A deafness mechanism of digenic Cx26 (GJB2) and Cx30 (GJB6) mutations: Reduction of endocochlear potential by impairment of heterogeneous gap junctional function in the cochlear lateral wall

Abstract

Digenic Connexin26 (Cx26, GJB2) and Cx30 (GJB6) heterozygous mutations are the second most frequent cause of recessive deafness in humans. However, the underlying deafness mechanism remains unclear. In this study, we created different double Cx26 and Cx30 heterozygous (Cx26^+/-/Cx30^+/-) mouse models to investigate the underlying pathological changes and deafness mechanism. We found that double Cx26^+/-/Cx30^+/- heterozygous mice had hearing loss. Endocochlear potential (EP), which is a driving force for hair cells producing auditory receptor current, was reduced. However, unlike Cx26 homozygous knockout (Cx26^-/-) mice, the cochlea in Cx26^+/-/Cx30^+/- mice displayed normal development and had no apparent hair cell degeneration. Gap junctions (GJs) in the cochlea form two independent networks: the epithelial cell GJ network in the organ of Corti and the connective tissue GJ network in the cochlear lateral wall. We further found that double heterozygous deletion of Cx26 and Cx30 in the epithelial cells did not reduce EP and had normal hearing, suggesting that Cx26^+/-/Cx30^+/- may mainly impair gap junctional functions in the cochlear lateral wall and lead to EP reduction and hearing loss. Most of Cx26 and Cx30 in the cochlear lateral wall co-expressed in the same gap junctional plaques. Moreover, sole Cx26^+/- or Cx30^+/- heterozygous mice had no hearing loss. These data further suggest that digenic Cx26 and Cx30 mutations may impair heterozygous coupling of Cx26 and Cx30 in the cochlear lateral wall to reduce EP, thereby leading to hearing loss.

Keywords: Cochlea; Connexin; Cx26; Cx30; Deafness; Endocochlear potential; Gap junction.

Fig. 1

Gap junction network in the cochlea and Cx26 and Cx30 expression in the Cx26^+/−/Cx30^+/− mouse cochlea. a–d: Immunofluorescent staining for Cx26 (green) and Cx30 (red) in the cochlea in Cx26^+/−/Cx30^+/− and WT mice. Both Cx26 and Cx30 labeling is visible in the Cx26^+/−/Cx30^+/− mouse cochlea but their intensities are weaker in comparison with those in WT mice. Scale bar: 50 μm. e: Quantitative measure of Cx26 and Cx30 labeling in the organ of Corti (OC) and the cochlear lateral wall (LW) in WT mice and Cx26^+/−/Cx30^−/+ mice. The intensity of labeling was normalized to the average value in WT mice. The expression of Cx26 and Cx30 in Cx26^+/−/Cx30^+/− mice was significantly reduced. **: P > 0.001, t-test. f: Schematic drawing of two gap junctional (GJ) networks in the cochlea: epithelial cell GJ (ECGJ) network in the cochlear sensory epithelium and connective tissue GJ (CTGJ) network in the cochlear lateral wall. Hair cells have neither gap junction nor connexin expression. SV, scala vestibuli; SM, scala media; ST, scala tympani.

Fig. 2

Hearing loss in double Cx26 and Cx30 heterozygous (Cx26^+/−/Cx30^+/−) mice. a: ABRs were evoked by click stimulations in WT, Cx26^+/−/Cx30^+/−, Cx26^+/−, Cx30^+/−, Cx26^−/−, and Cx30^−/− mice. b: Hearing loss in Cx26^+/−/Cx30^+/−, Cx26^−/−, and Cx30^−/− mice. ABR thresholds evoked by click stimulations were increased in Cx26^+/−/Cx30^+/−, Cx26^−/−, and Cx30^−/− mice. c: ABR thresholds in Cx26^+/−/Cx30^+/− and WT mice evoked by tone bursts. Hearing loss was observed in the whole-tested frequency range in Cx26^+/−/Cx30^+/− mice. WT littermates served as control. Mice were 6–8 weeks old. **: P<0.001 as determined by one-way ANOVA with a Bonferroni correction.

Fig. 3

Reduction of DPOAE in Cx26^+/−/Cx30^+/− mice. a: Spectrum of acoustic emission recorded from Cx26^+/−/Cx30^+/− and WT mice. Insets: Large scale plotting of 2f₁−f₂ and f₁ peaks. The peak of DPOAE (2f₁−f₂) in Cx26^+/−/Cx30^+/− mice was reduced but f₁ and f₂ peaks remained the same as those in WT mice. f₀=20 kHz, I₁/I₂=60/55 dB SPL. b: DPOAE thresholds in Cx26^+/−/Cx30^+/− mice were increased. WT littermates served as control. Data are expressed as mean ± S.D. **: P < 0.001 as determined by one-way ANOVA with a Bonferroni correction.

Fig. 4

Normal development of the cochlea and no hair cell loss in Cx26^+/−/Cx30^+/− mice. a–b: Cross-sections of the Cx26^+/−/Cx30^+/− mouse cochlea. The cochlea demonstrates normal shape and has no apparent cell degeneration. SV, scala vestibuli; SM, scala media; ST, scala tympani. c–d: whole-mounting of the cochlear sensory epithelium in Cx26^+/−/Cx30^+/− mice. Inner hair cells (IHCs) and outer hair cells (OHCs) were stained with phalloidin-Alexa 488 (green) and the cell nuclei were labeled with propidium iodide (PI, red). Mice were 2–3 months old. Scale bar: 100 μm in a & c, 40 μm in b, and 20 μm in d.

Fig. 5

Connexin expression in Prox1-Cx26^+/−/Cx30^+/− mice. a: Double immunofluorescent staining for Cx26 and Cx30 in Prox1-Cx26^+/−/Cx30^+/− mice. b: Schematic drawing shows target-deletion of Cx26 in Deiters cells (DCs) and outer pillar cells (OPCs) in the cochlear sensory epithelium in Prox1-Cx26 KO mice. c–d: Quantitative measure of Cx26 and Cx30 labeling in the organ of Corti (OC) and the cochlear lateral wall (LW) in WT mice and Prox1-Cx26^+/−/Cx30^−/+ mice. Littermates were used as controls. The intensity of labeling was normalized to the average value in WT mice. The expression of Cx26 and Cx30 in the OC in Prox1-Cx26^+/−/Cx30^+/− mice was significantly reduced. However, the expression of Cx26 in the lateral wall was not reduced. **: P > 0.001, t-test.

Fig. 6

No hearing loss in Prox1-Cx26^+/−/Cx30^+/− mice. WT littermates served as control. Mice were 4–8 weeks old. Data are expressed as mean ± S.D.

Fig. 7

EP reduction in Cx26^+/−/Cx30^+/− mice but not in Prox1- Cx26^+/−/Cx30^+/− mice. a. The EP was significantly reduced in Cx26^+/−/Cx30^+/− mice but not in Prox1- Cx26^+/−/Cx30^+/− mice. WT littermates served as control. **: P < 0.001 as determined by one-way ANOVA with a Bonferroni correction. b. EP reduction in homozygous Cx26^−/− KO mice and homozygous Cx30^−/− KO mice. Mice were 4–8 weeks old.

Fig. 8

Co-expression of Cx26 and Cx30 and EP generation in the cochlear lateral wall. The cross-section of the mouse cochlear lateral wall was stained for Cx26 (green) and Cx30 (red) in immunofluorescent staining. Empty triangles in panel b indicate heterozygous coupling between stria vascularis (basal cell layer) and spiral ligament (fibrocyte I cells). White errors in panel c indicate co-labeling of Cx26 and Cx30 at the same GJ plaques between fibrocyte cells in the spiral ligament. Most of GJ plaque-punctates have Cx26 and Cx30 co-labeling and show yellow color in the over-lap image. St. V, stria vascularis; SPL, spiral ligament. Scale bar: 20 μm in a & b and 10 μm in c.

Fig. 9

Schematic drawing of EP generation in the cochlear lateral wall. Based on the “two-cell” model, EP is generated by Kir4.1 in the apical membrane of intermediate cells in conjunction with Kir5.1 channels, Na⁺/K⁺ ATPases, and Na⁺, K⁺, 2Cl⁻-cotransporters in the fibrocytes through GJ coupling [40,45]. Heterozygous couplings of gap junctions between fibrocute I and II cells in the spiral ligament (SPL), between fibrocyte I cells and base cells in the stria vascularis (St. V), and between base cells and intermediate cells are important for eventually positive EP generation. MC: marginal cell; BC: basal cell; FC: fibrocyte; IC: intermediate cell; IS: intrastrial space; GJ: gap junction.