The 309th Hospital Animal Care and Use Committee approved the use of animal models in the present study. Cx26loxP/loxP Pax2Cre mice were generated by crossbreeding Cx26loxP/loxP and Pax2Cre mice. Both transgenic mice were obtained from the laboratory of Dr. Xi Lin. The genotype of the mice was defined via polymerase chain reaction (PCR). The loxP sequences were detected using the following primer pairs: forward 5'-ACAGAAATGTGTTGGTGATGG-3' and reverse 5'-CTTTCCAATGCTGGTGGAGTG-3'. The Cx26loxP/loxP and the wild-type mice generated 322 and 288 bp band, respectively. The larger band was caused by the insertion of the 34 bp loxP sequence around the GJB2 gene. The presence of CreER was detected using the following primer pair: forward 5'-AGCTAAACATGCTTCATCGTCGGTC-3' and reverse 5'-TATCCAGGTTACGGATATAGTTCATG-3'. This resulted in a 700 bp PCR product. The PCR protocol was as follows: 95℃ for 5 minutes, 35 cycles of 95℃ for 1 minute, 55℃ for 1 minute, and 72℃ for 1 minute.

Mice at various developmental stages (P8, P12, and P21) were used. They were anesthetized with intraperitoneal injections of chloral hydrate at 400 mg/kg (4% solution in phosphate-buffered saline [PBS], pH 7.4). The animals were decapitated; their cochleae were removed and fixed in glutaraldehyde (2.5% in phosphate buffer, 0.1 M) for 6 hours. The round window and apex of the cochlea were opened to provide access to the fixative. Samples were decalcified using 10% EDTA at 4℃ for 72 hours. The specimens were then rinsed with 0.1 M phosphate buffer (pH 7.4) and fixed again in osmium tetroxide (1% dissolved in 0.2 M sodium cacodylate buffer) for 2 hours at room temperature. Subsequently, they were perfused and were washed three times (15 minutes each) in phosphate buffer. Dehydration was carried out up to 70% alcohol (15 minutes each). The samples were dissected to remove the bony wall of the cochlea and expose the organ of Corti. The specimens were then prepared for SEM (S4800; Hitachi, Tokyo, Japan) after critical point drying and coating with gold-palladium.

Various developmental stages mice (P8, P14, P18, and P21) were used. The animals older than P8 were fixed via cardiac perfusion in 4% paraformaldehyde dissolved in PBS (pH 7.4). The dissected cochleae were fixed again in paraformaldehyde (4%) at 4℃ overnight and then decalcified in 10% EDTA solution for 72 hours. The organ of Corti was microdissected, rinsed in PBS, and immersed into 0.25% Triton X-100 for 5 minutes. It was then placed in fluorescein isothiocyanate (FITC)-conjugated phalloidin (Enzo, 1:200) in PBS for 30 minutes to label the stereocilium bundles and hair cell cuticular plates. The specimens were rinsed in PBS and then stained with 10 µg/mL propidium iodide (PI) for 10 minutes to label the nuclei selectively. The samples were subsequently rinsed with PBS, mounted on glass slides using Fluoromount-G™ (Southern Biotech, Birmingham, AL, USA), and examined under an LSM 510 confocal microscope (Zeiss, Jena, Germany). OHCs nuclei stained by PI showing shrunken, condensed, fragmented and their cuticular stained by FITC phollidin showing abnormal were considered apoptosis.

The approximate time point of death of the hair cell was first identified via SEM between P12 and P21. SEM observations were conducted on mice at P8, P12, and P21. The cochleae were observed throughout their length. The surface morphology of the basilar membrane on P8 in cCx26 mice developed normally without obvious differences from the wild type mice (Fig. 1A1-C1). In the basal turn, one row of inner hair cells (IHCs) and three rows of OHCs were observed. The stereocilia on the three rows of OHC were arranged in a W-shape whereas those on the IHCs formed a gently curving arc similar to those in normal ears. In the middle turn, an extra OHC row was observed. Kinocilia were seen on top of the OHCs and IHCs. In the apical turn, the IHCs did not fully develop and extra IHCs could be seen. Most of the OHCs ranked irregularly with kinocilia. The OHCs and IHCs were surrounded by microvilli.

The surface morphology of the organ of Corti seemed still normal at the P12 stage (Fig. 1A2-C2). However, extensive damage to the stereocilia and cuticular plate of OHC in the basal and middle turn of the cochlea was observed at the P21 stage (Fig. 2). The stereocilia and cuticular plate were absent or severely damaged on the OHC (Fig. 2A1, B1). Sporadic deletion of OHC was also observed (Fig. 2B1, B2). The damage was severe in the basal (Fig. 2A1-C1) turn and the middle turn (Fig. 2A2-C2) at this stage. The morphology of the apical turn (Fig. 2A3-C3) was intact with only irregular OHC stereocilia. In contrast to the massive damage to the OHCs, the IHCs developed fully. The IHCs showed no obvious pathologic changes regardless of their locations (Fig. 2C1-C3). These results indicate that the OHCs of the cochlea were damaged between P12 and P21. In addition, this damage is much more severe in the middle and basal turn of the cochlea than that in the apical turn.

The specimens were stained with PI to label the nuclei, counterstained with FITC-conjugated phalloidin, and evaluated under a confocal microscope to visualize the hair-cell nuclei and cuticular plates. Fig. 3 showed the surface preparation of the organ of Corti in wild type animals at P8 (Fig. 3B1), P14 (Fig. 3B2), and P21 (Fig. 3B3). The PI-labeled OHC nuclei were large, round, evenly stained, and arranged in three orderly rows. Phalloidin staining revealed a pale green cuticular plate and the intense, thin green ring at the boundary between the OHC and adjacent supporting cells. The appearance of basilar membrane in cCx26 mice (Fig. 3A1) showed normal morphologies as that of the wild-type mice did at the P8 stage. The cuticular plates during the P14 stage were a little loose on the apical turn but showed no obvious shrinkage and damage. The OHC nuclei in the P14 cochlea were less regularly arranged than those of the wild type. Some of their nuclei were deeply stained with PI and unevenly quasi-circular (Fig. 3A2). Obvious changes in surface morphology were observed during the P18 stage. The OHC nuclei in the middle and basal turns of the cCx26 cochlea were shrunken, condensed, fragmented (arrow), and in some cases, completely missing (arrow head) sporadically (Fig. 4). The FITC phalloidin labeling of the middle and basal turns of the cuticular plate around the OHCs appeared fuzzy and irregular (Fig. 4B, C), which may indicate cuticular plate damage. The aforementioned changes were more severe in the middle turn (Fig. 4B) than in the basal turn (Fig. 4C). In contrast to the basal and middle turns, the morphologic changes in the apical turn only showed different shapes in the stained OHC nuclei without obvious loss or damage (Fig. 4A). These morphological changes in the nuclei clearly indicate that hair cells are undergoing apoptosis in the cCx26 mice at the P18 stage. The cuticular plate around the OHC was fuzzy at this stage. Thus, the supporting cell surrounding the OHC, such as Dieter's cells, might begin to be damaged sporadically. More severe damage on the organ of Corti in the cCx26 mice was observed during the P21 stage. Loss of OHCs was detected in the whole length of the basilar membrane. The middle turn revealed segmental breakage of the basilar membrane, which indicates the deletion of OHCs and also the supporting cells surrounding them (Fig. 3A3).