OBJECTIVETo investigate the ototoxicity of co-administration of kanamycin and furosemide in mouse and establish a reliable model to induce a sensorineural hearing loss.

METHODSCBA/J mice strain was selected, with the age around 3-4 weeks old, to be received a single subcutaneous injection of kanamycin at dose of 1 g/kg and another single intraperitoneal injection of furosemide at dose of 0.4 g/kg 30 - 45 min afterward. The auditory brainstem response (ABR) threshold shift was tested. The series of experimental methods including propidium iodide, phalloidin staining, semithin section toluidine blue staining, TUNEL, scanning electron microscopy and transmission electron microscopy were applied to observe the characteristics of the lesion of cochlea and hair cells. The time course was set as following: before injection, 12, 24, 48 hours and 1, 2, 4, 12 weeks after injections, respectively.

RESULTSThe ABR threshold shift was firstly presented a significant increase at 12 h after injection at 2, 4, 8 kHz, then the ABR threshold kept going up during next 36 h until it was presented a stable level around 90 dB. Pathological examination showed an absence of outer hair cells at basal turn rapidly since 12 h after treatment, and then by 48 h the most commonly observed lesion, where all outer hair cells throughout the length of the cochlea were killed, in the contrast, however, the inner hair cells loss were delayed and mild. TUNEL-positive nuclei demonstrated that most hair cells died via an apoptotic pathway. In scanning electron microscopy abundance of necrotic outer hair cells were detected by 24 h after treatment, in which reticular lamina were collapsed. Then all outer hair cells were replaced by expansion of heads of supporting cells. At 48 h after treatment, marginal cells presented a swollen and some of them were observed to be fused. In addition, spherical cell extrusion appeared to leak out from some marginal cells. By 2 weeks, nearly all microvillus were lost and marginal cells presented a shape of stone-like change. A significant and progressive decrease in strial vascularis thickness was found, of which the reason probably related with a reduction in volume of marginal cells.

CONCLUSIONThis systemic protocol eliminates hair cells extensively in vivo, and it could be a reliable model to examine different aspects of cochlear pathology in transgenic or mutant mice strains.