All animal experiments examined in the present study were approved by the Mie University Review Board for Animal Investigation (no. 2020–19) and conformed to the ethical guidelines established by the committee. The clinical study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Mie University Graduate School of Medicine and Faculty of Medicine (no. 2021–170). Informed consent was obtained from participants through an opt-out method, because the data were collected retrospectively from electronic medical records.

Yo-Pro-1 (Catalog number: Y3603) was purchased from Invitrogen (Tokyo, Japan). CDDP, cimetidine (CMD), and LPZ were purchased from Tokyo Chemical Industry Co., Ltd (Tokyo, Japan). Stock solutions of Yo-Pro-1 and LPZ were prepared by dissolving in dimethyl sulfoxide. Other chemicals were dissolved in 0.3 × Danieau’s solution (19.3 mM NaCl, 0.23 mM KCl, 0.13 mM MgSO₄, 0.2 mM Ca(NO₃)₂, and 1.7 mM HEPES, pH 7.2). These solutions were diluted in 0.3 × Danieau’s solution to obtain the indicated concentrations. Lissamine-labeled control morpholino with no known target gene was purchased from Gene Tools (Philomath, OR, United States) and diluted with distilled water to prepare the stock solution.

Zebrafish were maintained as described previously (Nishimura et al., 2016). Briefly, zebrafish were raised at 28.5 ± 0.5°C with a 14 h/10 h light/dark cycle. Embryos were obtained via natural mating and cultured in 0.3 × Danieau’s solution.

There are two oct2 genes (zgc:64,076 and zgc:175,176) in zebrafish according to the database of genetic and genomic data specialized for zebrafish (https://zfin.org/). CRISPR RNA (crRNA) targeting zgc:64,076 or zgc:175,176, trans-activating crRNA (tracrRNA), and recombinant Cas9 protein were obtained from FASMAC (Kanagawa, Japan) (sequences shown in Supplementary Table S1). Crispant zebrafish for zgc:64,076 or zgc:175,176 were generated according to methods described previously (Adachi et al., 2021). Briefly, the larvae exhibiting bright fluorescence of lissamine were selected at 1 day-post fertilization (dpf). Genomic DNA was extracted from zebrafish embryos injected with crRNA targeting zgc:64,076 or zgc:175,176 (crispants) or without crRNA (wild type, WT). Short fragments of the zgc:64,076 and zgc:175,176 gene encompassing the sites targeted by crRNAs were amplified from genomic DNA using the primers shown in Supplementary Table S1. Three-step PCR was carried out, followed by electrophoresis in 10% polyacrylamide gels (FUJIFILM Wako Pure Chemical, Osaka, Japan) and visualization by ethidium bromide staining.

Several studies have demonstrated that the loss of sensory hair cells can be caused by CDDP exposure within 4 h at 1,000 μM and 24 h at 50 μM (Ou et al., 2007; Domarecka et al., 2020). In this study, we wanted to analyze the effects of protectants at relatively early stages of CIO. Zebrafish larvae at 5 dpf were incubated in 0.3× Danieau’s solution containing CDDP (0, 50, 100, 250 μM, 500 μM, or 1,000 μM) alone for 2 h or co-treated with CDDP (0 or 250 μM) and CMD (0 or 1 mM) or LPZ (0, 0.1, 0.25, or 0.5 μM) for 2 h.

Sensory hair cells of zebrafish areanalogous to those of mammals (Lush and Piotrowski, 2014). Sensory hair cells express purinergic receptor P2X ligand-gated ion channel 7 (P2X7), which is activated by ATP secreted from supporting cells under physiological conditions (Zhao et al., 2005). Yo-Pro-1, a fluorescent cyanine dye, can permeate the cells via activated P2X7 (Virginio et al., 1999). Therefore, sensory hair cells of zebrafish can be visualized by vital staining with Yo-Pro-1 (Santos et al., 2006; Sasagawa et al., 2016). In this study, zebrafish larvae at 5 dpf were immersed in medium containing 2 μM of Yo-Pro-1 for 30 min. After staining, zebrafish were transferred to 0.3 × Danieau’s solution containing 2-phenoxyethanol (500 ppm) to be anesthetized and then transferred to glass slides. A few drops of 3% low-melting agarose solution were placed over the larvae, and the zebrafish were immediately oriented sideways dorsal-side-up. The zebrafish were then observed using an epifluorescence microscope (SZX7, Olympus, Tokyo, Japan) with the SZX-MGFP filter (Ex 475/30 nm, Em 510 nm). The 16-bit fluorescence images were converted to 256-gray-scale (0–255) images in Fiji (Schindelin et al., 2012). Considering the previous studies (Chiu et al., 2008; Sasagawa et al., 2016), we measured the fluorescence signal of three neuromasts (MI1, MI2, and O2) (Raible and Kruse, 2000). Neuromasts are composed of different types of cells, including hair cells and supporting cells, which are important for hearing and balance in zebrafish (Froehlicher et al., 2009). The sum of the fluorescent areas of these neuromasts (shown in small rectangles in each figure) with signals over the threshold (25 in the 256-gray scale) was calculated and reported as the fluorescent area of hair cells in a zebrafish.

Zebrafish were harvested at 5 dpf and stored in RNA-later (Applied Biosystems, Foster City, CA, United States). Total RNA was extracted using an RNeasy Mini kit (Qiagen, Germantown, MD, United States) according to the manufacturer’s protocol. cDNA was generated using a ReverTra Ace qPCR RT Kit (Toyobo, Osaka, Japan) and qPCR was performed using QuantStudio3 real-time PCR machine (Applied Biosystems, Thermo Fisher, United States) with THUNDERBIRD SYBR qPCR Mix (Toyobo). The thermal cycling conditions were 95°C for 60 s, followed by 40 cycles of 95 °C for 15 s, 60°C for 15 s, and 72°C for 45 s. Expression of zgc:64,076 and zgc:175,176 mRNA was normalized to that of the eukaryotic translation elongation factor 1 alpha 2 (ef1a) to correct for variability in the initial template concentration and reverse transcription efficiency. qPCR primer sequences are shown in Supplementary Table S1.

Data on patient demographics and administration information, drug/biologic information, and adverse events from July 2014 to December 2020 were obtained from the FAERS database. Duplicate reports were excluded in accordance with the FDA recommendations. Data analyses were performed using ACCESS^® 2019 software (Microsoft, Redmond, WA). Data associated with CDDP administration were extracted. Disease names were defined using the Medical Dictionary for Regulatory Activities (MedDRA/J) version 24.0. The standardized MedDRA Query (SMQ) was used for searching for hearing impairment (SMQ code: 20,000,171). The effect of PPIs on CDDP-associated hearing impairment was evaluated using the reporting odds ratio (ROR). To calculate the ROR, CDDP-associated hearing impairment and all other reported adverse events associated with CDDP were defined as “cases” and “non-cases,” respectively. The RORs were calculated from two-by-two contingency tables of counts with or without PPI. RORs were expressed as point estimates with 95% confidence interval (CI). Although stratification based on age, gender, and primary diseases can be done using the FAERS database (Nagashima et al., 2016; Hosoya et al., 2022), we did not perform the stratified analysis due to the small number of the CIO cases treated with PPIs.

Clinical data were collected from EHRs of patients who received CDDP in the Department of Otorhinolaryngology-Head and Neck Surgery or the Department of Medical Oncology of Mie University Hospital during January 2010 and April 2021. Among 463 hospitalized adult patients treated with CDDP, 340 patients treated with ≥300 mg of total CDDP dose for cancers in the lung, esophagus, stomach, kidney, mammary glands, thymus, and reproductive tissues were enrolled in this study. Patients were excluded if they were defined as having hearing loss before CDDP treatment (n = 5), ear disease (n = 6), prescribed aminoglycoside antibiotics (n = 1), and PPIs prescribed other than LPZ (n = 39). Ototoxicity was defined as hearing loss, sensorineural hearing loss, or tinnitus following CDDP treatment.

The results are expressed as the mean ± standard deviation. Differences between two groups were assessed using unpaired t-tests. Statistical comparisons for multiple groups were performed using one-way analysis of variance followed by Dunnett’s test or Tukey’s multiple comparison test. Statistical comparisons between two groups in clinical data were performed using the Mann-Whitney U test and Fisher’s exact test for continuous and categorical variables, respectively. All statistical analyses were performed using GraphPad Prism 9 (GraphPad Software Inc. San Diego, CA). The significance was established at a p value < 0.05.

We first evaluated the ototoxicity of CDDP in zebrafish by in vivo fluorescence imaging of sensory hair cells stained with Yo-Pro-1. The in vivo imaging revealed that the fluorescence signals of hair cells in zebrafish were decreased dependent on the dose of CDDP (Figures 1A,B). Because the fluorescence of hair cells in zebrafish treated with 250 μM CDDP was around 50% (Figure 1B), we selected this dose to examine the otoprotective effects of CMD and LPZ against the ototoxicity of CDDP.

Previous studies have demonstrated that CMD inhibits OCT2 and suppresses both CIO and CIN in rodents (Okuda et al., 1996; Ciarimboli et al., 2010). CMD is a substrate of OCT2 and works as a competitive inhibitor of other OCT2 substrates (Urakami et al., 1998; Harper and Wright, 2013; Yin et al., 2016). Consistent with these studies, co-treatment with 1 mM of CMD significantly suppressed the decrease of fluorescence signals in hair cells in zebrafish treated with 250 μM of CDDP (Figures 1C,D).

We then examined the effect of LPZ on the decrease of fluorescence signals in hair cells in zebrafish treated with 250 μM of CDDP. As shown in Figures 1D,E, LPZ dose-dependently inhibited the decrease of fluorescence signals in hair cells in zebrafish treated with 250 μM of CDDP. These results suggest that LPZ ameliorates CIO through inhibition of OCT2.

To examine the role of oct2 in CIO in zebrafish, we generated two crispant zebrafish with KO of zgc:64,076 or zgc:175,176, zebrafish homologs of OCT2. To KO zgc:64,076, we designed crRNA targeting the sequence including the protospacer adjacent motif located behind the start codon in the first exon, and then injected the crRNA with tracrRNA and Cas9 protein into zebrafish embryos. At 1 dpf, the genomes were extracted, and PCR analysis was performed using primers to amplify the genomic regions around the sequences targeted by the crRNA. As shown in Figure 2A, extra bands (shown in red bracket) were observed below the main PCR products (black arrowhead) in the crispant but not WT zebrafish, suggesting that KO of zgc:64,076 were generated in the crispant zebrafish. There were no external abnormalities in the zgc:64,076 crispant zebrafish (Figure 2B). In vivo fluorescence imaging of zebrafish hair cells revealed that CDDP treatment caused a decrease of fluorescence signals in hair cells in WT but not zgc:64,076 crispant zebrafish (Figures 2C,D).

We also generated zgc:175,176 crispant zebrafish using CRISPR-Cas9 technology. Genomic PCR analysis confirmed KO of zgc:175,176 in the crispant zebrafish (Figure 2E). No external abnormalities were observed in the zgc:175,176 crispant zebrafish (Figure 2F). In vivo fluorescence imaging of zebrafish hair cells, however, revealed that CDDP treatment caused a decrease in fluorescence signals in hair cells in both WT and zgc:175,176 crispant zebrafish (Figures 2G,H). These results are consistent with the result showing that the expression of zgc:64,076 was significantly higher than that of zgc:175,176 in zebrafish at 5 dpf (Supplementary Figure S1) and suggest that zgc:64,076 is mainly involved in the CDDP-induced hair cell damage.

To examine the effect of LPZ on the incidence of CIO in humans, we analyzed the FAERS database that has been successfully used for drug repositioning (Zamami et al., 2021). We extracted 29,976 patients who experienced ototoxicity in relation to CDDP treatment. We then analyzed the reporting ratio of CIO, the ROR, and the 95% CI (Table 1). The reporting ratio of CIO in patients with LPZ (1.0%) was significantly lower than that in patients without LPZ (2.7%, p = 0.039). The ROR of CIO in patients treated with LPZ was 0.37 (95% CI: 0.14–0.99). In addition, the reporting ratio of CIO in patients treated with PPIs other than LPZ (i.e., omeprazole and esomeprazole) was also significantly lower than in that in patients treated without these PPIs (Table 1), suggesting that the protective effects of LPZ against CIO may be a class effect of PPIs.

We then validated the effect of LPZ on the incidence of CIO using the EHR of Mie University Hospital. According to the inclusion and exclusion criteria, we enrolled 289 patients in this study. Table 2 shows a comparison of patient characteristics between patients with ototoxicity and without ototoxicity. The rate of concomitant LPZ in patients without ototoxicity (88 out of 260 patients, 34%) was significantly higher than that in patients with ototoxicity (2 out of 29 patients, 7%, p = 0.002). Taken together, these results suggest that LPZ may be protective against CIO in humans.