Microtubule-associated protein 1 A and tubby act independently in regulating the localization of stereocilin to the tips of inner ear hair cell stereocilia

Tubby mice show obesity, blindness and deafness [1]. The tubby mutation is a G-to-T transversion that causes a splicing defect in the 3’-end of the Tub gene [2], but the molecular mechanisms underlying these phenotypes in tubby mice have been remained mysterious for a long time. We recently revealed the molecular mechanism by which tubby mice develop hearing impairment. Stereocilin, which should be localized to the tips of auditory stereocilia, is mislocalized in tubby mice [3]. Stereocilin is essential for maintaining the physical links between the outer hair cell (OHC) stereocilia and the tectorial membrane (TM), which is called TM-attachment links (TMALs). Strc knockout (KO) mice have defective TMALs, which are essential for mechanotransduction, leading to hearing impairment [4]. Naturally, tubby mice phenocopy Strc KO mice [3]. Map1a is a modifier of tubby hearing (moth1) [5] and, intriguingly, its wild-type allele, rather than the moth1 allele from C57BL/6 J mice, rescues the hearing impairment of tubby mice [3, 5]. Because tubby protein is highly expressed in neurons including spiral ganglion neurons that innervate the hair cells of the Organ of Corti, the mechanism of hearing rescue by microtubule-associated protein 1 A (MAP1A) was considered neuronal. For example, some expected it was related to MAP1A’s ability to bind the PSD95 postsynaptic protein [6, 7]. The recent progress showed that MAP1A possesses distinct and unexpected roles in the inner ear hair cell system; MAP1A regulates the stereociliary localization of stereocilin [3]. Due to the microscopic size and inaccessibility of the inner ear stereocilia, we have little information about the molecular interactions that support the localization of the stereociliary proteins (e.g., stereocilin, CDH23, PCDH15, TMC, etc.) required for normal hearing. These points suggest further study of MAP1A’s role will help clarify how the localization of stereociliary proteins is regulated.

First, we wanted to address whether MAP1A acts independently of mutant tubby protein to show that MAP1A itself regulates stereocilin localization. While we were unable to detect any tubby protein in the OHCs of tubby mice [3], the presence of Tub transcript and tubby protein were reported in the retina and the cochlear of tubby mice [5, 8]. To solve this problem, we generated Tub conditional KO mice in which exon 3 of the Tub gene is flanked by loxP sites. The resulting Tub^flox/flox mice were crossed with a global “Cre-deleter” mice (E2a-Cre) to produce homozygous null offspring (Fig. 1 A, B). In a previous study, targeted deletion of the Tub gene was found to causes retinal degeneration and obesity, suggesting the tubby mutation is a loss-of-function mutation at least in terms of these two tubby mouse phenotypes [9]. But, no one has reported whether Tub-null (Tub^−/−) mice also show hearing impairment. Immunofluorescence staining in Tub^−/− mice we generated revealed that stereocilin proteins in the OHC stereocilia disappeared (Fig. 1C). Accordingly, we found that Tub^−/− mice show elevated auditory brainstem response (ABR) thresholds like the ABR shifts observed in tubby mice [3] (Fig. 1D). Tub^−/− mice also showed elevated thresholds and reduced amplitudes of their distortion product otoacoustic emissions (DPOAEs), which are used to assess OHC integrity (Fig. 1E) (see Additional file 1 for the detailed methods and Additional file 2 for amplitude of DPOAEs). These findings confirm hearing impairment in Tub^−/− mice and suggest that the hearing impairment of tubby mice is due to a loss-of-function of Tub gene. We next examined whether the deletion of Tub gene in the inner ear causes hearing impairment using Tub^flox/flox; Pax2-Cre mice. We observed that deletion of Tub in the inner ear is responsible for hearing impairment in tubby mice (Fig. 1F, G; Additional file 2 for amplitude of DPOAEs and immunofluorescence staining of stereocilin), confirming that the site of action of tubby is cochlear. Ultimately, we explored whether wild-type MAP1A can rescue both hearing impairment and stereocilin mislocalization in Tub-deficient mice. The genetic background of Tub^−/− mice is C57BL/6 J that has a nonprotective variant of Map1a (Tub^−/−; Map1a^B6). Sequential mating Tub^−/− mice with AKR mice produced offspring that lack Tub but have a wild-type, protective Map1a allele (Tub^−/−; Map1a^AKR). We found that Tub^−/−; Map1a^AKR mice show the recovery of ABR threshold shifts and loss of DPOAE amplitudes (Fig. 1D, E) similar to that reported in tub/tub; Map1a^AKR mice [3] and restore the localization of stereocilin to the tips of the tallest OHC stereocilia (Fig. 1H). These findings indicate that MAP1A by itself may regulate the localization of stereocilin. Because we did not look for changes in the level of stereocilin protein in the OHCs, we cannot rule out the possibility that TUB or MAP1A might affect stereocilin expression or stability. This would obviously also affect its localization to the stereocilia. It also remains unclear whether the MAP1A-mediated localization of stereocilin to the tips of the stereocilia in Tub^−/−;Map1a^AKR mice shows an age-dependent decline and it is more vulnerable to aging or environmental stressors such as noise.

MAP1A plays important roles in stabilizing microtubules in neurons [10]. However, its roles in other cell types are obscure. The Map1a allele from AKR mice also reduces photoreceptor degeneration in Tulp1- and Tub-deficient mice [11]. However, we observed that Tub^−/−; Map1a^AKR mice still show obesity (unpublished data), suggesting that sensory cells may share pathologic mechanisms related to MAP1A. The detailed roles of MAP1A in the cell surface specializations of photoreceptors and cochlear hair cells are waiting to be discovered. Because there has been no report showing the presence of microtubule in the stereocilia, new function of MAP1A other than the conventional role of stabilizing microtubule can be expected. It is unclear why two seemingly unrelated proteins, MAP1A and tubby, contribute to the proper localization of a single stereociliary molecule. Stereocilia are nano-scale structures. However, they have fine cytoskeletal structures and complex protein interaction networks like primary cilia or neuronal postsynaptic densities. Stereocilin might be a core member of the stereociliary protein interactome whose localization is so important to normal hearing that two mechanisms are necessary. In addition, the independent contributions of MAP1A and tubby to the localization of stereocilin suggest it may be important to consider a “two-hit” mechanism when assessing the roles of stereociliary proteins in hearing impairment.