Extended time frame for restoring inner ear function through gene therapy in Usher1G preclinical model

Abstract

Neonatal gene therapy has been shown to prevent inner ear dysfunction in mouse models of Usher syndrome type I (USH1), the most common genetic cause of combined deafness-blindness and vestibular dysfunction. However, hearing onset occurs after birth in mice and in utero in humans, making it questionable how to transpose murine gene therapy outcomes to clinical settings. Here, we sought to extend the therapeutic time window in a mouse model for USH1G to periods corresponding to human neonatal stages, more suitable for intervention in patients. Mice with deletion of Ush1g (Ush1g-/-) were subjected to gene therapy after the hearing onset. The rescue of inner ear hair cell structure was evaluated by confocal imaging and electron microscopy. Hearing and vestibular function were assessed by recordings of the auditory brain stem response and vestibulo-ocular reflex and by locomotor tests. Up to P21, gene therapy significantly restored both the hearing and balance deficits in Ush1g-/- mice. However, beyond this age and up to P30, vestibular function was restored but not hearing. Our data show that effective gene therapy is possible in Ush1g-/- mice well beyond neonatal stages, implying that the therapeutic window for USH1G may be wide enough to be transposable to newborn humans.

Keywords: Gene therapy; Mouse models; Therapeutics.

Figure 1. AAV2/Anc80L65 efficiently transduces mature inner ear hair cells.

(A) The top panel shows a schematic representation of the inner ear (the cochlea and vestibular sensory epithelia in purple), illustrating viral vector injection through the round window membrane (RWM) of the cochlea. The vestibular hair cells (VHCs) are located in the crista ampullae (AC) of each semicircular canal and in the utricular and saccular macula (UM and SM) (left), whereas the sensory hair cells of the cochlea, inner hair cells (IHCs) and outer hair cells (OHCs), are harbored in the organ of Corti (right). OW, oval window; RW, round window membrane. (B) Organ of Corti, spanning from the middle to apical turns of the cochlea in a wild-type mouse, underwent AAV2/Anc80L65-CMV-GFP injection on P20 and was immunostained for myosin 6 (in red) and GFP (in green) on P25. (C and D) Maximum-intensity projections of confocal z-sections of IHCs from the cochlear middle turn (C) and SM (D). Almost all IHCs and most VHCs were transduced with AAV2/Anc80L65. Scale bars: 50 μm (B); 10 μm (C and D).

Figure 2. AAV2/Anc80L65 vector-mediated transfer of Sans cDNA fused to GFP restores protein expression and targeting.

(A) Schematic representation of the mechanoelectrical transduction (MET) machinery in the hair bundle of inner ear hair cells. The Ush1g protein (Sans protein) is located in the upper tip link density, together with the myosin 7A (MYO7A) and the harmonin. The tip link connecting 2 stereocilia is composed of the cadherin23 (CDH23) and the protocadherin15 (PCDH15) proteins. White arrowhead points the three rows of OHCs. (B and C) Confocal images of auditory (B) and vestibular (C) hair cells from P32 Ush1g^–/– mice following intracochlear injection of AAV2/Anc80L65-CMV-GFP-Sans-WPRE at P18, immunolabeled for actin (in orange) and GFP (in green), showing that the GFP-Sans exhibits a localization pattern mirroring the typical distribution of the protein Sans in the transduced inner hair cells (IHCs), outer hair cells (OHCs), and vestibular hair cells (VHCs). GFP-Sans is localized at the tip of the stereocilia (white arrow). The dotted circles delineate transduced (green) and nontransduced (white) VHCs. Scale bars: 5 μm (low-magnification images); 1 μm (high-magnification images).

Figure 3. Viral vector-mediated transfer of the Sans cDNA at a mature stage restores cochlear hair cell architecture.

(A) Confocal images of auditory hair cells from P45 wild-type (left), untreated Ush1g^–/– (middle), and treated Ush1g^–/– (right) mice, immunolabeled for actin (in orange). DAPI was used to stain the nuclei (in blue). White arrowheads points the three rows of the outer hair cells (OHCs) and one row of inner hair cells (IHCs). Scale bars: 10 μm. (B) A close-up view of the outer hair cells (OHCs) and inner hair cells (IHCs) at high magnification, showcasing the recovery of a near-normal stereocilia shape in the treated mice. Scale bars: 3 μm. (C) Low- and intermediate-magnification scanning electron micrographs of the organ of Corti of wild-type (left), untreated Ush1g^–/– (middle), and treated (at right) Ush1g^–/– mice, showing a partial restoration of hair-bundle architecture in IHCs and OHCs. Scale bars: 50 μm (low-magnification images); 1 μm (intermediate-magnification images). (D) Comparative analyses of stereocilia number and length in the tallest row in both OHCs and IHCs showing a significant improvement in Ush1g^–/– mice after gene therapy treatment (blue) relative to untreated mice (red), the measurements for wild-type mice are shown in black (1-way ANOVA). Number of stereocilia in the higher row of OHCs and IHCs and length of OHC stereocilia were assessed, respectively, in 10, 8, and 9 treated Ush1g^–/– mice; 8, 6, and 8 untreated Ush1g^–/– mice; and 10, 8, and 6 wild-type mice. *P < 0.05 and ****P < 0.0001.

Figure 4. Viral vector-mediated transfer of Sans cDNA restores vestibular hair cell architecture.

(A) Low- and high-magnification (inset) confocal microscopy images of utricular hair cells from P45 wild-type (left), untreated Ush1g^–/– (middle), and treated (right) Ush1g^–/– mice, immunolabeled for actin (in orange), demonstrating that the vestibular sensory epithelium of the treated Ush1g^–/– mouse is populated with hair cells with stereocilia of nearly normal shape, similar to those in the wild-type mouse. In contrast, the vestibular epithelium of the untreated Ush1g^–/– mouse is entirely devoid of hair cells with stereocilia. Scale bars: 10 μm (low-magnification images); 5 μm (high-magnification images). (B) Low- and high-magnification scanning electron micrographs of the utricular sensory epithelium of P112 wild-type (top), untreated Ush1g^–/– (middle), and treated Ush1g^–/– (bottom) mice. These results show that gene therapy treatment prevents vestibular hair cell (VHC) degeneration and restores the staircase pattern of hair bundles. (C) Comparative analyses of the hair cell stereocilia on P40, with treated Ush1g^–/– mice displaying a full recovery of stereocilia length (blue), compared with wild-type mice (black) and untreated Ush1g^–/– mice (red) (top, P < 0.0001, 1-way ANOVA; n = 13, 16, and 8, respectively, for treated Ush1g^–/–, untreated Ush1g^–/–, and wild-type mice). On P112, considerable heterogeneity in VHC stereocilium diameter is observed in untreated Ush1g^–/– mice; this heterogeneity was significantly decreased after gene therapy treatment (bottom, P = 0.009, 1-way ANOVA; n = 9, 13, and 4, respectively, for treated Ush1g^–/–, untreated Ush1g^–/–, and wild-type mice). **P < 0.01 and ****P < 0.0001.

Figure 5. AAV2/Anc80L65-Sans gene therapy at the P12–P21 stage restores hearing in Ush1g^–/– mice.

(A) Auditory brain stem response (ABR) traces for 15 kHz stimulation in wild-type, treated Ush1g^–/–, and untreated Ush1g^–/– mice. (B) ABR thresholds in P40 wild-type (n = 3), untreated Ush1g^–/– (n = 5), and treated Ush1g^–/– mice (n = 12), showing a partial recovery for the 10, 15, and 20 kHz frequencies (2-way ANOVA). ****P < 0.0001.

Figure 6. AAV2/Anc80L65-Sans gene therapy at the P12–P30 stage restores the balance function in Ush1g^–/– mice.

(A–C) Locomotor tests performed at P40, comparing wild-type (black), untreated Ush1g^–/– (red), and treated Ush1g^–/– mice after unilateral (orange) or bilateral (blue) AAV2/Anc80L65-Sans injection. (A) Video tracking, performed on 5, 6, 8, and 10 bilaterally, unilaterally treated, untreated Ush1g^–/–, and wild-type mice, respectively, showed a significant improvement after bilateral injection in circling behavior (P = 0.001, 1-way ANOVA) and (B) distance traveled (P < 0.0001, 1-way ANOVA) relative to untreated Ush1g^–/– mice. (C) In the platform test, Ush1g^–/– mice subjected to unilateral (P = 0.019, 1-way ANOVA, n = 9) or bilateral (P < 0.0001, 1-way ANOVA, n = 11) injections of gene therapy agent spent a longer time on the platform than untreated mice (n = 10); 9 wild-type mice were analyzed. (D and E) Angular vestibulo-ocular reflex gain (aVOR; D) and static ocular counter roll gain (OCR; E), determined by video-oculography after sinusoidal horizontal rotation (at different frequencies) and static head tilt roll, at P70, on 9, 3, 4, and 11 bilaterally treated (blue), unilaterally treated (orange), untreated Ush1g^–/– (red), and wild-type mice (black), respectively. Static OCR gain was significantly greater after bilateral injection than in untreated Ush1g^–/– mice (P = 0.0063, 1-way ANOVA). *P < 0.05, **P < 0.01, ****P < 0.0001.