Identification of small molecule inhibitors of cisplatin-induced hair cell death: results of a 10,000 compound screen in the zebrafish lateral line

Abstract

Hypothesis: The zebrafish lateral line can be used to identify small molecules that protect against cisplatin-induced hair cell death.

Background: Cisplatin is a commonly used chemotherapeutic agent, which causes hearing loss by damaging hair cells of the inner ear. There are currently no FDA-approved pharmacologic strategies for preventing this side effect. The zebrafish lateral line has been used successfully in the past to study hair cell death and protection.

Methods: In this study, we used the zebrafish lateral line to screen a library of 10,000 small molecules for protection against cisplatin-induced hair cell death. Dose-response relationships for identified protectants were determined by quantifying hair cell protection. The effect of each protectant on uptake of a fluorescent cisplatin analog was also quantified.

Results: From this screen, we identified 2 compounds exhibiting dose-dependent protection: cisplatin hair cell protectant 1 and 2 (CHCP1 and 2). CHCP1 reduced the uptake of a fluorescent cisplatin analog, suggesting its protective effects may be due to decreased cisplatin uptake. CHCP2 did not affect uptake, which suggests an intracellular mechanism of action. Evaluation of analogs of CHCP2 revealed 3 additional compounds that significantly reduced cisplatin-induced hair cell death, although none exceed the effectiveness or potency of the parent compound.

Conclusion: The zebrafish lateral line was used to identify 2 small molecules that protected against cisplatin-induced hair cell death.

Fig. 1

CHCP 1 and 2 cause concentration-dependent reduction in cisplatin-induced hair cell death. A) Chemical structures of CHCP1 and CHCP2. Dose-response functions for co-treatment with 50 μM cisplatin and either (B) CHCP1 or (C) CHCP2 for 24 h, demonstrating an increase in hair cell survival which reached significance at a concentration of 20 μM of the protectant (P < 0.05 for CHCP1 and P < 0.001 for CHCP2, Tukey-Kramer post-test). For all treatment groups n = 7-13 fish. Error bars = ±1 SD; ***P < 0.001, *P < 0.05, ^nsP > 0.05 by one-way ANOVA and Tukey-Kramer post-test.

Fig. 2

The protective effects of CHCP 1 and 2 are only partially maintained with increased cisplatin concentration. A) Co-treatment with 50 μM CHCP1 results in a significant increase in hair cell survival at cisplatin concentrations up to 50 μM (P < 0.01, Tukey-Kramer post-test). B) Co-treatment with 50 μM CHCP2 results in a significant increase in hair survival at cisplatin concentrations up to 100 μM (P < 0.001, Tukey-Kramer post-test). For all treatment groups n = 9-12 fish. Error bars = ±1 SD; ***P < 0.001, **P < 0.01, *P < 0.05, by one-way ANOVA and Tukey-Kramer post-test.

Fig. 3

Treatment with CHCP1, but not CHCP2, reduces hair cell uptake of Rho-Pt. A) Quantified Rho-Pt fluorescence after 1 h of treatment with the indicated protective compound normalized to untreated controls. Rho-Pt uptake is significantly reduced with CHCP1 treatment but unchanged with CHCP2. B) Representative neuromasts which received pre- and co-treatment with a protective compound or DMSO (control) as indicated. Rho-Pt uptake labeling of hair cells is reduced with CHCP1 compared with DMSO control and CHCP2. Scale bar for all figures = 10 μm. Error bars = ±1 SD; n = 22-35 neuromasts per treatment group; ^nsP > 0.05, *P < 0.05, by one-way ANOVA and Tukey-Kramer post-test.

Fig. 4

Effect of CHCP1 and CHCP2 on cytotoxic activity of cisplatin. Cytotoxic activity of cisplatin is maintained in the presence of protective compounds at some doses of cisplatin, but attenuated at others. A) CHCP1 demonstrated a small but significant inhibition of cisplatin toxicity in A549 cancer cells at 10 and 20 μM cisplatin (P < 0.001, one-way ANOVA). At 5 μM cisplatin there was no significant inhibition of tumor cell kill. CHCP1 alone (0 μM cisplatin) independently had significant toxicity to tumor cells. B) CHCP2 also demonstrated a small but significant inhibition of cisplatin toxicity in A549 cancer cells at 10 and 20 μM cisplatin (P < 0.001, one-way ANOVA). At 5 μM cisplatin there was no significant inhibition of tumor cell kill. CHCP2 alone (0 μM cisplatin) independently had significant toxicity to tumor cells. Similar results were obtained when these experiments were repeated with NCI-H23 cells (data not shown). Each condition was performed in quadruplicate. Data points are the mean A549 cancer cell survival as determined by ATP cell luminescence assay. Error bars = ±1 SD.

Fig. 5

Analogs of CHCP2 protect against cisplatin-induced hair cell death. A) Structures of CHCP2 and CHCP2 analogs. B) At a concentration of 20 μM, all three CHCP analogs demonstrate significant protection against cisplatin-induced hair cell death compared to DMSO treated controls. CHCP2.1 and 2.2 exhibit significantly less protection than CHCP2, while 2.3 achieves similar protection as CHCP2. For all treatment groups n = 8-11 fish. Error bars = ±1 SD; ***P < 0.001, *P < 0.05, ^nsP > 0.05 by one-way ANOVA and Tukey-Kramer post-test.