We previously demonstrated that an artificial protein, TAT-FNK, has antiapoptotic effects against cochlear hair cell (HC) damage caused by ototoxic agents when applied systemically. To examine the feasibility of topical protein therapy for inner ear disorders, we investigated whether gelatin sponge soaked with TAT-FNK and placed on the guinea pig round window membrane (RWM) could deliver the protein to the cochlea and attenuate aminoglycoside (AG)-induced cochlear damage in vivo. First, we found that the immunoreactivity of TAT-myc-FNK was distributed throughout the cochlea. The immunoreactivity was observed from 1–24 h after application. When Tat-FNK was applied 1 h before ototoxic insult (a combination of kanamycin sulfate and ethacrynic acid), auditory brainstem response threshold shifts and the extent of HC death were significantly attenuated. When cochlear organotypic cultures prepared from P5 rats were treated with kanamycin, TAT-FNK significantly reduced the extent of caspase-9 activation and HC death. These findings indicate that TAT-FNK topically applied on the RWM can enter the cochlea by diffusion and effectively prevent AG-induced apoptosis of cochlear HCs by suppressing the mitochondrial caspase-9 pathway.
Apoptosis is involved in cochlear sensory hair cell (HC) death caused by a variety of insults, which include acoustic trauma, loss of trophic factor support, ischemia–reperfusion, and exposure to ototoxic agents such as aminoglycoside (AG) antibiotics and the anti-neoplastic agent cisplatin.1, 2, 3 Protecting cells from apoptosis by controlling the balance of pro- and antiapoptotic proteins by techniques such as gene therapy is considered a good strategy for protection of HCs from ototoxic insults. Overexpression of Bcl-2 proteins by delivery of the Bcl-2 gene into HCs has been reported to prevent the degeneration of HCs exposed to AG or cisplatin.4 Injection of the Bcl-xL gene into mice cochlea also prevents HC degeneration induced by kanamycin.5 However, such gene transfer application cannot control the amount or exposure time of the target protein to achieve optimal prevention of cell death. In addition, gene transfer technology cannot avoid the possibility of detrimental insertion of transgenes. Therefore, injection of the target protein could be an alternative method. For example, several proteins such as granulocyte-colony stimulating factor6 have already been used in clinics. Such protein therapy, however, is not always applicable for treatment of inner ear disorders because the blood–labyrinth barrier may inhibit the delivery of high-molecular-weight proteins into the cochlea. This problem may be solved by using the protein transduction domain technology. When fused with a protein transduction domain such as the TAT domain of the HIV/Tat (transcription-transactivating) protein, a variety of high-molecular-weight proteins have been successfully introduced into cells both in vitro and in vivo.7, 8
We first constructed a powerful artificial antiapoptotic protein, FNK (originally designated Bcl-xFNK by Asoh et al.9), which has three amino-acid substitutions, Tyr-22 to Phe(F), Gln-26 to Asn(N) and Arg-165 to Lys(K), to strengthen the cytoprotective activity of Bcl-xL. We then demonstrated that fusion of FNK with TAT enabled FNK to penetrate highly negatively charged chondrocytes9, 10 and the blood–brain barrier,11 and that TAT-FNK showed an antiapoptotic effect in a model of brain and hepatic ischemia.11, 12 When injected intraperitoneally into guinea pigs in vivo, we observed that TAT-FNK was distributed widely in the cochlea and that it reduced the expression of cleaved poly-(ADP-ribose)-polymerase (PARP), auditory brainstem response (ABR) threshold shifts, and HC loss induced by a combination of ethacrynic acid (EA) and kanamycin sulfate (KM), in vivo.13
Another potential drug delivery system for treatment of cochlear disorders is topical drug application into the middle ear space. Compared with systemic injections, such local delivery is beneficial because it requires significantly lower amounts of drug and reduces systemic side effects. A major side effect after long-term administration of an antiapoptotic drug is a possibility of carcinogenesis. Overexpression of Bcl-xL, the original protein of FNK, is reported to have the potential to cause tetraploidization, which would result in neoplasia.14, 15 Schuknecht16 has developed a topical drug application technique for inner ear disorders: injection of streptomycin into the middle ear space of patients with Ménière's disease. Intra-tympanic dexamethasone injections have also been performed as primary treatment for sudden sensorineural hearing loss.17 Intra-tympanic drug application has also been used in animal studies to examine the effects on inner ear function or disorders. It is quite difficult, however, to achieve the delivery of high-molecular-weight proteins into the inner ear because these proteins cannot pass through the round window membrane (RWM), which is the main route into the inner ear.
In the current study, we examined whether the TAT fusion technique could make transtympanic protein therapy applicable for inner ear disorders. We investigated whether TAT-FNK applied topically on the RWM could be successfully delivered into the cochlea, and protect cochlear HCs from an ototoxic combination of KM and EA. We also investigated whether TAT-FNK could prevent HC death caused by KM by suppression of the mitochondrial caspase-9 pathway.
Transduction of TAT-myc-FNK into cochlear tissue
Immunohistochemical staining using an anti-myc-tag antibody revealed that TAT-myc-FNK was detectable in the cochlea from 1 to 24 h after the application onto the RWM. There was a statistically significant difference between the groups as determined by one-way analysis of variance (ANOVA) (F6,203=41.239, P<0.01). Scheffe's post hoc test revealed that the labeling indices (LIs) at 1, 3, 6, 12 and 24 h were significantly higher than that of the control (P<0.01). The LIs gradually increased from 1 to 6 h, but no differences were observed among the LIs at 1, 3 and 6 h. Beginning 12 h after the application, the LIs gradually decreased. The LIs at 6 and 12 h, 6 and 24 h, and 3 and 24 h were significantly different (P<0.01). No significant difference was observed between the LI of the control and that at 48 h (2a). High-power views of the organ of Corti (OC) and the spiral ganglion revealed that many spots consisting of TAT-myc-FNK were localized within the cells outside their nuclei (Figures 1b and c). The basal turn tended to show higher immunoreactivity than the upper turns, but there was no statistically significant difference between the cochlear turns at 1 and 6 h. Two-way ANOVA was conducted to examine the cochlear turns and the time course. There were no interactions between the two factors (F2,174=0.051, P=0.950). There was also no statistical difference in the main effect of the cochlear turns (F2,174=1.033, P=0.358; Figure 2b). In addition to the OC, the spiral ganglion cells (SGCs), the stria vascularis (SV) and spiral ligament (SL) also appeared to show greater immunoreactivity than the control 6 h after the application of TAT-myc-FNK onto the RWM. Because the background immunoreactivity in the control sections varied between the organs, the normalized LIs, that is, the ratio of the LIs in each organ to those of the control, of these organs were compared by one-way ANOVA. There was a statistically significant difference between the groups (F3,116=39.257, P<0.01). Scheffe's post hoc test revealed that immunoreactivity was strongest in the cells in the OC, followed by that in the SGCs. The normalized LI of the OC was significantly greater than that of the SGCs, the SV and the SL (P<0.01). The normalized LI of the SGCs was also significantly higher than that of the SV and the SL (P<0.01; Figure 2c). Specific immunoreactivity to myc was not observed in any control ears that were administered only myc-FNK (that is, without TAT) or in the ears of animals administered TAT-myc-FNK in the contralateral ear (Figure 1d).
Protective effects of TAT-FNK against ABR threshold shifts induced by ototoxic insults
The baseline ABR thresholds measured before ototoxic insult were statistically not different at all tested frequencies among animals (data not shown). Neither experimental nor drug control animals showed any signs of systemic illness, such as diarrhea or hair loss, until euthanasia. For drug control animals, a gelatin sponge soaked with TAT-FNK was placed on the RWM, but a combination of KM and EA was not given. These animals showed no ABR threshold shifts at any tested frequency (data not shown), indicating that topical application of TAT-FNK on the RWM is not harmful to cochlear function.
ABR threshold shifts 14 days after the ototoxic insult in the experimental animals are shown in Figure 3 (n=8 each). Two-way ANOVA was conducted to examine the effect of the TAT-FNK administration and the frequency on the ABR threshold shifts. There was no significant interaction between TAT-FNK administration and hearing frequency (F2,42=0.042, P=0.959). There was a main effect of TAT-FNK administration (F1,42=27.355, P<0.01) but no significant difference in frequency (F2,42=0.833, P=442), indicating that the ABR threshold shifts were significantly smaller at all the tested frequencies in the TAT-FNK-treated ears than in the untreated contralateral ears. This result suggests that the TAT-FNK treatment significantly attenuated the ABR threshold shifts induced by the ototoxic agents.
HC protective effects of TAT-FNK in vivo
Figures 4a and b show the average cytocochleograms in the TAT-FNK-treated and contralateral untreated ears, respectively, which were produced by plotting the average percentage of HC loss in every segment between 5 and 16 mm from the apex that was averaged across all subjects (n=6 each). Segments measuring under 5 mm and over 16 mm were excluded because the extent of HC damage could not be quantified owing to damage in some samples during surface preparation. The frequency map was added in the x-axis according to the data of Tsuji and Liberman.18
The ototoxic agents induced losses of 91.7±7.0% of the outer HCs (OHCs) and 13.8±5.9% of the inner HCs (IHCs) in the TAT-FNK-untreated ears, whereas the losses of the OHCs and the IHCs in the treated ears were reduced to 64.0±29.6% and 8.3±3.5%, respectively (Figure 4c). Two-way ANOVA was conducted to examine the effect of the TAT-FNK administration and the type of HC on HC loss. There was no significant interaction between TAT-FNK administration and type of HC (F1,20=3.055, P=0.096). There were main effects of TAT-FNK administration (F1,20=6.869, P=0.016) and type of HCs (F1,20=110.657, P<0.01), indicating that the TAT-FNK treatment significantly attenuated the HC damage induced by KM and EA. Drug control animals administered only TAT-FNK showed minimal HC loss throughout the cochlea.
In vitro effect of TAT-FNK on protection of HCs and caspase-9 activation
Figure 5c shows an intact, untreated cochlear explant that was double-labeled with rhodamine-conjugated phalloidin (red) and activated caspase-9 (green). The stereocilia bundles on the three rows of OHCs and one row of IHCs have normal morphology and negligible green staining. Figure 5a shows a cochlear explant treated with KM for 10 h. HCs are missing and caspase-9 labeling is present in the HC regions. These results indicate that KM treatment caused an increase in caspase-9 activation, leading to apoptosis of the HCs by a mitochondria-mediated pathway. Addition of TAT-FNK to the explants greatly suppressed caspase-9 activation (Figure 5b). The number of HCs with activated caspase-9 in the explants treated only with KM was 20.6±5.2 per 0.2-mm length, whereas the number was reduced to 7.6±3.2 per 0.2-mm length in the explants treated with KM and TAT-FNK (Figure 5d, n=4 each). The number of HCs with activated caspase-9 in explants treated with KM and FNK (without TAT) was 16.7±3.2. There was a statistically significant difference between the groups as determined by one-way ANOVA (F3,12=31.337, P<0.01). Scheffe's post hoc test revealed that there were significant differences between the KM with TAT-FNK-treated explants, and the KM-treated explants or the KM with FNK-treated explants (P<0.01). There was no statistically significant difference between the KM-treated explants and the KM with FNK-treated explants (P=0.429). Therefore, the TAT-FNK treatment significantly reduced the number of HCs entering the caspase-9-dependent apoptotic pathway after KM application.
We counted the number of viable HCs (n=6 each) after 12 h of culture. In the control that was not administered any additional agent such as KM, FNK or TAT-FNK, no or only few HCs were lost. When the number of viable HCs (n=6 each) was counted after 12 h of culture with KM (that is, in the absence of TAT-FNK), massive losses of the OHCs and the IHCs were induced, as only 25.2±9.1% and 28.5±11.9% of the cells survived, respectively. The TAT-FNK treatment attenuated OHC and IHC damages, as 80.1±14.9% and 74.1±20.6%, respectively, of the cells remained. In the explants treated with KM with FNK, the extent of survival was 27.6±5.9% for the OHCs and 38.3±15.9% for the IHCs. Two-way ANOVA was conducted to examine the effect of the drug administration and the type of HC on HC loss. There was a main effect of drug administration (F3,40=99.432, P<0.01). There was also a main effect of type of HC (F1,40=419.899, P<0.01). Finally, there was interaction between drug administration and type of HC (F3,40=47.846, P<0.01). The simple effects analysis revealed significant differences between the KM with TAT-FNK-treated explants, and the KM-treated explants or the KM with FNK-treated explants (P<0.01), in the OHCs. There was no significant difference between the KM-treated explants and the KM with FNK-treated explants. However, in the IHCs, there were no significant differences between the KM with TAT-FNK-treated explants, and the KM-treated explants or the KM with FNK-treated explants. These results indicate that the TAT-FNK treatment significantly attenuated damage to the OHCs; however, FNK alone did not protect the HCs against KM (Figures 6c and d). Drug control animals administered only TAT-FNK showed minor HC loss.
In the present study, we demonstrated that the TAT-fusion technique enabled the macromolecule FNK protein, which was infiltrated into a gelatin sponge and placed on the RWM, to successfully enter the cochlea because it allowed the protein to penetrate through the RWM. TAT-myc-FNK was distributed throughout all turns of the cochlea, but immunoreactivity was not observed in the contralateral ears, suggesting that, when topically applied, the distribution of TAT-FNK may be confined to the applied cochlea.
The LIs of TAT-myc-FNK gradually increased until 6 h, but there were no significant differences in the LIs at 1, 3 and 6 h. Beginning at 12 h, the LIs gradually decreased. At 48 h, the immunoreactivity disappeared. This suggests that TAT-myc-FNK was immediately distributed into the cochlea 1 h after administration onto the RWM and remained in high concentration until 6 h. It gradually decreased beginning at 12 h and disappeared by 48 h. When examining the entire cochlea at 6 h after administration of TAT-myc-FNK, the strongest immunoreactivity was present in the cytoplasm of the supporting cells and the HCs in the OC, followed by the SGCs. Immunoreactivity could also be observed in the SV and SL. This suggests that xTAT-myc-FNK was distributed most prominently in the OC followed by SGCs, and that it also reached the SV and SL at 6 h after administration on the RWM. The duration of FNK expression was much longer compared with when it was administered systemically.13 A single topical administration of TAT-FNK on the RWM effectively protected cochlear HCs from the combination of KM and EA in vivo. These findings imply that, when fused with TAT and soaked in a gelatin sponge macromolecular proteins can be applied on the RWM as an effective and selective therapeutic agent to function in the cochlea. Considering the adverse effects introduced by systemic injection, this technology is feasible as a novel treatment for inner ear disorders. TAT-FNK attenuated KM-induced HC death by suppressing the activation of pro-caspase-9 in vitro, suggesting that the antiapoptotic protein FNK has the potential to regulate the mitochondria-related apoptotic pathway in the inner ear.
The RWM is a main gate and barrier for various kinds of substances to enter from the middle ear into the inner ear.19 The membrane consists of three layers: an outer epithelium facing the middle ear, a core of connective tissue and an inner epithelium facing the inner ear.20, 21, 22, 16, 17, 18 The structure of the outer epithelium is such that substances can pass from the middle to the inner ear by selective absorption and secretion.23 The factors influencing permeability through the RWM include the molecular weight and configuration of the protein, its contact time and the concentration of the substances in the middle ear.23, 24, 25 Among these, molecular weight is the most important in determining permeability. Generally, low-molecular-weight compounds, such as antibiotics, corticosteroids and labeled ions, can easily pass through the RWM to enter the inner ear,26, 27 whereas penetration of high-molecular-weight substances, such as proteins and lipids, is limited.19, 27, 28, 29, 30 In the current study, myc-FNK, whose molecular weight is about 30 kDa, did not pass through the RWM, suggesting that it is too large to pass through. Many proteins related to apoptosis, such as p53, AKT and super oxide dismutase, have a molecular weight of 30–60 kDa. Thus, when considering the applicability of protein therapy from the middle ear space, new technology is required to overcome the difficulty of delivering large molecules to the inner ear. The Tat protein of HIV-1 is a protein of 101 residues. The Tat protein has the characteristic that it can cross the plasma membrane of neighboring cells.31 TAT comprises the short stretches of the Tat protein domain that are primarily responsible for their translocation ability, also referred to as protein transduction domains.32 Although the exact mechanism has not been elucidated, two models have been proposed: energy-dependent macropinocytosis, and direct uptake by electrostatic interactions and hydrogen bonding.33 By fusing FNK to the TAT domain, FNK was effectively absorbed into the outer epithelium cells and then secreted to the inner ear space, although further studies are needed to confirm this. When fused with TAT, various proteins were reported to be transported across cell membranes.7, 8 Even β-galactosidase, whose molecular weight is 120 kDa, has been reported to enter cells.33 Moreover, oligonucleotide, nucleic acids and liposome can also be conjugated with TAT to improve its penetration.34 Thus, protein transduction technology allows the use of macromolecules for therapeutic application in a variety of inner ear disorders.
In our previous study, we successfully delivered TAT-FNK protein to the inner ear by intraperitoneal injection.13 However, systemic administration may not be suitable for treatment of inner ear disorders because only relatively small amounts of drug can enter the inner ear, which therefore requires the application of high doses of drug to maintain a therapeutic concentration in the inner ear at an optimal time. Moreover, a single administration has a short half-life, as shown in our previous study,13 and thus repeated administration is required. High doses of drug may easily induce systemic toxicities and acute allergic side effects. In particular, high doses of this antiapoptotic protein may promote tumor, although this has not been reported. Therefore, injection of drugs topically into the middle ear space may be more appropriate to better control local drug delivery. Injection of drugs into the middle ear space, however, may result in a large portion of the drug being absorbed by the middle ear mucosa or drained into the epipharynx by the Eustachian tube.35
To deliver a drug more effectively into the inner ear, using an absorbable material as a drug carrier may be promising. These materials could secure the stability of drugs on the RWM, and thereby lengthen the contact time with the RWM, and reduce diffusion into the mucosa and drainage from the middle ear space.36 In the current study, we used a gelatin sponge as a carrier of TAT-FNK, as Husmann et al.37 used a gelatin sponge on the RWM to topically apply gentamicin, which then induced severe damage to the cochlea compared with a single application. Similarly, Okamoto et al.38 used a gelatin sponge containing bone morphogenetic protein-2, and demonstrated that bone morphogenetic protein-2 was slowly released and induced successful regeneration of cartilage in a canine tracheomalacia model. Compared with our previous systemic study,13 we could observe immunoreactivity of TAT-myc-FNK in the cochlea using approximately 1/600th the amount of TAT-myc-FNK. Immunoreactivity of TAT-myc-FNK could still be observed in the cochlea after 24 h, which is significantly longer than the expression periods observed in previous systemic injection studies targeting organs, including the cochlea.11, 13 Further, the amount of TAT-FNK we topically applied to protect HCs from ototoxicity was approximately 1/15th the dose we used systemically to protect from the same ototoxicity in our previous study.13 Thus, a gelatin sponge is considered to be an effective drug carrier for inner ear disorders.
When drugs are topically applied onto the RWM, they diffuse from the basal end of the cochlea and thus initially show a concentration gradient, decreasing toward the apex. The concentration gradient in the perilymph was investigated, and the greater concentration was demonstrated at the basal turn.39, 40 However, when the patterns of distribution of the drugs applied on the RWM were investigated by immunocytochemistry, the drugs appeared to be widely and rapidly distributed into the various organs of the inner ear. Imamura and Adams40 examined the distribution of gentamicin in the inner ear of guinea pig using a monoclonal antibody. When gentamicin was placed on the RWM, the entire cochlear cell was diffusely stained until 6 h after administration. Beginning at 6 h after application, staining was found to be localized mainly in the basal turn. Greater staining in the basal turn was also found when gentamicin was administered systemically. These results suggest that gentamicin can be diffused rapidly into the entire cochlea, and that the localization of staining in the basal turn is due to the nature of the cells in the basal turn to accumulate the drug, and not because of the predominant distribution of gentamicin at the basal turn. Zou et al.42 examined the distribution of lipid nanocapsules in cochlear cells after application on the RWM by using fluorescein isothiocyanate and rhodamine-B labeling. The lipid nanocapsules were present in the SGCs, OC and SV 30 min after application. Moreover, the nanocapsules were more strongly distributed in the SV in the second turn than in the basal turn. They assumed that after penetrating the RWM, the nanocapsules are rapidly diffused through the porous modiolar wall of the scala tympani, after which they enter the SGCs, and then are widely diffused through its nerve fibers. In the present study, although there was a trend of higher intensity of immunoreactivity of TAT-myc-FNK at the basal turn, we did not observe statistically significant differences among the turns. We assume that this rapid and relatively even distribution of TAT-FNK throughout the cochlea was achieved by this radial diffusion through the modiolus, and not by longitudinal perilymph diffusion. Pathways to uptake lipid nanocapsules and TAT-mediated particles into the tissue might be similar, as their high permeability is considered to accelerate their rapid diffusion.41, 42 High immunoreactivity in the SGCs compared with those in the SV and SL at 6 h can support this argument, although the cause of the higher immunoreactivity in the OC compared with that in the SGCs needs to be clarified in future research.
It is known that high-frequency hearing loss occurs initially after AG ototoxicity. However, when the damage by AG is severe, apical cells will be affected and hearing loss expands to lower frequencies.43 The doses of the KM and EA combination we chose were assumed to be sufficient to cause threshold shifts even at a low frequency. We observed some tendency that the OHCs in the basal parts are more susceptible to the combination of EA and KM than those the upper parts (Figures 4a and b). This finding is consistent with that in other studies.44 When we compared the extent of missing HCs in the region corresponding to the frequency at which ABR was measured, the percentages of missing OHCs was approximately 67% in the region corresponding approximately to 4 kHz and 80% in that corresponding to 20 kHz. In untreated ears, the percentages of missing OHCs in the regions comparable to 4 and 20 kHz were 95% and 100%, respectively. The differences in the extent of OHC loss between the 4- and 20-kHz regions were small, supporting the ABR findings that there was no significant difference in the threshold shifts among the tested frequencies, although the ABR threshold shifts were slightly greater at higher frequencies than at lower frequencies.
In the current study, we showed that caspase-9 was activated by KM in vitro. This finding suggests that KM-induced cochlear HC death is caspase-9-dependent, which is consistent with other studies.1, 45 We demonstrated that TAT-FNK suppressed the activation of caspase-9 in this study and reduced the extent of cleaved PARP in OHCs in our previous study,13 which suggests that TAT-FNK prevents the intrinsic apoptotic pathway, as does the parent protein Bcl-xL. It has also been shown that TAT-FNK affects the cytosolic movement of Ca2+ and protects neuronal cells from glutamate excitotoxicity.11 It has been shown in vitro that AG antibiotics cause an increase in intracellular calcium levels in avian HCs46 and in isolated OHCs of guinea pigs.47 Therefore, inhibition of Ca2+ homeostasis distribution may have a crucial role in the ability of TAT-FNK to prevent apoptotic cochlear HC death. The mechanism of how TAT-FNK prevents cochlear HC death remains to be fully elucidated.
In conclusion, we demonstrated that TAT-FNK infiltrated in gelatin sponge and placed on the guinea pig RWM could successfully deliver the protein to the cochlea by penetrating through the RWM, and that a single topical administration of TAT-FNK protected the cochlea against the combination of the ototoxic drugs KM and EA in vivo. An in vitro study demonstrated that TAT-FNK suppressed the activation of caspase-9 and protected cochlear HCs from KM-induced apoptosis. These findings suggest that topical administration of an antiapoptotic protein fused with TAT and soaked with a gelatin sponge is effective at preventing the apoptosis of cochlear HCs, and that such topical treatment is superior to systemic administration in terms of organ specificity and safety. Future studies using this technology may extend the feasibility of protein therapy for treatment of inner ear disorders.
Materials and methods
The experimental protocol was approved by the University Committee for the Use and Care of Animals at the University of Tokyo, and it conforms to the NIH Guidelines for the Care and Use of Laboratory Animals.
Construction and preparation of TAT-FNK and TAT-myc-FNK
We constructed FNK (originally designated as Bcl-xFNK) by introducing amino-acid substitutions into Bcl-xL using a two-step PCR mutagenesis method, as reported previously.9 The substituted codons were as follows: Tyr-22 (TAC) with Phe (TTC), Gln-26 (CAG) with Asn (AAC) and Arg-165 (CGG) with Lys (AAG). Among the mammalian antiapoptotic factors, FNK is the only mutant with a gain-of-function phenotype because, compared with Bcl-xL, FNK showed stronger antiapoptotic activity to protect cultured cells from death induced by various death stimuli, including oxidative stress, a calcium ionophore and withdrawal of growth factors.9 TAT-FNK and TAT-myc-FNK were then prepared as described previously.11 The gene constructed for FNK was fused with an oligonucleotide encoding TAT, and the resulting TAT-FNK gene encoded met-gly-TAT (consisting of 11 amino acids: YGRKKRRQRRR)-gly-FNK. An oligonucleotide encoding GEQKLISEEDLG (the myc TAG sequence is underlined) was inserted between the TAT and FNK sequences of TAT-FNK by PCR to obtain TAT-myc-FNK. To construct myc-FNK without the TAT domain, an oligonucleotide encoding met-gly-myc TAG-gly was also ligated to the FNK sequence by PCR. The TAT-FNK plasmid was introduced into Escherichia coli DH5a cells (Invitrogen, Life Technology, Carlsbad, CA, USA) and the TAT-FNK protein was overexpressed by treatment with 1 mM isopropyl 1-thio-β-D-galactoside for 5 h with vigorous shaking at 37 °C. Proteins were solubilized in buffer (7 M urea, 2% sodium dodecyl sulfate, 1 mM dithiothreitol, 62.5 mM Tris-HCl (pH 6.8) and 150 mM NaCl) and then subjected to sodium dodecyl sulfate-PAGE to remove contaminating proteins and endotoxins. The gel was treated with 1 M KCl and the transparent band corresponding to TAT-FNK was cut out. Proteins were electrophoretically extracted from the gel slice using extraction buffer (25 mM Tris, 0.2 M glycine and 0.1% sodium dodecyl sulfate) for in vitro and in vivo experiments. The extraction buffer was used as the vehicle. The concentration of the extracted TAT-FNK ranged from 1 to 6 mg ml−1.
Immunohistochemical detection of TAT-myc-FNK in the cochlea after tympanic administration
Eighteen male albino guinea pigs (Saitama Experimental Animals Supply Co. Ltd, Saitama, Japan) weighing 250–300 g were used. Under anesthesia with xylazine hydrochloride (10 mg kg−1; Bayer, Leverkusen, Germany) and ketamine hydrochloride (40 mg kg−1; Sankyo, Tokyo, Japan), a post-auricular incision was made and the bone posterior to the tympanic ring was exposed. A hole was drilled into the bulla exposing the middle ear space medial to the tympanic ring. The round window niche and the RWM were identified. The gelatin sponge (Spongel; Astellas Pharma Inc., Tokyo, Japan) was soaked in 3 μl of TAT-myc-FNK (0.5 mg ml−1) and placed on the RWM of the left ear. The animals were killed at 1, 3, 6, 12, 24 and 48 h (n=3 for each time point) after injection, while under deep anesthesia, using an overdose of xylazine hydrochloride (Bayer) and ketamine hydrochloride (Sankyo). Three animals that were killed 6 h after a similar tympanic administration of myc-FNK (3 μl; 0.5 mg ml−1) served as controls. The cochleae from both ears were perfused with 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS) at pH 7.4 through the oval and round windows, and immersed in the same fixative overnight at 4 °C. The specimens were decalcified in 10% EDTA acid for 14 days, dehydrated through a graded alcohol series and embedded in paraffin. The embedded tissues were cut into 5-μm-thick sections parallel to the modiolus and mounted on glass slides. The sections were deparaffinized, hydrated and rinsed with PBS. To detect TAT-myc-FNK and myc-FNK in situ, rabbit an anti-myc-tag polyclonal antibody was used (1:5000, 4 °C overnight; Upstate Biotechnology, Lake Placid, NY, USA) coupled with a DAKO Envision+ system (Dako Japan, Kyoto, Japan). Negative controls were established by replacing the primary antibody with blocking buffer.
The LI of the anti-myc antibody in the cochlear tissues was obtained by a modified Photoshop-based image analysis. The original method was developed by Lehr et al.48 In brief, an image was digitized on magnetic optical disks. Using the ‘Magic Wand’ tool in the ‘Select’ menu of Photoshop, the cursor was placed on a portion of the immunostained area. The tolerance level of the Magic Wand tool was adjusted so that the entire immunostained area was selected. Using the ‘Similar’ command in the ‘Select’ menu, all the immunostained areas were selected automatically. Subsequently, the image was transformed to an 8-bit grayscale format. An optical density plot of the selected areas was generated using the ‘Histogram’ tool in the ‘Image’ menu. The mean staining intensity and the number of pixels in the selected areas were quantified. Next, the background was selected using the ‘Inverse’ tool in the ‘Select’ menu. The mean background intensity was quantified using the ‘Histogram’ tool as mentioned above. The immunostaining intensity was calculated as the difference between the mean staining intensity and the mean background intensity. The immunostained ratio was calculated as the ratio of the number of pixels in all the immunostained areas to that in the entire image. LI was defined as the product of the immunostained ratio and the immunostaining intensity. The modiolar sections were obtained in every third section and five sections were randomly selected from each ear (10 sections from each animal). As a result, the LI was measured using 30 sections in each group by a technician naïve to the treatment, preparation techniques or the aims of the current study. To investigate differences among the cochlear turns, the LIs in the basal, second and third turns were also measured 1 and 6 h after application. The LIs of the SGCs, SV and SL 6 h after the application of TAT-myc-FNK onto the RWM, as well as those in the controls, were also measured. To compare the immunostaining intensity among the cells in these organs, the ratios of normalized immunoreactivity were calculated by dividing the LIs at 6 h by those of the control.
Tympanic injection of TAT-FNK in vivo
Eight male albino guinea pigs, weighing 250–300 g and showing ABR thresholds within normal limits based on our laboratory database, were used in this investigation. Only male animals were used because there are gender differences in the ability to detoxify reactive oxygen species and in the levels of endogenous antioxidants in the cochlea.49, 50
Animals were anesthetized with xylazine hydrochloride (10 mg kg−1) and ketamine hydrochloride (40 mg kg−1). Chloramphenicol sodium succinate (30 mg kg−1, intramuscular injection) was administered as a prophylactic. Under aseptic conditions, the bulla was exposed bilaterally from an occipitolateral approach and opened to allow visualization of the RWM. A gelatin sponge soaked with 3 μl of TAT-FNK (6 mg ml−1) was placed on the RWM in the left ear, whereas a gelatin sponge soaked with only a vehicle was placed on the RWM in the right ear. One hour after the wound was sutured, a single dose of KM (200 mg kg−1; Meiji, Tokyo, Japan) was injected subcutaneously. Then, 2 h after the KM injection, the jugular vein was exposed under general anesthesia and EA (40 mg kg−1; Sigma-Aldrich, Tokyo, Japan) was infused into the vein as described previously.51 An additional four animals served as drug controls: a gelatin sponge soaked with 3 μl of TAT-FNK (6 mg ml−1) was placed on the left RWM, but KM and EA were not administered.
ABRs were recorded using waveform storing and stimulus control using MEB-5504 (NIHON KOHODEN CO., Tokyo Japan) and DPS-725 (DIA MEDICAL CO., Tokyo, Japan). Sound stimuli were produced by the PT-R7 III ribbon-type speaker (PIONEER CO., Tokyo, Japan). Recordings were performed in a closed-field TRAACOUSTICS acoustic enclosure (TRACOUSTICS INC., Austin, TX, USA) and sound level calibration was performed using a sound-level meter (NA-28 RION, Tokyo, Japan). Pure tones (4, 8 and 20 kHz) were measured 3 days after the arrival of the animals to determine the baseline thresholds, and 14 days after the ototoxic insult (for experimental animals) or TAT-FNK application (for drug control animals) to determine the threshold shifts. The frequencies (4, 8 and 20 kHz) measured in this study were frequently used for other studies using guinea pigs, including our previous study.13, 52, 53 We have limited our investigation to these frequencies to evaluate hearing to minimize the stress on these animals. The method of ABR measurement has been described previously.54 In brief, animals were anesthetized with a mixture of xylazine hydrochloride (10 mg kg−1, intramuscular) and ketamine hydrochloride (40 mg kg−1, intramuscular), and needle electrodes were placed subcutaneously at the vertex (active electrode), beneath the pinna of the measured ear (reference electrode) and beneath the opposite ear (ground). The stimulus duration was 15 ms, with a presentation rate of 11 s−1, and the rise/fall time was 1 ms. Responses of 1024 sweeps were averaged at each intensity level (5-dB steps) to assess the threshold. The threshold was defined as the lowest intensity level at which a clear reproducible waveform was visible in the trace. When an ABR waveform could not be evoked, the threshold was assumed to be 5 dB greater than the maximum intensity produced by the system (105 dB SPL). Threshold shifts were calculated by subtracting the baseline thresholds from those observed before killing.
Assessment of extent of HC loss
After ABR measurements 14 days after ototoxic insults (experimental animals) or TAT-FNK application (drug control animals), animals were killed under deep anesthesia using xylazine hydrochloride and ketamine hydrochloride. The bilateral cochleae were perfused with 4% paraformaldehyde in 0.1 M PBS at pH 7.4 through the oval and round windows, and then immersed in the same fixative overnight at 4 °C. The cochleae were then washed with PBS, permeabilized with 0.3% Triton X-100 for 10 min and labeled with 1% rhodamine phalloidin (Molecular Probes, Eugene, OR, USA) for 30 min to stain F-actin. The tissues were processed as whole mounts using the surface preparation technique. The specimens were then mounted on glass slides using the Prolong Antifade kit (Molecular Probes) and observed. Reticules whose length (bin width) at × 40 was 0.45 mm were used to count the numbers of total and missing HCs. HCs that showed an identifiable cell body and cuticular plate were considered to be present. The presence of distinctive scar formations produced by convergence of adjacent phalangeal processes was regarded as an indicator of a missing HC. The percentage of HC loss for the IHCs and OHCs was calculated for each segment obtained from each animal. The average for each segment was then determined for each group and plotted from the apex to the base to produce an average cytocochleogram. Two animals were excluded because of tissue damage during surface preparation, leaving a total of six cochleae for the HC count study.
Assessment of the protective effects of TAT-FNK and caspase-9 detection for cultured HCs
Sprague–Dawley rats (Saitama Experimental Animals Supply Co. Ltd) were decapitated on postnatal day 5 (P5) and the cochlea was carefully dissected out. On the basis of the methods of Sobkowicz et al.,55 the SV, the SL and the spiral ganglion neurons were dissected away, leaving the OC. The cochlea used for analysis was prepared by cutting 2 mm from the basal end and 3 mm from the apical end of the cochlea (approximately half of the cochlea). Explants were maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum, 25 mM Hepes and 30 U ml−1 penicillin, and were cultured in an incubator at 37 °C under 5% CO2 and 95% humidity for 24 h. Explants were exposed to medium containing 20 nM TAT-FNK, 20 nM FNK without TAT or vehicle. Two hours after exposure, the medium was changed to one containing 6 mM KM and either 20 nM FNK, 20 nM TAT-FNK or the vehicle. Typically, 10 cultures were evaluated for each experimental condition: four were cultured for 10 h for detection of caspase-9 and six were cultured for 12 h for cell counting. Additional 10 cultures were evaluated for control, of which four were cultured for 10 h and six were cultured for 12 h, with the medium containing no KM.
Caspase-9 activity was examined by using the fluorescent caspase substrate fam-LEHD-fmk (caspase-9 substrate), which was obtained from Intergen (Purchase, NY, USA) and used according to the manufacturer's protocol. After culturing, the fluorescent substrate was added directly to the culture medium (final concentration, 5 μM) for the final hour in culture. After 1 h in this substrate, the OC was washed three times for 15 min each at 37 °C in the washing buffer supplied by the manufacturer. The cultures were then fixed overnight at 4 °C in the fixative supplied by the manufacturer. After fixation, the cochleae were washed with PBS, permeabilized with 0.3% Triton X-100 for 10 min and labeled with 1% rhodamine phalloidin (Molecular Probes) for 30 min to stain F-actin. Whole mounted cochleae were viewed with a confocal laser-scanning microscope (ZEISS LSM5 PASCAL). Caspase-9-positive cells were counted over a 0.2 mm longitudinal distance from four separate regions in each culture. A mean value was determined for each culture.
For HC counting, cultures were fixed overnight at 4 °C in the fixative supplied by the manufacturer. After fixation, the cochleae were washed with PBS, permeabilized with 0.3% Triton X-100 for 10 min and labeled with 1% rhodamine phalloidin (Molecular Probes) for 30 min to stain F-actin. To quantify HC loss in the cochlea after various treatments, IHCs and OHCs were counted over a 0.2 mm longitudinal distance from four separate regions of each culture. A mean value was determined for each culture.
The SPSS software was used for statistical analysis. The time course of the LI for TAT-myc-FNK in the OC was compared between groups by one-way and then pairwise comparisons, with statistical significance adjusted for multiple comparisons (Scheffe's test). The differences in the turns of the LI for TAT-myc-FNK in the OC at 1 and 6 h were compared by two-way ANOVA (the independent variables were cochlear turns and time course). The differences in the normalized LIs between the sub-sites in the cochlea were compared by one-way ANOVA, and then pairwise comparisons were performed by using Scheffe's test. The ABR thresholds at each frequency before and 14 days after the ototoxic insults were compared by two-way ANOVA (the independent variables were TAT-FNK administration and hearing frequency). The extent of missing HCs in vitro was also compared by two-way ANOVA (the independent variables were TAT-FNK treatment and type of HC). Caspase-9 activities between the groups were compared by one-way ANOVA followed by Scheffe's test. The extent of missing HCs in vitro after exposure to KM was compared by two-way ANOVA (the independent variables were type of HC and drug administration), and if a statistically significant interaction was observed, Bonferroni test was used for simple effects analysis. A level of P<0.05 was accepted as statistically significant.
Cheng AG, Cunningham LL, Rubel EW . Mechanisms of hair cell death and protection. Curr Opin Otolaryngol Head Neck Surg 2005; 13: 343–348.
Rybak LP, Whitworth CA, Mukherjea D, Ramkumar V . Mechanisms of cisplatin-induced ototoxicity and prevention. Hear Res 2007; 226: 157–167.
Rizzi MD, Hirose K . Aminoglycoside ototoxicity. Curr Opin Otolaryngol Head Neck Surg 2007; 15: 352–357.
Staecker H, Liu W, Malgrange B, Lefebvre PP, Van De Water TR . Vector-mediated delivery of bcl-2 prevents degeneration of auditory hair cells and neurons after injury. ORL J Otorhinolaryngol Relat Spec 2007; 69: 43–50.
Liu YH, Ke XM, Qin Y, Gu ZP, Xiao SF . Adeno-associated virus-mediated Bcl-xL prevents aminoglycoside-induced hearing loss in mice. Chin Med J 2007; 120: 1236–1240.
Haas R, Murea S . The role of granulocyte colony-stimulating factor in mobilization and transplantation of peripheral blood progenitor and stem cells. Cytokines Mol Ther 1995; 1: 249–270.
Schwarze SR, Ho A, Vocero-Akbani A, Dowdy SF . In vivo protein transduction: delivery of a biologically active protein into the mouse. Science 1999; 285: 1569–1572.
Becker-Hapak M, McAllister SS, Dowdy SF . TAT-mediated protein transduction into mammalian cells. Methods 2001; 24: 247–256.
Asoh S, Ohtsu T, Ohta S . The super antiapoptotic factor Bcl-xFNK constructed by disturbing intramolecular polar interactions in rat Bcl-xL. J Biol Chem 2000; 275: 37240–37245.
Asoh S, Mori T, Nagai S, Yamagata K, Nishimaki K, Miyato Y et al. Zonal necrosis prevented by transduction of the artificial anti-death FNK protein. Cell Death Differ 2005; 12: 384–394.
Asoh S, Ohsawa I, Mori T, Katsura K, Hiraide T, Katayama Y et al. Protection against ischemic brain injury by protein therapeutics. Proc Natl Acad Sci USA 2002; 99: 17107–17112.
Nagai S, Asoh S, Kobayashi Y, Shidara Y, Mori T, Suzuki M et al. Protection of hepatic cells from apoptosis induced by ischemia/reperfusion injury by protein therapeutics. Hepatol Res 2007; 37: 133–142.
Kashio A, Sakamoto T, Suzukawa K, Asoh S, Ohta S, Yamasoba T . A protein derived from the fusion of TAT peptide and FNK, a Bcl-x(L) derivative, prevents cochlear hair cell death from aminoglycoside ototoxicity in vivo. J Neurosci Res 2007; 85: 1403–1412.
Minn AJ, Boise LH, Thompson CB . Expression of Bcl-xL and loss of p53 can cooperate to overcome a cell cycle checkpoint induced by mitotic spindle damage. Genes Dev 1996; 10: 2621–2631.
Valera ET, Brassesco MS, Scrideli CA, Tone LG . Tetraploidization in Wilms tumor in an infant. Genet Mol Res 2010; 9: 1577–1581.
Schuknecht HF . Ablation therapy for the relief of Meniere's disease. Trans Am Laryngol Rhinol Otol Soc 1956; 66: 589–600.
Haynes DS, O’Malley M, Cohen S, Watford K, Labadie RF . Intratympanic dexamethasone for sudden sensorineural hearing loss after failure of systemic therapy. Laryngoscope 2007; 117: 3–15.
Tsuji J, Liberman MC . Intracellular labeling of auditory nerve fibers in guinea pig: central and peripheral projections. J Comp Neurol 1997; 381: 188–202.
Tanaka K, Motomura S . Permeability of the labyrinthine windows in guinea pigs. Arch Otorhinolaryngol 1981; 233: 67–73.
Bellucci RJ, Fisher EG, Rhodin J . Ultrastructure of the round window membrane. Laryngoscope 1972; 82: 1021–1026.
Carpenter AM, Muchow D, Goycoolea MV . Ultrastructural studies of the human round window membrane. Arch Otolaryngol Head Neck Surg 1989; 115: 585–590.
Miriszlai E, Benedeczky I, Horváth K, Köllner P . Ultrastructural organization of the round window membrane in the infant human middle ear. ORL J Otorhinolaryngol Relat Spec 1983; 45: 29–38.
Goycoolea MV, Lundman L . Round window membrane. Structure function and permeability: a review. Microsc Res Tech 1997; 36: 201–211.
Lundman LA, Holmquist L, Bagger-Sjoback D . Round window membrane permeability. An in vitro model. Acta Otolaryngol 1987; 104: 472–480.
Juhn SK, Hamaguchi Y, Goycoolea M . Review of round window membrane permeability. Acta Otolaryngol Suppl 1989; 457: 43–48.
Smith BM, Myers MG . The penetration of gentamicin and neomycin into perilymph across the round window membrane. Otolaryngol Head Neck Surg 1979; 87: 888–891.
Brady DR, Pearce JP, Juhn SK . Permeability of round window membrane to 22Na or RISA. Arch Otorhinolaryngol 1976; 214: 183–184.
Nomura Y . Otological significance of the round window. Adv Otorhinolaryngol 1984; 33: 1–162.
Morizono T, Hamaguchi Y, Juhn SK . Permeability of human serum albumin from chinchilla middle ear to inner ear. In XXIII workshop on inner ear biology 1986. GDR: Berlin.
Goycoolea MV, Paparella MM, Goldberg B, Carpenter AM . Permeability of the round window membrane in otitis media. Arch Otolaryngol 1980; 106: 430–433.
Richard JP, Melikov K, Vives E, Ramos C, Verbeure B, Gait MJ et al. Cell-penetrating peptides. A reevaluation of the mechanism of cellular uptake. J Biol Chem 2003; 278: 585–590.
Vivès E, Brodin P, Lebleu B . A truncated HIV-1 Tat protein basic domain rapidly translocates through the plasma membrane and accumulates in the cell nucleus. J Biol Chem 1997; 272: 16010–16017.
Wadia JS, Dowdy SF . Protein transduction technology. Curr Opin Biotechnol 2002; 13: 52–56.
Gupta B, Levchenko TS, Torchilin VP . Intracellular delivery of large molecules and small particles by cell-penetrating proteins and peptides. Adv Drug Deliv Rev 2005; 57: 637–651.
Ruel-Gariepy E, Chenite A, Chaput C, Guirguis S, Leroux J . Characterization of thermosensitive chitosan gels for the sustained delivery of drugs. Int J Pharm 2000; 203: 89–98.
Park AH, Jackson A, Hunter L, McGill L, Simonsen SE, Alder SC et al. Cross-linked hydrogels for middle ear packing. Otol Neurotol 2006; 27: 1170–1175.
Husmann KR, Morgan AS, Girod DA, Durham D . Round window administration of gentamicin: a new method for the study of ototoxicity of cochlear hair cells. Hear Res 1998; 125: 109–119.
Okamoto T, Yamamoto Y, Gotoh M, Huang CL, Nakamura T, Shimizu Y et al. Slow release of bone morphogenetic protein 2 from a gelatin sponge to promote regeneration of tracheal cartilage in a canine model. J Thorac Cardiovasc Surg 2004; 127: 329–334.
Salt AN . Simulation of methods for drug delivery to the cochlear fluids. Adv Otorhinolaryngol 2002; 59: 140–148.
Imamura S, Adams JC . Distribution of gentamicin in the guinea pig inner ear after local or systemic application. J Assoc Res Otolaryngol 2003; 4: 176–195.
Zou J, Saulnier P, Perrier T, Zhang Y, Manninen T, Toppila E et al. Distribution of lipid nanocapsules in different cochlear cell populations after round window membrane permeation. J Biomed Mater Res B Appl Biomater 2008; 87: 10–18.
Khalil IA, Kogure K, Akita H, Harashima H . Uptake pathways and subsequent intracellular trafficking in nonviral gene delivery. Pharmacol Rev 2006; 58: 32–45.
Chen Y, Huang WG, Zha DJ, Qiu JH, Wang JL, Sha SH, Schacht J . Aspirin attenuates gentamicin ototoxicity: from the laboratory to the clinic. Hear Res 2007; 226: 178–182.
Priuska EM, Schacht J . Mechanism and prevention of aminoglycoside ototoxicity: outer hair cells as targets and tools. J Ear Nose Throat J 1997; 76: 164–166, 168, 170-1.
Cunningham LL, Cheng AG, Rubel EW . Caspase activation in hair cells of the mouse utricle exposed to neomycin. J Neurosci 2002; 22: 8532–8540.
Hirose K, Westrum LE, Stone JS, Zirpel L, Rubel EW . Dynamic studies of ototoxicity in mature avian auditory epithelium. Ann NY Acad Sci 1999; 884: 389–409.
Li Y, Huang Y, Luo S . Alternation of intracellular Ca2+ caused by streptomycin in the isolated cochlear OHC. Zhonghua Er Bi Yan Hou Ke Za Zhi 1995; 30: 227–229.
Lehr HA, Mankoff DA, Corwin D, Santeusanio G, Gown AM . Application of Photoshop-based image analysis to quantification of hormone receptor expression in breast cancer. J Histochem Cytochem 1997; 45: 1559–1565.
Julicher RH, Sterrenberg L, Haenen GR, Bast A, Noordhoek J . Sex differences in the cellular defence system against free radicals from oxygen or drug metabolites in rat. Arch Toxicol 1984; 56: 83–86.
el Barbary A, Altschuler RA, Schacht J . Glutathione S-transferases in the organ of Corti of the rat: enzymatic activity, subunit composition and immunohistochemical localization. Hear Res 1993; 71: 80–90.
Yamasoba T, Kondo K . Supporting cell proliferation after hair cell injury in mature guinea pig cochlea in vivo. Cell Tissue Res 2006; 325: 23–31.
Suzuki M, Yagi M, Brown JN, Miller AL, Miller JM, Raphael Y . Effect of transgenic GDNF expression on gentamicin-induced cochlear and vestibular toxicity. Gene Ther 2000; 7: 1046–1054.
Lin CD, Oshima T, Oda K, Yamauchi D, Tsai MH, Kobayashi T . Ototoxic interaction of kanamycin and 3-nitropropionic acid. Acta Otolaryngol 2008; 128: 1280–1285.
Pourbakht A, Yamasoba T . Ebselen attenuates cochlear damage caused by acoustic trauma. Hear Res 2003; 181: 100–108.
Sobkowicz HM, Loftus JM, Slapnick SM . Tissue culture of the organ of Corti. Acta Otolaryngol Suppl 1993; 502: 3–36.
We thank Ms A Tsuyuzaki and Ms Y Kurasawa (Department of Otolaryngology and Head and Neck Surgery, Faculty of Medicine, University of Tokyo, Tokyo, Japan) for technical assistance. This work was supported by Grants (17659527 and 20390440) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan to TY and a Grant (15110201) from the Ministry of Health, Labor and Welfare of Japan to TY.
The authors declare no conflict of interest.
About this article
Journal of Biomolecular Structure and Dynamics (2019)
Engineering PLGA nano-based systems through understanding the influence of nanoparticle properties and cell-penetrating peptides for cochlear drug delivery
International Journal of Pharmaceutics (2017)
Topical application of superoxide dismutase mediated by HIV-TAT peptide attenuates UVB-induced damages in human skin
European Journal of Pharmaceutics and Biopharmaceutics (2016)
The autophagy pathway maintained signaling crosstalk with the Keap1–Nrf2 system through p62 in auditory cells under oxidative stress
Cellular Signalling (2015)
Cell and Tissue Research (2015)