FocusDEAFNESS

Gene therapy for deafness: How close are we?

See allHide authors and affiliations

Science Translational Medicine  08 Jul 2015:
Vol. 7, Issue 295, pp. 295fs28
DOI: 10.1126/scitranslmed.aac7545

Abstract

Virus-mediated transfer of genes encoding the mechanotransducer channel candidates TMC1 and TMC2 into hair cells of the ear partially restores hearing in animal models of human genetic deafness (Askew et al., this issue).

Approximately 360 million people—5% of the world’s population—suffer from disabling hearing impairment (HI), according to the World Health Organization (www.who.int/features/factfiles/deafness/en). HI commonly causes social isolation, depression, and reduction in professional capabilities. Therefore, understanding how hearing works and combating HI is of great importance. So far, despite major research efforts, a causal treatment based on pharmacology, gene therapy, or stem cells is not yet clinically available for its most common form: sensorineural HI, which arises from dysfunction of the cochlea or the auditory nerve. Whereas disorders such as noise-induced and age-related sensorineural HI prevail in the adult population, at least 50% of sensorineural HI in neonates is caused by defects in individual genes (monogenic deafness), of which ~100 have been identified so far (http://hereditaryhearingloss.org). Recent studies, including one in this issue of Science Translational Medicine by Askew et al. (1), raise hope that gene therapy approaches to HI will become available, at least for select forms of genetic deafness.

GENETIC MANIPULATION

For a few years now, adenovirus and adeno-associated virus (AAV) have become key for efficient gene transfer into the postmitotic hair cells, which are the mechanosensory receptor cells in the ear, and spiral ganglion neurons, which are nerve cells that carry the information from the cochlea to the brain (25). What has been a “workhorse” for analysis of structure and function of proteins in less complex experimental preparations, such as cultured cells, is becoming available much more slowly in auditory neuroscience because several challenges such as virus delivery to the complex anatomy and cell-type specificity of transduction had to be met. Some of the gene transfer approaches chose an elaborate transuterine procedure, injecting virus into the embryonic mouse ear (3, 5). Likewise, the application of virus to the postnatal ear requires sophisticated technique in order to avoid surgery-related damage to the middle or inner ear (2, 4, 6).

Nevertheless, hearing restoration has been demonstrated in a gene therapy study (2) in which rescue of hearing to near normal levels was observed in mice in which Vglut3 was knocked out upon reinstating the expression of the vesicular glutamate transporter 3 (VGLUT3) by using postnatal AAV-mediated delivery. VGLUT3 is required for the loading of glutamate into the synaptic vesicles of hair cells and, consequently, for encoding sensory information at their afferent synapses in the basolateral pole of the hair cell (Fig. 1, A and C). In humans, mutations in the coding gene VGLUT3 cause autosomal dominant HI—auditory synaptopathy. Human auditory synaptopathy also results from defects in the genes coding for otoferlin and the calcium ion (Ca2+) channel that both are too large to be carried by AAV but could use adenovirus or dual AAV approaches for gene therapeutic restoration of hearing.

Fig. 1 Gene therapy targets for deafness.

(A) Schematic drawing of a cochlear inner hair cell (IHC), with a hair bundle (rows of stereocilia) at the apical end and afferent ribbon synapses at the basolateral end. Mechanical vibrations deflect the hair bundle, exerting force on the mechanotransducer channel via so-called “tip links” (proteins connecting the tips of the stereocilia). The tip links pull open the channel, which is believed to reside in the tip of the next lower stereocilium. (B) Askew et al. (1) posit that TMC 1 and its ortholog, TMC2, form the mechanotransducer channels in hair cells. When TMC1 or TMC2 is absent, as in some forms of deafness, transduction fails. When TMC1 or TMC2 was reinstated via AAV-mediated gene transfer in mice in which Tmc1 was knocked out, transduction was restored in hair cells. (C) Genetic synaptopathy is another target for future gene therapy. The ribbon synapse is highly specialized in its molecular composition, and defects in glutamate uptake, Ca2+ influx, and vesicle exocytosis cause human HI.

CREDIT: H. McDONALD/SCIENCE TRANSLATIONAL MEDICINE

In their study in Science Translational Medicine, Askew et al. targeted the transduction machinery of the hair bundle at the apical pole of the hair cell–specific transmembrane channel–like 1 and -2 (TMC1 and TMC2) (Fig. 1B) (1). This transduction machinery is composed of numerous proteins, many of which are affected in nonsyndromic HI or in Usher syndrome (a combination of inner-ear and retinal disease). For many years, the field has searched for the molecular identity of the key component of the transduction machinery: the mechanotransducer channel. TMC1 and TMC2 have emerged as promising candidates for the mechanotransducer channel. They are required for hair cell mechanotransduction (Fig. 1B), which is defective in human deafness, and their genetic manipulation modulates the properties of mechanotransducer currents at the single-channel level.

Askew et al. now advance the field by demonstrating that reinstating TMC expression in the hair cells restores the mechanotransducer current and partial hearing (1). Mice lacking TMC1 and TMC2 were administered AAVs with either Tmc1 or Tmc2 postnatally. The animals demonstrated sizable mechanotransducer currents when expressing either TMC1 or TMC2. Interestingly, adaptation properties of the mechanotransducer current depended on which of the TMCs was expressed. At the systems level, the authors found that transgenic expression of either TMC1 or TMC2 enabled modest auditory activity, as reported by auditory brainstem responses (ABRs) and acoustic startle reflexes in otherwise deaf TMC1 mutant mice. One reason why hearing was only partially restored could be the minimal transduction of outer hair cells (~5%), whereas ~65% of inner hair cells expressed TMC1 at the stereocilia tip (Fig. 1B). Furthermore, mice carrying a point mutation in TMC1, known as Beethoven (Bth)—a model of human dominant-progressive hearing loss—also responded to Tmc2 gene therapy, with nearly half of the animals exhibiting a stronger ABR than that of baseline, but neither hearing improvement nor startle response was as dramatic as that seen in mice in which Tmc1 was knocked out, suggesting that Tmc2 expression could not sufficiently overcome the dominant Bth mutation.

However, definitive evidence that TMCs are pore-forming components of the hair cell mechanotransducer channel is still missing. For example, for proving that TMCs are the “channel proper,” changing the ionic selectivity of mechanotransducer current by means of site-directed mutagenesis of TMCs should ideally be demonstrated. Moreover, the possibility that deletion of TMCs would impair developmental assembly of the mechanotransduction machinery cannot be ruled out. However, the findings by Askew et al. (1) seem to argue against the latter hypothesis because postnatal expression of TMCs enabled transducer currents. Moreover, the work now paves the way for detailed analysis of TMC structure and function, permitting the field to scrutinize the role of TMCs in mechanotransduction—for instance, by searching for mutations that affect ionic selectivity.

Hearing restoration

The work of Askew et al. has implications for future clinical hearing restoration. AAV is considered to be nonpathogenic for humans, to have low immunogenicity, and to lack oncogenic potential, in contrast with the retrovirus (7, 8). However, although hearing was substantially rescued upon AAV-mediated expression of VGLUT3 in inner hair cells (Fig. 1C) (2), more work is required to improve the hearing restoration by AAV-mediated expression of TMCs so that it would at least suffice for rehabilitation with hearing aids in humans. In addition to transferring correct DNA, approaches to modulate or correct splicing have been used for tackling monogenic HI in animal models. Lentz et al. (9) systemically applied antisense oligonucleotides to correct defective splicing of pre-mRNA transcripts from the gene coding for harmonin, which is required for hair cell mechanotransduction and hearing.

Gene therapeutic approaches are also used beyond gene replacement or transcript correction. For example, an adenovirus-based clinical hearing restoration trial has recently been approved by the U.S. Food and Drug Administration for transdifferentiation of nonsensory cells into sensory hair cells [ClinicalTrials.gov identifier: NCT02132130], which are typically lost in HI. Based on previous work in animals (6), this trial will use adenovirus for ectopic expression of the transcription factor Atoh1 in nonsensory cells of the human cochlea. Last, even when physiological hearing cannot be restored, gene therapy might help improve artificial hearing provided by cochlear implants. This approach will be clinically relevant, as monogenic deafness that is amenable to gene replacement only constitutes the minority of sensorineural HI. To date, more than 450,000 subjects currently use the cochlear implant worldwide, and most of them achieve open speech understanding. Future combination of cochlear implants with the expression of neurotrophic factors will likely improve the survival of spiral ganglion neurons and enable greater proximity of their peripheral neurites to the electrode contacts (10). Optical stimulation of the spiral ganglion neurons is spatially more confined (3) and promises to improve frequency resolution in future implants, which is a major bottleneck in the rehabilitation with the current cochlear implants. Optogenetics requires ectopic expression of light-sensitive channels in neurons and, thus, also greatly profits from advances in AAV-based gene therapeutic approaches.

The study by Askew et al. falls in a time when AAV-mediated gene replacement therapy has proven successful for various genetic disorders. A promising example of gene therapy of sensory disorders is the AAV-mediated expression of the enzyme RPE65 in Leber’s congenital amaurosis (7). Here, a single AAV injection led to a long-lasting improvement of visual function. The pioneering gene therapy studies provide hope that restoration of hearing will become available for select forms of deafness within the next decade; also—and not less important—the strategies developed for genetic manipulation of inner-ear cells aid fundamental research on hearing impairment.

References and Notes

  1. Acknowledgment: The author thanks T. Pangršič for critical reading.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article