Targeted ablation of connexin26 in the inner ear epithelial gap junction network causes hearing impairment and cell death

Abstract

Mutations in the gene encoding the gap junction protein connexin26 (Cx26) are responsible for the autosomal recessive isolated deafness, DFNB1, which accounts for half of the cases of prelingual profound hereditary deafness in Caucasian populations. To date, in vivo approaches to decipher the role of Cx26 in the inner ear have been hampered by the embryonic lethality of the Cx26 knockout mice. To overcome this difficulty, we performed targeted ablation of Cx26 specifically in one of the two cellular networks that it underlies in the inner ear, namely, the epithelial network. We show that homozygous mutant mice, Cx26(OtogCre), have hearing impairment, but no vestibular dysfunction. The inner ear developed normally. However, on postnatal day 14 (P14), i.e., soon after the onset of hearing, cell death appeared and eventually extended to the cochlear epithelial network and sensory hair cells. Cell death initially affected only the supporting cells of the genuine sensory cell (inner hair cell, IHC), thus suggesting that it could be triggered by the IHC response to sound stimulation. Altogether, our results demonstrate that the Cx26-containing epithelial gap junction network is essential for cochlear function and cell survival. We conclude that prevention of cell death in the sensory epithelium is essential for any attempt to restore the auditory function in DFNB1 patients.

Figure 1. Schematic Representation of a Transverse Section of the Cochlear Duct

The cochlear duct is filled with a K⁺-rich liquid, the endolymph (in yellow), and is immersed in a Na⁺-rich liquid, the perilymph (in blue). The endolymphatic compartment and the intrastrial compartment are delimited by a tight junction network (red line). The organ of Corti, the cochlear sensory epithelium, lies on the basilar membrane (bm). The cochlear epithelial gap junction network appears by embryonic day 16 (E16) and comprises interdental cells (i), inner sulcus cells (is), Claudius cells (c), outer sulcus cells (os), and root cells (r) (all in pink). It also comprises supporting cells of the organ of Corti (all in blue), which include border cells (b), inner phalangeal cells (ip), inner and outer pillar cells (p), Deiters’ cells (d), and Hensen’s cells (h), but not the sensory hair cells [7, 8, 33]. The cochlear connective tissue gap junction network begins to develop at birth and comprises fibrocytes (f) of the spiral ligament (slg) and spiral limbus (sl), mesenchymal cells lining the scala vestibuli (m), and the basal and intermediate cells (but not the marginal cells) of the stria vascularis (sv) (in green) [7, 8, 34]. The stria vascularis is a bilayered epithelium, which secretes the K⁺ into the endolymph [35], and is responsible for the generation of the endocochlear potential [36]. rm, Reissner’s membrane; ohc, outer hair cells; ihc, inner hair cells; tm, tectorial membrane; cn, cochlear nerve; sp, spiral prominence.

Figure 2. Generation of Mice Deficient for Cx26 in the Epithelial Gap Junction Network of the Inner Ear

(A–C) Generation of Cx26^loxP/loxP mice. (A) Structure of the wild-type Cx26 allele, the gene targeting vector, and the floxed allele (from the top to the bottom). The coding region of Cx26 exon II is marked by an arrow. Black triangles represent the loxP sites. The probe used for the Southern blot analysis and the EcoRI (E) restriction fragments detected with this probe (12 kb and 4 kb) are indicated. Small arrows mark the positions of the PCR primers, Cx26R and Cx26F, used for genotyping. TK, thymidine kinase selection cassette; Neo, neomycin-resistance selection cassette. (B) Southern blot analysis of properly targeted ES cell clones. Wild-type (wt) and mutated (fl) restriction fragments are indicated. (C) Genotyping of mice by PCR amplification. M, DNA size ladder. (D) Structure of the Otog-Cre transgene. Cre was inserted into a BAC containing Otog by homologous recombination. polyA, bovine growth hormone polyadenylation signal; nls, nuclear localization signal. The white box represents the 5′-untranslated region of Otog exon 1 (left arrow). Gray boxes represent the coding sequences of Otog exons 1 (right arrow) and 2. “A” and “B” indicate the sequences used for homologous recombination in the original BAC. (E and F) Immunohistofluorescence analysis of Cx26 distribution at P15 on transverse sections of the cochlear duct. White lines indicate the position of the organ of Corti. (E) Cx26^OtogCre mice. (F) Cx26^loxP/loxP mice. In Cx26^OtogCre mice, Cx26 is no longer expressed in the supporting cells of the organ of Corti or in the flanking epithelial cells. In contrast, Cx26 is still expressed by the fibrocytes of the spiral limbus and spiral ligament, as well as by the basal and intermediate cells of the stria vascularis. The scale bars in (E) and (F) represent 40 μm.

Figure 3. Histological Analysis of the Cochlear Structure in Cx26^OtogCre Mice

(A–D) Longitudinal sections of the cochlea and details of the organ of Corti in two Cx26^OtogCre homozygous mutants at (A and B) P12 and at (C and D) P33. The structure of the organ of Corti at P12 is normal. The number and shape of the cells are normal. The arrows show the three rows of OHCs and Deiters’ cells. The arrowhead indicates the IHC. (D) In contrast, the organ of Corti of a Cx26^OtogCre mutant at P33 (auditory threshold value: 75 dB at 8 kHz and 16 kHz, and 85 dB at 32 kHz) shows signs of degeneration: arrows indicate the normal emplacement of the three rows of OHCs and Deiters’ cells in which some cells of both types are now missing compared to P12. The IHC is conserved in this mutant (arrowhead). (C) The Reissner’s membrane remains well positioned. The scale bars in (A) and (C) represent 200 μm, and the scale bars in (B) and (D) represent 20 μm.

Figure 4. Electron Microscopy of the Organ of Corti in Cx26^OtogCre Cochlea

(A–C) Transmission electron microscopy at P30. (A) OHC in a heterozygous mouse with normal tight junctions on both sides of the cell. Note the presence of the dense cuticular plate at the base of the stereocilia. (B and C) Degenerated OHCs and Deiters’ cells (D) in a Cx26^OtogCre cochlea. Note the vacuolization of the cytoplasm and the absence of a cuticular plate and stereocilia in the OHC in (B) and in the left OHC in (C). (B and C) The processes of Deiters’ cells are damaged and no longer stick to the OHCs, thus resulting in the disruption of the reticular lamina at several emplacements indicated by arrows. P, outer pillar cell. The scale bars represent 2.3 μm in (A), 1.25 μm in (B), and 0.8 μm in (C).

Figure 5. Detection of Dying Cells by TUNEL and Activated Caspase 3 Immunodetection in Cx26^OtogCre Cochlea

(A–D) Four sections of the organ of Corti in Cx26^OtogCre mice at P14, P16, P30, and P60. Myosin VIIA immunofluorescent staining of the hair cells (in red) was performed prior to the TUNEL reaction (in green). White arrows indicate dying cells, namely, an IHC-supporting cell (border cell) in (A), an OHC-supporting cell in (B), an OHC in (C), and an interdental cell in (D). Note the normal structure of the organ of Corti at P14 compared to P16. The picture framed in (B) shows the immunofluorescent detection of the activated form of caspase 3 in a Cx26^OtogCre section of the organ of Corti at P17. Punctiform labeling is detected in the three rows of Deiters’ cells. The arrowheads indicate the hair cell emplacements. The scale bars represent 10 μm in (A)–(C) and 40 μm in (D).