test). Cisplatin treatment: 5 µM for 72 h. Emricasan concentration: 10 µM

Emricasan significantly reduces cell death in cisplatin-treated phoenix auditory cells and primary SGN

As cisplatin damages not only HCs but also SGNs, we next analyzed the effect of Emricasan in neuronal in vitro models, i.e., using phoenix auditory cells and primary rat SGN cultures. The phoenix auditory neuroprogenitors are derived from the A/J mouse cochlea and form spheres of different forms and sizes when cultured in a proliferation medium containing growth factors [37, 38]. In contrast to UT and Emricasan-only cells, which showed normal morphology, the NI culture showed less sphere formation, with more single or dead cells as well as more cell debris after 72 h of cisplatin treatment (Fig. 5A). Notably, cisplatin-treated cultures that received Emricasan appeared healthy and still formed spheres of varying sizes.

Fig. 5

Emricasan significantly reduces cisplatin-induced toxicity in auditory neuronal cells. Representative photographs showing the morphology of phoenix auditory neuroprogenitors (A) and phoenix auditory neurons (B) in untreated (UT), cisplatin-treated (No inhibitor, NI), Emricasan-only, and cisplatin-treated Emricasan cells. Scale bar = 100 µm. C Analysis of the percentage of Annexin V+ (early apoptotic), Annexin V+/DAPI+ (late apoptotic/dead), and DAPI+ (dead) cells in phoenix auditory neuroprogenitor (left), phoenix auditory neuron (middle), and primary SGN (right) cultures that were untreated NI, Emricasan- and cisplatin co-treated. N = 3 independent experiments. The data is depicted as mean ± standard deviation (SD) (P ≤ 0.05 (*), P ≤ 0.01 (**), and ns = non-significant, determined using one-way ANOVA together with Dunnett's post-hoc test). Cisplatin treatment: 5 µM for phoenix auditory neuroprogenitors and 20 µM for phoenix auditory neurons or primary SGN for 72 h. Emricasan was applied at a dose of 10 µM

The phoenix auditory neuroprogenitors can differentiate into phoenix auditory neurons with a bipolar morphology (see UT and Emricasan-only cultures, Fig. 5B) upon replacement of the growth factors in the culture medium with neurotrophic factors. NI phoenix auditory neurons displayed fewer protrusions and more dead cells after cisplatin application. In contrast, cisplatin-treated phoenix auditory neurons co-treated with Emricasan exhibited a morphology similar to UT controls.

Emricasan’s beneficial effect was confirmed by cell death analyses on the phoenix auditory neuroprogenitor (Fig. 5C, left graph) and neuron cultures (Fig. 5C, middle graph). Strikingly, Emricasan significantly reduced the percentage of Annexin V+/DAPI+ cells in cisplatin-treated cultures to a level comparable to the UT control cells. Similar results were obtained in dissociated primary rat SGN cultures, in which Emricasan significantly reduced the percentage of Annexin V+ and/or DAPI+ cells compared to NI cultures (Fig. 5C, right graph).

These results suggest that Emricasan can protect neuronal cell types from cisplatin-induced cytotoxicity.

Emricasan provides greater protection from cisplatin-induced toxicity than STS in neuronal cells

As the FDA approved the antioxidant STS as a preventative treatment regimen to protect pediatric cancer patients from cisplatin-induced ototoxic effects, we next aimed to compare the protective effects of Emricasan to STS. Since previous in vitro studies used a broad range of STS concentrations (15 µg/mL to 2 mg/mL), we included both a high (STShigh; 2 mg/mL) and a low concentration of STS (STSlow; 25 µg/mL). The high STS concentration reduced ROS generation in cisplatin-treated HEI-OC1 cells (Supplementary Fig. 2A). Moreover, STShigh achieved preservation of mitochondrial membrane potential in cisplatin-treated cultures as evidenced by the formation of JC-1 aggregates, which were reduced in NI counterparts, indicating healthy mitochondria and, thus, vital cells in the presence of STShigh (Supplementary Fig. 2B). Additionally, STShigh decreased cell death in HEI-OC1 cells following cisplatin treatment (Supplementary Fig. 2C). Emricasan also reduced ROS generation, increased the percentage of JC-1 aggregates, and enhanced the viability in cisplatin-treated HEI-OC1 cultures. In direct comparison, STShigh was significantly more effective, bringing almost all parameters tested to levels as observed in UT cells. Interestingly, the combination of Emricasan and STShigh was most effective in reducing the occurrence of Annexin V+ and/or DAPI+ cells.

Compared to STSlow, Emricasan protected HEI-OC1 cells from cisplatin-induced death to a similar extent (Fig. 6A). However, in the phoenix auditory cells and primary SGN, Emricasan showed slightly stronger effects than STSlow (Fig. 6B–D). The combination of Emricasan and STSlow was slightly more effective at reducing cisplatin-induced toxicity in HEI-OC1 cells and phoenix auditory cells, suggesting a mild additive effect of the two small molecules.

Fig. 6

The reduction of cisplatin-induced cytotoxicity in neuronal cell types through Emricasan is superior to STSlow. The indicated cell culture types were left untreated (UT), treated with cisplatin only (No inhibitor, NI) or co-treated with cisplatin and Emricasan (10 µM) and/or STSlow (25 µg/mL). A–D Cell death analysis based on determination of the percentage of early apoptotic (Annexin V+), late apoptotic/dead (Annexin V+/DAPI+), and dead (DAPI+) HEI-OC1 cells (A; 5 µM cisplatin), phoenix auditory neuroprogenitors (B; 5 µM cisplatin), phoenix auditory neurons (C; 20 µM cisplatin), and primary SGNs (D; 20 µM cisplatin). N = 3 independent experiments. The data is depicted as mean ± standard deviation (SD) (P ≤ 0.05 (*), P ≤ 0.01 (**), and ns = non-significant, determined using one-way ANOVA together with Dunnett’s post-hoc test)

In conclusion, the performance of STS is concentration-dependent, with STShigh slightly more effective at reducing cisplatin-induced toxicity in HEI-OC1 cells as compared to Emricasan, but with Emricasan being as effective as STSlow in HEI-OC1 cells and, importantly, being superior to STSlow in the neuronal cell cultures.

The beneficial effect of Emricasan appears superior to that of STS in neomycin-treated cells

Because cisplatin and aminoglycosides, like neomycin, cause ototoxicity through similar mechanisms, such as increasing ROS generation and inducing apoptosis of cochlear cells, we next aimed to investigate whether Emricasan and/or STS could be beneficial for patients receiving aminoglycoside treatment. Hence, we tested Emricasan and both concentrations of STS either alone or in combination in different types of neomycin-treated cells (Fig. 7 and Supplementary Fig. 3). Strikingly, in HEI-OC1 cells (Fig. 7A), phoenix auditory neuroprogenitors (Fig. 7B), phoenix auditory neurons (Fig. 7C), and primary SGNs (Fig. 7D), Emricasan alone was found to be more effective at preventing cell death than STSlow. Combining STSlow and Emricasan in HEI-OC1 cells and phoenix auditory neurons resulted in a slightly more efficient reduction of Annexin+/DAPI+ cells upon neomycin treatment than Emricasan alone.

Fig. 7

The reduction of neomycin-induced cytotoxicity in otic cell types through Emricasan is superior to STSlow. The indicated cell culture types were left untreated (UT), treated with cisplatin only (No inhibitor, NI), or co-treated with cisplatin and Emricasan (10 µM) and/or STSlow (25 µg/mL). A–D Cell death analysis based on determination of the percentage of early apoptotic (Annexin V+), late apoptotic/dead (Annexin V+/DAPI+), and dead (DAPI+) HEI-OC1 cells (A), phoenix auditory neuroprogenitors (B), phoenix auditory neurons (C) and primary SGNs (D) upon treatment with 1 mM neomycin for 72 h (A–C) or 48 h (D). N = 3 independent experiments. The data is depicted as mean ± standard deviation (SD) (P ≤ 0.05 (*), P ≤ 0.01 (**), P ≤ 0.001 (***), and ns = non-significant, determined using one-way ANOVA together with Dunnett’s post-hoc test)

Remarkably, the higher concentration of STS failed to reduce neomycin-induced toxicity in HEI-OC1 cells (Supplementary Fig. 3). Consequently, STShigh could not reduce ROS generation (Supplementary Fig. 3A) or maintain the mitochondrial membrane potential (Supplementary Fig. 3B). Moreover, the administration of STShigh did not reduce cell death rates in neomycin-treated HEI-OC1 cells (Supplementary Fig. 3C). In contrast, Emricasan significantly decreased the percentage of Annexin V+/DAPI+ cells compared to STS cultures, suggesting that Emricasan holds promise as a potential therapeutic drug candidate to prevent aminoglycoside-induced ototoxicity.