Trauma to the inner ear can result in death of sensory cells within the cochlea if a damage threshold is exceeded. Outer hair cells (OHCs) are usually the first to be damaged during physical trauma, followed by the inner hair cells (IHCs) and lastly the supporting cells of the organ of Corti. Several mechanisms of cell death are thought to be involved: necrosis, necrosis-like programmed cell death (PCD; type 2 PCD), and apoptosis (type 1 PCD). In this article, we present the role of cell death signal cascades in two common types of physical trauma to the cochlea, i.e., electrode insertion trauma- and noise trauma-induced hearing losses. We also discuss some innovative therapeutic strategies applicable to the protection of hearing from trauma-initiated forms of hearing loss.
Cochlear implantation trauma and noise-induced hearing loss both involve a physical disruption of the organ of Corti and may involve several mechanisms of cell death at the molecular level, i.e., necrosis, necrosis-like programmed cell death (PCD; type 2 PCD), and apoptosis (type 1 PCD). This article reviews several promising therapeutic strategies that are currently being developed. One of these promising new strategies involves the use of a highly effective peptide inhibitor of the c-Jun N-terminal kinase cell death signal cascade (i.e., D-JNKI-1) to prevent apoptosis of injured auditory hair cells. Our recent studies showed prevention of cochlear implantation-induced hearing loss by infusing this peptide into the cochlea of guinea pigs. Another otoprotective therapy under investigation is the application of mild hypothermia to protect the cochlea from the development of a hearing loss that follows exposure to a physical trauma, e.g., electrode array insertional trauma. These forward-looking strategies have the potential of improving hearing outcomes after cochlear implantation and providing novel means of otoprotection from noise-induced trauma. Anat Rec Part A, 2006. © 2006 Wiley-Liss, Inc.
Apoptosis appears to play a critical role in the pathogenesis of hearing loss that develops following physical trauma to the cochlea, i.e., sound trauma or electrode insertion trauma. Apoptosis is associated with an active cell death process and can be identified by distinct morphologic features such as nuclear condensation and fragmentation of nuclear DNA within the affected cell, whereas necrosis is a passive consequence of an overwhelming injury to a cell and is marked by nuclear swelling and lyses of the affected cell. Necrosis-like PCD has some of the features of apoptosis, but usually does not involve the activation of procaspase molecules (Hertz et al.,2005). While acute hearing loss posttrauma can involve necrosis, necrosis-like PCD, and apoptosis, the progressive hearing loss that develops following an initial insult to the cochlea appears to be due predominantly to apoptosis of damaged hair cells. Apoptosis is an active biochemical process that is controlled by the damaged cell once a threshold of damage has been exceeded and it is meant to protect the overall health and survival capability of the whole organism by the elimination of pathological and potentially dangerous cells. Not all the cell death signaling pathways that direct damaged sensory cells to undergo apoptosis within a traumatized cochlea are precisely known, and they are currently under active investigation (Hu et al.,2000,2002a,2002b,2006; Nicotera et al.,2003; Van De Water et al.,2004; Wang et al., 2004; Yang et al.,2004).
Apoptosis is controlled by the expression of several genes and proteins, which include members of the Bcl-2 family of pro- and antiapoptotic proteins, some of the members of the family of caspase proteases, and cytochrome c (Herz et al.,2005; Kim et al.,2005). Caspases are a family of aspartate-specific cysteine proteases, which exist as latent intracellular zymogens (Van De Water et al.,2004). More than 14 caspases have been identified and are divided into two groups: upstream initiator caspases (caspases 2, 8, 9, and 10) and downstream effector caspases (caspases 3, 5, 6, and 7).
Effector caspases, once activated, selectively cleave distinct intracellular substrates that lead to the dismantling of a cell's architecture, DNA, signaling apparatus, and restorative repair mechanisms. The sequence of caspase activation shows that distinct cascades are activated depending on the specific pathology, conditions employed, and the cell type. At present, caspases 8, 9, and 3 are known to be involved in the apoptosis of physically damaged hair cells (Nicotera et al.,2003) and caspases 5 and 6 by the results of an inhibitor study are suggested to participate (Do et al.,2004), while caspases 7 and 10 are thought to be involved but not yet proven to participate in the process of apoptosis of physical trauma-damaged hair cells (Fig. 1) (Van De Water et al.,2004). The underlying events include loss of mitochondrial transmembrane potential, release of cytochrome c from the damaged mitochondria into the cytoplasm, formation of an apotosome, sequential activation of activator and then effector procaspases, and a subsequent increase in lipid peroxidation of cellular membranes. The apoptosis-suppressor gene products of Bcl-2 and Bcl-xL play important protective functions in preventing the release of cytochrome c into the cytosol of an injured cell, thereby preventing the activation of procaspase 9, which would in turn activate procaspase 3, or other effector caspases (e.g., procaspase 6). The precise molecular mechanism leading to the induction of the cell death in the cochlea is not yet entirely clear; one signal cascade includes activation of c-jun N-terminal kinase (JNK), also known as stress-activated protein kinase (SAPK) of the mitogen-activated protein kinase (MAPK) pathway (Pirvola et al.,2000). Figure 2 shows the MAPK/JNK cell death signal cascade and the site of action of many pharmacologic inhibitors that have been demonstrated either to block or partially to block different points within this signaling pathway and Figure 3 shows the mechanism of action by which D-JNKI-1 blocks the actions of JNK, with c-Jun being used as the example of a major downstream target (Zine and Van De Water,2004). An example of the ability of the D-JNKI-1 peptide molecule to prevent the apoptosis of auditory hair cells is seen in Figure 4, where organ of Corti explants were exposed to an ototoxic level of neomycin and if these explants were not treated with D-JNKI-1, there was extensive loss of hair cells via apoptosis (Wang et al.,2003). The mode of action of the D-JNKI-1 inhibitor peptide in preventing hair cell loss is partially understood because phosphorylation of c-Jun by JNK within hair cell nuclei was prevented in the inhibitor-treated cultures but occurred in the neomycin-exposed untreated explants (Fig. 5). In addition, the treatment of the neomycin-exposed explants with D-JNKI-1 suppressed the expression of c-fos transcripts (Fig. 4), which is due to the blocking of the action of JNK on activation of a downstream target, i.e., Elk-1 (Fig. 2). Another signal cascade may involve the focal adhesion kinase pathway, which is known to disrupt the extracellular matrix below the outer hair cells by modulating the binding of actin to integrins following acoustic trauma. Information regarding the activation of downstream caspase 3 and its relationships to JNK activation is not yet fully understood. An understanding of these events is essential for understanding the mechanisms that govern cell death and cell survival during and following traumatic injury to the cochlea.
COCHLEAR IMPLANT INSERTION TRAUMA
Histological evaluation of the cochlea after insertion of a cochlear implant electrode into the scala tympani of temporal bones obtained from cadavers has demonstrated in multiple studies that there can be immediate damage to different structures of the cochlea such as the basilar membrane, spiral ligament, and osseous spiral lamina (Fayad et al.,1991; Welling et al.,1993; Nadol et al.,2001; Eshraghi et al.,2003). In a previous cryohistological study of human temporal bones, we analyzed both electrode position and overt structural damage to the cochlea and based on these observations have formulated a numerical scale of values ranging from 0 to 4 to rate the extent and severity of initial macroscopic trauma to the cochlea by the insertion of full-size electrode arrays that in an implanted patient would act to stimulate directly the auditory neurons present in Rosenthal's canal (Fayad et al.,1991). We used this trauma scale (Table 1) to demonstrate that the severity of direct trauma to cochlear structures can be lessened by the design and use of less traumatic electrodes and by modification of the surgical technique used for the insertion of the electrode array (Balkany et al.,2002; Eshraghi et al.,2004). However, using objective electrophysiological testing of the hearing threshold (auditory brainstem response (ABR) and distortion products of otoacoustic emissions (DPOAE)) in a rat model of cochlear implant electrode trauma, we have demonstrated that despite a very soft technique of electrode insertion and a grade 0 of macroscopic trauma to the cochlear duct, we still observed the development of a progressive loss of hearing for up to 1 week after the initial physical trauma caused by electrode insertion. These observations lead us to postulate that the insertion of a cochlear implant electrode array causes tissue trauma on a molecular level that cannot be rated on our scale of direct physical trauma, which can kill auditory hair cells via either necrosis, necrosis-like PCD, apoptosis, or any combination of these mechanisms of cell death. Do et al. (2004) discussed in their mouse model of implantation trauma that a hydraulic trauma induced by the insertion of an electrode and postoperative inflammation represent alternative pathways of damage to the cochlea that can result in the development of a hearing loss. The results in their study demonstrate that discrete changes in the volume of cochlear perilymph result in repeatable hearing losses and that the use of cell death inhibitors (e.g., pancaspase inhibitor-zVAD-FMK and specific inhibitors to caspase 3, 5, and 6) significantly protected the hearing from a loss of threshold in response to hydraulic trauma from injection of a bolus of artificial perilymph that was equal to 30% of the original perilymph volume (Do et al.,2004).
|Grade 1: No observable macroscopic trauma*|
|Grade 2: Elevation of basilar membrane|
|Grade 3: Dislocatioon of electrode to scala vestibule|
|Grade 4: Fracture of osseous spiral lamina or modiolus, or tear in tissues of stria vascularis/spiral ligament complex|
In a recent study, we have characterized the pattern of hearing loss that occurs in the guinea pig following electrode insertion trauma with the same objective measurements of hearing that was used in our rat model (i.e., DPOAEs and ABRs) of electrode insertion trauma-induced hearing loss (Eshraghi et al.,2005a). In this model, we performed a cochleostomy to insert the electrode into the cochlea, while in the rat model, we used a round window membrane route for electrode insertion. We have confirmed with this new electrode insertion trauma model in the guinea pig that immediately following insertion trauma, there is an acute loss of hearing (approximately 20 dB SPL), followed by a progressive loss of hearing of the same amount that developed over 7 days after electrode insertion trauma in our rat model of cochlear implantation trauma. In the operated cochlea, we observed an initial increase of ABR thresholds following electrode insertion trauma (P > 0.05) across all frequencies tested (which included the apical part of the cochlea that is far from the site of electrode insertion trauma). There was also a progressive and significant decrease in the amplitudes of the DPOAEs (P > 0.05) after electrode insertion trauma.
Cochleae from control and electrode insertion trauma guinea pigs were fixed and organ of Corti surface preparations were microdissected from the lower middle turns. The tectorial membrane was carefully removed prior to FITC-phalloidin/propidium iodide staining. The double-stained organ of Corti specimens were observed using confocal microscopy that allows for identification of OHC and IHC cuticular plates with their stereocilliary bundles and on a deeper level of focus the nuclei of the IHCs and OHCs. We observed changes in nuclear staining in these auditory hair cells that are consistent with apoptosis, i.e., nuclear DNA condensation and formation of apoptotic bodies. TUNEL labeling of the middle turn organ of Corti surface preparations from contralateral control and electrode insertion traumatized cochlea at 12, 24, and 36 hr posttrauma showed a progressive increase in TUNEL-labeled hair cell nuclei in the traumatized cochleae over time posttrauma while compared to the contralateral control cochleae.
These observations lead us to propose a modification to our previous classification of cochlear implant trauma because the specimens were categorized as grade 0 trauma. Grade 0 is now considered as no observable macroscopic damage, but with the possibility of damage on a molecular level (Table 1).
D-JNKI-1 Inhibitor Therapy
The apoptosis of auditory neurons as a consequence of oxidative stress damage has been shown to involve the downstream target of the JNK signal cascade, i.e., c-Jun (Scarpidis et al.,2003). A previous study using D-JNKI-1, which competitively binds to JNK molecules, which prevents interaction with their downstream targets and signaling capacity (Fig. 3), has demonstrated that this inhibitory peptide can prevent loss of both hearing capacity and hair cells in animals challenged with exposure to either a damaging level of sound trauma or to an ototoxic level of an aminoglycoside antibiotic (Wang et al.,2003).
Using our guinea pig model of cochlear implant trauma and the same objective measurements (i.e., DPOAE and ABR tests) used to define hearing function after inner ear trauma, we studied the otoprotective effect of direct delivery of the synthetic inhibitory peptide of JNKs [i.e., perfusion of 10 μM solution of D-JNKI-1 in artificial perilymph (AP) into the scala tympani]. We have demonstrated that there was no posttrauma increase in the ABR thresholds or decrease in DPOAE amplitudes in electrode insertion-traumatized cochleae while traumatized cochleae that were either not treated or treated with perfusion with only AP all experienced an ongoing loss of hearing function for the duration of this experiment, i.e., 7 days (Van De Water et al.,2005).
D-JNKI-1 peptide acts by interrupting the MAPK/JNK signal cascade at the level of JNK molecules preventing the phosphorylation of c-Jun, which disrupts the formation of an AP-1 transcription factor; by preventing the JNKs from disrupting the activity of the antiapoptotic members of the Bcl-2 family; and by JNK activation of other downstream targets (e.g., ATF-2). Thus, D-JNKI-1 is a promising molecule in the arsenal of otoprotective molecules for the development of therapeutic strategies for the prevention of hearing loss during electrode insertion in partial hearing patients. In sum, the delayed progressive component of electrode insertion trauma-initiated hearing loss can be prevented by treating the cochlea immediately after electrode insertion with a peptide inhibitor of the c-Jun N-terminal kinase cell death pathway.
Mild Hypothermia Therapy
Hypothermia has been shown to have a protective effect in the brain following a traumatic injury (Busto et al.,1987; Dietrich et al.,1994,1996; Kil et al.,1996). The beneficial effect of hypothermia on neuronal injury has been attributed to a variety of mechanisms. These include a reduction in metabolic rate, reduced tissue oxygen consumption, decreased metabolic acidosis, a suppression of calcium influx into neurons, diminished nitric oxide production, and a reduction in the level of glutamate excitotoxicity (Hyodo et al.,2001; Eshraghi et al.,2005b).
Hypothermia has also been demonstrated to reduce brain damage following ischemia by inhibiting the formation of reactive oxygen species (ROS), thereby limiting the extent of oxidative stress (Zhao et al.,1996). More recently, it was reported that hypothermia could protect against noise-induced threshold elevation of hearing thresholds (i.e., ABR measures) in laboratory mice (Henry,2003).
In a recent study, we evaluated the otoprotective effect of mild hypothermia in our rat model of cochlear implantation trauma (Balkany et al.,2005). One ear was randomly chosen to undergo electrode insertion while the contralateral ears served as nonoperative controls. In group 1, subjects were kept at 37°C throughout the procedure using a heating blanket and heat lamp. The core temperature of the animals was monitored continuously using a rectal probe with digital readout. In group 2, subjects were cooled from 37°C to 34°C using a cooling blanket for 30 min before surgery and kept at this state of mild hypothermia during surgery and electrode insertion trauma and for a period of 30 min after trauma.
Recordings were performed immediately prior to surgery and on days 0 (immediately after surgery), 3, 5, and 7 using the same objective test of hearing previously reported (ABRs and DPOAEs). The result of this study demonstrates that mild hypothermia can reduce the immediate component of trauma-induced hearing loss and prevent the progressive component of loss of auditory function following cochlear electrode insertion in the rat model of cochlear implantation trauma-induced hearing loss.
NOISE-INDUCED HEARING LOSS
It has been known for some time that noise-induced hair cell death in the cochlea and loss of hearing function continue well after the termination of a noise exposure. However, the underlying mechanisms leading to the expansion of a noise-induced cochlear lesion are not yet fully understood. Hu et al. (2002a) have reported involvement of the apoptotic pathway in the progression of OHC death in the chinchilla cochlea following exposure to a 4 kHz narrow band noise at 110 dB SPL for 1 hr. Morphological examination of OHC nuclei with propidium iodide-stained DNA revealed nuclear condensation and fragmentation, typical morphological features of apoptosis. Apoptosis of the OHCs developed asymmetrically toward the apical and basal parts of the cochleae following sound exposure. Two days after the noise exposure, there was still active OHC pathology with condensed and fragmented nuclei in the basal part of the cochlea. Detection of caspase 3 activation in this study, an intracellular marker for a late stage of apoptosis, showed in their study a spatial agreement between the apoptotic nuclei and activated caspase 3. This observation was confirmed and extended in a more recent study by this group (Nicotera et al.,2003), where the participation of caspase 3 activation in hair cell (HC) apoptosis was confirmed and the participation of activated caspases 8 and 9 was added to the process on sound trauma-induced apoptosis of damaged HCs. Another study has shown that activation of apoptotic changes in the nucleus of a damaged HC correlates with a decrease in F-actin staining of the affected HC linking nuclear changes with alterations in the distribution of cellular proteins that are necessary for function (Hu et al.,2002b). A very recent finding has been that in the early time after sound trauma, apoptosis is a prominent feature within the sound-damaged cochlea and only later is there an equal distribution of both necrosis and apoptosis among the sound-damaged cells of the traumatized cochleae (Yang et al.,2004). Apoptosis therefore appears to play a critical role in the pathogenesis of noise-induced hearing loss. In contrast, Yoshida and Liberman (2000) have demonstrated that sound conditioning (pretreatment with a low-level of nondamaging sound) protects the cochlea against hair cell death and thereby preserves hearing after a subsequent acoustic trauma that exposes the cochlea to a damaging level of sound. Using a combination of immunolabel and Western blotting techniques, these investigators have shown that acoustic trauma causes the release of cytochrome c from the mitochondria into the cytoplasm and a decrease in bcl-2 immunostaining of the outer hair cells. Sound conditioning was shown to trigger a protective effect against these detrimental changes. These results demonstrate that bcl-2 plays an important role in the regulation of hair cell death and provides evidence that the neuroprotective bcl-2 gene acts as an inducible gene that can be unregulated by a sound conditioning paradigm and thereby protect the cochlea from noise-induced hearing loss.
D-JNKI-1 Inhibitor Therapy
Hearing loss can be caused by acoustic trauma, which principally affects the viability of sensory hair cells via the MAPK/JNK cell death signaling pathway that incorporates JNK molecules (i.e., JNKs 1, 2, and 3). Wang et al. (2003) evaluated the otoprotective efficacy of D-JNKI-1, a cell-permeable peptide (Bonny et al.,2001) that blocks JNK signaling (Figs. 2 and 3). The experimental studies included organ cultures of neonatal mouse cochlea exposed to an ototoxic drug (Figs. 4 and 5) and cochleae of adult guinea pigs that were exposed to acoustic trauma. Local delivery of D-JNKI-1 prevented acoustic trauma-induced permanent hearing loss in a dose-dependent manner. Figure 6 shows the results of D-JNKI-1 protection against hair cell loss depicting the surface ultrastructure of a protected and an unprotected organ of Corti as well as quantitative data on hair cell loss as determined by counts of intact hair cells of SEM-prepared cochlear ducts. Treatment of sound trauma-exposed cochlea with D-JNKI-1 also prevented the conversion of a temporary threshold shift (TTS) into a permanent threshold shift (PTS) as depicted in Figure 7. These results indicate that the MAPK-JNK signal pathway is involved in acoustic trauma-induced hair cell loss and also in the development of a permanent hearing loss after exposure to a damaging level of sound or noise. Blocking this signal pathway with D-JNKI-1 is of potential therapeutic value for long-term protection of both the morphological integrity and the physiological function of the organ of Corti during times of physical trauma-initiated oxidative stress, which can lead to the apoptosis of damaged hair cells.
Henry and Chole (1984) have studied the thresholds of the cochlear action potential from rodents at euthermic (38°C) and hypothermic (30°C and 25°C) conditions as determined by rectal temperatures. When subjects susceptible to PTS at low and middle frequencies (anesthetized immature mice) were exposed to 115 dB noise, hypothermia reduced the level of the PTS at these most susceptible frequencies (2–16 kHz). When alert adult mice were exposed to this level of damaging noise, hypothermia protected them from the development of a PTS at their most vulnerable frequency (i.e., 32 kHz).
More recently, Henry (2003) has studied the scalp-recorded cochlear nerve envelope response (CNER). The CNER measure reflects the ability of high-frequency cochlear nerve axons to fire in a phase-locked fashion to low-frequency modulations of the acoustic envelope of high-frequency stimuli. This property is useful in evaluating the adverse effects of noise exposure on the ability of the ear to detect acoustic changes characteristic of vocalizations and speech. Hypothermia (30°C) elevated CNER thresholds elicited by high-frequency stimuli. Mice exposed to noise when hypothermic had smaller threshold elevations than those exposed when euthermic (36°C) (Henry,2003).
Cochlear implant insertion trauma and noise-induced hearing loss both involve physical disruption of the organ of Corti and involve both necrosis and apoptosis pathways at the molecular level. Promising new therapeutic strategies are currently under investigation. One of these promising new strategies involves the use of a highly effective peptide inhibitor of the c-Jun N-terminal kinase cell death pathway (i.e., D-JNKI-1) and the other protective therapy is the use of mild hypothermia to protect the cochlea from the deleterious effect of physical trauma-initiated apoptosis.