Potential Treatments for Genetic Hearing Loss in Humans: Current Conundrums

REVIEW Potential treatments for genetic hearing loss in humans: current conundrums

R Minoda, T Miwa, M Ise and H Takeda

Genetic defects are a major cause of hearing loss in newborns. Consequently, hearing loss has a profound negative impact on human daily living. Numerous causative genes for genetic hearing loss have been identiﬁed. However, presently, there are no truly curative treatments for this condition. There have been several recent reports on successful treatments in mice using embryonic gene therapy, neonatal gene therapy and neonatal antisense oligonucleotide therapy. Herein, we describe state-of-the-art research on genetic hearing loss treatment through gene therapy and discuss the obstacles to overcome in curative treatments of genetic hearing loss in humans.

Gene Therapy (2015) 22, 603–609; doi:10.1038/gt.2015.27

INTRODUCTION NORMAL INNER EAR ANATOMY AND DEVELOPMENT Sensorineural hearing loss (SNHL) is the most common congenital The inner ear has two basic functions: hearing, which occurs in the disease in humans. The incidences of profound SNHL at birth in cochlea, and balancing, which occurs in the semicircular canals the United Kingdom and United States are 133 per 100,000 and and vestibule. The cochlea is divided into three compartments: 186 per 100,000 births, respectively.1 A genetic defect is the most the scala vestibule, scala tympani and scala media. The scala common cause of hearing loss at birth and in childhood. More media, a part of the endolymphatic space, contains the organ of than half of all neonates with SNHL have inherited hearing loss. Corti (OC). The OC contains three cell populations: inner hair cells Approximately 70% of hereditary hearing loss cases are non- (IHCs), outer hair cells (OHCs) and supporting cells (SCs) (Figure 1). syndromic and ~ 30% are syndromic SNHL.2,3 Hereditary forms of Hair cells have stereocilia that emerge from their apical surface. fl genetic SNHL are as follows: 80% autosomal recessive, ~ 20% Receptor potentials, generated by de ection of the stereocilia autosomal dominant, ~ 1% X linked and ⩾ 1% via mitochondrial within the IHCs, induce neurotransmitter release at the synaptic ends.15 Therefore, sound waves are transmitted via the outer and inheritance.4,5 Autosomal dominant SNHL often takes a post- middle ear to the inner ear fluid in the cochlea and transduced to lingual and progressive form, whereas autosomal recessive SNHL 5 electrical signals via IHCs. These signals are subsequently takes a prelingual form. The most common genetic cause of non- transmitted to the brain via efferent neurons and perceived syndromic SNHL is a mutation within the GJB2 gene, which 6–8 as sound. encodes connexin 26 (Cx26). The second most frequent cause The mammalian inner ear and its sensory neurons develop from of non-syndromic SNHL is a mutation in the GJB6 gene, the otic placode, a thickened patch of head ectoderm.16–18 8,9 which encodes Cx30. The most common form of syndromic Subsequently, one otocyst (per side) is formed by invagination of SNHL is Pendred syndrome, which is associated with mutations the otic placode at the level of the hindbrain at 4 weeks gestation within the solute carrier family 26 (SLC26A4) gene.10,11 Mutations in humans19 and embryonic day 9.5 (E9.5) in mice. Soon after, in mitochondrial DNA can also induce syndromic and non- formation of the otocyst and neuroblasts delaminate from the syndromic hearing loss.5,12 There are many reported causative ventral region of the otocyst. These neuroblasts will coalesce genes of genetic hearing loss in addition to those described adjacent to the developing inner ear and begin to form the above.4 statoacoustic ganglion.20 By E12.5 in mice, the positions of the Although numerous causative genes for genetic hearing loss developing sensory patches, which form from a single common – have been identified, there are no truly curative treatments for this patch in the otocyst, can be identified.20 22 The cochlear part of condition yet. At present, treatments for SNHL only include the otocyst then begins to elongate into a spiral structure. The two hearing aids and/or cochlear implants. While hearing aids and and one-half turns of the coiled cochlea are not completed until 19 cochlear implants are helpful treatments to compensate for 25 weeks gestation in humans, whereas the mouse cochlear hearing loss, they do not restore hearing to normal levels. There is duct has completed three-quarters of one turn around E13.5. still an urgent need for truly curative treatments. Recently, there have been several reports on successful treatment of genetic hearing loss caused by Cx30, vesicular glutamate transporter 3 A THEOREM OF GENETIC HEARING LOSS TREATMENTS (VGLUT3) and Usher syndrome 1c (USH1C) gene mutations, which During development, appropriate spatiotemporal control of gene are all known genetic causes of hearing loss in humans.4,13,14 expression is necessary for normal development of the inner ear.

Departments of Otolaryngology-Head and Neck Surgery, Kumamoto University, Graduate School of Medicine, Kumamoto City, Kumamoto, Japan. Correspondence: Dr R Minoda, Departments of Otolaryngology-Head and Neck Surgery, Kumamoto University, Graduate School of Medicine, 1-1-1 Honjo Chuoku, Kumamoto City, Kumamoto 860-0811, Japan. E-mail: [email protected] Received 16 December 2014; revised 24 January 2015; accepted 12 February 2015; accepted article preview online 17 March 2015; advance online publication, 9 April 2015 Potential genetic hearing loss treatments R Minoda et al 604 Data from neonatal hearing screening tests have demonstrated that the majority of SNHL patients can be detected using this method.5 Indeed, hearing loss caused by Cx26 mutations, the most common genetic cause of non-syndromic SNHL, also usually presents with a congenital onset.24 Therefore, embryonic treatments are inevitable if the treatment must be administered before hearing loss phenotype occurrence in such congenital genetic hearing loss patients. One important issue regarding embryonic treatments is the need to treat embryonic inner ears in the maternal uterus, which is technically feasible, but not facile. Thus, experiments targeting embryonic mouse inners ears are very complex and involved. Because of this intricateness, most animal studies on genetic hearing loss treatments have not targeted embryonic inner ears, but rather neonatal mouse inner ears. When we use rodents, particularly mice, to study genetic hearing loss treatments, we need to be aware of differences in the development and maturation of auditory functions between rodents and humans. Auditory function initiates around postnatal day 13 (P13) in mice, but at 20 weeks gestation in humans.25,26 Figure 1. Transverse section image of the cochlea. The adult Even if an effective treatment is administered during the mouse mammalian cochlea is divided into three compartments: the scala neonatal period, the respective neonatal treatment would likely vestibule, scala tympani and scala media. This image represents a be ineffective in humans, as mouse neonatal treatment equates to cross-section of the scala media, which contains the OC. The OC human embryonic treatment. contains three cell populations: IHCs, OHCs and SCs. The two types When considering treatment for genetic hearing loss caused by of auditory hair cells (IHCs and OHCs) have critical roles as gain-of-function mutations, treatments in which new molecular mechanoelectrical transducers for hearing. Auditory hair cells are covered by the tectorial membrane. The stria vascularis, located in functions are suppressed through RNA interference or degradation the lateral wall of the scala media, is responsible for the secretion of of the mutated gene product would likely be successful. As K+ into the endolymph and for production of the endocochlear mentioned previously, timing is an important factor for determin- potential. ing the simplicity of a treatment and subsequent methods effective for loss-of-function mutations. This relationship in loss- of-function mutations is generally the same as those in gain-of- During this process, expression of a gene begins; subsequently, function mutations. One difference between loss-of-function and the gene expression gradually becomes more widespread, and gain-of-function mutations is that postnatal treatments may be reaches a spatiotemporal maximum. Disruption of this process by more feasible in cases of genetic SNHL caused by a gain-of- a genetic mutation during inner ear development can cause function mutation, which typically occurs postlingually;5 specifi- genetic hearing loss. There are two major classes of genetic cally, gain-of-function mutations usually do not present as hearing mutations: loss-of-function and gain-of-function. In loss-of- loss during the neonatal period. Additionally, neonatal hearing function mutations, the most common form, the protein product screening test data have demonstrated that ~ 15% of preschool of a gene is either missing, non-functional or reduced in level. children with SNHL show progressive hearing loss.27 In such These are typically recessive mutations, because a wild-type allele genetic SNHL patients who do not present with hearing loss can usually compensate for the non-functional allele. In contrast, during the neonatal period, neonatal treatment may be more the altered gene product takes on a new molecular function in feasible for preventing subsequent hearing loss. gain-of-function mutations, which usually follows dominant inheritance, because the presence of a normal allele is not capable of preventing the mutant allele from behaving abnor- TREATMENT FOR GENETIC HEARING LOSS CAUSED BY A LOSS- mally. One gain-of-function mutation subtype is the dominant- OF-FUNCTION MUTATION negative mutation, whereby the product of the mutant gene can Embryonic treatments compete with or inhibit the function of the wild-type product.23 There are three important time points to be cognizant in Although it is an extremely intricate period of development, the embryonic stage is the ideal treatment period for genetic hearing genetic hearing loss treatments: (1) the target gene initiation point 28 at which target gene expression begins in normal individuals loss. Miwa et al. reported successful treatments via transuterine (hereafter referred to as ‘normal gene initiation time’), (2) the time gene transfer to the embryonic inner ear in Cx30-knockout mice. Cx proteins, which assemble to form vertebrate gap junctions, are point at which the normal gene expression matures (hereafter crucial for auditory function.29 A large deletion within the Cx30 referred to as ‘normal gene mature expression time’) and (3) the gene is the second most frequent cause of non-syndromic SNHL, time point at which phenotypes begin to manifest after target 9,30 fi as reported in nine countries. Homozygous Cx30 deletion mice gene de ciency (Figure 2). Among treatments for genetic hearing have severe hearing impairment and demonstrate a complete loss loss caused by loss-of-function mutations, the most effective of endocochlear potential; endocochlear potential represents the would be a gene redeeming treatment, which is matched transepithelial difference in electric potential between the ‘ ’ precisely to the normal gene initiation time. However, before endolymphatic and perilymphatic compartments, and it is crucial manifestation of the hearing loss phenotype, we might be able to for normal hearing function.31,32 treat genetic hearing loss by recovering reduced or missing Miwa et al.28 aimed to determine whether embryonic gene functions via gene or protein transfer. Moreover, when the transfer into the developing inner ear of Cx30-deficient mice could intention is to treat genetic hearing loss after phenotype prevent manifestation of a subsequent hearing loss phenotype. manifestation, patients can likely be treated by redeeming genetic They used electroporation-mediated transuterine gene transfer defects, in addition to regenerating damaged cochleae. Therefore, into otocysts at E11.5 and induced robust transgene expression in treatments initiated before hearing loss phenotype manifestation the cochleae of developing inner ears. Consequently, Miwa et al.28 would be simpler and more effective. showed that gene supplementation to insert the wild-type Cx30

Figure 2. To treat genetic hearing loss caused by loss-of-function mutations, there are three important time points: (1) the target gene initiation point at which target gene expression begins in normal individuals (‘normal gene initiation time’), (2) the time point at which the normal gene expression matures (‘normal gene mature expression time’) and (3) the time point at which phenotypes begin to manifest after target gene deficiency. Miwa et al.28 demonstrated that otocystic gene transfer before the ‘normal gene initiation time’ can prevent subsequent hearing loss in Cx30-deficient mice. Yu et al.39 reported that hearing loss in conditional Cx26-knockout mice was untreatable by AAV vector-based gene therapy in the neonatal period, which may be after the ‘normal gene mature expression time’ but before phenotype manifestation. Akil et al.13 reported that AAV-mediated VGLUT3 gene transfer at P1–P12, which corresponds to the time after initial VGLUT3 gene expression in wild-type mice and probably also corresponds to the time around or after onset of the hearing loss phenotype manifestation, is effective at restoring hearing in VGULT3-deficient mice. Choi et al.42 reported that forced expression of Slc26a4 before the ‘normal gene initiation time’ completely prevented subsequent hearing loss, which was diminished after the ‘normal gene initiation time’. The solid blue lines with arrows indicate successful functional recovery treatment time windows; the solid orange line with arrows indicates a period during which treatment is partially effective. The dotted black lines with arrows indicate the periods during which treatments are ineffective. gene into the otocysts of E11.5 Cx30-knockout mice prevented transfer via the round window membrane at P10–P12 exhibited postnatal hearing loss (Figure 3). Their results demonstrated that a normal hearing threshold as above, but hearing levels were the induction of a target gene before the ‘normal gene initiation maintained in only 1 out of 19 mice after 28 weeks postnatal. time’ can prevent postnatal hearing loss. The Cx30 gene is first These findings suggest that AAV1-mediated VGLUT3 gene transfer expressed around E14 in wild-type mice,33 and Cx30-deficient at P1–P12 is effective at restoring hearing in VGULT3-deficient mice auditory functions begin to degenerate around P4.31 These mice. Furthermore, earlier treatments are more effective than later findings suggest that transfection of wild-type Cx30 before the treatments for maintaining ameliorated hearing levels. ‘normal gene initiation time’ is effective at preventing subsequent Generally, when considering genetic hearing loss treatments, hearing loss phenotype manifestation. These findings also suggest one important issue to address is the determination of an that precise timing of treatment with wild-type gene supplemen- appropriate strategy to induce gene transfection into the tation and ‘normal gene initiation time’ may be unnecessary for cochleae. Very few strategies are able to induce gene transfection 35 achieving positive therapeutic effects. effectively in the cochleae without any consequential damage. One feature of AAV vectors is that they are able to transfect IHCs via administration of AAV vectors into the scala tympani in the Neonatal treatments 36,37 13 cochlea. Administration into the scala tympani is advanta- Akil et al. reported that gene transfer of an adeno-associated geous in that it is less traumatic compared with administration – virus serotype 1 (AAV1) vector at P1 P12 in the cochleae using the into the scala media, because the scala tympani does not contain wild-type VGLUT3 gene significantly ameliorated auditory function 36,38 13 sensory cells, whereas the scala media does (Figure 1). in VGLUT3-knockout mice. VGLUT3 deficiency has been shown Thus, AAV vectors could be the most efficient carrier as a to cause severe hearing loss by P10–P12 in mice because of the treatment for hearing loss caused by VGLUT3 deficiency, as AAV loss of glutamate release at IHC afferent synapses in the vectors are able to effectively transfer VGLUT3 to IHCs less 34 cochleae. In wild-type mice, VGLUT3 is not expressed in IHCs traumatically via scala tympanic administration. 34 at E15, but is expressed by E19. Although there is no further Recently, Yu et al.39 reported that hearing loss in conditional detailed information about the timing of gene expression Cx26-knockout mice was untreatable by AAV vector-based 13 maturation, P1–P12 (the time at which Akil et al. performed neonatal gene therapy.39 They performed Cx26 gene transfer gene transfer), corresponds to the timing after initial VGLUT3 gene using AAV serotype 2/1 hybrid vectors into the scala media of expression in wild-type mice and probably also corresponds to the conditional Cx26-knockout mice at either P0 or P1, and found that timing around or after the onset of the hearing loss phenotype Cx26 gene transfer via AAV vectors induced extensive Cx26 manifestation. Akil et al.13 also reported that all mice that expression in cells lining the scala media of the cochleae. underwent gene transfer into the scala tympani via the round However, auditory brainstem responses did not show significant window membrane at P1–P3 exhibited a normal hearing thresh- hearing amelioration. In wild-type mice, Cx26 is not detectable at old, and hearing levels were maintained in 5 out of 19 mice at E12 or E14 in the developing cochleae but is detectable by E16 in 9 months postnatal. Additionally, all mice that underwent gene developing cochlea ducts, and by E18, this distribution is more

Figure 3. The embryonic stage is the ideal period for genetic hearing loss treatment. Miwa et al.28 reported successful treatment via transuterine gene transfer into the embryonic inner ear in Cx30-knockout mice. Gene transfer was performed by embryonic gene transfer into the developing inner ear in Cx30-deﬁcient mice, which successfully prevented subsequent hearing loss phenotypic manifestation. Electroporation-mediated transuterine gene transfer was performed in otocysts (EUGO) in Cx30-deﬁcient mice at E11.5. Embryos were delivered via C-section at E18.5, and the pups that underwent gene transfer at E11.5 were passed to surrogate dams to raise said embryos; these pups did not demonstrate hearing loss at P30.

widespread.40 In conditional Cx26-knockout mice, degeneration of the cochlea is first detectable in Claudius cells around P8 and in OHCs around P13.41 Therefore, the neonatal treatment timing period adopted by Yu et al.39 may be after ‘normal gene mature expression time’ but before phenotype manifestation. Thus, there were significant differences in the efficacy of AAV vector-based neonatal additive gene therapies between the reports by Akil et al.13 and Yu et al.39 that we will discuss in the next section.

TREATMENT TIMING WINDOW Choi et al.’s42 recent temporal expression study of doxycycline- inducible expression of Slc26a4 using transgenic mice revealed that forced expression of Slc26a4 from E0 to E16.5 completely prevented subsequent hearing loss, which diminished after E16.5 (Figures 2 and 4). Slc26a4, which encodes the anion exchanger, − − − 43,44 pendrin, a transporter of anions such as Cl ,I and HCO3, is a causative gene of Pendred syndrome, which involves development of thyroid goiters and SNHL. Normally, Slc26a4 protein (pendrin) expression begins in the cochlea at E16.5 at the basal turn of the cochlea and is detectable throughout the cochlea by E17.5.45 Although Choi et al.42 did not intend to evaluate genetic hearing Figure 4. Choi et al.42 reported that temporal expression of loss treatment per se, their results do provide us with valuable – information on the appropriate treatment timing window for loss- doxycycline-inducible solute carrier family 26 (Slc26a4) at E0 E16.5 in transgenic mice prevented subsequent hearing loss. The line in of-function genetic hearing loss. Their data also suggest that the fi ‘ the gure indicates the relationship between onset of forced induction of normal target gene expression, before normal gene Slc26a4 expression and the consequent hearing threshold at initiation time’, can prevent putative postnatal hearing loss, and its 1 month after birth. The efficacy of forced expression for preventing treatment efficacy diminished rapidly after the ‘normal gene putative postnatal hearing loss decreased rapidly after E16.5, at initiation time’. Their results further suggest that precise timing of which Slc26a4 expression begins at the basal turn of the cochlea. treatment and normal gene initiation is probably unnecessary to achieve positive therapeutic effects. These results are consistent with data from Miwa et al.,28 which demonstrated that otocystic Consequently, all mice showed severe to profound hearing loss. gene transfer at E11.5 can prevent subsequent hearing loss in If we apply Choi’s finding to treatments of genetic hearing loss Cx30-deficient mice; normal Cx30 initial expression begins caused by loss-of-function type mutations, their results suggest around E14. that gene redeeming therapy after ‘normal gene initiation time’ Choi et al.42 also reported that forcible expression of Slc26a4 appears to rapidly decrease its effectiveness. Thus, Choi et al.’s42 after E18.5 was unable to prevent hearing loss (Figure 4). finding is almost consistent with Yu et al.’s39 results in conditional

Gene Therapy (2015) 603 – 609 © 2015 Macmillan Publishers Limited Potential genetic hearing loss treatments R Minoda et al 607 Cx26-knockout mice, in which additive gene therapy after ‘normal 5ʹ splice site in exon 3 of its premessenger RNA. Splicing mutations gene mature expression time’ was ineffective at hearing cause incorrect translation, thereby generating aberrant harmonin amelioration.41 However, Choi et al.’s42 findings are inconsistent proteins and subsequently causing a loss-of-function type SNHL. with those of Akil et al.13 in VGLUT3-deficient mice, in which Lentz et al.14 reported that SNHL in Ush1c216AA knock-in mice, additive gene therapy around or even after the onset of hearing which are useful animal models for Usher syndrome type 1C loss phenotype manifestation was effective for hearing amelioration. caused by the USH1C 216G4A mutation, is treatable using Although the exact reason for the discrepancy between Choi et al.’s42 intraperitoneal administration of an antisense oligonucleotide and Akil et al.’s13 respective data remain elusive, there may be a (ASO).14 ASOs are currently being tested in a number of clinical significant difference regarding appropriate treatment timing trials as treatments for muscular dystrophy, Crohn’s disease, and windows depending on each specific gene. Each genetic hearing others.47,48 Lentz et al. selected the most effective ASO for loss causative gene induces hearing loss via a different mechan- correcting splicing mutations from several candidates via in vitro ism, and the treatable time window range for each gene might be studies. ASOs were intraperitoneally injected into P3-P5 Ush1- affected by its respective mechanism. c216AA knock-in mice, and their auditory and vestibular functions To summarize, precise timing of treatment and normal gene were assessed at 1, 2 and 3 months of age. While control knock-in initiation is probably unnecessary to achieve positive therapeutic mice that were administered a scrambled ASO showed profound effects. Normal gene redeeming therapy before the ‘normal gene hearing loss and vestibular dysfunction, mice that received an initiation time’ is effective at preventing hearing loss caused by effective ASO showed significant amelioration of auditory function loss-of-function mutations, and its efficacy probably diminishes at both low and mid frequencies, as well as of vestibular functions. rapidly after the ‘normal gene initiation time’. Normal gene Therefore, ASO treatment appears to be a very promising method redeeming therapy after the ‘normal gene mature expression for congenital hearing loss caused by splicing mutations. However, time’ is probably ineffective. However, the efficacy of normal gene there are several issues that must be clarified and others that need redeeming therapy after the ‘normal gene initiation time’ and/or to be resolved. First, we do not know the mechanism by which the the ‘normal gene mature expression time’ might vary depending intraperitoneally injected ASOs are transmitted past both the on the causative genes. blood-labyrinthine and blood-cochlea barriers into the inner ears. Second, we need to develop better curative treatments, since ASO treatment at P3-P5 demonstrated no amelioration at high LOCATIONS THAT ARE CRUCIAL FOR GENETIC HEARING LOSS frequencies.14 It has also been reported that harmonin, which is TREATMENT encoded by Ush1c, is detectable by E15 at the basal turn of the As mentioned, during development, appropriate spatiotemporal cochlea and is detectable at the apical turn as late as P30.49 control of gene expression is necessary for normal development The absence of amelioration at high frequencies observed in Lentz of the inner ear. The pattern of gene expression largely varies by et al.’s study may therefore be attributable to belated treatments location. Ideally, precise matching of gene expression pattern by following ‘normal harmonin initial gene expression’ at the basal location would be preferable to achieve curative functional turn. In addition, to achieve better curative treatments utilizing recovery without any harmful effects. There are several previous ASOs, we should consider that the development and maturation reports that provide valuable information regarding this issue. of auditory functions in rodents occur later than do those in The Slc26a4 protein (pendrin) is first expressed in the normal humans.25,26 Therefore, to achieve more curative treatments, murine inner ear at the following times: endolymphatic sac at the treatments should be performed at earlier stages, such as E11.5; cochlea at 16.5 in the basal turn; saccule and utricle by 14.5; the embryonic stage.14 and ampullae by 16.5.45 Li et al.46 generated Slc26a4 transgenic mice, which expressed Slc26a4 only at the endolymphatic sac without detectable expression in the cochlea or vestibular organs. FUTURE DIRECTIONS They showed that Slc26a4 expression in the endolymphatic sac The above results regarding congenital genetic hearing loss successfully prevented anatomical abnormalities of the inner ear suggest that normal gene redeeming treatments initiated prior to and postnatal auditory and vestibular dysfunction, which are the ‘normal gene initiation time’ are probably most effective at commonly observed in Slc26a4-deleted mice. Their results also preventing subsequent hearing loss. Additionally, after the imply that pendrin supplementation at the endolymphatic sac is ‘normal gene initiation time’, the treatment efficacy probably sufficient to treat Pendred syndrome in humans, which is caused diminishes significantly.42 However, the efficacy of normal gene by a Slc26a4 deficiency. redeeming therapy after the ‘normal gene initiation time’ might Therapeutic gene transfers may cause gene expression in vary depending on the causative genes.13 unintended areas, at locations lacking endogenous expression of Considering the results of Lentz et al.,14 intraperitoneal the causative gene. Miwa et al.28 reported that electroporation- administration of an embryonic ASO may be a simple and mediated transuterine gene transfer into otocysts induced a effective method to treat genetic SNHL caused by splicing broader range of Cx30 gene expression than its original area of mutations. However, wild-type gene supplementation during the expression and caused no harmful effects on inner ear functions.28 embryonic stage is critical for curative treatment of the majority of Thus, it is likely unnecessary to match the ectopic location of genetic SNHL cases caused by loss-of-function mutations. a gene precisely with its endogenous area of expression. However, To achieve this, we need to clarify and resolve several issues. we do need to clarify the locations essential for curative First, there are limited methods available that can achieve gene treatments using gene transfer, and whether gene transfer in transfer in embryonic inner ears at specific periods and locations. excessive areas causes harmful effects on inner ear functions for New transfer methods satisfying such conditions are necessary. each causative gene. Second, although access to the developing inner ear is necessary to achieve gene transfer, currently the only embryonic stage at which we can access developing mouse inner ears is E11.5 due to ANTISENSE OLIGONUCLEOTIDE TREATMENT anatomical location. There are no reported methods that have Splicing mutations in the Ush1C gene, which encodes the accessed developing human inner ears successfully. Therefore, protein harmonin, causes Usher syndrome type 1C, which there is an urgent need to develop reliable and safe methods to involves congenital SNHL, vestibular dysfunction and retinitis manipulate developing inner ears at any stage, in both mice and pigmentosa.14 The USH1C 216G4A mutation, a special cause of humans. An ultrasound imaging system would likely be useful for splicing mutations in Usher syndrome type 1C, creates an aberrant this purpose (Figure 5). Third, treatable time windows and

Figure 5. (a) Sagittal section of E14.5 mouse embryos in the uterus, observed via a small animal ultrasound imaging system with spatial resolution up to 30 μm (Prospect; S-Shape Corporation, New Taipei, Taiwan). The outline of the embryo is clearly detectable. The asterisk indicates the mouth. Arrowheads indicate the uterine wall. The dotted line indicates the plane on which the embryo’s head was observed in (b). (b) Horizontal section images of the head of an E14.5 mouse embryo in the uterus. The arrow indicates the eye. The asterisk indicates the auricle. The circled area represents the cochlea. The inset image in the right corner shows a magniﬁed image of the cochlea; arrowheads indicate the cochlea. Thus, detection of the cochlea using the ultrasound imaging system may be feasible, but the approach to the developing inner ear does not appear to be facile.

essential locations of curative treatments for each gene have yet develop clinically relevant treatments.52 Although we did not discuss to be determined. Additionally, we have summarized the relationship treatments using stem cells, they are likely the best option for genetic between normal gene expression time in the cochlea and efficacy SNHL treatment in patients following phenotypic manifestation. of normal gene redeeming therapy from the data from previous reports (Figure 3). However, as mentioned above, Li et al.’s data46 suggest that Slc26a4 expression in the cochlea might not be CONFLICT OF INTEREST essential for curative treatment. We need to clarify precisely the The authors declare no conflict of interest. relationship between the locations essential for curative treatments and treatment timing. Fourth, to perform embryonic treatment, we must be able to precisely diagnose genetic hearing REFERENCES loss during the early embryonic stages. In humans, the otocysts 1 Morton CC, Nance WE. Newborn hearing screening—a silent revolution. N Engl J are generated at 5 weeks gestation. Therefore, to perform Med 2006; 354: 2151–2164. embryonic treatment, precise information on the patient’s genetic 2 Van Camp G, Willems PJ, Smith RJ. Nonsyndromic hearing impairment: unpar- – mutations and functional outcomes caused by the genetic alleled heterogeneity. Am J Hum Genet 1997; 60:758 764. 3 ACMG. Genetics Evaluation Guidelines for the Etiologic Diagnosis of Congenital mutation must be obtained at the earliest gestational stage fl Hearing Loss. Genetic Evaluation of Congenital Hearing Loss Expert Panel. ACMG possible. Genetic testing using amniotic uid, which currently is statement. Genet Med 2002; 4:162–171. being performed clinically, would be highly useful for this 4 Smith RJ, Bale Jr JF, White KR. Sensorineural hearing loss in children. Lancet 2005; purpose. In addition, it has been reported that fetal cell-free 365:879–890. DNA can be isolated from maternal blood as early as the fifth 5 Schrijver I. Hereditary non-syndromic sensorineural hearing loss: transforming week of gestation.50 Genetic testing using maternal blood could silence to sound. J Mol Diagn 2004; 6:275–284. also be an attractive option for early genetic diagnosis and 6 Kenneson A, Van Naarden Braun K, Boyle C. GJB2 (connexin 26) variants and non- – treatments in the future.51 syndromic sensorineural hearing loss: a HuGE review. Genet Med 2002; 4:258 274. 7 Matos TD, Simoes-Teixeira H, Caria H, Goncalves AC, Chora J, Correia Mdo C, et al. There is a ray of hope for genetic hearing loss treatments, despite Spectrum and frequency of GJB2 mutations in a cohort of 264 Portuguese non- the many issues to overcome. Accumulating knowledge on the syndromic sensorineural hearing loss patients. Int J Audiol 2013; 52: 466–471. cellular and molecular mechanisms involved in inner ear function 8 Hoppman N, Aypar U, Brodersen P, Brown N, Wilson J, Babovic-Vuksanovic D. and morphology, in both mice and humans, will be important to Genetic testing for hearing loss in the United States should include deletion/

Gene Therapy (2015) 603 – 609 © 2015 Macmillan Publishers Limited Potential genetic hearing loss treatments R Minoda et al 609 duplication analysis for the deafness/infertility locus at 15q15.3. Mol Cytogenet 32 Michel V, Hardelin JP, Petit C. Molecular mechanism of a frequent genetic form of 2013; 6: 19. deafness. N Engl J Med 2003; 349:716–717. 9 del Castillo I, Villamar M, Moreno-Pelayo MA, del Castillo FJ, Alvarez A, Telleria D, 33 Sun J, Ahmad S, Chen S, Tang W, Zhang Y, Chen P, et al. Cochlear gap junctions et al. A deletion involving the connexin 30 gene in nonsyndromic hearing coassembled from Cx26 and 30 show faster intercellular Ca2+ signaling than impairment. N Engl J Med 2002; 346: 243–249. homomeric counterparts. Am J Physiol Cell Physiol 2005; 288: C613–C623. 10 Everett LA, Glaser B, Beck JC, Idol JR, Buchs A, Heyman M, et al. Pendred syndrome 34 Seal RP, Akil O, Yi E, Weber CM, Grant L, Yoo J et al. Sensorineural deafness is caused by mutations in a putative sulphate transporter gene (PDS). Nat Genet and seizures in mice lacking vesicular glutamate transporter 3. Neuron 2008; 57: 1997; 17: 411–422. 263–275. 11 Ito T, Choi BY, King KA, Zalewski CK, Muskett J, Chattaraj P, et al. SLC26A4 gen- 35 Fukui H, Raphael Y. Gene therapy for the inner ear. Hear Res 2013; 297:99–105. otypes and phenotypes associated with enlargement of the vestibular aqueduct. 36 Minoda R, Hayashida M, Yumoto E. Gene transfer into the organ of Corti utilizing Cell Physiol Biochem 2011; 28: 545–552. adeno-associated viral vectors via the scala tympani. Otol Jpn 2006; 16:397. 12 Sylvester JE, Fischel-Ghodsian N, Mougey EB, O'Brien TW. Mitochondrial ribosomal 37 Liu Y, Okada T, Sheykholeslami K, Shimazaki K, Nomoto T, Muramatsu S, et al. proteins: candidate genes for mitochondrial disease. Genet Med 2004; 6:73–80. Specificandefficient transduction of cochlear inner hair cells with recombinant 13 Akil O, Seal RP, Burke K, Wang C, Alemi A, During M, et al. Restoration of hearing in adeno-associated virus type 3 vector. Mol Ther 2005; 12:725–733. the VGLUT3 knockout mouse using virally mediated gene therapy. Neuron 2012; 38 Minoda R, Yumoto E. Trans-uterine gene transfer into mouse otocysts. Toukeibu 75: 283–293. Juritsushinkei 2006; 20:33–37. 14 Lentz JJ, Jodelka FM, Hinrich AJ, McCaffrey KE, Farris HE, Spalitta MJ, et al. Rescue 39 Yu Q, Wang Y, Chang Q, Wang J, Gong S, Li H, et al. Virally expressed connexin26 of hearing and vestibular function by antisense oligonucleotides in a mouse restores gap junction function in the cochlea of conditional Gjb2 knockout mice. model of human deafness. Nat Med 2013; 19:345–350. Gene Therapy 2014; 21:71–80. 15 Jia S, Dallos P, He DZ. Mechanoelectric transduction of adult inner hair cells. 40 Frenz CM, Van De Water TR. Immunolocalization of connexin 26 in the developing J Neurosci 2007; 27: 1006–1014. mouse cochlea. Brain Res Brain Res Rev 2000; 32:172–180. 16 Fritzsch B, Barald KF, Lomax MI. Early Embryology of the Vertebrate Ear. Springer: 41 Cohen-Salmon M, Ott T, Michel V, Hardelin JP, Perfettini I, Eybalin M, et al. Tar- New York, NY, USA, 1998. geted ablation of connexin26 in the inner ear epithelial gap junction network 17 Fekete DM. Development of the vertebrate ear: insights from knockouts and causes hearing impairment and cell death. Curr Biol 2002; 12:1106–1111. mutants. Trends Neurosci 1999; 22: 263–269. 42 Choi BY, Kim HM, Ito T, Lee KY, Li X, Monahan K, et al. Mouse model of enlarged 18 Baker CV, Bronner-Fraser M. Vertebrate cranial placodes I. Embryonic induction. vestibular aqueducts defines temporal requirement of Slc26a4 expression for Dev Biol 2001; 232:1–61. hearing acquisition. J Clin Invest 2011; 121: 4516–4525. 19 Carlson BM. Ebook Library. Human Embryology and Developmental Biology,5thedn. 43 Scott DA, Wang R, Kreman TM, Sheffield VC, Karniski LP. The Pendred syndrome Elsevier/Saunders: Philadelphia, PA, USA, 2014, pp1 online resource xiii, 506pp. gene encodes a chloride-iodide transport protein. Nat Genet 1999; 21:440–443. 20 Morsli H, Choo D, Ryan A, Johnson R, Wu DK. Development of the mouse inner ear 44 Soleimani M, Greeley T, Petrovic S, Wang Z, Amlal H, Kopp P, et al. Pendrin: an and origin of its sensory organs. J Neurosci 1998; 18: 3327–3335. apical Cl-/OH-/HCO3- exchanger in the kidney cortex. Am J Physiol Renal Physiol 21 Kelley MW. Regulation of cell fate in the sensory epithelia of the inner ear. Nat Rev 2001; 280: F356–F364. Neurosci 2006; 7: 837–849. 45 Kim HM, Wangemann P. Epithelial cell stretching and luminal acidification lead to 22 Bryant J, Goodyear RJ, Richardson GP. Sensory organ development in the inner a retarded development of stria vascularis and deafness in mice lacking pendrin. ear: molecular and cellular mechanisms. Br Med Bull 2002; 63:39–57. PLoS One 2011; 6: e17949. 23 Strachan T, Read AP. Chapter 16, Molecular pathology. In: Human Molecular 46 Li X, Sanneman JD, Harbidge DG, Zhou F, Ito T, Nelson R, et al. SLC26A4 targeted Genetics, 2nd edn. Wiley-Liss: New York, NY, USA, 1999. to the endolymphatic sac rescues hearing and balance in Slc26a4 mutant mice. 24 Smith RJH, Van Camp G. GeneReviews. Nonsyndromic Hearing Loss and Deafness. PLoS Genet 2013; 9: e1003641. DFNB1, University of Washington: Seattle, USA, 1993. 47 Miller TM, Pestronk A, David W, Rothstein J, Simpson E, Appel SH, et al. An 25 Ng L, Goodyear RJ, Woods CA, Schneider MJ, Diamond E, Richardson GP et al. antisense oligonucleotide against SOD1 delivered intrathecally for patients with Hearing loss and retarded cochlear development in mice lacking type 2 iodo- SOD1 familial amyotrophic lateral sclerosis: a phase 1, randomised, first-in- thyronine deiodinase. Proc Natl Acad Sci USA 2004; 101:3474–3479. man study. Lancet Neurol 2013; 12: 435–442. 26 Hall JW 3rd. Development of the ear and hearing. J Perinatol 2000; 20: S12–S20. 48 Monteleone G, Fantini MC, Onali S, Zorzi F, Sancesario G, Bernardini S, et al. Phase 27 Hone SW, Smith RJ. Medical evaluation of pediatric hearing loss. Laboratory, I clinical trial of Smad7 knockdown using antisense oligonucleotide in patients radiographic, and genetic testing. Otolaryngol Clin N Am 2002; 35:751–764. with active Crohn's disease. Mol Ther 2012; 20:870–876. 28 Miwa T, Minoda R, Ise M, Yamada T, Yumoto E. Mouse otocyst transuterine gene 49 Boeda B, El-Amraoui A, Bahloul A, Goodyear R, Daviet L, Blanchard S, et al. transfer restores hearing in mice with connexin 30 deletion-associated Myosin VIIa, harmonin and cadherin 23, three Usher I gene products that hearing loss. Mol Ther 2013; 21: 1142–1150. cooperate to shape the sensory hair cell bundle. EMBO J 2002; 21: 29 Wang WH, Liu YF, Su CC, Su MC, Li SY, Yang JJ. A novel missense mutation in the 6689–6699. connexin 30 causes nonsyndromic hearing loss. PLoS One 2011; 6: e21473. 50 Guibert J, Benachi A, Grebille AG, Ernault P, Zorn JR, Costa JM. Kinetics of SRY 30 Del Castillo I, Moreno-Pelayo MA, Del Castillo FJ, Brownstein Z, Marlin S, Adina Q, gene appearance in maternal serum: detection by real time PCR in early preg- et al. Prevalence and evolutionary origins of the del(GJB6-D13S1830) mutation in nancy after assisted reproductive technique. Hum Reprod 2003; 18: 1733–1736. the DFNB1 locus in hearing-impaired subjects: a multicenter study. Am J Hum 51 Lo YM. Non-invasive prenatal testing using massively parallel sequencing of Genet 2003; 73: 1452–1458. maternal plasma DNA: from molecular karyotyping to fetal whole-genome 31 Teubner B, Michel V, Pesch J, Lautermann J, Cohen-Salmon M, Sohl G, et al. sequencing. Reprod Biomed Online 2013; 27:593–598. Connexin30 (Gjb6)-deficiency causes severe hearing impairment and lack of 52 Minoda R, Izumikawa M, Kawamoto K, Raphael Y. Strategies for replacing lost endocochlear potential. Hum Mol Genet 2003; 12:13–21. cochlear hair cells. NeuroReport 2004; 15: 1089–1092.