|Home | About | Journals | Submit | Contact Us | Français|
Mice lacking the vesicular glutamate transporter-3 (VGLUT3) are congenitally deaf due to loss of glutamate release at the inner hair cell afferent synapse. Cochlear delivery of VGLUT3 using adeno-associated virus-1 (AAV1) leads to transgene expression in only inner hair cells (IHC), despite broader viral uptake. Within two weeks of AAV1-VGLUT3 delivery, acoustic brainstem response (ABR) thresholds normalize, along with partial rescue of the startle response. Lastly, we demonstrate partial reversal of the morphologic changes seen within the afferent IHC ribbon synapse. These findings represent the first successful restoration of hearing by gene replacement in mice, which is an important step towards gene therapy of human deafness.
Hearing loss is one of the most common human sensory deficits, with congenital hearing loss occurring in approximately 1.5 in 1000 children (Smith et al., 2005). Of these, about half are attributed to a genetic basis (Di Domenico et al., 2010). 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., 2010). To date, a majority of the research in this arena has focused on cochlear hair cell regeneration, 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., 2009; 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 successfully 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 KO mouse using virally-mediated gene delivery.
Our first goal was to determine the extent of transfection with the adeno-associated virus type 1 (AAV1) within the cochlea. Using 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 (Figure 1A), in a pattern similarly described by other investigators (Jero et al., 2001; Konishi et al., 2008). Subsequently, virus containing 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) (Figure 1B–E). Following delivery, RT-PCR of inner ear tissue (Figure 1C) demonstrated strong VGLUT3 mRNA expression in the rescued whole cochlea, organ of Corti, stria vascularis, vestibular neuroepithelium, and very weakly in the spiral ganglion. Non-injected cochleas of knock-outs do not demonstrate VGLUT3 expression as noted (Figure 1C, KO −/+RT). In contrast, under immunofluorescence, 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 µl or 1µl of AAV1-VGLUT3 (2.3×1013 virus genomes (vg)/ml) into the cochlea (Figure 1D–E). Microinjecting 1µl of virus resulted in 100% of IHCs labeled with anti-VGLUT3 antibody; in contrast, microinjecting 0.6µl resulted in only ~ 40% of IHCs labeled by the antibody.
We next sought to determine if earlier viral delivery would result in more robust VGLUT3 expression (Figures 1D–E, ,2).2). As shown, delivery of virus via the RWM at post-natal day 10 (P10) results in ~40% of the IHCs expressing VGLUT3 (Figures 1D–E, ,2),2), whereas similar doses (0.6 µl) of virus injected at the P1-3 results in 100% of IHC transfected in all animals (Figures 1D–E, ,22).
After verifying successful transgene expression within the IHC without significant organ of Corti injury, we next sought to determine whether the reintroduction of VGLUT3 would lead to measureable hearing recovery (Figure 3). In these studies only 0.6µl of AAV1-VGLUT3 was delivered at P10-12. Acoustic brainstem response (ABR) thresholds were first measurable within 7 days following viral delivery, with near normalization of thresholds to wild-type (WT) levels within 2 weeks (P24-26) (Figure 3A–C). Initially a cochleostomy (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 subsequently changed to a round window membrane (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 > 10db above WT levels, between the CO and RWM methods (Figure 3D). In both groups, all rescued KO mice maintained hearing for at least 7 weeks. At 28 weeks post-delivery 40% of the mice who achieved successful CO delivery still had hearing within 10db 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 10db of WT thresholds, is described in the legend of figure 3D.
We subsequently measured hearing recovery in mice injected via the RWM at P1-3 (Figure 3D). Due to the small size of the cochlea, only 0.6 µl 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.
For an additional assay of hearing recovery we studied the startle response at approximately 3 weeks following viral delivery (Figure 4). In these experiments, the AAV1-VGLUT3 delivery was done via the RWM at age P10-12. As expected, VGLUT3 KO mice 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 120dB, 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 differences 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.
As 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 P10-12) 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, counts were also taken at P5months, but again, no significant differences in SG counts or cell size were seen in the KO vs. rescued mice at this later time point (data not shown). Subsequently, spiral ganglion cell counts were also undertaken in mice that underwent virus delivery at P1-3. However, despite a robust IHC transfection and early hearing recovery (see Figures 1D–E, ,2),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 non-rescued mice) and the stria vascularis (data not shown).
As originally reported, VGLUT3 KO mice demonstrate abnormally thin, elongated ribbons in IHC synapses, though the number of synaptic vesicles tethered to ribbons or docked at the plasma membrane were normal (Seal et al., 2008). We thus sought to determine if these morphologic abnormalities could be reversed with hearing rescue. As shown (Figure 6, Table 1), in the rescued mice, ribbon synapses are normal in appearance, taking on a more rounded shape similar to the WT, while the non-rescued mice continue to demonstrate abnormally thin and elongated ribbons. The rescued mice also displayed a significantly larger number of synaptic vesicles associated with the ribbon (19 rescued vs. 14 WT, p=0.02) (Table 1). Interestingly, within individual hair cells, the synaptic vesicles themselves demonstrated a mixture of elongated and circular morphology, 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 either the WT or 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 the WT levels in the rescued KO mice.
These studies document the successful rescue of the deafness phenotype in a mouse model of inherited deafness. With viral delivery of VGLUT3 at P10-12 in the KO mouse, ABR thresholds normalize within 7–14 days, and remain in this range for at least 7 weeks, with 2 mice maintaining auditory thresholds for as long as a year and a half in this current study. Earlier delivery, at P1-3, results in an even more robust IHC transfection and long-lived hearing recovery in this mouse model.
One unexpected result from these investigations was the differential finding of more widespread expression of GFP protein following AAV1-GFP transfection as compared to isolated VGLUT3 expression restricted to the IHC following AAV1-VGLUT3 transfection (Figure 1A–C, ,2).2). RT-PCR results demonstrate that following AAV1-VGLUT3 delivery, there is also more widespread VGLUT3 mRNA transcription than in just IHCs (Figure 1C). These results suggest that there is a post-transcriptional regulatory mechanism acting on VGLUT3 mRNA which leads to selective expression of the protein only within IHCs. Several types of post-transcriptional regulation have been described within the cochlea, and whether this specific mechanism 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 appropriately 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 cochleostomy (CO) as compared to the round window membrane (RWM) delivery technique. As noted, we initially started with an apical CO delivery method, but abandoned it due to the low success rate of hearing restoration (17% of animals). Subsequently, we changed to a 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 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 separate hole into the scala through bone must be created which by its nature is traumatic, despite our best efforts to minimize trauma. In contrast, a RWM injection simply involves piercing the membrane and sealing it with fascia following viral delivery. However, histologically we were unable to see any obvious differences between the ears of animals with and without 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 P1-3, as opposed to P10-12, resulted in hearing recovery that was more consistently long-lived, with all mice 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 P10-12 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 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 (0.6µl), delivered at P10-12, caused VGLUT3 expression in only ~40% of IHCs (Figure 1D–E), and yet this was enough to restore ABR thresholds to WT levels for click responses, and near normal for pure tone thresholds (Figure 3A–C). Similar results have been documented in other models of hearing recovery following noise-exposure (Kujawa and Liberman, 2009; Lin et al., 2011), where 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 regard 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 transmission (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 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 the WT and all elongated in the KO mice, implying that there may still be differences in transmitter release in the rescued vs. WT mice.
Another interesting finding with regard to the ribbon synapse was the larger numbers of synaptic vesicles that were associated 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 the WT mice (Figure 1C RT-PCR data), and suggests, though does not prove, an association between increased mRNA levels and vesicle number. We believe that circular vesicles represent properly packaged vesicles, while the elongated vesicles are improperly packaged vesicles. Perhaps continuous production of VGLUT3 by the constitutive CBA promoter driving transfected VGLUT3 production 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 P10-12 (Figure 4) was an unexpected finding. While none of the animals had complete normalization of ABR amplitude and startle-response levels, the amplitude growth does imply that bilateral input increases the auditory response centrally. An analogous phenomenon is seen with ‘binaural summation’, 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 P10-12 is not surprising (Figure 5), given that hair cell activity and afferent stimulation can provide a trophic effect on SG survival. This is likely at least partly due to the fact that virus was delivered at ~P10 with subsequent ABR thresholds recovery at ~P17-24, 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 likely have to occur earlier. Further, with only ~40% of IHCs expressing VGLUT3 (using the lower concentration of virus, delivered at P10-12), there are still many spiral ganglion neurons not receiving afferent input, which also likely impacts this result as well. We were thus surprised that even earlier delivery of virus, at P1-3, 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 ultimately be beneficial to patients with DFNA25 remains unclear. Despite these differences, as the first study to document 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.
VGLUT3 null mutant mice were generated as described in a C57 (Seal et al., 2008) strain then backcrossed with FVB mice (> 7 generations) to obtain a homogeneous genetic background. P1 – 12 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 (2mg/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 24hrs postoperatively animals were given subcutaneous Carprofen analgesia (2mg/kg) to manage inflammation and pain. Animals were closely monitored for signs of distress and abnormal weight loss postoperatively.
Mouse VGLUT3 cDNA was sub-cloned into the multiple cloning site of vector AM/CBA-WPRE-BGH (kindly provided by R. Palmiter). Human embryonic kidney 293 cells were co-transfected with three plasmids—AAV-mVGLUT3 plasmid, appropriate helper plasmid encoding rep and AAV1 cap genes, and adenoviral helper pF Δ6—using standard CaPO4 transfection. Cells were harvested 60 hours following transfection, cell pellets lysed with sodium deoxycholate and AAV vectors purified from the cell lysate by ultracentrifugation through an iodixanol density gradient, then concentrated and dialyzed against phosphate-buffered saline (PBS), as previously described (Cao et al., 2009; Cao et al., 2010; Lawlor et al., 2009). Vectors were titered using real-time PCR (ABI Prism 7700; Applied Biosystems, Foster City, CA), and purity of vector stocks was confirmed by running a 10 µl sample on sodium dodecyl sulfate polyacrylamide gel electrophoresis and staining with Coomassie blue.
Animals were anesthetized, the left ear was approached via a dorsal incision as described by Duan et al (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 µl or 1.0 µl of a 2.3 1013 virus genomes/ml) previously drawn into the fine pipette was gently injected through RWM over 1–2 minutes. After pulling out the pipette, the RW niche was quickly sealed with fascia and adipose tissue. The bulla was sealed with dental cement (Dentemp, Majestic Drug Company, South Fallsburg, NY) and the wound was sutured in layers with a 5-0 absorbable chromic suture (Ethicon, New Jersey).
The right ear was approached via ventral, paramedian incision in the neck as described by Jero et al(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. Using a 0.5 mm drill pit to thin gently the bone of the otic capsule where the stria vascularis could be slightly visualized as a brownish stripe. Once enough bone was shaved a slight fluid interface became visible, 0.6 µl VGLUT3-AAV1 (2.3 1013 virus genomes/ml) was pipetted into the hole over a period of 1–2 minutes. Following application the hole in the cochlea was sealed with a small amount of bone wax. After dried a small amount of sterile tissue glue is applied to the bone wax and the bulla was sealed and the wound was sutured as described above.
Acoustic startle responses of VGLUT3 KO (n = 5), WT littermate (n = 5), rescued VGLUT3 KO one ear (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 5 min and then exposed to sound intensities of 100 dB, 110 dB and 120 dB (each with a 0 ms rise time, 40 ms plateau, 0 ms 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 calculated 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 defines as p<0.05.
Sound were presented and ABRs were tested in a free field conditions as previously described in a sound-proofed chamber (Akil et al., 2006; Fremeau et al., 2004). ABR thresholds were determined postoperative at varying time-points, as early as 4 days following viral delivery for P10-12 mice. The mean value of thresholds checked by visual inspection and computer analysis was defined as ABR hearing threshold for click and 8, 16 and 32 KHz tones stimulus.
For the CAP recording, a ventral surgical approach (Jero et al., 2001) was used to expose the right cochlea 7–14 days following AAV1-VGLUT3 delivery to the inner ear of the P10-12 mice, including KO (n=5), rescued KO (n=8) and WT littermates (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, Alachua, FL, USA).
Immunofluorescence studies were conducted similarly for whole-mount and cochlear sections with the following differences:
Mice cochleae were perfused with 4% PFA in 0.1 M PBS (phosphate buffered saline), pH 7.4, and incubated in the fixative for 2 hours at 4°C. The cochleae were subsequently rinsed with PBS three times for 10 min, and then decalcified with 5% EDTA in 0.1 M PBS. The otic 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 pre-incubated 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 specific marker) (Proteus Biosciences, INC Cat # 25-6790) at a dilution of 1:50 in blocking buffer and Guinea pig anti-VGLUT3 antibody (a gift from Dr.Robert Edwards, Dept. of Neurology, UCSF) at 1:5000. After an overnight incubation at 4°C, the cochlear whole mount was rinsed twice for 10 min with PBS and then incubated for 2 h in goat anti-rabbit IgG conjugated to Cy2 and goat anti guinea pig IgG conjugated to Cy3 diluted to 1:4000 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 cells counts the cochlear whole mounts were visualized under a microscope equipped with epifluorescence, using a 40X objective. To quantify the number of IHC transfected with AAV1-VGLUT3 specimens were labeled with anti-VGLUT3 antibody, and IHCs were manually counted in the cochlear whole mount and in the base, mid-turn and apex. . For GFP labeling, surface preparation (cochlea whole mount) were incubated with a rabbit anti GFP antibody (Invitrogen # A11122) at 1:250. After an overnight incubation at 4°C, the cochlea sections were rinsed twice for 10 min with PBS and then incubated for 2 h in goat anti-rabbit IgG conjugated to Cy2 diluted to 1:4000 in PBS. Then rinsed in PBS twice for 10 min, and mounted on glass slides in a mounting solution containing DAPI and observed under an Olympus microscope with confocal immunofluorescence.
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 h. After fixation, the cochleae were rinsed with PBS and immersed in 5% EDTA in 0.1M PB for decalcification. 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 µm 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 were used as controls
Cochleae were isolated from deeply anesthetized WT, VGLUT3 KO and rescued KO mice, perfused through oval and round windows with 2.5% paraformaldehyde 1.5% glutaraldehyde in 0.1M 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 post fixed in 1% osmium tetroxide and 1.5% potassium ferricyanide (for improved contrast) for 2 h. The cochleae subsequently were 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, Fort Washington, PA) and sectioned at 5µm. After sections were stained with Toluidine Blue, they were mounted in Permount (Fisher Scientific, Houston, TX) on microscope slides.
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 KO were applied when examining and visualizing the synaptic ribbons and vesicles. The morphological assessment of ribbons and vesicles was performed as described by Roux et al (Roux et al., 2006) using 50–61 IHCs and 17–20 different IHCs ribbons synapse from three WT and three KO and three rescued KO mice. Sections were stained with uranyl acetate and lead citrate and examined under a 60 kV in a JEOL-JEM 100S transmission electron microscope. The number of vesicles tethered to the ribbon included all the vesicles within 30nm of the ribbon. All the vesicles clearly located immediately below the ribbon were considered to be docked in our 2D estimation.
Spiral ganglion cell numerical density assessment was calculated as described by Leake et al (Leake et al., 2011) to accurately estimate the number of nuclei in a given volume of tissue. For this analysis 3 sets of 3 serial sections (5 µm thickness, were collected from the base, mid-turn and the apex of 4 WT, 3 KO and 4 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.
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 Corp) and reverse transcribed with superscript II RNase H− (Invitrogen) for 50 min at 42°C, using oligodT primers (Akil et al., 2006). Reactions without the reverse transcriptase enzyme (−RT) were performed as control. Two microliters of RT reaction product was 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 10 min 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 µg/ml ethidium bromide. Gels were visualized using a digital Camera and image processing system (Kodak, Rochester NY, USA). Candidate bands were cut out and the DNA was extracted (Qiaquick gel extraction kit, Qiagen) and sequenced (Elim Biopharmaceuticals, Inc. Hayward, CA, USA). The PCR product was then compared directly to the full VGLUT3 sequence for identity.
We thank Dr. Diana Bautista and Dr. Makoto Tsunozaki (UC Berkeley) for critical advice and the use of their startle response chamber.
The authors would like to acknowledge the financial support provided by Hearing Research Inc.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Omar Akil, Department of Otolaryngology- Head & Neck Surgery, University of California San Francisco, San Francisco, CA, 94143-0449. Phone: 415-476-0728. Email: oakil/at/ohns.ucsf.edu.
Rebecca P. Seal, Department of Neurology- University of Pittsburgh, Pittsburgh, PA 15213-3301. Phone: 412-624-5183. Email: rpseal/at/pitt.edu.
Kevin Burke, Department of Otolaryngology- Head & Neck Surgery, University of California San Francisco, San Francisco, CA, 94143-0449. Phone: 415-476-0728. Email: kburke/at/ohns.ucsf.edu.
Chuansong Wang, Comprehensive Cancer Center, The Ohio State University, Columbus, Ohio. Phone: 614-247-4351.
Aurash Alemi, Department of Otolaryngology- Head & Neck Surgery, University of California San Francisco, San Francisco, CA, 94143-0449. Phone: 415-476-0728. Email: AAlemi/at/ohns.ucsf.edu.
Matthew During, Comprehensive Cancer Center, The Ohio State University, Columbus, Ohio. Phone: 614-247-4351. Email: Matthew.During/at/osumc.edu.
Robert H. Edwards, Department of Neurology, University of California San Francisco, San Francisco, CA, 94143-2140. Phone: 415-502-5687. Email: robert.edwards/at/ucsf.edu.