Skip to main content
NCBI home page
Molecular Therapy logoLink to Molecular Therapy
. 2015 Oct 20;24(1):17–25. doi: 10.1038/mt.2015.150

Gene Therapy Restores Hair Cell Stereocilia Morphology in Inner Ears of Deaf Whirler Mice

Wade W Chien 1,2,*, Kevin Isgrig 3, Soumen Roy 3, Inna A Belyantseva 4, Meghan C Drummond 4, Lindsey A May 3, Tracy S Fitzgerald 5, Thomas B Friedman 4, Lisa L Cunningham 3
PMCID: PMC4754541  PMID: 26307667

Abstract

Hereditary deafness is one of the most common disabilities affecting newborns. Many forms of hereditary deafness are caused by morphological defects of the stereocilia bundles on the apical surfaces of inner ear hair cells, which are responsible for sound detection. We explored the effectiveness of gene therapy in restoring the hair cell stereocilia architecture in the whirlin mouse model of human deafness, which is deaf due to dysmorphic, short stereocilia. Wild-type whirlin cDNA was delivered via adeno-associated virus (AAV8) by injection through the round window of the cochleas in neonatal whirler mice. Subsequently, whirlin expression was detected in infected hair cells (IHCs), and normal stereocilia length and bundle architecture were restored. Whirlin gene therapy also increased inner hair cell survival in the treated ears compared to the contralateral nontreated ears. These results indicate that a form of inherited deafness due to structural defects in cochlear hair cells is amenable to restoration through gene therapy.

Introduction

Deafness is the most common inherited sensory disorder, affecting 1 in every 1,000 births.1 Although there are hundreds of syndromes that include deafness as one feature of a complex phenotype, approximately 70% of congenital deafness is nonsyndromic. More than 130 genetic loci have been linked to nonsyndromic hereditary hearing loss in humans, and more than 60 genes with causative mutations have been identified at these loci.2 Children with hereditary hearing loss are often diagnosed early via infant hearing screening, and the current management options for these children include hearing amplification and cochlear implantation. While these interventions are effective in many cases, some children receive minimal benefits from them and continue to struggle with the physiologic and psychosocial impacts of deafness.

The delivery of corrective DNA into the inner ear offers the potential for restoring hearing in patients with hereditary hearing loss caused by mutations that affect the development and organization of stereocilia bundles protruding on the mechanosensory hair cells of the inner ear (Figure 1).3,4 For example, mutations of MYO15A, which encodes the unconventional myosin 15A, result in non-syndromic deafness DFNB3.5,6,7 MYO15A, an actin-based motor protein, is required for proper transport of several proteins, including whirlin, to the tips of stereocilia.8,9,10 Whirlin is a PDZ-containing scaffold protein encoded by WHRN, and it is required for the development of the stereocilia bundle from microvilli that elongate and thicken on the apical surface of differentiating hair cells.10,11,12,13 Whirlin is also widely expressed in a variety of tissues, including the eye.8,13,14 In humans, depending upon the allele, mutations of WHRN result either in nonsyndromic deafness DFNB31 or type 2 Usher syndrome, characterized by retinitis pigmentosa and sensorineural hearing loss.15,16 Two major isoforms of whirlin are found in the cochlea: the long and the short isoforms.12 During early postnatal development, whirlin together with other type 2 Usher proteins is transiently present at the ankle-link location in stereocilia.17,18 In mature hair cells, whirlin is localized to the tips of stereocilia.8,9,19

Figure 1.

Figure 1

Open in a new tab

The inner ear and whirlin gene therapy. Sound is transmitted through the outer and middle ear into the inner ear, where the cochlea resides. The cochlea contains hair cells, which are mechanosensory cells that transduce sound energy into neural input to the brain. The organ of Corti contains three rows of outer hair cells and a single row of inner hair cells. The inner hair cells synapse with the primary afferent auditory nerve fibers, which send axons via the VIIIth cranial nerve to the auditory brainstem. Many forms of hereditary hearing loss are caused by genetic mutations that lead to abnormalities of the hair cells. In this study, we explore whether cochlear gene therapy can be used to restore hair cell stereocilia morphology and function in whirler mutant mice, which are a model for human DFNB31 nonsyndromic deafness.

The whirler (whrnwi/wi) mouse is a naturally-occurring model of DFNB31.20 Whirler mice have profound hearing loss as well as vestibular defects that result in circling and head-tossing behaviors.20 The whrnwi mutation is a 592 bp deletion that creates a translation frameshift resulting in a complete loss of whirlin function.12 Examination of inner ear morphology in whirler mice revealed abnormalities in hair cell stereocilia.21,22,23 Hair cells from whirler (whrnwi/wi) mice have significantly shorter and wider stereocilia compared to homozygous wild-type (whrn+/+) and heterozygous (whrnwi/+) mice.22 In addition, hair cells of whrnwi/wi mice have aberrantly arranged supernumerary rows of stereocilia. Hair cell degeneration in whirler mice is first evident at the cochlear base approximately 30 days after birth (P30).21 Thus, there is a 1-month window after birth during which the defect could be corrected prior to the permanent loss of hair cells. In this study, we examined whether a gene therapy approach in which wild-type copies of whirlin cDNA delivered using an adeno-associated virus serotype 2/8 (AAV8) vector to cochleas of whirler mice can restore normal hair cell stereocilia architecture and auditory function in these animals (Figure 1).

Results

Whirlin gene therapy restores whirlin expression in infected hair cells

In the normal adult mouse cochlea, whirlin is localized to the stereocilia tips of inner hair cells (IHCs; Figure 2a). In contrast, the absence of whirlin in whrnwi/wi cochleas results in short and dysmorphic stereocilia bundles (Figure 2b). To determine if whirlin gene therapy is sufficient to rescue stereocilia architecture and morphology in whirler mice, we surgically delivered AAV8-whirlin through the cochlear round window in adult mice. Our surgical approach is illustrated in Figure 1. AAV8 was used as the vector for whirlin cDNA delivery because it infects inner and outer hair cells of the organ of Corti and is nonpathogenic in both mice and humans.24 The long isoform of whirlin cDNA was used because it was previously shown to restore stereocilia morphology in vitro.8 When AAV8-whirlin was delivered to cochleas of adult (~P30) whrnwi/wi mice, inner hair cells were primarily infected, as evidenced by the presence of whirlin expression at the stereocilia tips of a subset of these cells (Figure 2c). In contrast, whirlin expression was not detected in the contralateral nonsurgical ear of the same animal (Figure 2d). Despite the presence of whirlin expression at the stereocilia tips in adult animals, the infected whrnwi/wi stereocilia did not demonstrate improved morphology (Supplementary Figure S1). For this reason, we decided to deliver AAV8-whirlin gene therapy to cochleas of neonatal (P1-P5) mice. The cochleas were examined by confocal microscopy 30–90 days later (Figure 2e). All subsequent experiments were done in neonatal animals. Similar to our results in adult animals, AAV8-whirlin infected primarily inner hair cells, as evidenced by the presence of whirlin expression (Figure 2e and Supplementary Figure S2). Again whirlin expression was not detected in cells of contralateral whrnwi/wi ears that did not receive AAV8-whirlin gene therapy (Figure 2f). In AAV8-whirlin-infected hair cells, whirlin was observed at stereocilia tips (Figure 2e), consistent with whirlin localization in control mice expressing endogenous functional whirlin (Figure 2a). Whirlin expression was examined in infected whrnwi/wi inner hair cells at P30 (Figure 2g), and it persisted until at least P90 (Figure 2h). In whrnwi/wi cochleas that received AAV8-whirlin gene therapy during the neonatal period, 15.3% of IHCs demonstrated whirlin localization at stereocilia tips in the basal turn, 16.2% in the middle turn, and 11.8% in the apical turn (Figure 2i). The differences in the number of IHCs with whirlin expression among the cochlear turns were not statistically significant (P > 0.59). These data indicate that AAV8-whirlin-mediated gene therapy restored whirlin expression and localization at the tips of stereocilia in AAV8-whirlin-infected inner hair cells.

Figure 2.

Figure 2

Open in a new tab

AAV8-whirlin gene therapy restores whirlin expression in infected IHCs. (a) Inner hair cells in control mice (whrn+/wi) show whirlin (green) localized at the tips of the stereocilia (red). Examples of whirlin expression in stereocilia of IHCs are indicated by thin white arrows. (b) Hair cells in whirler (whrnwi/wi) mice lack whirlin expression at the stereocilia tips of IHCs. In addition, whrnwi/wi inner hair cells have aberrantly short stereocilia and supernumerary rows of stereocilia (white arrowhead). (c) AAV8-whirlin gene therapy delivered to cochleas of adult whrnwi/wi mice restored whirlin expression at the stereocilia tips in infected IHCs (white arrow). A neighboring noninfected IHC is also shown (white arrowhead). (d) The contralateral (nonsurgery) ear of the same mouse in (c) showed absence of whirlin expression in all IHCs. (e) AAV8-whirlin gene therapy delivered to cochleas of neonatal whrnwi/wi mice restored whirlin expression in infected IHCs. Infected IHCs show whirlin immunolocalization at the stereocilia tips (white arrow), while adjacent noninfected IHCs do not show whirlin expression (white arrowhead). Infected IHCs appear to have longer stereocilia than adjacent noninfected IHCs. (f) IHCs in the control (contralateral, nonsurgery) ear (same mouse as in e) show no whirlin expression. F-actin is labeled red with phalloidin. (g) A whrnwi/wi mouse cochlea that received AAV8-whirlin at P2 and was examined at P30. An infected IHC and a neighboring non-infected IHC are indicated by white arrow and white arrowhead, respectively. (h) A whrnwi/wi mouse cochlea that received AAV8-whirlin gene therapy at P2 and was examined at P90. An infected IHC and a neighboring non-infected IHC are indicated by white arrow and white arrowhead, respectively. Whirlin expression at the tips of the stereocilia in infected IHCs persisted until at least P90. (i) The percentage of IHCs infected (IHCs demonstrating whirlin expression at stereocilia tips) in whrnwi/wi ears that underwent AAV8-whirlin gene therapy during the neonatal period. Error bars represent standard errors. All images were taken from the apical turn of the cochlea and were acquired at 63×. Scale bar = 10 μm.

Whirlin gene therapy in neonatal mice improves hair cell stereocilia morphology

We measured inner hair cell stereocilia lengths in cochleas of whrnwi/wi mice that underwent AAV8-whirlin gene therapy at P1-P5. Stereocilia length is precisely controlled in normal control mouse cochleas, and stereocilia rows are arranged in an orderly fashion. Only stereocilia in the tallest row were measured. Because AAV8-whirlin did not infect every inner hair cell in any surgical cochlea, we were able to compare stereocilia lengths between AAV8-whirlin-infected versus neighboring noninfected inner hair cells in the same cochlea (Figure 3a). Whrnwi/wi IHCs that expressed whirlin at the stereocilia tips had longer stereocilia than neighboring IHCs that did not express whirlin at the stereocilia tips (Figures 2e,g,h and 3a). Some of the whirlin-positive stereocilia appeared fully restored in length while others were partially restored. The differences in stereocilia lengths between infected whrnwi/wi IHCs and their neighboring non-infected whrnwi/wi IHCs were significant in all regions of the cochlea (P < 0.00001). However, on average, the stereocilia length in infected whrnwi/wi IHCs remained shorter than that of normal control stereocilia (P < 0.001). The average length of IHC stereocilia that lacked whirlin expression at the stereocilia tips in the treated ears was similar to the stereocilia length in the contralateral nonsurgery ears throughout the cochlea (P > 0.05). The distribution of stereocilia lengths in infected whrnwi/wi IHCs in all regions of the cochlea is shown in Figure 3b. In normal controls, IHC stereocilia in the tallest row were >2 μm in length, and in the treated whrnwi/wi ears, 63% of the infected IHCs had stereocilia that were >2 μm in length. In contrast, stereocilia from neighboring noninfected whrnwi/wi IHCs (as well as virtually all stereocilia from IHCs in whrnwi/wi ears without surgery) were <2 μm in length (100 and 99%, respectively). These data indicate that whirlin gene therapy partially restores stereocilia length in infected whrnwi/wi IHCs, and in some cases, the stereocilia length was restored to normal.

Figure 3.

Figure 3

Open in a new tab

AAV8-whirlin gene therapy restores stereocilia length and reduces supernumerary stereocilia rows. (a) Measurements of the length of IHC stereocilia in the tallest row. The number of IHCs counted and the number of animals used for each group in each cochlear region are shown in the table. The “n” in figure legend indicates the total number of animals used for each group. (b) Stereocilia length distributions in all four groups. The average stereocilia lengths are sorted into 1 μm bins. The percentage of measured IHCs whose stereocilia lengths fall into each bin is shown. The labels are the same as in panel A. (c) IHC stereocilia row numbers. The labels are the same as in panel a. Error bars represent standard error. Asterisks indicate statistical significance (P < 0.05).

We next analyzed stereocilia bundle architecture in cochleas of whrnwi/wi mice that received AAV8-whirlin gene therapy. Stereocilia in wild-type IHCs are arranged in two to three rows (Figure 2a; also visible in Figure 5a), whereas whrnwi/wi IHCs contain supernumerary rows of stereocilia (Figure 2b, arrowhead; also visible in Figure 5b). In the control (wild-type and heterozygous) littermates, IHCs had two rows of stereocilia on average in the base, middle, and apex of the cochlea (Figure 3c), whereas in whrnwi/wi mice, IHCs had on average five to six rows of stereocilia, depending on the cochlear region (Figure 3c). Quantification of the stereocilia rows showed that after AAV8-whirlin gene therapy, whrnwi/wi IHCs with whirlin expression at the stereocilia tips had partial restoration of the stereocilia row number compared to the neighboring noninfected IHCs (Figure 3c). This restoration of stereocilia bundle architecture was statistically significant throughout the cochlea (P < 0.0001). The average stereocilia row number in whrnwi/wi IHCs without whirlin expression at the stereocilia tips in the treated ears was similar to that of the contralateral nonsurgery ears throughout the cochlea (P > 0.05, Figure 3c). These results indicate that AAV8-whirlin gene therapy is effective at reducing aberrant supernumerary stereocilia rows in whrnwi/wi IHCs, thus partially restoring stereocilia bundle architecture. Experiments in which a control virus (AAV8-GFP) was injected into whirler cochleas did not result in any restoration of stereocilia length or bundle architecture (Figure 4). Thus, the changes in whrnwi/wi IHCs in response to AAV8-whirlin were a result of AAV8-whirlin gene therapy and were not due to AAV8 infection alone, the expression of GFP in hair cells, or a general effect of protein overexpression.

Figure 4.

Figure 4

Open in a new tab

AAV8-GFP does not alter stereocilia morphology in whrnwiwi IHCs. (a) Injection of the control virus AAV8-GFP into the whirler cochlea did not alter stereocilia morphology or bundle architecture in infected IHCs. Infected IHCs (arrow) showed GFP expression that colabeled with myosin 7a. GFP was not localized to the stereocilia tips. The neighboring noninfected IHCs (arrowheads) showed no GFP expression. Both GFP-infected and noninfected IHCs had short stereocilia and supernumerary rows. Images were acquired at 63×. Scale bar represents 10 μm. (b) Quantification of stereocilia lengths in IHCs infected with the control virus AAV8-GFP (GFP) versus neighboring noninfected IHCs (control). The stereocilia lengths are similar between the two groups (1.14 versus 1.18 μm, respectively). The differences are not statistically significant (P > 0.5). n = the number of animals. Error bars represent standard error.

Scanning electron microscopy (SEM) was used to examine stereocilia surface structures in more detail and at greater resolution. In normal control ears, the stereocilia are arranged into two to three rows in a staircase pattern (Figure 5a). In whrnwi/wi ears, the stereocilia are short and arranged in supernumerary rows (Figure 5b). When AAV8-whirlin was delivered to the whrnwi/wi cochleas, we observed IHCs with longer stereocilia compared to neighboring IHCs, which had short stereocilia similar to those seen in whirler mice without gene therapy (Figure 5c,d). In addition, the IHCs with longer stereocilia appeared to have restored stereocilia bundle shape and architecture with two to three stereocilia rows and a reduced number of rows of supernumerary stereocilia. In order to determine if the IHCs with longer stereocilia were infected with AAV8-whirlin, we sequentially performed fluorescent immunohistochemistry (visualized with confocal microscopy) and high-resolution structural analysis (SEM) on the same specimen (correlative SEM).25 By comparing images of inner hair cells collected on the confocal microscope (Figure 5e) with subsequent imaging of these same inner hair cells by SEM (Figure 5f), we confirmed that whirler IHCs with long stereocilia were infected with AAV8-whirlin and had whirlin expression at their stereocilia tips.

Figure 5.

Figure 5

Open in a new tab

Correlative scanning electron microscopy (SEM) and confocal microscopy of inner hair cells after AAV8-whirlin gene therapy. (a) A row of IHCs from a normal control (whrn+/wi) cochlea (middle turn) showing stereocilia arranged in staircase pattern with two to three rows (white arrows). (b) A row of IHCs from a whrnwi/wi cochlea (no AAV8-whirlin gene therapy, basal turn) showing all whrnwi/wi IHCs have short stereocilia bundles arranged in supernumerary rows (white arrowheads). (c) and (d) Examples of whrnwi/wi cochleas that received AAV8-whirlin gene therapy (middle and basal turns, respectively). Some IHCs have longer stereocilia (white arrow), while the adjacent IHC has short stereocilia (white arrowhead) that are reminiscent of the whrnwi/wi cochlea without gene therapy (compare to a and b). (e) and (f) Correlative confocal and SEM images of a single whrnwi/wi specimen after AAV8-whirlin gene therapy (from apical turn). The middle IHC (white arrow) was infected with AAV8-whirlin and has elongated stereocilia arranged in fewer rows (resembling the wild-type cochlea), while the adjacent two IHCs (white arrowheads) were not infected with AAV8-whirlin and have shorter stereocilia arranged in supernumerary rows (resembling bundles from whrnwi/wi mice). This confirms that the whrnwi/wi IHCs with elongated stereocilia bundles were infected with AAV8-whirlin and expressed whirlin at the stereocilia tips. Scale bars represent 5 μm.

Whirlin gene therapy increases inner hair cell survival

Hair cells in cochleas of whrnwi/wi mice undergo progressive degeneration and death starting at approximately P30.12,21 Hair cell death in whrnwi/wi mice occurs first at the cochlear base and then progresses toward the apex. Hair cell death in whrnwi/wi mice occurs by an unknown mechanism, but it indicates that whirlin is required for hair cell homeostasis. IHC counts in whrnwi/wi cochleas that underwent AAV8-whirlin gene therapy revealed increased IHC survival in the whrnwi/wi cochleas that received AAV8-whirlin compared to contralateral ears that did not receive gene therapy (Figure 6). The difference in IHC survival between the treated and nontreated ears was significant at the cochlear base (P < 0.001). Quantification revealed that over the entire cochlea ~13% of the surviving IHCs in whrnwi/wi ears that received AAV8-whirlin gene therapy showed whirlin expression at the stereocilia tips. However, when the cochleas from whirler mice that underwent AAV8-whirlin gene therapy were examined at P90, significant IHC loss in the basal turn was seen in both treated and nontreated ears. These data indicate that whirlin gene therapy increased IHC survival in the treated whrnwi/wi ears at P30, but this protection was not permanent.

Figure 6.

Figure 6

Open in a new tab

AAV8-whirlin gene therapy increases IHC survival. (a) Confocal image of the contralateral cochlea from a whrnwi/wi mouse at P30 (base) that did not undergo AAV8-whirlin gene therapy. Significant IHC degeneration has begun with no surviving IHC in this image. (b) Confocal image of a whrnwi/wi cochlea (base) after AAV8-whirlin gene therapy. Myosin 7a, which labels the hair cells, is shown in magenta, and phalloidin, which labels F-actin in the stereocilia, is shown in red. Whirlin is shown in green. Even though IHC degeneration has begun, many IHCs are still intact (thin white arrows). There is one IHC with abundant whirlin expression in the cytoplasm (thick white arrow). Images were taken at 63×. Scale bar represents 10 μm. (c) A normal control cochlea is shown for comparison. (d) IHC survival in whrnwi/wi ears injected with AAV8-whirlin gene therapy (whrnwi/wi with AAV8-whirlin) vs. the contralateral noninjected ears (Contralateral). The whrnwi/wi ears that underwent AAV8-whirlin gene therapy had increased IHC survival relative to the contralateral (no AAV8-whirlin) ears. This difference was significant at the cochlear base (P < 0.001). Error bars represent standard errors. n = the number of animals.

Auditory function in whrnwi/wi mice after whirlin gene therapy

Whirler mice fail to develop normal hearing postnatally and are deaf throughout life.20 Hearing function was assessed by auditory brainstem response (ABR) measurements in whrnwi/wi mice that received AAV8-whirlin gene therapy (Figure 7). An ABR threshold refers to the lowest sound pressure level (SPL) that can generate a reliable far-field electrical response. All ABR testing was done in animals between P30-P90. Hearing sensitivity in control mice was normal, as indicated by ABR thresholds below 40 dB SPL at all stimulus frequencies26 (Figure 7). Whirler mice without gene therapy are profoundly deaf, as indicated by absence of ABRs at the maximum stimulus intensity (90 dB SPL). Whirler mice that received whirlin gene therapy did not demonstrate improved hearing sensitivity at any frequency (Figure 7). We performed ABR testing in these animals as early as P16 and did not observe measurable ABR thresholds. To assess whether AAV8-whirlin or the surgery itself caused an elevation in ABR thresholds, control and sham surgeries were performed in which AAV8-whirlin or carrier media alone (without AAV8-whirlin) was delivered to ears of wild-type and heterozygous littermates. In all controls, a small elevation of ABR thresholds (10–20 dB SPL relative to normal controls without surgery) was detected (Figure 7). Similar ABR thresholds were observed in animals receiving AAV8-whirlin versus carrier media alone, suggesting that the mild ABR threshold elevation in these animals is largely caused by surgical trauma and not by the AAV8 and/or transgene.

Figure 7.

Figure 7

Open in a new tab

AAV8-whirlin did not restore hearing in whirler mice. In both the whrnwi/wi ears injected with AAV8-whirlin (whrnwi/wi with AAV8-whirlin) and untreated whrnwi/wi ears (whrnwi/wi no surgery), no auditory brainstem responses (NR) were detected even at 90 dB sound pressure level. Control mice that underwent AAV8-whirlin gene therapy or sham surgery had a small elevation in ABR thresholds in all three frequencies tested. n = the number of animals. All ABR testing was done at age P30 to P90.

Discussion

Cochlear gene therapy offers the promise of a new treatment option for hereditary hearing loss. In this study, we demonstrated that infection of neonatal auditory hair cells in vivo using AAV8-whirlin cDNA is sufficient to restore the normal morphology of the stereocilia bundles in whirler mutant mice. To date, most studies of gene therapy of the inner ear have focused on delivering normal copies of the gene of interest to replace the underlying mutated gene.27,28 For example, AAV1 was used to deliver vesicular glutamate transporter-3 (Vglut3) cDNA to IHCs in the Vglut3 knock-out mouse.27 The authors found that Vglut3 gene therapy restored hearing in the Vglut3 knock-out mouse.27 In another study, AAV2/1 was used to deliver Gjb2 cDNA to a conditional Gjb2 knock-out mouse.28 Gjb2 gene therapy was found to restore gap junction function in infected cochlear supporting cells.28 Improvement in auditory function was not observed in treated animals. While both of these studies demonstrate the promise of cochlear gene therapy, the hair cells in the mutant mice used in both of these studies have no intrinsic structural defects affecting stereocilia. Our study is the first to demonstrate the feasibility of gene therapy as a viable way of improving and sometimes restoring structural defects in stereocilia bundles of cochlear hair cells.

We found that when AAV8-whirlin gene therapy was delivered to cochleas of adult whrnwi/wi mice in vivo, whirlin expression was restored in infected IHCs; however, restoration of stereocilia length and row number was much less effective relative to mice that received gene therapy during the neonatal period. This decrease in the effectiveness of gene therapy in the adult mouse cochlea has been reported previously27 and suggests that there may be a critical developmental period during which the mouse cochlea is amenable to gene therapy to repair structural defects in the stereocilia bundle.

One of the major challenges in cochlear gene therapy is achieving high rates of gene delivery (i.e., high infection efficiency). In this study, ~15% of IHCs were infected with AAV8-whirlin and showed whirlin expression at the stereocilia tips. We did not observe any OHCs with obvious AAV8-whirlin infection. The predilection of AAV8 for infecting IHCs is consistent with our previous study and the findings from others,24,29 and may be caused by the presence of a receptor and/or coreceptor on IHCs that is not expressed on OHCs. In addition, a previous study has shown that the promoter used can determine AAV infection specificity.30 However, we believe that this is a less likely explanation for our study since we used the universal CMV promoter that is reported active in all cell types. Higher rates of IHC and OHC infection may be possible using other AAV serotypes and/or alternative promoters.

We utilized a method of sequential confocal microscopy followed by SEM microscopy to examine the same hair cells at both the protein expression/fluorescence and ultrastructural levels. Using this correlative SEM technique, we confirmed that the whrnwi/wi IHCs with elongated stereocilia expressed whirlin at the stereocilia tips. The whrnwi/wi IHCs with elongated stereocilia had normal architecture of the stereocilia bundle as well as a reduction in the number of supernumerary stereocilia rows. Combining these two methods allowed us to evaluate the structural details of the stereocilia bundles using SEM, while confirming whirlin expression using confocal microscopy.

We found that AAV8-whirlin gene therapy prolonged IHC survival at the cochlear basal turn in treated ears. This was assessed at approximately P30, when IHC degeneration is underway in the basal turn of whirler cochleas. Even though we observed higher IHC numbers at the middle and apical turns in the treated ears compared to the nontreated ears, the differences were not statistically significant. Although the treated ears had more surviving IHCs compared to the contralateral non-treated ears, not all surviving IHCs in the treated ears showed whirlin expression at the stereocilia tips. It is possible that AAV8-whirlin gene therapy caused variable levels of infection in the IHCs, with some showing high levels of infection and robust whirlin expression at the stereocilia tips, whereas others showing much lower levels of infection with no observable whirlin expression at the stereocilia tips. Despite the lack of whirlin expression at the stereocilia tips, this low level of AAV8-whirlin infection may be sufficient to prolong IHC survival. When we examined IHC survival at P90, there was significant IHC degeneration regardless of whether the ear received AAV8-whirlin or not. These data suggest that while AAV8-whirlin prolongs IHC survival, it does not prevent eventual degeneration. Increased cell survival was also observed in the Gjb2 gene therapy study, where the treated ears had prolonged spiral ganglion survival compared to the nontreated ears.28 However, Gjb2 gene therapy had no effect on improving hair call survival. To the best of our knowledge, our study is one of the first to show increased hair cell survival using a gene therapy approach in the setting of hereditary hearing loss.

Examination of auditory function in wild-type and heterozygous littermates that underwent AAV8-whirlin injections revealed a small ABR threshold shift (8–20 dB SPL) compared to normal controls that did not undergo cochlear gene delivery. Animals that underwent sham surgery demonstrated similar ABR threshold shifts, indicating that the hearing loss in these animals is likely caused by surgical trauma instead of AAV8 or the transgene. It may be possible to reduce the surgical trauma associated with cochlear gene delivery by injecting smaller volumes of viral solution to the cochlea, modifying the injection rate, and/or providing otoprotective agents such as dexamethasone in the perioperative period.31

Although AAV8-whirlin gene therapy restored stereocilia architecture in the infected whrnwi/wi IHCs and prolonged IHC survival, we did not see any improvement in hearing sensitivity in these animals. Several factors may have contributed to the lack of functional recovery after AAV8-whirlin gene therapy. First is the low infection efficiency we observed: only 10–15% of whrnwi/wi IHCs expressed whirlin at the stereocilia tips after gene therapy. This infection rate may be insufficient for hearing recovery. Second, primarily the IHCs were infected by AAV-8, whereas functional inner and outer hair cells are required for normal auditory function.32 Third, while AAV8-whirlin was delivered to the whrnwi/wi cochlea as early as P1 in our study, irreversible damage to cochlear hair cells may have already occurred, such that even the restoration of whirlin expression by the AAV8-whirlin was not sufficient for hearing restoration. Fourth, our study utilized the long isoform of whirlin cDNA, whereas the whirler mutation affects both long and short whirlin isoforms.12 The long isoform of whirlin was selected for our study because a previous study had shown that the long isoform is sufficient to restore stereocilia length in whrnwi/wi hair cells in vitro.8 The short isoform of whirlin may also be important for either maintaining the normal morphology of stereocilia bundles or it may have some other function that is critical for hearing. Finally, it is possible that whirlin is required elsewhere in the auditory pathway, and restoration of expression in the cochlea is not sufficient to recover normal function.

Gene therapy using AAV8-whirlin restored whirlin expression at the stereocilia tips in infected whrnwi/wi IHCs. The infected IHCs demonstrated restoration of stereocilia length and bundle architecture. In addition, our data indicate that whirlin gene therapy prolonged IHC survival in the treated whrnwi/wi ears. This proof-of-concept study demonstrates that cochlear gene therapy can be effective at restoring structural defects in IHCs in a mouse model of human hereditary hearing loss.

Materials and Methods

Study design. We examined whether AAV8-whirlin gene therapy can restore stereocilia defects in the whrnwi/wi cochlea. AAV8-whirlin was injected into the left ears of whrnwi/wi mice. Normal controls include wild-type (whrn+/+) and heterozygous (whrnwi/+) littermates. The contralateral nonsurgery ears in mice are used as whrnwi/wi nonsurgery controls. The mice were examined up to 90 days after gene therapy delivery. In each animal, hearing sensitivity was assessed using ABR at P30 to P90. Whirlin expression and localization were examined using confocal microscopy. Stereocilia morphology and architecture were examined using both confocal microscopy and SEM.

AAV vector construction. Purified serotype 2/8 AAV vectors (AAV8) carrying the long isoform of whirlin cDNA (NCBI accession number AY739114) driven by a CMV promoter (AAV8-whirlin) were constructed by Vector Biolabs (Malvern, PA). The long isoform of whirlin cDNA is 2,724 bp, which is within the limit of the AAV carrying capacity of ~4,500 bp. The AAV8 vector was selected based on its high hair cell infection efficiency.24 The stock concentration of purified AAV8-whirlin was approximately 1 × 1013 genome copies (gc) per ml. The AAV8-GFP vector was purchased from Vector Biolabs. The GFP DNA insert was driven by a CMV promoter.

Animal surgery for cochlear gene therapy. All aspects of the animal research are approved by the Animal Care and Use Committee at the National Institute on Deafness and Other Communication Disorders (NIDCD ASP#1378-15). The whrnwi/wi mice and control homozygous wild-type and heterozygous littermates between P1-5 were used in this study. Heterozygous (whrn+/wi) animals have normal stereocilia morphology and normal auditory and vestibular functions, and they have been utilized as normal controls in previous studies.12,21 The surgical approach we used to access the neonatal mouse cochlea was modified slightly from that of Akil et al.27 Neonatal mice were anesthetized using hypothermia, where their body temperature was temporarily lowered to 6 °C. Once the appropriate level of anesthesia was confirmed, a 5-mm postauricular incision was made from approximately 2 mm posterior to the auricular crease. The submandibular gland was reflected gently to expose the underlying neck muscles. The facial nerve was identified and traced to the area of the auditory bulla. During the neonatal period, the bulla is cartilaginous and transparent and the round window membrane is easily identified. A glass micropipette containing AAV8-whirlin was advanced through the bulla into the round window membrane of the cochlea (Figure 1) using a Sutter MM-33 micromanipulator (Sutter Instrument, Novato, CA). A Nanoliter Microinjection System (model Nanoliter2000, World Precision Instruments, Sarasota, FL) was used to deliver 5 × 109 gc of AAV8-whirlin into the scala tympani at an injection speed of 40–50 nl/second. A total of 10 injections (400–500 nl) were delivered to each ear. We have tried doubling the volume of AAV8-whirlin injections to ~1,000 nl (~1 × 1010 gc) but did not observe any increase in the number of IHCs infected. After injections were completed, the glass micropipette was carefully withdrawn from the cochlea using the micromanipulator, and the skin was closed with sutures. The animals were placed on a covered heating pad to recover from hypothermia. The total surgery time was approximately 10 minutes per animal.

We also performed AAV8-whirlin injections in adult whrnwi/wi mice and control homozygous wild-type and heterozygous littermates (~P30). Anesthesia in adult mice was done using isoflurane. The surgical approach used in adult mice was similar to the approach described above for neonatal mice, except the bulla was opened to expose the round window membrane for injections.

Auditory testing. Auditory brainstem response (ABR) testing was used to evaluate hearing sensitivity between postnatal day 30 (P30) and P90. Animals were anesthetized with ketamine (100 mg/kg) and xylazine (10 mg/kg) via intraperitoneal injections and placed on a warming pad inside a sound booth (ETS-Lindgren Acoustic Systems, Cedar Park, TX). The animal's temperature was maintained using a closed feedback loop and monitored using a rectal probe (CWE Incorporated, TC-1000, Ardmore, PN). Subdermal needle electrodes were inserted at the vertex (+) and test-ear mastoid (−) with a ground electrode under the contralateral ear. Stimulus generation and ABR recordings were completed using Tucker Davis Technologies hardware (RZ6 Multi I/O Processor, Tucker-Davis Technologies, Gainesville, FL) and software (BioSigRx, v.5.1). ABR thresholds were measured at 4, 8, 16, and 32 kHz using 3-ms, Blackman-gated tone pips presented at 29.9/second with alternating stimulus polarity. At each stimulus level, 512–1,024 responses were averaged. Thresholds were determined by visual inspection of the waveforms and were defined as the lowest stimulus level at which any wave could be reliably detected. A minimum of two waveforms was obtained at the threshold level to ensure repeatability of the response. Physiological results were analyzed for individual frequencies, and then averaged for each of these frequencies from 4 to 32 kHz.

Immunohistochemistry and stereocilia length measurements. After completion of ABR testing, mice were euthanized by CO2 asphyxiation followed by decapitation. Temporal bones were harvested and fixed overnight with 4% paraformaldehyde followed by decalcification in 120 mmol/l EDTA for 4 days. The cochlear sensory epithelia were micro-dissected, blocked, and labeled with mouse anti-myosin 7a antibody to label hair cells (1:250, Developmental Studies Hybridoma Bank), and rabbit anti-whirlin antibody at 1:300, as previously reported.8 Primary and secondary antibodies were diluted in PBS. Rhodamine-conjugated phalloidin was used to label filamentous actin in stereocilia (1:50; Life Technologies, Carlsbad, CA). Images were obtained using a Zeiss LSM780 confocal microscope at 10× and 63× using z-stacks (~0.3 and 3 μm thickness, respectively). Stereocilia length measurements were performed on rhodamine-phalloidin stained stereocilia using Volocity 3D image analysis software (PerkinElmer, Waltham, MA). This allowed for 3D rendering of stereocilia to facilitate stereocilia length measurement. On each IHC, five stereocilia were measured and the values were averaged. Only stereocilia from the tallest row were measured because it is the most prominent and easiest to observe on confocal images, and therefore the measurements are most reliable and precise.

SEM. The microdissected cochlea sensory epithelia fixed in 4% paraformaldehyde in 1× HBSS supplemented with calcium and magnesium and processed for immunocytochemistry were carefully dismounted from the slides and washed in 0.1 M sodium cacodylate buffer followed by immersion in fixative solution containing 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer pH 7.3 with 2 mmol/l CaCl2 for 2 hours at room temperature. Then samples were washed in cacodylate buffer three times 5 minutes each, rinsed in distilled water and dehydrated in graded series of ethanol (EtOH). Specimens were then transferred from 100% EtOH to 100% acetone, placed into metal mesh baskets (Ted Pella, Redding, CA), critical point dried from liquid CO2 (CPD030 Critical Point Dryer, BAL-TEC AG, Florida), and finally sputter-coated with 4-nm thick platinum using turbo-pumped sputter coater Q150T (Quorum Technologies, UK). Samples were mounted on aluminum studs (Electron Microscopy Sciences, Hatfield, PA) and imaged using a field emission scanning electron microscope (S-4800, Hitachi, Japan). When samples were not immunostained prior to SEM analysis, the microdissected organ of Corti samples were immersed directly in fixative solution mentioned above and the rest of the processing was identical.

Statistics. Student's t-test was used to analyze data on stereocilia length, row number, and IHC survival. A one-way analysis of variance was used for analyses of infection efficiency and ABR data. Statistical significance was defined as a P < 0.05.

SUPPLEMENTARY MATERIAL Figure S1. Stereocilia length and row number are not restored in whirler mice that received AAV8-whirlin gene therapy as adults. Figure S2. AAV8-whirlin infected primarily IHCs.

Acknowledgments

This work was supported by the intramural NIDCD surgeon-scientist training program and the Triological Society Research Career Development Award (to W.W.C.). There are no conflicts of interest. We thank Andrew J. Griffith (NIDCD) and Carmen C. Brewer (NIDCD) for critiquing the manuscript.

Supplementary Material

Supplementary Figures
Click here for additional data file. (100.7KB, pdf)

References

  1. Morton, NE (1991). Genetic epidemiology of hearing impairment. Ann N Y Acad Sci 630: 16–31. [DOI] [PubMed] [Google Scholar]
  2. Lenz, DR and Avraham, KB (2011). Hereditary hearing loss: from human mutation to mechanism. Hear Res 281: 3–10. [DOI] [PubMed] [Google Scholar]
  3. Drummond, MC, Belyantseva, IA, Friderici, KH and Friedman, TB (2012). Actin in hair cells and hearing loss. Hear Res 288: 89–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Richardson, GP, de Monvel, JB and Petit, C (2011). How the genetics of deafness illuminates auditory physiology. Annu Rev Physiol 73: 311–334. [DOI] [PubMed] [Google Scholar]
  5. Belyantseva, IA, Labay, V, Boger, ET, Griffith, AJ and Friedman, TB (2003). Stereocilia: the long and the short of it. Trends Mol Med 9: 458–461. [DOI] [PubMed] [Google Scholar]
  6. Friedman, TB, Hinnant, JT, Fridell, RA, Wilcox, ER, Raphael, Y and Camper, SA (2000). DFNB3 families and Shaker-2 mice: mutations in an unconventional myosin, myo 15. Adv Otorhinolaryngol 56: 131–144. [DOI] [PubMed] [Google Scholar]
  7. Bird, JE, Takagi, Y, Billington, N, Strub, MP, Sellers, JR and Friedman, TB (2014). Chaperone-enhanced purification of unconventional myosin 15, a molecular motor specialized for stereocilia protein trafficking. Proc Natl Acad Sci USA 111: 12390–12395. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Belyantseva, IA, Boger, ET, Naz, S, Frolenkov, GI, Sellers, JR, Ahmed, ZM et al. (2005). Myosin-XVa is required for tip localization of whirlin and differential elongation of hair-cell stereocilia. Nat Cell Biol 7: 148–156. [DOI] [PubMed] [Google Scholar]
  9. Delprat, B, Michel, V, Goodyear, R, Yamasaki, Y, Michalski, N, El-Amraoui, A et al. (2005). Myosin XVa and whirlin, two deafness gene products required for hair bundle growth, are located at the stereocilia tips and interact directly. Hum Mol Genet 14: 401–410. [DOI] [PubMed] [Google Scholar]
  10. Manor, U, Disanza, A, Grati, M, Andrade, L, Lin, H, Di Fiore, PP et al. (2011). Regulation of stereocilia length by myosin XVa and whirlin depends on the actin-regulatory protein Eps8. Curr Biol 21: 167–172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Fleming, J, Rogers, MJ, Brown, SD and Steel, KP (1994). Linkage analysis of the whirler deafness gene on mouse chromosome 4. Genomics 21: 42–48. [DOI] [PubMed] [Google Scholar]
  12. Mburu, P, Mustapha, M, Varela, A, Weil, D, El-Amraoui, A, Holme, RH et al. (2003). Defects in whirlin, a PDZ domain molecule involved in stereocilia elongation, cause deafness in the whirler mouse and families with DFNB31. Nat Genet 34: 421–428. [DOI] [PubMed] [Google Scholar]
  13. Wang, L, Zou, J, Shen, Z, Song, E and Yang, J (2012). Whirlin interacts with espin and modulates its actin-regulatory function: an insight into the mechanism of Usher syndrome type II. Hum Mol Genet 21: 692–710. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. de Nooij, JC, Simon, CM, Simon, A, Doobar, S, Steel, KP, Banks, RW et al. (2015). The PDZ-domain protein Whirlin facilitates mechanosensory signaling in mammalian proprioceptors. J Neurosci 35: 3073–3084. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Mustapha, M, Chouery, E, Chardenoux, S, Naboulsi, M, Paronnaud, J, Lemainque, A et al. (2002). DFNB31, a recessive form of sensorineural hearing loss, maps to chromosome 9q32-34. Eur J Hum Genet 10: 210–212. [DOI] [PubMed] [Google Scholar]
  16. Ebermann, I, Scholl, HP, Charbel Issa, P, Becirovic, E, Lamprecht, J, Jurklies, B et al. (2007). A novel gene for Usher syndrome type 2: mutations in the long isoform of whirlin are associated with retinitis pigmentosa and sensorineural hearing loss. Hum Genet 121: 203–211. [DOI] [PubMed] [Google Scholar]
  17. Grati, M, Shin, JB, Weston, MD, Green, J, Bhat, MA, Gillespie, PG et al. (2012). Localization of PDZD7 to the stereocilia ankle-link associates this scaffolding protein with the Usher syndrome protein network. J Neurosci 32: 14288–14293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Michalski, N, Michel, V, Bahloul, A, Lefèvre, G, Barral, J, Yagi, H et al. (2007). Molecular characterization of the ankle-link complex in cochlear hair cells and its role in the hair bundle functioning. J Neurosci 27: 6478–6488. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kikkawa, Y, Mburu, P, Morse, S, Kominami, R, Townsend, S and Brown, SD (2005). Mutant analysis reveals whirlin as a dynamic organizer in the growing hair cell stereocilium. Hum Mol Genet 14: 391–400. [DOI] [PubMed] [Google Scholar]
  20. Lane, PW (1963). Whirler mice: a recessive behavior mutation in linkage group VIII. J Hered 54: 263–266. [DOI] [PubMed] [Google Scholar]
  21. Holme, RH, Kiernan, BW, Brown, SD and Steel, KP (2002). Elongation of hair cell stereocilia is defective in the mouse mutant whirler. J Comp Neurol 450: 94–102. [DOI] [PubMed] [Google Scholar]
  22. Mogensen, MM, Rzadzinska, A and Steel, KP (2007). The deaf mouse mutant whirler suggests a role for whirlin in actin filament dynamics and stereocilia development. Cell Motil Cytoskeleton 64: 496–508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mustapha, M, Beyer, LA, Izumikawa, M, Swiderski, DL, Dolan, DF, Raphael, Y et al. (2007). Whirler mutant hair cells have less severe pathology than shaker 2 or double mutants. J Assoc Res Otolaryngol 8: 329–337. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kilpatrick, LA, Li, Q, Yang, J, Goddard, JC, Fekete, DM and Lang, H (2011). Adeno-associated virus-mediated gene delivery into the scala media of the normal and deafened adult mouse ear. Gene Ther 18: 569–578. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liv, N, Zonnevylle, AC, Narvaez, AC, Effting, AP, Voorneveld, PW, Lucas, MS et al. (2013). Simultaneous correlative scanning electron and high-NA fluorescence microscopy. PLoS One 8: e55707. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Zheng, QY, Johnson, KR and Erway, LC (1999). Assessment of hearing in 80 inbred strains of mice by ABR threshold analyses. Hear Res 130: 94–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Akil, O, Seal, RP, Burke, K, Wang, C, Alemi, A, During, M et al. (2012). Restoration of hearing in the VGLUT3 knockout mouse using virally mediated gene therapy. Neuron 75: 283–293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Yu, Q, Wang, Y, Chang, Q, Wang, J, Gong, S, Li, H et al. (2014). Virally expressed connexin26 restores gap junction function in the cochlea of conditional Gjb2 knockout mice. Gene Ther 21: 71–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Chien, WW, McDougald, DS, Roy, S, Fitzgerald, TS and Cunningham, LL (2015). Cochlear gene transfer mediated by adeno-associated virus: Comparison of two surgical approaches. Laryngoscope (epub ahead of print). [DOI] [PubMed]
  30. Luebke, AE, Foster, PK, Muller, CD and Peel, AL (2001). Cochlear function and transgene expression in the guinea pig cochlea, using adenovirus- and adeno-associated virus-directed gene transfer. Hum Gene Ther 12: 773–781. [DOI] [PubMed] [Google Scholar]
  31. Ye, Q, Tillein, J, Hartmann, R, Gstoettner, W and Kiefer, J (2007). Application of a corticosteroid (Triamcinolon) protects inner ear function after surgical intervention. Ear Hear 28: 361–369. [DOI] [PubMed] [Google Scholar]
  32. Cheatham, MA, Huynh, KH, Gao, J, Zuo, J and Dallos, P (2004). Cochlear function in Prestin knockout mice. J Physiol 560(Pt 3): 821–830. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Figures
Click here for additional data file. (100.7KB, pdf)

Articles from Molecular Therapy are provided here courtesy of The American Society of Gene & Cell Therapy

RESOURCES