11/8/2013
1
Lawrence R. Lustig, MDDepartment of Oto-HNS
University of California San Francisco
Cochlear Gene Therapy: Is It Time?
UCSF
> 1700 Gene Therapy Trials
• Retinal Congenital Amaurosis• X-linked SCID• ADA-SCID• Adrenoleukodystrophy• Parkinson’s Disease• HIV Is It Time for the Ear?
11/8/2013
2
Vector
Promoter
Gene
Which Vector?
Vector Advantages Disadvantages
Adenovirus Easy to make<10kb inserts
ImmunogenicTransient expression
Adeno-Assoc Virus(AAV)
Non-pathologicCell division not req’d
<5kb insertsDifficult to make
Variable transfection
Herpesvirus 100kb insertsStable expression
Human diseasecytopathic
Lentivirus Stable expression Insertional mutagenesisLow transfection rate
Liposomes Easy to makeNo insert size limit
Non-pathologic
Low and transient expression
Which Vector?
Vector Advantages Disadvantages
Adenovirus Easy to make<10kb inserts
ImmunogenicTransient expression
Adeno-Assoc Virus Non-pathologicCell division not req’d
<5kb insertsDifficult to make
Variable transfection
Herpesvirus 100kb insertsStable expression
Human diseasecytopathic
Lentivirus Stable expression Insertional mutagenesisLow transfection rate
Liposomes Easy to makeNo insert size limit
Non-pathologic
Low and transient expression
Which Promoter?
Promoter
Gene
• “Constitutive” = Always on
• Gene-specific / Cell Specific
11/8/2013
3
Which Gene ???
Promoter
Gene
• Hair Cell Regeneration?
• Neuronal Growth Factors?
• Antioxidant / Antiapoptic factors?
• AAV-β-gal, Ad2 major late promoter
• Osmotic minipump
• Signal in spiral limbus, spiral ligament, spiral ganglion cells and the organ of Corti
• Weaker, similar pattern in contra ear
Development of in vivo gene therapy for hearing disorders: introduction of adeno-associated virus into the cochlea of the guinea pig.
Lalwani et al., Gene Therapy, 1996
•AAV-GFP
•CMV (constitutive) Promoter
•5 Serotypes studied (1, 2, 5, 6, 8)
•+/- deafeningKilpatrick et al, 2011
•Transduced:Hair cells, supporting cellsauditory nervespiral ligament
•IHCs most effectively transduced
•All 5 serotypes transducedauditory nerve
•AAV-8 most efficient vector
11/8/2013
4
2. RoundWindow Injection
Routes of Delivery
3. Cochleostomy
1. RoundWindow Diffusion
RWM Diffusion
•Variable carrier (ie gelfoam)
•Poor uptake alone
•Improved w/ Hyaluronic AcidShibata et al, Hum Gene Therapy 2011
•Enzymatic Digestion of RWWang et al, Gene Therapy, 2011
RWM Injection vs CochleostomyRWM Injection Cochleostomy
Less traumaticEasier approach
Theoretic access to SM
Direct access to SM
Used in human CI Used in human CI
Cochlear Gene Therapy Applications
11/8/2013
5
Animal Models of Cochlear Gene Therapy
•Hair Cell Regeneration
•Ototoxicity
•Spiral Ganglion Preservation
•Autoimmune Hearing Loss
•Genetic Hearing Loss
Animal Models of Cochlear Gene Therapy
•Hair Cell Regeneration
•Prevention of Ototoxicity
•Spiral Ganglion Preservation
•Autoimmune Hearing Loss
•Genetic Hearing Loss
Animal Models of Cochlear Gene Therapy
•Hair Cell Regeneration
•Prevention of Ototoxicity
•Spiral Ganglion Preservation
•Autoimmune Hearing Loss
•Genetic Hearing Loss
Math1 KO MiceBermingham et al, Science, 1999
• Failure of cochlear & vestibular HC’s to differentiate
• Cochlea/vestib endorgansotherwise nl
11/8/2013
6
Math1 gene transfer generates new cochlear
hair cells in mature guinea pigs in vivo
Kawamoto K et al,J Neurosci. 2003
Normal (Ad-perilymph)cochleaNormal (Ad-perilymph)cochlea
Kawamoto K et al
J Neurosci June 1 200323(1):4395-4400
Math1 Gene Transfer
Izumikawa et al. Nature Medicine 11, 271 - 276 (2005)Auditory hair cell replacement and hearing improvement by Atoh1 gene therapy in deaf mammals
Izumikawa et al. (2005)Auditory hair cell replacement and
hearing improvement by Atoh1 gene therapy
in deaf mammals
11/8/2013
7
Vestibular Hair Cell Regeneration and Restoration of Balance Function Induced by Math1 Gene Transfer.Staecker H et al. Otol Neurotol 2007
Control
Neomycin
Neomycin+Admath1
Neomycin+Admath1
Swim time
HVOR Recovery
Staecker H et al. Otol Neurotol 2007.
•Human Elongation factor 1-α promoter
•Math-1 insert
•In utero gene injection w/ electroporation
•Gain of function
Gubbels et al, 2008
“Acquired”
Genetic
11/8/2013
8
•GJB2=Connexin26 = autosomal recessive non-syndromic HL
•Lipid-mediated transfection of dominant allele causes HL
•Co-transfection of siRNA reduces hearing loss
• Single IP injection of ASO vs. defective pre-mRNA Ush 1C transcript
• Partially corrects splicing, increases protein expression, improves stereocilia organization, and rescues cochlear hair cells, vestibular function and low-frequency hearing
2013 Mar;19(3):345-50
2008
DAPI Synaptophysin VGLT3
VGLUT3 in the Organ of Corti
IHC OHC
TC
Seal et al, Neuron, 2008
11/8/2013
9
Am J Hum Genet 2008
•DFNA25 = autosomal-dominant progressive, high-frequency non-syndromic hearing loss
•2 unrelated families = a heterozygous missensemutation, c.632C/T was found to segregate with DFNA25 Neuron 75, 283–293, July 26, 2012
11/8/2013
10
Delivery P10-12
Vglut3 ABR P40
01020304050
60708090
100
Click 8 KHz 16 KHz 32 KHz
Stimulus
AB
R T
hre
sho
ld d
B S
PL
Rescued KOWild TypeKO
Delivery P1-3
VGLUT3 Rescued MiceAcknowledgements
UCSF: Lustig LabOmar Akil, PhDSean AlemiKevin BurkeEdwards LabRobert Edwards
Ohio State UnivMatthew During,PhD
Univ PittsburghRebecca Seal, PhD
Johns Hopkins UnivElisabeth Glowatzki, PhD
nature medicine volume 19 | number 3 | march 2013 345
Rescue of hearing and vestibular function by antisense oligonucleotides in a mouse model of human deafnessJennifer J Lentz1,6, Francine M Jodelka2,6, Anthony J Hinrich2,6, Kate E McCaffrey2, Hamilton E Farris1, Matthew J Spalitta1, Nicolas G Bazan3, Dominik M Duelli4, Frank Rigo5 & Michelle L Hastings2
Hearing impairment is the most common sensory disorder, with congenital hearing impairment present in approximately 1 in 1,000 newborns1. Hereditary deafness is often mediated by the improper development or degeneration of cochlear hair cells2. Until now, it was not known whether such congenital failures could be mitigated by therapeutic intervention3–5. Here we show that hearing and vestibular function can be rescued in a mouse model of human hereditary deafness. An antisense oligonucleotide (ASO) was used to correct defective pre-mRNA splicing of transcripts from the USH1C gene with the c.216G>A mutation, which causes human Usher syndrome, the leading genetic cause of combined deafness and blindness6,7. Treatment of neonatal mice with a single systemic dose of ASO partially corrects Ush1c c.216G>A splicing, increases protein expression, improves stereocilia organization in the cochlea, and rescues cochlear hair cells, vestibular function and low-frequency hearing in mice. These effects were sustained for several months, providing evidence that congenital deafness can be effectively overcome by treatment early in development to correct gene expression and demonstrating the therapeutic potential of ASOs in the treatment of deafness.
Usher syndrome is characterized by hearing impairment combined with retinitis pigmentosa and, in some cases, vestibular dysfunction. The fre-quency in the general population may be as high as 1 in 6,000 (ref. 8). Type 1 Usher syndrome is characterized by profound hearing impair-ment and vestibular dysfunction at birth and the development of retinitis pigmentosa in early adolescence. Approximately 6–8% of type 1 Usher syndrome cases are caused by mutations in the USH1C gene9, which encodes the protein harmonin. The USH1C216G>A (216A) mutation accounts for all cases of type 1 Usher syndrome in Acadian popula-tions9–11 and creates a cryptic 5ʹ splice site that is used preferentially over the authentic 5ʹ splice site of exon 3 (Fig. 1a), resulting in a frameshift and truncated harmonin protein12.
We used a mouse model of Usher syndrome based on the human 216A mutation13 to investigate a treatment for deafness and vestibular dysfunc-tion using ASOs (Supplementary Fig. 1) designed to redirect cryptic
splicing of 216A pre-mRNA to the authentic site (Fig. 1a). To screen for ASOs that block 216A cryptic splicing, we transfected a minigene expression plasmid composed of exons 2–4 and the intervening introns of human USH1C (wild type) or USH1C 216A into HeLa cells with 47 different individual ASOs surrounding the mutation and quantified splicing correction (Fig. 1b and Supplementary Table 1). Several ASOs blocked cryptic splicing and promoted correct splicing (Fig. 1b,c) in a dose-dependent manner (Fig. 1d). The ASOs also blocked cryptic splicing, promoted correct splicing of the endogenous Ush1c 216A gene transcript and increased harmonin protein expression in a mouse kidney cell line derived from mice homozygous for the Ush1c 216A mutation (216AA mice) (Supplementary Fig. 2). ASOs induced correct splicing in vivo after a series of intraperitoneal injections of 50 mg per kg body weight of ASO in adult 216AA mice (Fig. 1e). ASO-29 promoted the highest amount of correct splicing of the ASOs tested (Fig. 1e) and also corrected splicing and increased harmonin protein expression (Fig. 1f,g and Supplementary Fig. 3) in a dose-dependent manner.
216AA mice are deaf and have severe vestibular dysfunction, as indi-cated by their auditory brainstem response (ABR) and head-tossing and circling behavior13,14. Neonatal 216AA mice were injected intra-peritoneally with 300 mg per kg body weight of ASO-C or ASO-29 to test whether ASO-29 can correct vestibular and hearing defects. 216AA mice untreated or treated with a mismatched ASO (ASO-C) showed circling behavior, whereas 216AA mice treated with ASO-29 did not circle, and showed similar behavior to heterozygous (216GA) or wild-type (216GG) mice (Fig. 2a,b and Supplementary Video 1). We did not observe any circling behavior in mice treated with ASO-29 at postembryonic day 3 (P3), P5, P10 or P13, whereas 216AA mice treated on P16 showed circling behavior similar to untreated or ASO-C–treated 216AA mice (Fig. 2b). ASO-29–treated 216AA mice had no vestibular dysfunction at 6 months of age (Supplementary Fig. 4a). We performed trunk-curl, contact-righting and swim tests on 2- to 3-month-old and 6- to 9-month-old mice to further quantify ves-tibular function15. The younger and older 216AA mice all performed poorly in these tests, whereas 216AA mice treated with ASO-29 at P5 performed similarly to untreated or ASO-C–treated 216GA mice and showed no vestibular dysfunction (Supplementary Fig. 4b). Our
1Neuroscience Center and Department of Otorhinolaryngology & Biocommunications, Louisiana State University Health Sciences Center (LSUHSC), New Orleans, Louisiana, USA. 2Department of Cell Biology and Anatomy, Chicago Medical School, Rosalind Franklin University of Medicine and Science, North Chicago, Illinois, USA. 3Neuroscience Center and Department of Ophthalmology, LSUHSC, New Orleans, Louisiana, USA. 4Department of Cellular and Molecular Pharmacology, Chicago Medical School, Rosalind Franklin University of Medicine and Science, North Chicago, Illinois, USA. 5Isis Pharmaceuticals, Carlsbad, California, USA. 6These authors contributed equally to this work. Correspondence should be addressed to M.L.H. ([email protected]) or J.J.L. ([email protected]). Received 12 September 2012; accepted 28 January 2013; published online 4 February 2013; doi:10.1038/nm.3106
l e t t e r snp
g©
201
3 N
atur
e A
mer
ica,
Inc.
All
right
s re
serv
ed.
346 volume 19 | number 3 | march 2013 nature medicine
We recorded responses to different sound frequencies (8–32 kHz and broad band noise, BBN) at different intensities (18–90 dB SPL). Hearing thresholds represent the lowest sound intensity at which a recognizable ABR wave (neural response) is observed. We compared ABR thresholds in P30 216AA mice treated with ASO-29, treated and untreated 216GG wild-type and 216GA heterozygous mice, and 216AA mutant mice treated with ASO-C. 216GG and 216GA mice had thresholds typical of mice with normal hearing (Fig. 2c,d). Similar to untreated 216AA mice14, 216AA mice treated with ASO-C had abnormal or absent ABRs (Fig. 2c,d and Supplementary Fig. 6). 216AA mice treated between P3 and P5 with a single dose of ASO-29
results suggest that ASOs can prevent vestibular dysfunction associated with Usher syndrome in mice when delivered neonatally.
Treatment of 216AA mice with ASO-29 also rescued hearing. Startle responses to high-amplitude sound are similar in ASO-29–treated 216AA and either ASO-C–treated or untreated 216GA mice (Supplementary Video 2 and Supplementary Fig. 5). 216AA mice treated with ASO-C, however, showed neither an ini-tial startle response, defined as an ear twitch and rapid head and body movement, nor a subsequent freezing response after acoustic stimulus (Supplementary Video 2). We recorded auditory-evoked brainstem responses to quantitatively assess hearing function.
3 4
Correct splicing
216G>A
2
Cryptic splicing(frameshift)
135 aaTruncated protein
PDZ1
PDZ1
PDZ1
a
b
c
PDZ2
PDZ2
PDZ2
CC1
CC1
CC1
CC2
PDZ3
PST PDZ3
Full-length harmonin
552 aa
899 aa
403 aa
USH1C
Exon 2 Exon 3 Exon 4216A
12 3 4 5–8 9–12 13–16 17–20 21–24 25–28 29–32 33–36 37–40 41–44
454647
UnsplicedCorrectCryptic
Skip
ASO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 C C
10 1 1 0 0 0 0 0 0 0 0 0 0 1 0 0 00 0 1 0 1 1 11 1 4 2 2 0 0 0 0 0 0 0 0 1 2 4 2 1 0 1 0 0 0 0
13 6 10 32 35 64 24 35 42 12 37 26 22 30 64 16 58 5 4 10 22 17 22 24 15 2 12 3 3 20 99 99 100 91 98 100
100 98 96 95 97 98 98 98 94 98 98 98 98
Correct (%)Cryptic (%)
ASO-49ASO-48ASO-29ASO-28USH1C
Authentic 5′ splice site216ACryptic 5′ splice site
MinigeneASO(nM)
28 29 48 49
USH1C 216A
UnsplicedCorrectCryptic
Skip
ASO(nM)
28 29 48 49
5 10 20 40 80 5 10 20 40 80 5 10 20 40 80 5 10 20 80
5 10 20 40 80 5 10 20 40 80 5 10 20 40 80 5 10 20 80
Correct Cryptic
020406080
100
Splic
ing
(%)
Mice 216AAASO
28 29 48 49―
CorrectCryptic
Skip
28 29 48 49―ASO0
10
20
30
Corre
ct s
plici
ng (%
)
―ASO ASO-29
CorrectCryptic
Skip
(mg per kg BW) 0 50 100 200
0 50 100 2000
102030
Corre
ct
splic
ing
(%)
(mg per kg BW)
―ASO ASO-29(mg per kg BW) 0 50 100
b
a
c
Harmonin
β-actin17
2433405572
100135
–
–
MW (kDa)
******
**
**
*
a
b
c
d
e
f g
Figure 1 Correction of USH1C 216A splicing using ASOs. (a) Gene structure of USH1C exons 2–4 and RNA splicing and protein products. Boxes represent exons and lines are introns. Diagonal lines indicate splicing. The locations of the 216A mutation and the cryptic splice site are labeled. aa, amino acids; PDZ, PSD95-Dlg1-zo1-domain; CC, coiled-coil domain; PST, proline-serine-threonine–rich domain. (b) Top, diagram of ASOs tested below, mapped to their position of complementarity on USH1C. Bottom, radioactive RT-PCR of RNA isolated from HeLa cells transfected with a USH1C 216A minigene and the indicated ASO at a final concentration of 50 nM. RNA spliced forms are labeled. ‘Unspliced’ refers to transcripts with intron 3 retained, and ‘skip’ indicates exon 3 skipping. (Below) Quantification of the percentage splicing in graph is calculated as: Correct (%) = [(correct/(correct + cryptic + skip)] × 100, and Cryptic (%) = [(cryptic/(correct + cryptic + skip)] × 100. ‘C’ indicates mock-treated control. (c) Sequence and USH1C target region of ASOs. Exonic and intronic sequences are in capital and lower-case letters, respectively. (d) Analysis of USH1C 216A minigene transcript splicing in HeLa cells treated with increasing concentrations of the indicated ASOs from c. Quantification of percentage correct and percentage cryptic splicing is shown in graph (below). (e) RT-PCR analysis of RNA isolated from kidneys of adult 216AA mice 24 h after final injection of 50 mg per kg body weight (BW) of different ASOs. Samples from three individual mice are shown. Ush1c spliced products are indicated and quantified in graph (below) as described above. Error bars represent s.e.m. (*P ≤ 0.05, **P ≤ 0.01, n = 3, two-tailed Student’s t-test compared to vehicle treatment). (f) RT-PCR analysis of RNA isolated from kidneys of adult Ush1c 216AA mice treated with different doses of ASO-29. Samples from three individual mice are shown. Ush1c spliced products are indicated and quantified as described above. Error bars represent s.e.m. (***P ≤ 0.001, n = 3, two-tailed Student’s t-test compared to vehicle). (g) Immunoblot analysis of harmonin protein in lysates from the kidneys of adult 216AA mice analyzed in f. Blots were also probed with a b-actin–specific antibody for a loading reference. MW, molecular weight.
l e t t e r snp
g©
201
3 N
atur
e A
mer
ica,
Inc.
All
right
s re
serv
ed.
nature medicine volume 19 | number 3 | march 2013 347
216AA mice (P ≤ 0.05) (Supplementary Fig. 9a,b). These results sug-gest that mice injected with a single ASO treatment early in life can hear at 6 months of age, indicating a long-term, if slowly declining, therapeutic rescue of hearing.
To determine the effect of ASOs on Ush1c mRNA splicing and harmo-nin expression, we analyzed the cochleae from mice injected at P5 with ASO-29 or ASO-C. At P30, we observed a low amount of correct exon 3 splicing in the ASO-29–treated 216AA mice (Fig. 3a). Correct splicing peaked at P120, and expression of the full-length mRNA transcript at P180 was similar to that at P30, indicating that the effect of the ASO on splicing is stable and corresponds with ABR results at this time point (Supplementary Fig. 9c). Harmonin protein abundance was also higher in cochleae from ASO-29–treated 216AA mice compared to 216AA mice treated with ASO-C and similar to harmonin levels at P32–P35 in the cochleae of control 216GA mice (Fig. 3b and Supplementary Fig. 10).
The harmonin b isoform (Fig. 1a) localizes to the developing and mature stereocilia bundle of cochlear hair cells16–19, where it is hypothe-sized to scaffold the molecular components of the mechanotransduction machinery20. We examined the expression and localization of harmonin b in microdissected organs of Corti from P30 216AA mice injected at P5 with either ASO-29 or ASO-C and untreated 216GA heterozygous lit-termates. Harmonin b was abundantly expressed in the tips of outer hair cell stereocilia bundles of 216GA mice, whereas the 216AA mice had less expression and it was mislocalized in the atypical bundles (Fig. 3c,d).
had recognizable waveforms and near-normal thresholds to BBN and 8- and 16-kHz pure tones when compared to 216GG and 216GA control mice (Fig. 2c,d and Supplementary Fig. 6). The thresholds of ASO-29–treated 216AA mice at 32 kHz were not significantly dif-ferent than those of control ASO-C–treated 216AA mice, indicating that treatment was not effective at rescuing high-frequency hearing (Fig. 2d). These data show rescue of low- and mid-frequency hear-ing at P30 in ASO-29–treated 216AA mice. 216AA mice treated with ASO-29 at P10 had significantly higher thresholds than those treated at P3–P5 but significantly lower thresholds in response to BBN and an 8-kHz tone than untreated mutants or mutants treated with ASO-C (P ≤ 0.05) (Fig. 2c,d and Supplementary Fig. 6), implying a develop-mental window of therapeutic efficacy in mice. At 2 months of age, 216AA mice treated at P3–P5 had ABR thresholds to BBN and 8 kHz and 16 kHz, but not 32 kHz, equivalent to 216GG and 216GA mice (P ≤ 0.05, Fig. 2e and Supplementary Fig. 7a–d). At 3 months, there was no significant difference in ABR thresholds at 8 kHz between control 216GA heterozygous mice and 216AA mutant mice treated with ASO-29 at P3–P5 (P ≤ 0.05), but the data show some loss of sen-sitivity in ASO-29–treated 216AA mice to 16 kHz and BBN (Fig. 2f and Supplementary Fig. 8a–d). At 6 months of age, there were signifi-cant differences in ABR thresholds between control 216GA or ASO-C–treated 216GG mice and ASO-29–treated 216AA mice at all frequencies (P ≤ 0.05), though ASO-29–treated 216AA mice showed ABR thresh-olds that were significantly lower than those of ASO-C–treated
a b
c
d e f
Mutant (216AA)ASO-C
Mutant (216AA)ASO-29
Heterozygote (216GA)ASO-C
*** *** *** *** NS ***
***NS NS NS NS
AA AA AA AA AA AA AA GA GA GG– P3-P16 P3 P5 P10 P13 P16 P5 – –
ASO-29 – –ASO-C–
Rot
atio
ns p
er 1
20 s
0
10
20
30
40
50
Treatment
Treatment day
Ush1c216 genotype
Mutant (216AA)ASO-C, P5
Mutant (216AA)ASO-29, P5
Mutant (216AA)ASO-29, P10
Heterozygote (216GA)
8 kH
z2 uV
1 m
90 dB SPL78 dB SPL66 dB SPL54 dB SPL48 dB SPL42 dB SPL30 dB SPL24 dB SPL
1 month 2 months 3 months
AA, ASO-C, P5 AA, ASO-29, P5 AA, ASO-29, P10 GG/GA, controlSound frequency (kHz)
Thre
shol
d (d
B SP
L)
5 8 16 32 BBN0
20
406080
100
5 8 16 32 BBN0
20
406080
100
5 8 16 32 BBN0
20
406080
100
n = 39 n = 21 n = 9 n = 32 n = 7 n = 4 n = 8 n = 17 n = 30 n = 5
Figure 2 ASOs correct vestibular function and rescue hearing in 216AA mice. (a) Representative open-field pathway traces (120 s) from a P22 mouse in each group are shown. (b) Quantification of the number of rotations in 120 s from P22–P35 mice. Error bars represent s.e.m., the number (n) of mice analyzed is indicated within the individual bars. Significance (***P ≤ 0.001 or not significant, NS) was calculated using one-way analysis of variance (ANOVA) and Tukey-Kramer post test. (c) Representative ABR waveforms at 8-kHz stimulus from a 216AA mouse injected at P5 with ASO-C (left), a 216AA mouse injected at P5 or P10 with ASO-29 (middle) and a 216GA mouse (right). Colored lines (red = AA, ASO-C; blue = AA, ASO-29 injected at P5; orange = AA, ASO-29 injected at P10; gray = GA control) represent thresholds detected. (d–f) Average ABR thresholds (dB SPL) to pure tones ranging from 8 to 32 kHz or BBN in 216AA mutant (AA) and 216 GG wild-type (GG) or 216 GA heterozygous (GA) mice at 1 month of age (n = 11, 8, 5 and 11 for 216AA, ASO-C; 216AA, ASO-29 at P5; 216AA, ASO-29 at P10 and 216GG/GA, respectively (wild-type and heterozygotes were grouped together because we found no difference between them)) (d), 2 months of age (n = 4, 6 and 5 for 216AA, ASO-C; 216AA, ASO-29; and 216GG/GA, respectively) (e) and 3 months of age (n = 3, 4, and 4 for 216AA, ASO-C; 216AA, ASO-29; and 216GG or GA, respectively) (f). Error bars represent s.e.m. * indicates a significant difference between ASO-29–treated 216AA and ASO-C–treated 216AA mice, and # indicates a significant difference between ASO-29–treated 216AA and 216GA control mice (P ≤ 0.05; two-way ANOVA with Tukey-Kramer post test).
l e t t e r snp
g©
201
3 N
atur
e A
mer
ica,
Inc.
All
right
s re
serv
ed.
348 volume 19 | number 3 | march 2013 nature medicine
stereocilia bundles with typical ‘U’ or ‘W’ bundle shapes (Fig. 4a,b). The 216AA mice treated with ASO-29 had significantly fewer atypical bun-dles in the regions that detect 8 (0.8–1.2 mm) and 16 kHz (1.8–2.2 mm) (P ≤ 0.05) but not in the region detecting 32 kHz (3.8–4.2 mm) (Fig. 4c). This pattern of anatomical stereocilia rescue is consistent with the ABR results showing a rescue of hearing in the 8- and 16-kHz range and less robust rescue in the 32-kHz range. Together, our results indicate a change in the bundle structure and number of hair cells at the apical–mid regions in ASO-29–treated 216AA mice, providing an anatomical correlate to function.
Our study demonstrates that a human disease–causing mutation can be corrected to treat deafness and vestibular dysfunction in a mouse model. Treatment during the critical hair cell developmental period is probably necessary and perhaps sufficient for long-term rescue of hearing. Notably, a modest correction of Ush1c mRNA splic-ing and expression of full-length harmonin is sufficient to rescue hearing in mice for 6 months. This enduring effect of the ASOs is consistent with the duration of action of ASOs observed in other mouse models of disease such as spinal muscular atrophy23 and myo-tonic dystrophy24.
ASO-29–treated 216AA mice had elevated harmonin b expression with localization at the tips of stereocilia similar to that in the 216GA mice (Fig. 3c,d). These results suggest proper localization of harmonin with ASO treatment.
Frequency place-mapping in the mouse cochlea21,22 suggests that the region corresponding to 8–16 kHz (the frequencies most robustly rescued by ASO-29) is located approximately 1–2 mm from the apex tip (Fig. 3f). To assess the relationship between hair cell number and ABR threshold, we counted hair cells labeled with parvalbumin. At P35, 216AA mice had significant outer hair cell loss from approxi-mately 0.8–2.0 mm from the apex, corresponding to hearing at 6–20 kHz (P ≤ 0.05) (Fig. 3g,h). In contrast, the number of outer hair cells in this region of 216AA mice treated with ASO-29 at P3–P5 did not differ from that in 216GA mice, which is consistent with rescued physi-ological function (Fig. 3g,h).
We also assessed changes in hair cell morphology that may reflect the rescue of hearing in the regions of the cochlea sensitive to 8 and 16 kHz. By P35, 216AA mice had significant hair cell loss in this region (Fig. 3); therefore, we analyzed subcellular structures in P18 216AA and 216GA mice before the loss of hair cells. We quantified the number of
Heterozygote (216GA)ASO-C
Mutant (216AA)ASO-29
Mutant (216AA)ASO-C
GA, ASO-C AA, ASO-29 AA, ASO-C
IHCs OHCs
Distance from apex tip (mm)
Avg.
no.
hai
r cel
lspe
r 100
μm
20
25
30
35
40
0 0.5 1.0 1.5 2.0 0 0.5 1.0 1.5 2.00
25
50
75
100
125
Heterozygote (216GA)ASO-C
Mutant (216AA)ASO-29
Mutant (216AA)ASO-C
Dis
tanc
e fro
m a
pex
tip (m
m)
1.8–
2.0
0.8–
1.0
16 kHz
1.0 mm
8 kHz
OHCs
IHCsApex tip
ParvalbuminDAPI
b
a
c
Harm
oninβ-actin
172433405572
100135
Mouse MW (kDa)TreatmentGenotype
1 2 3 4 5 6 7 8 9ASO-29— —
AA GA
CorrectCryptic
Skip
Correct (%)
Mouse
TreatmentGenotype
1 2 3 4 5 6 7 8 9
ASO-29— —AA GA GG
0.25± 0.08n = 10
1.6± 0.68n = 6
41± 13.1n = 4
100± 0
n = 3
a c
bd
e
f g
h
Figure 3 ASO-29 treatment corrects mRNA splicing and harmonin protein expression and prevents cochlear hair cell loss in 216AA mice. (a) RT-PCR analysis of cochlear RNA isolated at P32–P35 from mice treated with control (ASO-C) or ASO-29 at P3–P5. Spliced products are labeled. (b) Immunoblot analysis of harmonin expression in cochlea isolated at P32–35 from mice that were treated at P5. Different isoforms of harmonin expressed from Ush1c are labeled. Blots were also probed with a b-actin–specific antibody for a loading reference. (c) Immunofluorescence staining of harmonin b (green) and F-actin (red, phalloidin) in outer hair cell (OHC) bundles in the region of the basilar membrane that corresponds to hearing at 8 kHz (0.8–1.5 mm from apex tip). Images are from 216GA mice (left), 216AA mice treated with ASO-29 at P5 (middle) or 216AA mice treated with ASO-C at P5 (right). Scale bar, 3 mm. (d) Digital magnification (2× zoom) of images shown in c. Scale bar, 2 mm. (e) Immunofluorescence image of a primary antibody isotype control from a 216GA mouse taken from a similar OHC bundle location. (f) Immunofluorescence image of the regions of the basilar membrane that are represented in g and h. Hair cells are labeled with parvalbumin (red), and nuclei are counterstained with DAPI (blue). Scale bar, 50 mm. IHCs, inner hair cells. (g) Cochleogram showing inner (left) and outer (right) hair cell counts from regions progressively distant from the apex tip. Error bars represent the s.e.m. (*P ≤ 0.05; n = 3 mice). At least 100 cells from each experimental group were evaluated for each region. (h) Immunofluorescence images of representative regions along the basilar membrane 0.8–1.0 mm (top) or 1.8–2.0 mm (bottom) from the extreme apex from P35 216GA mice treated with ASO-C at P5 (left), 216AA mice treated with ASO-29 at P5 (middle) or 216AA mice treated with ASO-C at P5 (right). Scale bar, 20 mm.
l e t t e r snp
g©
201
3 N
atur
e A
mer
ica,
Inc.
All
right
s re
serv
ed.
nature medicine volume 19 | number 3 | march 2013 349
occurs at the base of the cochlea, this result may suggest that Ush1c is expressed tonotopically during development and that splicing correction at P5 benefits only the 30. apical-mid regions of the cochlea. The develop-ment of the ear and hearing in humans occurs in utero27. Thus, treatment in humans would probably require delivery to the fetus via approaches such as intrauterine transfusion28.
Individuals with Usher syndrome present with both hearing and vision loss, and the correction of one of these sensory deficits may have a considerable positive impact. Although the retinitis pigmentosa associated with Usher syndrome is recapitulated in the 216AA mice, their retinal cell loss occurs later, at approximately 1 year of life14. Thus, our analysis of vision in these mice will require further investigation at later time points. The rescue of hearing in this study offers a model for studying the development of hearing and vestibular function and for developing approaches to correct these processes when they are impaired.
METHODSMethods and any associated references are available in the online ver-sion of the paper.
Note: Supplementary information is available in the online version of the paper.
ACKNOWLEDGMENTSWe gratefully acknowledge support from the Hearing Health Foundation, Midwest Eye-Banks, the National Organization for Hearing Research Foundation, Capita Foundation and the US National Institutes of Health. We thank D. Cunningham and E. Rubel for assistance with scanning electron microscopy analysis; A. Rosenkranz, R. Marr and M. Oblinger for use of equipment; J. Huang for assistance with open-field analysis, L. Ochoa for assistance with ABR analysis computer support, G. MacDonald for assistance with confocal imaging and deconvolution analysis, U. Wolfrum (Johannes Gutenberg University of Mainz) for harmonin b–specific antibodies, H. Thompson for statistical analysis, and A. Case, B. Keats and M. Havens for discussions and comments on the manuscript.
AUTHOR CONTRIBUTIONSThe project was conceived of by M.L.H. Experiments were designed and performed by A.J.H., F.M.J., J.J.L., K.E.M., M.L.H. and D.M.D., and were analyzed by M.L.H., J.J.L., F.R., F.M.J., A.J.H. and K.E.M. Animal work was carried out by M.L.H., D.M.D., K.E.M., A.J.H., F.M.J., J.J.L. and M.J.S. Molecular experiments were performed by A.J.H., F.M.J., K.E.M. and M.L.H. J.J.L. carried out the immunofluorescence analysis. J.J.L., M.J.S. and H.E.F. performed auditory brainstem response experiments, and J.J.L. and N.G.B. interpreted the results. M.L.H. and J.J.L. wrote the paper.
COMPETING FINANCIAL INTERESTSThe authors declare competing financial interests: details are available in the online version of the paper.
Reprints and permissions information is available online at http://www.nature.com/reprints/index.html.
1. Morton, C.C. & Nance, W.E. Newborn hearing screening—a silent revolution. N. Engl. J. Med. 354, 2151–2164 (2006).
2. Dror, A.A. & Avraham, K.B. Hearing loss: mechanisms revealed by genetics and cell biology. Annu. Rev. Genet. 43, 411–437 (2009).
3. Conde de Felipe, M.M., Feijoo Redondo, A.F., Garcia-Sancho, J., Schimmang, T. & Alonso, M.B. Cell- and gene-therapy approaches to inner ear repair. Histol. Histopathol. 26, 923–940 (2011).
4. Di Domenico, M. et al. Towards gene therapy for deafness. J. Cell. Physiol. 226, 2494–2499 (2011).
5. Bermingham-McDonogh, O. & Reh, T.A. Regulated reprogramming in the regeneration of sensory receptor cells. Neuron 71, 389–405 (2011).
6. Bitner-Glindzicz, M. et al. A recessive contiguous gene deletion causing infantile hyper-insulinism, enteropathy and deafness identifies the Usher type 1C gene. Nat. Genet. 26, 56–60 (2000).
7. Verpy, E. et al. A defect in Harmonin, a PDZ domain-containing protein expressed in the inner ear sensory hair cells, underlies Usher syndrome type 1C. Nat. Genet. 26, 51–55 (2000).
8. Kimberling, W.J. et al. Frequency of Usher syndrome in two pediatric populations: Implications for genetic screening of deaf and hard of hearing children. Genet. Med. 12, 512–516 (2010).
The rescue of hearing in mice by ASO-29 treatment demonstrates that deafness can be treated if intervention occurs early in development. Treatment at P10 corrects vestibular function and partially restores hear-ing, whereas treatment at P3–P5 rescues vestibular function and hearing with ABRs comparable to those of wild-type mice (Supplementary Fig. 1). Although Ush1c is expressed as early as embryonic day 15 in mice25,26, there is a peak of developmental expression in the cochlea that occurs after P4 and before P16 (https://shield.hms.harvard.edu/viewgene.html?gene=Ush1c). Our results are consistent with this expression pat-tern, suggesting that high expression before P5 is not required for the development of low- and mid-frequency hearing, but expression between P5 and P10 may be crucial. Hearing at high frequencies (32 kHz) is not rescued to the same level as that at the lower frequencies, and the rescue is more transient (Fig. 2). Because detection of high-frequency sound
Heterozygote(216GA)
Mutant(216AA)ASO-29
Mutant(216AA)ASO-C
0.8–
1.2
mm
1.8–
2.2
mm
3.8–
4.2
mm
Dis
tanc
e fro
m a
pex
tip
16 kHz
Apextip
8 kHz32 kHz1 mm
3 mm
4 mm
2 mm
100 μm
AA, ASO-29 AA, ASO-C
10 100Frequency (kHz)At
ypic
al s
tere
ocilia
bun
dles
(%)
0
20
40
60
80
100
b
a
c
Figure 4 Restoration of hair cell stereocilia bundle shape in mice. (a) Scanning electron micrographs of outer hair cell bundles from P18 216GA and 216AA mice treated at P5 with ASO-C or ASO-29. Distance (mm) was measured from apex tip. Scale bars represent 1 mm and 10 mm for the high- and low-magnification images, respectively. (b) Scanning electron micrograph illustrating the regions of the cochlea that are represented in a. Scale bar, 100 mm. (c) Quantification of atypical bundles shown as a percentage of total cells counted at different positions along the basilar membrane in 216AA mice treated either with ASO-29 (blue line) or ASO-C (red line) (*P ≤ 0.05, **P ≤ 0.005; two-tailed unpaired t-test; n = 3 or 4 mice per region). Error bars represent s.e.m. At least 200 cells from each experimental group were evaluated with at least 60 hair cells from each region. 216GA control mice have no atypical bundles (data not shown).
l e t t e r snp
g©
201
3 N
atur
e A
mer
ica,
Inc.
All
right
s re
serv
ed.
350 volume 19 | number 3 | march 2013 nature medicine
links of the hair bundle in its cohesion, orientation and differential growth. Development 135, 1427–1437 (2008).
19. Boëda, B. et al. Myosin VIIa, harmonin and cadherin 23, three Usher I gene products that cooperate to shape the sensory hair cell bundle. EMBO J. 21, 6689–6699 (2002).
20. Peng, A.W., Salles, F.T., Pan, B. & Ricci, A.J. Integrating the biophysical and molecular mechanisms of auditory hair cell mechanotransduction. Nat. Commun. 2, 523 (2011).
21. Müller, M., von Hunerbein, K., Hoidis, S. & Smolders, J.W. A physiological place-frequency map of the cochlea in the CBA/J mouse. Hear. Res. 202, 63–73 (2005).
22. Greenwood, D.D. A cochlear frequency-position function for several species—29 years later. J. Acoust. Soc. Am. 87, 2592–2605 (1990).
23. Hua, Y. et al. Peripheral SMN restoration is essential for long-term rescue of a severe spinal muscular atrophy mouse model. Nature 478, 123–126 (2011).
24. Wheeler, T.M. et al. Targeting nuclear RNA for in vivo correction of myotonic dystrophy. Nature 488, 111–115 (2012).
25. El-Amraoui, A. & Petit, C. Usher I syndrome: unravelling the mechanisms that underlie the cohesion of the growing hair bundle in inner ear sensory cells. J. Cell Sci. 118, 4593–4603 (2005).
26. Petit, C. & Richardson, G.P. Linking genes underlying deafness to hair-bundle develop-ment and function. Nat. Neurosci. 12, 703–710 (2009).
27. Hall, J.W. III. Development of the ear and hearing. J. Perinatol. 20, S12–S20 (2000).28. Uhlmann, R.A., Taylor, M., Meyer, N.L. & Mari, G. Fetal transfusion: the spectrum of
clinical research in the past year. Curr. Opin. Obstet. Gynecol. 22, 155–158 (2010).
9. Ouyang, X.M. et al. Characterization of Usher syndrome type I gene mutations in an Usher syndrome patient population. Hum. Genet. 116, 292–299 (2005).
10. Ebermann, I. et al. Deafblindness in French Canadians from Quebec: a predominant founder mutation in the USH1C gene provides the first genetic link with the Acadian population. Genome Biol. 8, R47 (2007).
11. Ouyang, X.M. et al. USH1C: a rare cause of USH1 in a non-Acadian population and a founder effect of the Acadian allele. Clin. Genet. 63, 150–153 (2003).
12. Lentz, J. et al. The USH1C 216G→A splice-site mutation results in a 35-base-pair deletion. Hum. Genet. 116, 225–227 (2005).
13. Lentz, J., Pan, F., Ng, S.S., Deininger, P. & Keats, B. Ush1c216A knock-in mouse survives Katrina. Mutat. Res. 616, 139–144 (2007).
14. Lentz, J.J. et al. Deafness and retinal degeneration in a novel USH1C knock-in mouse model. Dev. Neurobiol. 70, 253–267 (2010).
15. Hardisty-Hughes, R.E., Parker, A. & Brown, S.D. A hearing and vestibular phenotyping pipeline to identify mouse mutants with hearing impairment. Nat. Protoc. 5, 177–190 (2010).
16. Michalski, N. et al. Harmonin-b, an actin-binding scaffold protein, is involved in the adaptation of mechanoelectrical transduction by sensory hair cells. Pflugers Arch. 459, 115–130 (2009).
17. Grillet, N. et al. Harmonin mutations cause mechanotransduction defects in cochlear hair cells. Neuron 62, 375–387 (2009).
18. Lefèvre, G. et al. A core cochlear phenotype in USH1 mouse mutants implicates fibrous
l e t t e r snp
g©
201
3 N
atur
e A
mer
ica,
Inc.
All
right
s re
serv
ed.
tests were performed by placing mice in a tub of room-temperature water and observing their swimming behavior for 10 s. Contact-righting reflex testing was performed by placing the mouse into a closed clear tube or box and measuring the time it took to right when turned upside down. The trunk-curl test was performed by holding the tail and observing whether the mouse reached for a nearby surface or curled toward the base of its tail.
Auditory-evoked brain stem response. Hearing thresholds of treated and untreated Ush1c 216GG, 216GA and 216AA mice were measured by auditory-evoked brain stem response (ABR). Mice were anesthetized (intraperitoneal ket-amine, 100 mg per kg body weight; xylazine, 6 mg per kg body weight), and body temperature was maintained near 38 °C with a heat pad. All recordings were con-ducted in a sound-proof room. Stimuli consisted of 5-ms pulses of broad-band noise, 8, 16 and 32 kHz, with 0.5-ms linear ramps. Although these tone stimuli encompass low, medium and high regions of mouse spectral sensitivity, BBN was included to confirm that responses are representative of the whole cochlear response. The stimuli were broadcast through a Motorola piezoelectric speaker (model no. 15D87141E02) fitted with a plastic funnel and 2-mm-diameter tubing over the speaker front, producing an acoustic wave guide which was positioned in the external meatus approximately 0.5 cm from the tympanum. Using continuous tones, stimulus amplitude was calibrated at the end of the tubing with a Bruel and Kjaer 2610 measuring amplifier (fast, linear weighting), 4135 microphone (grid on) and 4230 pistonphone calibrator. All stimulus amplitudes were dB SPL (re. 20 mPa). Total harmonic distortion was –40 dB (Hewlet Packard 3562A Signal Analyzer). Stimuli were generated (195 kHz srate) and responses digitized (10 kHz srate) using TDT System III hardware and software (BioSig). ABRs were recorded with a 30-gauge subdermal steel electrode placed subcutaneously behind the left ear, with indifferent and ground electrodes (steel wire, 30-gauge) placed subcutaneously at the vertex and hind limbs, respectively. After amplifi-cation (60 dB, Grass P5 AC), filtering (0.3 Hz–1 kHz; TDT PF1), and averaging (n = 600–1,024), thresholds (± 6 dB) were determined by eye as the minimum stimulus amplitude that produced an ABR wave pattern similar to that produced for the highest intensity stimulus (90 dB).
Scanning electron microscopy. The scanning electron microscopy analysis was performed as has been previously described14. Specifically, intralabyrinthine per-fusion with fixative (2.5% glutaraldehyde/1% paraformaldehyde/1.5% sucrose in 0.12 M phosphate buffer (pH 7.4)) was performed on whole cochleae dis-sected from mice at P18. Cochleae were post-fixed by immersion in the same fixative for 1 d at 4 °C with gentle rotation followed by three washes in 0.12 M phosphate-buffered saline (PBS) and stored for 1 week at 4 °C. Cochleae were next fixed in 1% OsO4 in PBS for 40 min and washed in PBS. Specimens were then serially dehydrated in ethanol, dried in a critical point drier (Autosamdri-814, Tousinis Research Corporation) and mounted on aluminum stubs. The bony capsule of the cochlea, spiral ligament, stria vascularis and Reissner’s membrane were removed, and the whole organ of Corti was exposed with fine dissecting instruments. Specimens were coated in gold/palladium with a Hummer VIA sputter coater (Anatech) and viewed on a JEOL JSM 6300 F scanning electron microscope. At least three individual animals representative of each experimental paradigm were analyzed.
The cochlear place-frequency map relating distance from cochlear apex and frequency is based on a tonotopic map of mice with an average basilar membrane distance from apex to base of 5.13 mm21. Distances were calculated using the equation: d = 156.5 – 82.5 × log(f); d is the normalized distance from the base (%) and f the frequency in kHz21.
Immunofluorescence. Fluorescent labeling of microdissected preparations of the organ of Corti was used to study the hair cells of 1-month-old treated and untreated mutant and control mice as described previously14,31. Briefly, cochleae were isolated from the auditory bulla, and a small opening was created in the apex. The stapes was removed from the oval window, and the cochleae were gen-tly perfused with 2% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4 and post-fixed by immersion for 2 h at 4 °C with gentle rocking. Tissues were washed twice with PBS following fixation and processed for immunohistochemistry. For harmonin analysis, the tectorial membrane was removed with a fine forceps, and the stria vascularis was trimmed. Tissues were blocked for 1 h at room tempera-ture or overnight at 4 °C (harmonin analysis) in a blocking solution consisting of 10% normal donkey serum, 0.5% bovine serum albumin, 0.1% Triton X–100 and
ONLINE METHODSOligonucleotide synthesis. The synthesis and purification of all 2ʹO–methoxyethyl– modified oligonucleotides with phosphorothioate backbone and all 5–methyl cytosines, was performed as described29. The oligonucleotides were dissolved in 0.9% saline and stored at –20 °C. Sequences are shown in Supplementary Table 1.
Plasmids. The minigene expression plasmids pCI–Ush1C_216G and pCI–Ush1C_216A were constructed by amplifying genomic DNA from lymphoblast cell lines derived from an individual with Usher syndrome homozygous for the USH1C c.216G>A mutation USH1C.216AA (GM09458, Coriell Institute) or a healthy individual (GM09456, Coriell Institute). PCR primers specific for the 5ʹ end of exon 2 with restriction sites for XhoI and for the 3ʹ end of exon 4 with a restriction site for NotI at the 3ʹ end were used to amplify by PCR the USH1C 216A minigene fragment. The PCR product was purified and digested with XhoI and NotI and ligated into the expression plasmid pCI expression vector (Promega) digested with the same restriction enzymes.
Cell culture. Plasmids (1 mg) expressing a minigene of human USH1C 216A exons 2–4 and ASOs (50 nM final concentration) were transfected into HeLa cells using Lipofectamine 2000 (Life Technologies). Forty-eight hours after transfec-tion, RNA was isolated using Trizol reagent (Life Technologies) and analyzed by radioactive RT-PCR with the primers pCI FwdB and pCI Rev (Supplementary Table 1) to plasmid sequences flanking exon 2 and exon 4.
Mice. Ush1c 216A knock-in mice were obtained from LSUHSC13 and bred and treated at Rosalind Franklin University of Medicine and Science (RFUMS). All procedures met the NIH guidelines for the care and use of laboratory animals and were approved by the Institutional Animal Care and Use Committees at RFUMS and LSUHSC. Mice were genotyped using ear punch tissue and PCR as described previously14. For all studies, both male and female mice were used in approximately equal proportions. For studies in adult mice (Fig. 1e,f), homozy-gous 216AA mice (2–4 months of age) were injected intraperitoneally with the indicated ASO at indicated dose twice a week for 2 weeks (4 doses). RNA was isolated from different tissues 24 hrs after final injection using Trizol reagent (Life Technologies) and analyzed by radioactive RT-PCR using primers musU-SH1Cex2F and musUSH1Cex5F (Supplementary Table 1) of the Ush1c 216A transgene. Products were separated on a 6% nondenaturing polyacrylamide gel and quantified using a Typhoon 9400 phosphorimager (GE Healthcare). For studies in neonate mice, pups were injected with 300 mg per kg body weight of 2ʹMOE ASOs at different ages, post-natal day 3–16 (P3–P16), as indicated, by intraperitoneal injection. After ABR analysis, mice were killed and tissues were collected. For ABR analysis, mice were shipped to LSUHSC 2–3 weeks after treat-ment. ABR was carried out at least 3 d after mice arrived.
Splicing and protein analysis. Retinae and inner ears were isolated, and cochleae and vestibules were separated and immediately frozen in liquid nitrogen or stored in Trizol reagent. For immunoblot analysis, proteins were obtained from homogenization in a modified RIPA buffer30 or isolated from Trizol reagent (Life Technologies) according to the manufacturer’s instructions. Proteins were separated on 4–15% Tris-glycine gradient gels, transferred to membrane and probed with USH1C- (20900002, Novus Biologicals) or b-actin– (Sigma-Aldrich) specific antibodies. Blots were quantified using ImageJ v1.45s software (NIH). RNA was isolated from different tissues using Trizol reagent (Life Technologies) and analyzed by radioactive RT-PCR using primers musUSH1Cex2F and musU-SH1Cex5F of the Ush1c 216A transgene. Briefly, 0.25–1 mg of RNA was reverse transcribed using GoScript Reverse Transcriptase (Promega, Fitchburg, WI), and 1 ml of cDNA was used in PCR reactions with GoTaq Green (Promega) supple-mented with primers and 0.1–.25 ml of a-32P-dCTP. Products were separated on a 6% nondenaturing polyacrylamide gel and quantified using a Typhoon 9400 phosphorimager (GE Healthcare).
Behavioral analysis. Behavioral tests were performed according to previously established protocols15. Investigators were blinded except in cases where the phenotype made the treatment/genotype status obvious. To quantify circling behavior, mice were placed in an open-field chamber, and behavior was analyzed using ANY-maze behavioral tracking software (Stoelting Co). Ear-twitch, startle and freezing behavior in response to a high-amplitude sound was measured by observing mouse activity following a short whistle (Supplementary Fig. 4). Swim
nature medicine doi :10.1038/nm.3106
npg
© 2
013
Nat
ure
Am
eric
a, In
c. A
ll rig
hts
rese
rved
.
0.03% saponin in PBS to reduce nonspecific binding of primary and secondary antibodies. Primary antibody incubations were then performed at 4 °C in PBS containing 5% normal donkey serum, 0.05% bovine serum albumin, 0.1% Triton X-100 and 0.03% saponin in PBS. For counting cells, a mouse monoclonal anti-parvalbumin antibody (parv19, P3088, Sigma, 1:250) was used to label cochlear hair cells32. To analyze harmonin b expression, polyclonal rabbit anti-harmonin antibodies specific to isoform b (gift from U. Wolfrum, 1:100) were used. For mouse antibodies against parvalbumin, the M.O.M. kit was used as specified by the manufacturer (Vector Labs). Tissues were washed (three times for 10–15 min each) after primary and secondary antibody (donkey anti-mouse Alexa555 (A31570) and Donkey anti-rabbit Alexa488 (A21206), 1:400, Life Technologies). incubations in 0.1% Tween-20 in PBS, and nuclei were counterstained with DAPI (1 mg/ml; D9542, Sigma-Aldrich). F-actin was labeled with rhodamine phalloidin (Life Technologies) according to the manufacturer’s instructions. For counting hair cells, specimens were dehydrated through an ethanol series, cleared with methyl salicylate:benzl benzoate (5:3) and examined by confo-cal fluorescence microscopy. For harmonin b analysis, labeled specimens were mounted and stored in Prolong Gold (Life Technologies). All samples were imaged with a Zeiss motorized system operated with LSM software (Zeiss) and equipped with 405-, 543- and 633-nm diodes along with a multiline argon laser (457 nm, 488 nm, 515 nm); an XYZ stage; and several objectives that include the Plan-NEOFLUAR 10× (NA = 0.3), Plan-NEOFLUAR 40× (NA = 1.3 oil) and Plan-APOACHROMAT 100× (NA = 1.4 oil) used. Scans were performed through a sequential (line) mode and PMT voltages dynamically regulated to compensate for signal loss due to scatter and depth limitations. Planes were captured at a resolution of 2048 × 2048 pixels and speeds of 1–2 ms per pixel. Optical volumes were deconvolved with a constrained maximum likelihood estimation algorithm and a calculated point spread function using Huygens
Professional 4.1 (Scientific Volume Imaging) running on a Mac Pro computer (Apple). Z-stack images were reconstructed and analyzed using ImageJ, Fuji and Photoshop softwares.
Statistical analyses. Data were analyzed by ANOVA with post hoc tests and Student’s t-test (SAS Institute Inc, NC or Prism 5 Graphpad Software) as noted in the figure legends. Hair cell counts were analyzed as the dependent variable separately for both inner and outer hair cells in a nested ANOVA with a two-level factorial arrangement of treatments33. The nested effect was the mice within each genotype treatment combination, the two main effect factors were cell location in the cochlea and genotype/treatment combination (combined into one variable with three levels, see Fig. 4). Adjustment for multiple comparisons conducted to separate interaction means was by a simulation method34. All data management and analysis was performed using programs and procedures in the Statistical Analysis System (SAS Institute).
29. Baker, B.F. et al. 2ʹ-O-(2-methoxy)ethyl–modified anti-intercellular adhesion molecule 1 (ICAM-1) oligonucleotides selectively increase the ICAM-1 mRNA level and inhibit formation of the ICAM-1 translation initiation complex in human umbilical vein endo-thelial cells. J. Biol. Chem. 272, 11994–12000 (1997).
30. Hastings, M.L. et al. Tetracyclines that promote SMN2 exon 7 splicing as therapeutics for spinal muscular atrophy. Sci. Transl. Med. 1, 5ra12 (2009).
31. Hardie, N.A., MacDonald, G. & Rubel, E.W. A new method for imaging and 3D recon-struction of mammalian cochlea by fluorescent confocal microscopy. Brain Res. 1000, 200–210 (2004).
32. Sage, C., Venteo, S., Jeromin, A., Roder, J. & Dechesne, C.J. Distribution of frequenin in the mouse inner ear during development, comparison with other calcium-binding proteins and synaptophysin. Hear. Res. 150, 70–82 (2000).
33. Milliken, G.A. & Johnson, D.E. Analysis of Messy Data Volume I: Designed Experiments Ch. 30, 413–423 (Lifetime Learning Publications, Belmont, California, 1984).
34. Edwards, D. & Berry, J.J. The efficiency of simulation-based multiple comparisons. Biometrics 43, 913–928 (1987).
doi:10.1038/nm.3106 nature medicine
npg
© 2
013
Nat
ure
Am
eric
a, In
c. A
ll rig
hts
rese
rved
.
Competing financial interests
F.R. may materially benefit financially through stock options in Isis Pharmaceuticals. M.L.H. and F.R. have patents pending with the United States Patent and Trademark Office for the ASOs and the targeting approach.
npg
© 2
013
Nat
ure
Am
eric
a, In
c. A
ll rig
hts
rese
rved
.
Neuron
Article
Restoration of Hearingin the VGLUT3 Knockout MouseUsing Virally Mediated Gene TherapyOmar Akil,1 Rebecca P. Seal,3 Kevin Burke,1 Chuansong Wang,4 Aurash Alemi,1 Matthew During,4 Robert H. Edwards,2
and Lawrence R. Lustig1,*1Department of Otolaryngology, Head and Neck Surgery2Department of NeurologyUniversity of California San Francisco, San Francisco, CA, 94143, USA3Department of Neurology, University of Pittsburgh, Pittsburgh, PA 15213, USA4Comprehensive Cancer Center, The Ohio State University, Columbus, OH 43210, USA
*Correspondence: [email protected]://dx.doi.org/10.1016/j.neuron.2012.05.019
SUMMARY
Mice lacking the vesicular glutamate transporter-3(VGLUT3) are congenitally deaf due to loss of gluta-mate release at the inner hair cell afferent synapse.Cochlear delivery of VGLUT3 using adeno-associ-ated virus type 1 (AAV1) leads to transgene expres-sion in only inner hair cells (IHCs), despite broaderviral uptake. Within 2 weeks of AAV1-VGLUT3delivery, auditory brainstem response (ABR) thresh-olds normalize, alongwith partial rescue of the startleresponse. Lastly, we demonstrate partial reversal ofthe morphologic changes seen within the afferentIHC ribbon synapse. These findings represent asuccessful restoration of hearing by gene replace-ment in mice, which is a significant advance towardgene therapy of human deafness.
INTRODUCTION
Hearing loss is one of the most common human sensory deficits,
with congenital hearing loss occurring in approximately 1.5 in
1,000 children (Smith et al., 2005). Of these, about half are attrib-
uted to a genetic basis (Di Domenico et al., 2011). While our
understanding of the causes of genetic hearing loss has
advanced tremendously over the past 30 years (Petersen and
Willems, 2006), treatments have advanced little over this same
time period and currently consist of hearing amplification for
mild to severe losses and cochlear implantation for severe to
profound losses (Kral and O’Donoghue, 2010). Though cochlear
implantation has profoundly influenced our treatment of children
with congenital deafness, there are still significant limitations in
function with an implant, and these results cannot compare to
native hearing (Kral and O’Donoghue, 2010). Thus, there remains
intense interest in restoring normal organ of Corti function
through techniques such as hair cell regeneration and gene
therapy (Di Domenico et al., 2011). To date, a majority of the
research in this arena has focused on cochlear hair cell regener-
ation, applicable to the most common forms of hearing loss
including presbycusis, noise damage, infection, and ototoxicity.
Several studies have now demonstrated regeneration of hair
cells in injured mice cochlea and improvement of both hearing
and balance with virally mediated delivery of Math1 (Baker
et al., 2009; Husseman and Raphael, 2009; Izumikawa et al.,
2008; Kawamoto et al., 2003; Praetorius et al., 2010; Staecker
et al., 2007). While these efforts in wild-type animals are quite
important, they still do not address the problem of an underlying
causative genetic mutation. In such a scenario, even success-
fully regenerated hair cells will still be subject to the innate
genetic mutation that led to hair cell loss in the first place. To
date, efforts to restore hearing in this type of hearing loss with
gene therapy have been met with limited success (Maeda
et al., 2009), and no study has reported the reversal of deafness
in an animal model of genetic deafness.
Previous reports have described a mouse model of hereditary
deafness, which occurs as a result of a null mutation in the
gene coding for the vesicular glutamate transporter-3 (VGLUT3)
(Obholzer et al., 2008; Ruel et al., 2008; Seal et al., 2008).
Synaptic transmission mediated by glutamate requires transport
of the excitatory amino acid into secretory vesicles by a family of
three vesicular glutamate transporters (Fremeau et al., 2004;
Takamori et al., 2002). We previously demonstrated that inner
hair cells of the cochlea express VGLUT3 and that mice lacking
this transporter are congenitally deaf (Seal et al., 2008). Hearing
loss in these mice is due to the elimination of glutamate release
by inner hair cells and hence to the loss of synaptic transmission
at the IHC-afferent nerve synapse. Subsequent studies have
shown that a missense mutation in the human gene SLC17A8,
which encodes VGLUT3, might underlie the progressive high-
frequency hearing loss seen in autosomal dominant DFNA25
(Ruel et al., 2008). Here we report the successful restoration of
hearing in the VGLUT3 knockout (KO) mouse using virally
mediated gene delivery.
RESULTS
AAV1-VGLUT3 Transfection Results in LocalizedExpression in Inner Hair CellsOur first goal was to determine the extent of transfection with the
adeno-associated virus type 1 (AAV1) within the cochlea. Using
Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc. 283
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
an AAV1-GFP construct, there appeared to be labeling of
a variety of cell types within the cochlea, including the inner
hair cells and supporting cells using an anti-GFP antibody (Fig-
ure 1A), in a pattern similarly described by other investigators
(Jero et al., 2001; Konishi et al., 2008). Subsequently, virus con-
taining the VGLUT3 gene (AAV1-VGLUT3) was microinjected
into the cochlea using two different techniques: initially via an
apical cochleostomy (CO) and subsequently by direct injection
through the round window membrane (RWM) (Figures 1B–1E).
After delivery, RT-PCR of inner ear tissue (Figure 1C) demon-
strated strong VGLUT3 mRNA expression in the rescued whole
cochlea, organ of Corti, stria vascularis, vestibular neuroepithe-
lium, and very weakly in the spiral ganglion. Noninjected
cochleas of knockouts do not demonstrate VGLUT3 expression
as noted (Figure 1C, KO �/+RT). In contrast, under immunofluo-
rescence, inner hair cells were the only cells labeled with anti-
VGLUT3 antibody (Figure 1B).
To determine the dose dependence of VGLUT3 expression
in the IHCs, we injected either 0.6 ml or 1 ml of AAV1-VGLUT3
(2.3 3 1013 virus genomes [vg]/ml) into the cochlea (Figures 1D
and 1E). Microinjecting 1 ml of virus resulted in 100% of IHCs
labeled with anti-VGLUT3 antibody; in contrast, microinjecting
0.6 ml resulted in only �40% of IHCs labeled by the antibody.
We next sought to determine whether earlier viral delivery
would result in more robust VGLUT3 expression (Figures 1D,
1E, and 2). As shown, delivery of virus via the RWM at postnatal
day 10 (P10) results in �40% of the IHCs expressing VGLUT3
(Figures 1D, 1E, and 2), whereas similar doses (0.6 ml) of virus
injected at P1–P3 results in 100% of IHC transfected in all
animals (Figures 1D, 1E, and 2).
Delivery of AAV1-VGLUT3 Restores Normal ABR andCAP Thresholds within 14 DaysAfter verifying successful transgene expression within the IHC
without significant organ of Corti injury, we next sought to deter-
mine whether the reintroduction of VGLUT3 would lead to meas-
ureable hearing recovery (Figure 3). In these studies, only 0.6 ml
of AAV1-VGLUT3 was delivered at P10–P12. Auditory brainstem
response (ABR) thresholds were first measurable within 7 days
after viral delivery, with near normalization of thresholds
to wild-type (WT) levels within 2 weeks (P24–P26) (Figures
3A–3C). Initially a CO technique was used for viral delivery.
However, this method restored hearing in only �17% of animals
(n = 5 out of 30 animals attempted), presumably because it was
more technically challenging and due to the trauma of the
approach (see Discussion). As a result, the method was subse-
quently changed to an RWM delivery, which resulted in hearing
restoration in 100% of mice (n = 19 out of 19 mice). The time
course of hearing recovery was similar for the CO (when
successful in 17%) and the RWM delivery techniques (100% of
mice). Compound action potentials (CAPs) were also restored
within 7–14 days of viral delivery (Figure 3A). Since the loss of
VGLUT3 affects only glutamate release at the IHC synapse
(Ruel et al., 2008; Seal et al., 2008), restoration of normal ABRs
and CAPs also implies restoration of synaptic function. We
also compared the longevity of hearing recovery, defined as
the period of time between onset of hearing recovery and
when ABR thresholds become elevated >10 dB aboveWT levels,
284 Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc.
between the CO and RWMmethods (Figure 3D). In both groups,
all rescued KO mice maintained hearing for at least 7 weeks. At
28 weeks postdelivery, 40% of the mice who achieved suc-
cessful CO delivery still had hearing within 10 dB of WT mice
(n = 2/5), while only 5% of the RWM mice had the same level
of hearing (n = 1/19). Interestingly, some rescued mice in each
group, CO and RWM, maintained normal ABR thresholds up to
1.5 years. The number of animals for each rescued group at
each time point, within 10 dB of WT thresholds, is described in
the legend of Figure 3D.
We subsequently measured hearing recovery in mice injected
via the RWM at P1–P3 (Figure 3D). Due to the small size of the
cochlea, only 0.6 ml of virus could be delivered at this time point.
However, 100% of mice recovered normal ABR thresholds by
P14 (n = 19 mice). Five mice were followed for 9 months and still
maintained normal ABR thresholds at this later time point. Earlier
delivery thus not only appears to be more efficient (100% of
animals recover hearing) but also leads to greater longevity of
hearing recovery.
Bilateral AAV1-VGLUT3 Rescue Results in a LargerRecovery of Behavioral and Electrical Measuresof HearingFor an additional assay of hearing recovery, we studied the
startle response at approximately 3 weeks after viral delivery
(Figure 4). In these experiments, the AAV1-VGLUT3 delivery
was done via the RWM at age P10–P12. As expected, VGLUT3
KOmice show no startle response due to the absence of hearing.
When hearing was rescued in one ear (‘‘unilat,’’ Figure 4A), at the
loudest presentation level of 120 dB, the startle response
improved to 8% of normal, while if both ears were rescued
(‘‘bilat,’’ Figure 4A), the startle response increased to 33% of
normal, both measures being statistically different than the KO
response. Interestingly, similar amplitude growth was observed
with ABR wave I amplitudes when both ears, as opposed to
a single ear, were rescued (Figure 4B). ABR wave I latency was
also studied (Figure 4C), and while there appeared to be a trend
for reduced latency in the unilateral-rescued mice, the differ-
ences between unilateral- and bilateral-rescued and WT mice
were not significant. Thus, while ABR thresholds can be brought
to normal, ‘‘behavioral’’ thresholds and ABR amplitudes can be
improved, but not normalized, to the WT level with this rescue
technique.
AAV1-VGLUT3 Rescue Partially Restores SynapticMorphology, but Not Spiral Ganglion Cell Counts,in the Knockout MouseAs we previously demonstrated (Seal et al., 2008), at P21,
VGLUT3 KO mice show a 10%–18% decrease in spiral ganglion
(SG) neurons compared to WT mice. This decrease was still
observed in the AAV1-VGLUT3 rescued mice (RWM delivery at
P10–P12) at P21 (Figure 5A). Further, rescued mice showed no
significant differences in spiral ganglion cell size as compared
to KO mice (Figure 5B), though both were significantly less
than WT mice. To determine whether long-term hearing would
reverse this trend, we also took counts at 5 months after birth,
but again, no significant differences in SG counts or cell size
were seen in the KO versus rescued mice at this later time point
Figure 1. AAV1-GFP Transduction in Mice Organ of Corti
(A) AAV1-GFPwas used to assess viral delivery to the cochlea on organ of Corti surface preparations. AAV1 transfects a number of cell types including inner (IHC)
and supporting cells (white arrows, left) in the organ of Corti. The VGLUT3 KO (V3KO, middle) documents IHC labeling of transfected cells only. In the WTmouse
(right), VGLUT3 colabels IHCs (red) along AAV1-GFP-transfected cells (green).
(B) VGLUT3 expression after AAV1-VGLUT3 delivery via the RWM, delivered at P10–P12, and stained with anti-myosin 7A antibody (green, used as a hair cell
marker, left column) and anti-VGLUT3 antibody (red, middle column) and merged (right column). WTmice show IHCs labeled with both anti-Myo7A and VGLUT3
as expected (top row, right), whereas KO mice only show Myo7a expression (middle row, right). After AAV1-VGLUT3 delivery, the IHC is colabeled by both
anti-Myo7a and anti-VGLUT3 antibodies (bottom row, right) (TC, tunnel of Corti; DC, Deiter’s cells; OHC, outer hair cells; IHC, inner hair cells).
(C) RT-PCR was used to verify VGLUT3 mRNA expression in the transfected KO (first lane) and WT (second lane) mice. Rescued whole cochlear extract (R-Co)
demonstrates strong VGLUT3 mRNA expression, as does the stria vascularis + organ of Corti lane (R-SV+OC) and vestibular neuroepithelium (R-Vest). Spiral
ganglion only shows weak mRNA expression (R-SG).
(D and E) Inner hair cells labeled with anti-VGLUT3 antibody after transfection were counted, tabulating both the entire cochlea and within the base, midturn, and
apex to determine differences in regions (at P10–P12 delivery, WT n = 5, CO n = 5, RWM n = 6 and at P1–P3 delivery, RWM n = 6). At P10–P12, when 1 ml of virus
was microinjected, 100% of IHCs were labeled and similar results were seen at P1–P3 viral delivery (D) when injected with 0.6 ml. In contrast, when 0.6 ml of virus
was injected at P10–P12, approximately 40% of IHCs were labeled, with no significant differences seen between the apex, midturn, or base in the variability of
IHC labeling with the delivery technique (E).
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc. 285
Figure 2. VGLUT3 IHC Transfection: Early versus Late Delivery
AAV1-VGLUT3 delivery at P1–P3 versus P10–P12 are compared using similar amount of virus (0.6 ml) and examined at �P30. Anti-Myo7a antibody (green) and
anti-VGLUT3 antibody (red) are used for staining, and the merged images (yellow) are shown. As expected, IHC from WT mice show both anti-Myo7a and anti-
VGLUT3 staining (row 1), whereas KO mice only show anti-Myo7a label (row 2). Delivery of virus via the RWM at P10–P12 results in fewer IHCs expressing
VGLUT3 (row 3), whereas similar doses of virus injected at P1–P3 results in 100% of hair cells transfected in all animals (row 4) (IHC, inner hair cells).
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
(data not shown). Subsequently, spiral ganglion cell counts were
also undertaken in mice that underwent virus delivery at P1–P3.
However, despite a robust IHC transfection and early hearing
recovery (see Figures 1D, 1E, and 2), again, no differences in
SG cell counts were noted between KO and rescued mice
(data not shown). Additionally, histology (Figure 5C) documents
no obvious cochlear trauma as a result of viral delivery in the
rescued mice, as evidenced by normally appearing organ of
Corti structures with preservation of inner and outer hair cells,
supporting cells, spiral ganglion neurons (though similarly
reduced in number as nonrescuedmice), and the stria vascularis
(data not shown).
As originally reported, VGLUT3 KO mice demonstrate abnor-
mally thin, elongated ribbons in IHC synapses, though the
number of synaptic vesicles tethered to ribbons or docked at
the plasma membrane was normal (Seal et al., 2008). We thus
sought to determine whether these morphologic abnormalities
could be reversedwith hearing rescue. As shown (Figure 6, Table
1), in the rescued mice, ribbon synapses are normal in appear-
ance, taking on a more rounded shape similar to the WT, while
the nonrescued mice continue to demonstrate abnormally thin
and elongated ribbons. The rescuedmice also displayed a signif-
icantly larger number of synaptic vesicles associated with the
ribbon (19 rescued versus 14 WT, p = 0.02) (Table 1). Interest-
286 Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc.
ingly, within individual hair cells, the synaptic vesicles them-
selves demonstrated a mixture of elongated and circular mor-
phology, as opposed to all circular in the WT and all elongated
in the KO mice. However, when analyzing the average number
of docked synaptic vesicles at a ribbon synapse, rescued
animals did not show a significant difference between the WT
and KO mice (Table 1). While these results demonstrate only
a partial reversal of the synaptic changes seen in the KO mouse
ribbon synapse, it is enough to recover ABR thresholds to theWT
levels in the rescued KO mice.
DISCUSSION
These studies document the successful rescue of the deafness
phenotype in a mouse model of inherited deafness. With viral
delivery of VGLUT3 at P10–P12 in the KO mouse, ABR thresh-
olds normalize within 7–14 days and remain in this range for at
least 7 weeks, with two mice maintaining auditory thresholds
for as long as a year and a half in this current study. Earlier
delivery, at P1–P3, results in an even more robust IHC transfec-
tion and long-lived hearing recovery in this mouse model.
One unexpected result from these investigations was the
differential finding of more widespread expression of green fluo-
rescent protein (GFP) after AAV1-GFP transfection as compared
Figure 3. Hearing Restoration in the VGLUT3 KO Mice
(A) Representative ABR and CAP tracings fromWT, KO, and rescued KO mice after delivery of 0.6 ml AAV1-VGLUT3. Waveforms from the WT and rescued mice
appear similar, while KO mice show no ABR and no CAP responses. ‘‘I’’ indicates the location of ABR wave I.
(B) Rescued mice begin to show hearing recovery within 7 days postdelivery (P17–P19), with near normalization of ABR thresholds by 14 days postdelivery
(P24–P26). Hearing recovery is seen with both CO and RWM delivery through 69 weeks. Black arrows indicate absence of ABR threshold of the KO above 92 dB
(the maximum level that our recording equipment can measure).
(C) At 40 days postdelivery (P50–P52), similar levels of recovery are noted at 8 and 16 kHz, while at 32 kHz, CO delivery appears to result in slightly elevated
thresholds, though still significantly better than KO.
(D) Hearing longevity, as defined as the number of mice with ABR threshold levels within 10 dB of WT levels at each time point, is measured for RWM at P10–P12,
CO at P10–P12, and RWM at P1–P3. For viral delivery at P10–P12, CO delivery results in longer-lasting hearing recovery on average than RWM delivery, with
some rescued KO in each group maintaining ABR thresholds to within 10 dB of WT animals beyond 28 weeks. However, 100% of rescued KO mice with RWM
delivery recover hearing, while only 17% of CO-rescued mice recover hearing. The number of animals with viral delivery at P10–P12 at each time point is the
following: RWM 0–9 weeks n = 19, 10–13 weeks n = 12, 14–24 weeks n = 8, 24–28 weeks n = 1, 28+ weeks n = 1; CO 0–7 weeks n = 5, 8–14 weeks n = 4,
15–20 weeks n = 3, 21–28 weeks n = 2, 28+ weeks n = 1. In contrast, in mice undergoing viral delivery at P1–P3, all animals maintain ABR threshold recovery at
least through 9 months (n = 5).
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
to isolated VGLUT3 expression restricted to the IHC after
AAV1-VGLUT3 transfection (Figures 1A–1C and 2). RT-PCR
results demonstrate that after AAV1-VGLUT3 delivery, there is
also more widespread VGLUT3 mRNA transcription than in just
IHCs (Figure 1C). These results suggest that there is a posttran-
scriptional regulatory mechanism acting on VGLUT3 mRNA,
which leads to selective expression of the protein only within
IHCs. Several types of posttranscriptional regulation have been
described within the cochlea, and whether this specific mecha-
nism involves microRNA inactivation (Elkan-Miller et al., 2011),
transcription factor regulation (Masuda et al., 2011), or another
process remains to be determined. Such a mechanism, if appro-
priately elucidated and exploited, could theoretically allow the
expression (or conversely suppression) of a number of different
proteins within the inner ear to alter function in pursuit of hearing
preservation.
Another interesting finding was the variable success with the
CO as compared to the RWM delivery technique. As noted, we
initially started with an apical CO delivery method but aban-
doned it due to the low success rate of hearing restoration
(17% of animals). Subsequently, we changed to an RWM
delivery technique for several reasons; this would be the most
likely method of delivery in any future human studies, and it
was less likely to be traumatic, as evidenced by a number of
recent human studies looking at hearing preservation with round
window insertion of cochlear implants (von Ilberg et al., 2011). In
Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc. 287
Figure 4. Behavioral Measures and Physiologic Growth after AAV1-VGLUT3 Rescue
(A) Startle responses were studied as an additional behavioral measure of hearing recovery 3 weeks after viral delivery (P31–P33). While none of the rescuedmice
recover startle responses to WT levels, they nonetheless develop a response, which increases when both ears are rescued; the louder the sound delivered (100,
110, and 120 dB presentation levels), themore robust the startle response was, with rescue of both ears (bilat) creating a larger response than a single ear (unilat).
These differences were statistically significant.
(B) Similar growth was seen for ABR wave I amplitudes when both ears (bilat) as compared to a single ear (unilat) were rescued, though again not as robust as in
the WT mice.
(C) ABRwave I latency wasmeasured, which showed no statistically significant differences between unilateral, bilateral, rescued, andWTmice, though there was
a trend toward an increased latency in the rescued compared to WT mice.
(D) Representative ABR tracings documenting increasing wave I amplitudes from the KO, unilaterally rescued KO (unilat), bilaterally rescued KO (bilat), and WT
mice.
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
fact, the change in technique resulted in hearing restoration in
100% of animals attempted. We believe the likely difference in
success between the two techniques relates to the degree of
trauma induced by each method. With a cochleostomy, a sepa-
rate hole into the scala through bone must be created, which by
its nature is traumatic, despite our best efforts to minimize
trauma. In contrast, an RWM injection simply involves piercing
the membrane and sealing it with fascia after viral delivery.
However, we were histologically unable to see any obvious
differences between the ears of animals with andwithout hearing
rescue in the cochleostomy group (data not shown) and there
may be other reasons for the variable success between the
two techniques. Further, we noted that even earlier delivery via
the RWM at P1–P3, as opposed to P10–P12, resulted in hearing
recovery that was more consistently long lived, with all mice
288 Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc.
followed out through 9 months showing ongoing normal ABR
thresholds (Figure 3D).
Transgene expression with AAV1 should theoretically last for
a year or longer (Henckaerts and Linden, 2010). However, it is
not entirely clear why there is a variable loss of hearing after
7 weeks, regardless of delivery technique at the later P10–P12
delivery time point (Figure 3D). We analyzed individual cochleae
but did not see histologic evidence of active inflammation in
those animals that lost hearing and IHCs still expressing VGLUT3
protein. Further, spiral ganglion counts did not significantly differ
in animals with and without hearing. One possibility could be due
to the trauma of viral delivery, with gradual reopening of the
delivery site (RWM or cochleostomy) leading to a perilymphatic
leak with resulting hearing loss. Such a lesion might not be
detectable on histology. Another possible explanation may be
Figure 5. Spiral Ganglion Cell Counts
In these studies, the AAV1-VGLUT3 delivery was done at P10–P12 and at P1–P3. Spiral ganglion (SG) cell counts in rescued mice versus KO and WT mice were
undertaken at P21 (A and B) and at 5 months after birth (data not shown) for P10–P12 virus delivery and at twomonths after birth (data not shown) for P1–P3 virus
delivery. AAV1-VGLUT3 transfection did not result in a reversal of the SG neuronal loss in either age group. No significant differences were seen between the
groups. These results also document the lack of trauma (C) associated with virally mediated delivery of VGLUT3 through the RWM, with normal organ of Corti
morphology, including inner and outer hair cells, spiral ganglion neurons, and stria vascularis (data not shown).
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
due to transgene inactivation, by a hypothetical mechanism such
as microRNA inactivation or methylation. Clearly, if one hopes to
consistently achieve long-term transgene expression within the
ear, which will be critical for application of this technique in
humans, this variable will need to be better understood and
controlled, particularly at later ages of delivery.
It is interesting to note that the lower dose of virus used for
most of the studies performed (0.6 ml), delivered at P10–P12,
caused VGLUT3 expression in only �40% of IHCs (Figures 1D
and 1E), and yet this was enough to restore ABR thresholds to
WT levels for click responses and to near normal for pure tone
thresholds (Figures 3A–3C). Similar results have been docu-
mented in other models of hearing recovery after noise exposure
(Kujawa and Liberman, 2009; Lin et al., 2011), in which even
‘‘reversible’’ noise exposure with recovery of auditory thresholds
leads to long-term afferent nerve terminal degeneration while
retaining ‘‘normal’’ auditory thresholds. Similar findings with re-
gard to the discrepancy of ABR threshold and amplitudes have
also been shown from mutant mice lacking synaptic ribbons
(Buran et al., 2010). However, correlative studies in human
temporal bones suggest that cochlear implants in humans can
still function very effectively despite significant spiral ganglion
neuron loss, allowing for meaningful speech and sound trans-
mission (Gassner et al., 2005; Khan et al., 2005). Thus, complete
normalization of all cellular abnormalities may ultimately not be
required for the technique to be successful in humans, though
this should remain a goal for animal studies going forward.
The KO mice develop an unusual appearing ribbon that is thin
and elongated, as noted here and previously (Seal et al., 2008). A
similar ribbon morphologic pattern, flat and plate-like, is seen in
the Otoferlin KO mouse (Roux et al., 2006). As Otoferlin is also
critical in glutamate release at the IHC synapse, this implies
that lack of physiologic activity of the synapse results in such
a flat ribbon appearance. In the rescued mice, while the ribbon
itself appeared normal, we did still see a mixture of elongated
and circular vesicles within the transfected IHCs, as opposed
to all circular in theWT and all elongated in the KOmice, implying
that there may still be differences in transmitter release in the
rescued versus WT mice.
Another interesting finding with regard to the ribbon synapse
was the larger numbers of synaptic vesicles that were associ-
ated with the ribbon seen in the rescued mice, as well as the
mixture of elongated and circular vesicles observed. The data
shows that much larger quantities of VGLUT3 mRNA are being
produced in the rescued as compared to theWTmice (Figure 1C,
RT-PCR data) and suggests, though does not prove, an associ-
ation between increased mRNA levels and vesicle number. We
believe that circular vesicles represent properly packaged vesi-
cles, while the elongated vesicles are improperly packaged
vesicles. Perhaps continuous production of VGLUT3 by the
Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc. 289
Figure 6. Partial Normalization of Synaptic Ribbon Morphology in the Rescued Mice
Electron microscopy was performed on broken serial thin sections of the synaptic region of the IHCs, which were cut in a horizontal plane parallel to the basilar
membrane in the same orientation for the WT, KO, and rescued KO mice. As previously reported (Seal et al., 2008), VGLUT3 KO mice demonstrate abnormally
thin, elongated ribbons in IHC synapses (middle column), as compared to WT littermates (left column). In the rescued mice (right column, single ribbon shown on
top, double on the bottom), ribbon synapses are normal appearing, taking on amore rounded shape similar to theWT. Black arrows point to the normal or circular
synaptic vesicles and white arrows point to elongated synaptic vesicles.
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
constitutive CBA promoter driving transfected VGLUT3 pro-
duction prevents the IHC from properly packaging the vesicles
at a normal rate, leading to a higher number as well as a mixture
of regular and irregular-appearing vesicles. Another possibility
is that the incomplete transfection rate of IHCs (40% of IHCs
labeled at the doses used for these morphology studies) led to
the heterogeneity of the ribbon morphology seen.
The observed growth on behavioral and electrical measures
seen with bilateral, as opposed to unilateral, rescue (RWM
delivery at P10–P12; Figure 4) was an unexpected finding. While
none of the animals had complete normalization of ABR ampli-
tude and startle-response levels, the amplitude growth does
imply that bilateral input increases the auditory response cen-
trally. An analogous phenomenon is seen with ‘‘binaural summa-
tion’’ and clinically in patients who wear two hearing aids as
opposed to one and report lower levels of amplification required
(Noble, 2010; Steven Colburn et al., 2006) and suggests that the
response seen in these studies is physiologic. Recent studies
have localized VGLUT3 to various structures in the brainstem,
including cochlear nucleus (Fyk-Kolodziej et al., 2011) as well
as the LSO and MNTB (Lee et al., 2011). It is certainly possible
that deficits within auditory brainstem signal pathways could
be contributing to the inability to restore the startle response to
WT levels.
The failure of the technique to reverse the spiral ganglion cell
loss seen in the VGLUT3 KO mice when delivered at P10–P12
is not surprising (Figure 5), given that hair cell activity and afferent
290 Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc.
stimulation can provide a trophic effect on SG survival. This is
probably at least partly due to the fact that virus was delivered
at �P10 with subsequent ABR threshold recovery at �P17–
P24, after spiral ganglion neuronal degradation has begun
(Seal et al., 2008). This also implies that in order for SG neurons
to be preserved at normal levels, intervention would probably
have to occur earlier. Further, with only �40% of IHCs express-
ing VGLUT3 (using the lower concentration of virus, delivered at
P10–P12), there are still many spiral ganglion neurons not
receiving afferent input, which also probably impacts this result
as well. Wewere thus surprised that even earlier delivery of virus,
at P1–P3, which resulted in relatively early onset of hearing,
measureable by P14, with 100% of IHC expressing VGLUT3,
also did not lead to restoration of SG cell counts to WT levels.
Perhaps there are in utero factors that also help maintain SG
numbers, or even a small delay in hearing onset can lead to
SG loss.
Lastly, while the mutant mouse in the current study and the
hearing loss described in patients with DFNA25 are both due
to mutations in the gene coding for VGLUT3, the comparison
may not be straightforward. First, it is not certain that the
missense mutation described in SLC17A8 is the cause of the
hearing loss seen in DFNA25, though a strong correlation was
observed (Ruel et al., 2008). Second, the null mutation studied
in these experiments would represent a much more severe
phenotype than the missense mutation described as potentially
causative for DNFA25. Thus, whether this technique could
Table 1. Transmission Electron Micrographic Analysis of the
Ribbon Synapse
WT KO Rescued KO
Number of inner hair
cells examined
58 50 61
Total number of
ribbons
18 20 18
Number of floating
ribbons
3 3 4
Number of normal
ribbons
18 0 17
Number of elongated
ribbons
0 20 1
Average number of
synaptic vesicles
associated with the
ribbon
14.44 ± 1.92 16.26 ± 2.85 19.17 ± 4.52
Shape of the synaptic
vesicles
All circular All elongated Mixture of
circular +
elongated
Average number of
docked synaptic
vesicles
2.28 ± 1.27 1.56 ± 0.92 2 ± 0.84
This analysis compares the ribbon synapse and vesicles inWT versus KO
versus AAV1-VGLUT3-rescued KO mice. Electron microscopy was per-
formed on broken serial thin sections of the synaptic region of the
IHCs, which were cut in a horizontal plane parallel to the basilar
membrane with identical handling and orientation for the WT, KO, and
rescued KO. When differentiating vesicles as either ‘‘circular’’ or ‘‘elon-
gated,’’ a vesicle was defined as ‘‘circular’’ when the perpendicular diam-
eters were found to bewithin 50%of each other. Any vesicle with unequal
perpendicular dimensions (>50% difference) was counted as ‘‘elon-
gated,’’ and any ribbon synaptic body that had a length greater than three
times its greatest width was considered to be ‘‘abnormal’’ in appearance.
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
ultimately be beneficial to patients with DFNA25 remains
unclear. Despite these differences, as our study documents
restoration of normal ABR levels in such a null mutant model, it
nonetheless represents an important initial step for the potential
treatment of inherited deafness.
EXPERIMENTAL PROCEDURES
Animals
VGLUT3 null mutant mice were generated as described in a C57 (Seal et al.,
2008) strain then backcrossed with FVB mice (less than seven generations)
to obtain a homogeneous genetic background. P1–P12 mice were used for
AAV1-VGLUT3 delivery. All procedures and animal handling complied with
NIH ethics guidelines and approved protocol requirements at the University
of California, San Francisco (IACUC).
All surgical procedures were done in a clean, dedicated space. Instruments
were thoroughly cleaned with 70% ETOH and autoclaved prior surgery.
Surgery was carried out under a Leica MZ95 dissecting scope and animals
were situated with neck extended over solid support. Mice were anesthetized
by intraperitoneal injection of a mixture of ketamine hydrochloride (Ketaset,
100 mg/kg), xylazine hydrochloride (Xyla-ject, 10 mg/kg), and acepromazine
(2 mg/kg) and boosted with one-fifth the original dose as required. Depth
of anesthesia was continuously checked by deep tissue response to toe
pinch. Body temperature was maintained with a heating pad and monitored
with a rectal probe throughout procedures. Preoperatively and every 24 hr
postoperatively, animals were given subcutaneous carprofen analgesia
(2 mg/kg) to manage inflammation and pain. Animals were closely monitored
for signs of distress and abnormal weight loss postoperatively.
AAV1-VGLUT3 Viral Construct
Mouse VGLUT3 cDNA was subcloned into the multiple cloning site of vector
AM/CBA-WPRE-BGH (kindly provided by R. Palmiter). Human embryonic
kidney 293 cells were cotransfected with three plasmids—AAV-mVGLUT3
plasmid, appropriate helper plasmid-encoding rep and AAV1 cap genes,
and adenoviral helper pF D6—using standard CaPO4 transfection. Cells
were harvested 60 hr after transfection, cell pellets were lysed with sodium
deoxycholate, and AAV vectors were purified from the cell lysate by ultracen-
trifugation through an iodixanol density gradient, then concentrated and
dialyzed against phosphate-buffered saline (PBS), as previously described
(Cao et al., 2009, 2010; Lawlor et al., 2009). Vectors were titered using real-
time PCR (ABI Prism 7700; Applied Biosystems), and purity of vector stocks
was confirmed by running a 10 ml sample on sodium dodecyl sulfate polyacryl-
amide gel electrophoresis and staining with Coomassie blue.
Surgical Procedures
Round Window Membrane Injection
Animalswere anesthetized, the left ear was approached via a dorsal incision as
described by Duan et al. (2004). A small hole was made in the bulla with an
18G needle and expanded as necessary with forceps and the round window
membrane (RWM) was identified. The RWM was gently punctured with a
borosilicate capillary pipette and remained in place until efflux stabilized. A
fixed volume of AAV1-VGLUT3 (0.6 ml or 1.0 ml of a 2.33 1013 vg/ml) previously
drawn into the fine pipette was gently injected through RWM over 1–2 min.
After pulling out the pipette, the RWM niche was quickly sealed with fascia
and adipose tissue. The bulla was sealed with dental cement (Dentemp,
Majestic Drug Company) and the wound was sutured in layers with a 5-0
absorbable chromic suture (Ethicon).
Cochleostomy Delivery
The right ear was approached via ventral, paramedian incision in the neck as
described by Jero et al. (2001). The injection method was similar to the RWM
except that the hole in bulla was made slightly more anterior and larger, to
directly approach the space above the stapedial artery. Injection of virus
was made into the apical turn. We used a 0.5 mm drill bit to gently thin the
bone of the otic capsule where the stria vascularis could be slightly visualized
as a brownish stripe. Once enough bonewas shaved and a slight fluid interface
became visible, 0.6 ml VGLUT3-AAV1 (2.3 3 1013 vg/ml) was pipetted into the
hole over a period of 1–2 min. After application, the hole in the cochlea was
sealed with a small amount of bone wax. After it dried, a small amount of sterile
tissue glue was applied to the bone wax and the bulla was sealed and the
wound was sutured as described above.
Auditory Testing
Acoustic Startle Response Testing
Acoustic startle responses of VGLUT3KO (n = 5),WT littermate (n = 5), rescued
VGLUT3 KO unilateral (n = 5), and rescued VGLUT3 KO bilateral (n = 5) mice
were measured as previously described (Seal et al., 2008). In brief, in darkened
startle chambers (SR-LAB hardware and software, San Diego Instruments),
piezoelectric sensors located under the chambers detect and measure the
peak startle response. Mice were acclimatized to the startle chambers by
presentation of a 70 dB white noise for 5min and then exposed to sound inten-
sities of 100 dB, 110 dB, and 120 dB (each with a 0ms rise time, 40ms plateau,
and 0ms fall time), presented in pseudorandom order with intersound intervals
of 10–50 s. Each run was repeated eight times. Average peak startle amplitude
at each sound level was calculated from eight runs. Final results were calcu-
lated as a percentage of WT mice at the 120 dB presentation level. Statistical
significance between measures was determined using a Student’s t test with
significance defined as p < 0.05.
Auditory Brainstem Response Recording
Sounds were presented and ABRs were tested in free-field conditions as
previously described in a sound-proofed chamber (Akil et al., 2006; Fremeau
et al., 2004). ABR thresholds were determined postoperatively at varying
time points, as early as 4 days after viral delivery for P10–P12 mice. The
mean value of thresholds checked by visual inspection and computer analysis
Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc. 291
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
was defined as ABR hearing threshold for click and 8, 16, and 32 kHz tone
stimuli.
Compound Action Potential Recording
For the CAP recording, a ventral surgical approach (Jero et al., 2001) was used
to expose the right cochlea 7–14 days after AAV1-VGLUT3 delivery to the inner
ear of the P10–P12mice, including KO (n = 5), rescued KO (n = 8), andWT litter-
mates (n = 5). A fine Teflon-coated silver wire recording electrode was placed
in the round window niche, and the ground electrode was placed in the soft
tissue of the neck. The sound stimulus was generated with Tucker-Davis
System II hardware and software (Tucker-Davis Technologies).
Immunofluorescence
Immunofluorescence studies were conducted similarly for whole-mount and
cochlear sections with the following differences.
Cochlear Whole Mount
Mice cochleaewere perfusedwith 4%PFA in 0.1MPBS (pH7.4) and incubated
in the fixative for 2 hr at 4�C. The cochleae were subsequently rinsed with PBS
three times for 10min and thendecalcifiedwith 5%EDTA in 0.1MPBS. Theotic
capsule, the lateral wall, tectorial membrane, and Reissner’s membrane were
removed in that order. The remaining organ of Corti was further dissected into
a surface preparation (microdissected into individual turns), then preincubated
for 1 hr in PBS containing 0.25% Triton X-100 and 5% normal goat serum
(blocking buffer). The whole mount was then incubated with rabbit anti-myosin
VIIa antibody (a hair cell-specificmarker) (Proteus Biosciences Cat 25-6790) at
a dilution of 1:50 in blocking buffer and guinea pig anti-VGLUT3 antibody (a gift
from Dr. Robert Edward, Department of Neurology, UCSF) at 1:5,000. After an
overnight incubation at 4�C, the cochlear whole mount was rinsed twice for
10 min with PBS and then incubated for 2 hr in goat anti-rabbit IgG conjugated
to Cy2 and goat anti-guinea pig IgG conjugated to Cy3 diluted to 1:4,000 in
PBS. Specimens were next rinsed in PBS twice for 10 min and mounted on
glass slides in a mounting solution containing DAPI (nucleus stain) and
observed under an Olympus microscope with confocal immunofluorescence.
For inner hair cell counts, the cochlear whole mounts were visualized under
a microscope equipped with epifluorescence, using a 403 objective. To quan-
tify the number of IHC transfected with AAV1-VGLUT3, we labeled specimens
with anti-VGLUT3 antibody, and IHCs were manually counted in the cochlear
whole mount and in the base, midturn, and apex. For GFP labeling, surface
preparation (cochlea whole mount) was incubated with a rabbit anti-GFP anti-
body (Invitrogen A11122) at 1:250. After an overnight incubation at 4�C, thecochlea sections were rinsed twice for 10 min with PBS and then incubated
for 2 hr in goat anti-rabbit IgG conjugated to Cy2 diluted to 1:4,000 in PBS.
They were then rinsed in PBS twice for 10 min and mounted on glass slides
in amounting solution containingDAPI and observed under anOlympusmicro-
scope with confocal immunofluorescence.
Cochlear Sections
Mice were anesthetized and their cochleae were isolated, dissected, perfused
through oval and round windows by 2% paraformaldehyde in 0.1 M PB at
pH 7.4, and incubated in the same fixative for 2 hr. After fixation, the cochleae
were rinsed with PBS and immersed in 5% EDTA in 0.1 M PB for decalcifica-
tion. When the cochleae were completely decalcified, they were incubated
overnight in 30% sucrose for cryoprotection. The cochleae then were
embedded in OCT Tissue Tek Compound (Miles Scientific). Tissues were
cryosectioned at 10–12 mm thickness, mounted on Superfrost microscope
slides (Erie Scientific), and stored at �20�C until use. Sections were then
double labeled as described above (see cochlear whole mount). Slides were
then mounted in a 1:1 mixture of PBS and glycerol before being coverslipped.
Slides treated with the same technique but without incubation with the primary
antibody used as a control.
Histology
Light Microscopy
Cochleae were isolated from deeply anesthetized WT, VGLUT3 KO, and
rescued KO mice, perfused through oval and round windows with 2.5% para-
formaldehyde and 1.5% glutaraldehyde in 0.1 M PB at pH 7.4, and incubated
overnight at 4�C with slow agitation in fixative. The cochleae were rinsed with
0.1 M PB and postfixed in 1% osmium tetroxide and 1.5% potassium ferricy-
anide (for improved contrast) for 2 hr. The cochleae subsequently were
292 Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc.
immersed in 5% EDTA (0.2 M). The decalcified cochlea were dehydrated in
ethanol and propylene oxide and embedded in Araldite 502 resin (Electron
Microscopy Sciences) and sectioned at 5 mm. After sections were stained
with toluidine blue, they were mounted in Permount (Fisher Scientific) on
microscope slides.
Electron Microscopy
Electron microscopy was performed as previously described (Akil et al., 2006)
on broken serial thin sections of the synaptic region of the IHCs, which were
cut in a horizontal plane parallel to the basilar membrane. In this study, the
cochleae were all handled and cut exactly the same and the same protocol
and orientation for the WT, KO, and rescued KOwere applied when examining
and visualizing the synaptic ribbons and vesicles. The morphological assess-
ment of ribbons and vesicles was performed as described by Roux et al. (2006)
using 50–61 IHCs and 17–20 different IHC ribbon synapses from three WT,
three KO, and three rescued KO mice. Sections were stained with uranyl
acetate and lead citrate and examined under 60kV in a JEOL-JEM 100S trans-
mission electron microscope. The number of vesicles tethered to the ribbon
included all the vesicles within 30 nm of the ribbon. All the vesicles clearly
located immediately below the ribbon were considered to be docked in our
two-dimensional (2D) estimation.
Spiral Ganglion Cell Counts
Spiral ganglion cell numerical density assessmentwas calculated as described
by Leake et al. (2011) to accurately estimate the number of nuclei in a given
volumeof tissue. For this analysis, threesets of three serial sections (5mmthick-
ness)were collected from the base,midturn, and the apex of fourWT, threeKO,
and four rescued KO cochlea. Adjacent serial sections were compared, and
new nuclei of spiral ganglion neurons that appear in the second section were
counted. Statistical differences were measured using a Student’s t test.
RT-PCR of the Cochlea
Cochlea from WT, VGLUT3 KO, and rescued KO were dissected. The total
RNA was extracted from the whole cochlea, organ of Coti + stria vascularis,
spiral ganglion, and vestibular epithelium (Trizol, Invitrogen) and reverse tran-
scribed with superscript II RNase H� (Invitrogen) for 50 min at 42�C, usingoligodT primers (Akil et al., 2006). Reactions without the reverse transcriptase
enzyme (�RT) were performed as control. Two microliters of RT reaction
product were used for subsequent polymerase chain reaction (PCR; Taq
DNA Polymerase, Invitrogen) of 35 cycles using the following parameters:
94�C for 30 s, 60�C for 45 s, 72�C for 1 min, followed by a final extension of
72�C for 10min and storage at 4�C. Primers were designed to amplify a unique
sequence of VGLUT3 isoform of 759 bp. The PCR primers that were used for
mouse include VGLUT3 (GenBank accession number AF510321.1: forward-
[gctggaccttctatttgctctta] and reverse- [tctggtaggataatggctcctc]). Analysis of
each PCR sample was then performed on 2% agarose gels containing
0.5 mg/ml ethidium bromide. Gels were visualized using a digital camera and
image processing system (Kodak). Candidate bands were cut out and the
DNA was extracted (Qiaquick gel extraction kit, QIAGEN) and sequenced
(Elim Biopharmaceuticals). The PCR product was then compared directly to
the full VGLUT3 sequence for identity.
ACKNOWLEDGMENTS
We thank Dr. Diana Bautista and Dr. Makoto Tsunozaki (UC Berkeley) for
critical advice and the use of their startle response chamber. The authorswould
like to acknowledge the financial support provided by Hearing Research.
Accepted: May 1, 2012
Published: July 25, 2012
REFERENCES
Akil, O., Chang, J., Hiel, H., Kong, J.H., Yi, E., Glowatzki, E., and Lustig, L.R.
(2006). Progressive deafness and altered cochlear innervation in knock-out
mice lacking prosaposin. J. Neurosci. 26, 13076–13088.
Baker, K., Brough, D.E., and Staecker, H. (2009). Repair of the vestibular
system via adenovector delivery of Atoh1: a potential treatment for balance
disorders. Adv. Otorhinolaryngol. 66, 52–63.
Neuron
Hearing Restoration in the VGLUT3 KO Mouse
Buran, B.N., Strenzke, N., Neef, A., Gundelfinger, E.D., Moser, T., and
Liberman, M.C. (2010). Onset coding is degraded in auditory nerve fibers
from mutant mice lacking synaptic ribbons. J. Neurosci. 30, 7587–7597.
Cao, L., Lin, E.J., Cahill, M.C., Wang, C., Liu, X., and During, M.J. (2009).
Molecular therapy of obesity and diabetes by a physiological autoregulatory
approach. Nat. Med. 15, 447–454.
Cao, L., Liu, X., Lin, E.J., Wang, C., Choi, E.Y., Riban, V., Lin, B., and During,
M.J. (2010). Environmental and genetic activation of a brain-adipocyte
BDNF/leptin axis causes cancer remission and inhibition. Cell 142, 52–64.
Di Domenico, M., Ricciardi, C., Martone, T., Mazzarella, N., Cassandro, C.,
Chiarella, G., D’Angelo, L., and Cassandro, E. (2011). Towards gene therapy
for deafness. J. Cell. Physiol. 226, 2494–2499.
Duan, M., Venail, F., Spencer, N., and Mezzina, M. (2004). Treatment of
peripheral sensorineural hearing loss: gene therapy. Gene Ther. 11 (Suppl
1 ), S51–S56.
Elkan-Miller, T., Ulitsky, I., Hertzano, R., Rudnicki, A., Dror, A.A., Lenz, D.R.,
Elkon, R., Irmler, M., Beckers, J., Shamir, R., and Avraham, K.B. (2011).
Integration of transcriptomics, proteomics, and microRNA analyses reveals
novel microRNA regulation of targets in the mammalian inner ear. PLoS ONE
6, e18195.
Fremeau, R.T., Jr., Kam, K., Qureshi, T., Johnson, J., Copenhagen, D.R.,
Storm-Mathisen, J., Chaudhry, F.A., Nicoll, R.A., and Edwards, R.H. (2004).
Vesicular glutamate transporters 1 and 2 target to functionally distinct synaptic
release sites. Science 304, 1815–1819.
Fyk-Kolodziej, B., Shimano, T., Gong, T.W., and Holt, A.G. (2011). Vesicular
glutamate transporters: spatio-temporal plasticity following hearing loss.
Neuroscience 178, 218–239.
Gassner, H.G., Shallop, J.K., and Driscoll, C.L. (2005). Long-term clinical
course and temporal bone histology after cochlear implantation. Cochlear
Implants Int. 6, 67–76.
Henckaerts, E., and Linden, R.M. (2010). Adeno-associated virus: a key to the
human genome? Future Virol 5, 555–574.
Husseman, J., and Raphael, Y. (2009). Gene therapy in the inner ear using
adenovirus vectors. Adv. Otorhinolaryngol. 66, 37–51.
Izumikawa, M., Batts, S.A., Miyazawa, T., Swiderski, D.L., and Raphael, Y.
(2008). Response of the flat cochlear epithelium to forced expression of
Atoh1. Hear. Res. 240, 52–56.
Jero, J., Mhatre, A.N., Tseng, C.J., Stern, R.E., Coling, D.E., Goldstein, J.A.,
Hong, K., Zheng, W.W., Hoque, A.T., and Lalwani, A.K. (2001). Cochlear
gene delivery through an intact round window membrane in mouse. Hum.
Gene Ther. 12, 539–548.
Kawamoto, K., Ishimoto, S., Minoda, R., Brough, D.E., and Raphael, Y. (2003).
Math1 gene transfer generates new cochlear hair cells in mature guinea pigs
in vivo. J. Neurosci. 23, 4395–4400.
Khan, A.M., Handzel, O., Burgess, B.J., Damian, D., Eddington, D.K., and
Nadol, J.B., Jr. (2005). Is word recognition correlated with the number of
surviving spiral ganglion cells and electrode insertion depth in human subjects
with cochlear implants? Laryngoscope 115, 672–677.
Konishi, M., Kawamoto, K., Izumikawa, M., Kuriyama, H., and Yamashita, T.
(2008). Gene transfer into guinea pig cochlea using adeno-associated virus
vectors. J. Gene Med. 10, 610–618.
Kral, A., and O’Donoghue, G.M. (2010). Profound deafness in childhood.
N. Engl. J. Med. 363, 1438–1450.
Kujawa, S.G., and Liberman, M.C. (2009). Adding insult to injury: cochlear
nerve degeneration after ‘‘temporary’’ noise-induced hearing loss.
J. Neurosci. 29, 14077–14085.
Lawlor, P.A., Bland, R.J., Mouravlev, A., Young, D., and During, M.J. (2009).
Efficient gene delivery and selective transduction of glial cells in the mamma-
lian brain by AAV serotypes isolated from nonhuman primates. Mol. Ther. 17,
1692–1702.
Leake, P.A., Hradek, G.T., Hetherington, A.M., and Stakhovskaya, O. (2011).
Brain-derived neurotrophic factor promotes cochlear spiral ganglion cell
survival and function in deafened, developing cats. J. Comp. Neurol. 519,
1526–1545.
Lee, J.H., Pradhan, J., Maskey, D., Park, K.S., Hong, S.H., Suh, M.W., Kim,
M.J., and Ahn, S.C. (2011). Glutamate co-transmission fromdevelopingmedial
nucleus of the trapezoid body—lateral superior olive synapses is cochlear
dependent in kanamycin-treated rats. Biochem. Biophys. Res. Commun.
405, 162–167.
Lin, H.W., Furman, A.C., Kujawa, S.G., and Liberman, M.C. (2011). Primary
neural degeneration in the guinea pig cochlea after reversible noise-induced
threshold shift. J. Assoc. Res. Otolaryngol. 12, 605–616.
Maeda, Y., Sheffield, A.M., and Smith, R.J. (2009). Therapeutic regulation of
gene expression in the inner ear using RNA interference. Adv.
Otorhinolaryngol. 66, 13–36.
Masuda, M., Dulon, D., Pak, K., Mullen, L.M., Li, Y., Erkman, L., and Ryan, A.F.
(2011). Regulation of POU4F3 gene expression in hair cells by 50 DNA in mice.
J. Neurosci. 197, 48–64.
Noble, W. (2010). Assessing binaural hearing: results using the speech, spatial
and qualities of hearing scale. J. Am. Acad. Audiol. 21, 568–574.
Obholzer, N., Wolfson, S., Trapani, J.G., Mo, W., Nechiporuk, A., Busch-
Nentwich, E., Seiler, C., Sidi, S., Sollner, C., Duncan, R.N., et al. (2008).
Vesicular glutamate transporter 3 is required for synaptic transmission in
zebrafish hair cells. J. Neurosci. 28, 2110–2118.
Petersen, M.B., and Willems, P.J. (2006). Non-syndromic, autosomal-
recessive deafness. Clin. Genet. 69, 371–392.
Praetorius, M., Hsu, C., Baker, K., Brough, D.E., Plinkert, P., and Staecker, H.
(2010). Adenovector-mediated hair cell regeneration is affected by promoter
type. Acta Otolaryngol. 130, 215–222.
Roux, I., Safieddine, S., Nouvian, R., Grati, M., Simmler, M.C., Bahloul, A.,
Perfettini, I., Le Gall, M., Rostaing, P., Hamard, G., et al. (2006). Otoferlin,
defective in a human deafness form, is essential for exocytosis at the auditory
ribbon synapse. Cell 127, 277–289.
Ruel, J., Emery, S., Nouvian, R., Bersot, T., Amilhon, B., Van Rybroek, J.M.,
Rebillard, G., Lenoir, M., Eybalin, M., Delprat, B., et al. (2008). Impairment of
SLC17A8 encoding vesicular glutamate transporter-3, VGLUT3, underlies
nonsyndromic deafness DFNA25 and inner hair cell dysfunction in null mice.
Am. J. Hum. Genet. 83, 278–292.
Seal, R.P., Akil, O., Yi, E., Weber, C.M., Grant, L., Yoo, J., Clause, A., Kandler,
K., Noebels, J.L., Glowatzki, E., et al. (2008). Sensorineural deafness and
seizures in mice lacking vesicular glutamate transporter 3. Neuron 57,
263–275.
Smith, R.J., Bale, J.F., Jr., andWhite, K.R. (2005). Sensorineural hearing loss in
children. Lancet 365, 879–890.
Staecker, H., Praetorius, M., Baker, K., and Brough, D.E. (2007). Vestibular hair
cell regeneration and restoration of balance function induced by math1 gene
transfer. Otol. Neurotol. 28, 223–231.
Steven Colburn, H., Shinn-Cunningham, B., Kidd, G., Jr., and Durlach, N.
(2006). The perceptual consequences of binaural hearing. Int. J. Audiol. 45
(Suppl 1 ), S34–S44.
Takamori, S., Malherbe, P., Broger, C., and Jahn, R. (2002). Molecular cloning
and functional characterization of human vesicular glutamate transporter 3.
EMBO Rep. 3, 798–803.
von Ilberg, C.A., Baumann, U., Kiefer, J., Tillein, J., and Adunka, O.F. (2011).
Electric-acoustic stimulation of the auditory system: a review of the first
decade. Audiol. Neurootol. 16 (Suppl 2 ), 1–30.
Neuron 75, 283–293, July 26, 2012 ª2012 Elsevier Inc. 293