Gene Therapy for Deaf Mice Goes Viral

Agroundbreaking new study has demonstrated recovery of auditory function in a mouse model of genetic deafness following gene therapy mediated by an adeno-associated viral vector type 1 (AAV1).¹ The vector carried the coding sequence for the vesicular glutamate transporter type 3 (VGLUT3), which is required for synaptic transmission between the sensory cells of the inner ear and the primary afferent neurons that convey auditory information to the brain. VGLUT3 knockout mice lack the capability to efficiently package the neurotransmitter glutamate into synaptic vesicles, leading to loss of auditory function. Remarkably, in the treated mice, AAV1 was able to transduce 100% of the inner hair cells, which were the therapeutic target for the therapy. This level of gene transfer efficiency is unprecedented in the inner ear, and expression of the exogenous VGLUT3 was shown to be sufficient to restore some auditory functions to normal levels in the otherwise deaf mice. This proof-of-concept study elevates the prospects for gene transfer in the human inner ear as a potential treatment strategy for restoration of hearing and balance function in deaf and dizzy patients.

Hearing loss is the most common sensory deficit, affecting ~30 million Americans, about 250 million people worldwide.^2,3 These numbers are projected to double over the next 20 years as the population expands and ages. Nearly half of the cases are suspected to be of genetic origin, the remainder being the result of overexposure to noxious environmental factors, such as loud sounds, ototoxic drugs, and infectious agents. Of the inherited forms of hearing loss, more than 300 genetic loci have been defined with the causative gene identified for about 70 (ref. 4). Other than mechanical devices, such as hearing aids and cochlear implants, there are currently no viable treatments for hearing loss. Although the success of cochlear implants for some patients cannot be overstated, they are indicated for only a subset of patients and provide only a partial restoration of sound sensitivity.² As such, development of biological therapies for hearing loss is a high priority for inner ear biologists and patients who suffer from hereditary and acquired hearing loss. Although some progress has been made with gene therapies stimulating the regeneration of hair cells,^5,6 rescue of a gene defect in the inner ear has not been previously achieved.

A number of recent successes using gene therapy in the eye to treat blindness have raised hopes that analogous strategies might be translated into success stories for treating deafness. Similar to the eye, the ear is compartmentalized in fluid-filled spaces that bathe the sensory cells, which limits dose requirements and biodistribution, thereby reducing safety concerns. Because loss or damage to primary sensory cells is a major cause of deafness, strategies that target sensory hair cells for repair or replacement may be fruitful. However, unlike in the eye, the sensory cells of the inner ear are embedded deep within the temporal bone, often recognized as the hardest bone of the body. As such, access to the inner ear for vector delivery presents a significant technical barrier. Finally, whereas compartments in the eye benefit from a partial immune privilege, the thresholds for destructive immune activation to vector and transgene antigens in the inner ear are anticipated to be lower and have not been well studied in the context of gene delivery.⁷ Recent work by Akil et al., the focus of this Commentary, has demonstrated significant progress toward overcoming the challenges of targeted gene delivery to the inner ear and restoration of function at the cellular, systems, and behavioral levels.¹

To model human deafness, Akil et al. used a knockout mouse that lacked expression of VGLUT3. The mice are deaf from birth but the sensory cells remain, albeit without the ability to transmit sensory signals to the postsynaptic neurons. To investigate the prospects of gene therapy for hearing loss, the investigators administered AAV1 encoding VGLUT3 to the inner ear in VGLUT3 knockout mice early after birth (P1–P10). Of two surgical approaches investigated, they found that microinjection through the round window of the cochlea allowed introduction of 0.6 to 1 μl containing 1.3 to 2.3 × 10¹⁰ particles of the AAV1 vector without compromising the integrity of the sensory organ (Figure 1). The injection targeted the perilymphatic spaces of the cochlea, where the cell bodies of the sensory hair cells are bathed with fluid. Importantly, the round window, which functions as a thin membrane to relieve sound pressure, was sealed over with fascia and adipose tissue, without ill effect. The study reported successful delivery and expression of VGLUT3 in 100% of the targeted inner hair cells examined. Remarkably, although a similar green fluorescent protein vector targeted a variety of cell types, the AAV1 encoding VGLUT3 driven from a constitutive chicken β-actin promoter demonstrated exclusive inner hair cell expression. This observation led the investigators to speculate that VGLUT3 expression is regulated post-transcriptionally through an unknown mechanism, but may also be due to strong activity of this specific promoter in the inner hair cells.¹

Figure 1.

Schematic representation of the anatomy of the ear, injection routes, therapeutic target cells, and mechanism of functional recovery in VGLUT3 cochlear gene augmentation. AAV, adeno-associated virus; CO, cochleostomy method; RWM, round-window method. (Steven Moskowitz, Advanced Medical Graphics)

The injected animals were assayed for recovery of presynaptic morphology in the transfected cells and recovery of auditory function using measurements of auditory brainstem responses (ABRs). The ABR data were remarkable, as they showed near-normal thresholds, suggesting restoration of sound sensitivity at two weeks after treatment. The longevity of the rescue, however, was impacted dramatically by the time of vector injection after birth, with only a very early injection leading to a long-lived hearing recovery in all mice up to nine months. Functionally, mice treated in a single ear had partially restored auditory startle reflexes, which improved up to 33% when both ears received vector. Although the ABR threshold recovery was nearly complete, an incomplete recovery of ABR amplitude and latency, reflecting the synchrony and speed of signal transmission, may underlie the limited behavioral rescue.

These successes bode well for the future of inner ear gene therapy, yet a number of important basic and translational questions remain that must be addressed before these murine proof-of-concept results can be translated into the clinic. For direct translation of this approach, the contribution of VGLUT3 to human disease must be further understood; although it has become clear that the human homo­logue of VGLUT3 is associated with progressive hearing loss in humans, the inheritance is autosomal dominant.⁸ It will also be essential to evaluate in vivo vector biology in the inner ear further, particularly in larger animals that better model the scale, surgical access, and host immunity of the human setting. Furthermore, it will be essential to study the longevity of expression, and what determinants may have led to the loss of rescue in animals injected at P10–P12 noted by Akil et al., as this limits the therapeutic window significantly. The need for early intervention—and therefore a short window of opportunity—obviously negatively impacts clinical feasibility of any future gene therapy, emphasizing the importance of understanding whether this observation is a function of surgery, vector, gene delivery, disease progression, or several of these. Hopefully, then, ways can be identified to overcome these underlying factors to allow for a later and more long-lived therapeutic intervention. Finally, perfect gene targeting to the inner hair cells with complete recovery of a critical electrical measure of hearing (ABR thresholds) but without full recovery of ABR latency, amplitude, and behavioral responses suggests a missing link pivotal for making this strategy fully efficacious. This is important in light of the relative success and wide use of cochlear implants, as a future gene therapy will probably have to present a better safety or efficacy profile to be competitive. Because cochlear implants are capable of restoring sensitivity to only a limited number of sound frequencies, biological therapies such as gene therapy carry the hope of extending hearing restoration over the entire frequency range of human auditory perception.

In conclusion, the work by Akil et al. represents a significant proof-of-concept advance for the field, as it highlights the need for more research in the basic biology of auditory function, dysfunction, and repair, and loudly proclaims a sound future of inner ear gene therapy.