Pharmaceutical otoprotection strategies to prevent impulse noise-induced hearing loss

Pharmaceutical otoprotection from noise-induced hearing loss (NIHL) has been a major topic of research for well over two decades. Among the many challenges associated with pharmaceutical otoprotection from NIHL is the heterogeneity of the acoustic properties of the noise exposures and the subsequent heterogeneity in the cochlear injuries that can result. Impulse noises result from the abrupt release of energy into the atmosphere (Hamernik and Hsueh, 1991), with an example being gunfire. Impulses can be contrasted from blasts. Blasts are also abrupt releases of energy but are characterized by a shock wave (Ahroon et al., 1996) and require a shock tube, instead of a speaker, for laboratory simulation. Acoustically, the impulses will vary considerably depending on the source and the propagation medium, but as damaging noise sources, they are characterized by their short durations and high peak sound pressure levels (SPLs). Transient exposures, whether they are impulses and impacts (resulting from the collision of objects), induce a broad set of pathologies to the cell populations in the cochlea that present unique challenges for pharmaceutical otoprotection.

When considered as a damaging noise exposure, impulse noise exposures contribute to a higher kurtosis level of the exposure by increasing the moment-to-moment variability of SPLs. While the sound pressure level (SPL) and duration of exposure are strong predictors of NIHL from Gaussian noise exposures with variation over time, the kurtosis factor is an important predictive variable for noise with a high number of short-duration impulses (Hamernik et al., 2003). Thus, the equal energy hypothesis, which focuses on SPL and duration, does not provide a complete picture of the hearing hazard imposed by noise exposures that include transients (Xie et al., 2016). The impulse exposures vary depending on the peak SPL, the number of impulses in the exposure, and the rate of presentation of the impulses. These variables will contribute heavily to the magnitude of the cochlear injury, the types of pathologies that occur in the cochlea, and whether the injury will consist of temporary threshold shifts (TTS), permanent threshold shifts (PTS), or compound threshold shifts (CTS). With TTS, the recovery of threshold is complete, and there is no PTS. With PTS, there is a permanent change in hearing that is measured after the time period for recovery has ended, and thus no recovery will occur beyond that time point. With CTS, there is an acute TTS component, but without complete recovery, so that there remains a PTS component as well (please see Ryan et al., 2016 for a more detailed review of noise-induced threshold shifts). Of particular concern when considering protection strategies is the relative contribution of mechanical versus metabolic injury. Impulse exposures have been documented to cause significant degrees of mechanical injury to the cochlea, including: shearing of the reticular lamina causing leakage of the endolymph into the compartments filled with cortilymph (Geyer et al., 1978), loss of the mooring of the basilar membrane to the modiolus, disconnection of Hensen's and Deiters' cells (Hamernik et al., 1984), stereocilia damage (Slepecky et al., 1981), and detachment of the outer hair cells (OHCs) from the Deiters' cells (Henderson et al., 2006). These mechanical injuries have several consequences for cochlear physiology and hearing loss. Acutely, they can result in threshold shift due to reduced cochlear amplification and mechanical transduction efficiency. Further, they can trigger longer-term metabolic damage, including oxidative stress (Xiong et al., 2011), inflammation (Kirkegaard et al., 2006), and dysfunction of the mitochondria (Hu, 2007). The combined mechanical and metabolic injuries from impulse noise act as triggers for apoptotic and necrotic cell death (Hu et al., 2006).

From the perspective of pharmaceutical otoprotection, the focus is largely on prevention of metabolic cell death. Reducing the mechanical trauma to the cell populations requires acoustic protection measures that limit the amount of air and bone conducted energy to the cochlea. The goal of pharmaceutical protection is to preserve as many of the cochlear cells as possible after the mechanical trauma. Preservation of cochlear cells will promote retention of function in the acute phase after the exposure, or will permit a restoration of as much function as possible in the chronic recovery phase after the exposure. Prevention of injury will manifest in reduced CTS in the acute period after the exposure, and restoration of function will result in a larger recovery of thresholds (shifting more of the threshold shift from PTS to TTS). While the preservation of cells in the cochlea does not guarantee restoration of function after mechanical disruption, death of hair cells or spiral ganglion neurons in the cochlea will certainly limit the amount of recovery that is currently possible.

Impulse noise exposure, particularly from firearm use (Hoffman et al., 2017; Meinke et al., 2017) represents a significant risk factor for acquired hearing loss. While many occupational NIHLs occur gradually over periods of months or years, the transient nature of impulse noise creates a different time course for injury and intervention. In cases of high-level impulse exposures, organ of Corti degeneration and cell death occur over a window of time that can extend through 30 days after the exposure (Hamernik et al., 1984). This 30-day window allows for post-exposure rescue treatments aimed at maximizing recovery. As the impulse presentation rate (in impulses/s) increases, the peak SPL decreases, and/or the number of exposures increases, it is reasonable to expect that the pattern of injury would start to resemble a standard occupational exposure more closely. This, in turn, could minimize opportunities for post-exposure rescue, since more of the cochlear cell death would potentially be occurring during the noise exposure as opposed to after the exposure. Reviewing the literature on otoprotection from impulse NIHL, the focus has been on discrete impulse exposures that comprised no more than a few minutes of exposure time and were designed to evoke combinations of mechanical and metabolic injuries. Successful otoprotection from impulse NIHL has been demonstrated across several animal models and exposure paradigms for four primary classifications of compounds, each of which will be discussed separately: (1) magnesium supplements, (2) antioxidants, (3) glucocorticoids, and (4) anti-apoptotic compounds. The compounds will be used to group the experiments together, but critical attention will be paid to acoustic properties of the impulse exposures to provide consideration of the influence that the exposures may have on the otoprotection results that were obtained in the experiments.

Magnesium (Mg²⁺) has a long history in research as an otoprotective compound against NIHL and other ototoxins. Despite the extensive literature correlating Mg²⁺ and acquired hearing losses, the mechanisms through which Mg²⁺ reduces susceptibility to cochlear injury are not fully elucidated. One hypothesis is that Mg²⁺ acts as an antagonist for calcium (Ca²⁺), preventing it from entering the hair cells (Cevette et al., 2003; Sendowski, 2006). This, in turn, could have two effects in the organ of Corti. The first would be to prevent over-excitation of the spiral ganglion neurons. Since Ca²⁺ influx is the triggering event for glutamate vesicle release, antagonism with Mg²⁺ can act to block that vesicle release, thus reducing the amount of glutamate in the synapses. Mg²⁺ can also act as to block current flow through NMDA receptors (Mayer et al., 1984), which could serve to limit over-excitation of spiral ganglion cells from excessive glutamate stimulation. Both limiting glutamate release and preventing spiral ganglion neuron excitotoxicity could have the effect of attenuating noise-induced synaptopathy, which may be linked to TTS (Roberston, 1983) and the permanent auditory de-afferentation detected in animal models (Kujawa and Liberman, 2006, 2009; Lin et al., 2011; Fernandez et al., 2015). The second possible effect of Mg²⁺ in the organ of Corti is blockade of Ca²⁺ influx into the OHCs, which could prevent over-stressing of the cells that can lead to cell death through metabolic disruption. This protection of the OHCs would be more likely than prevention of synaptopathy to reduce PTS.

For NIHL, deficiency in Mg²⁺ has been shown to increase susceptibility from long-duration noise exposures (Ising et al., 1982). Similar results were found with Mg²⁺ deficiency and susceptibility to short-duration, high-level impulse noise exposures (Scheibe et al., 2000). Guinea pigs were fed a low Mg²⁺ diet, and then were either given Mg²⁺ supplementation in the water or were maintained on the low Mg²⁺ condition with regular tap water. The guinea pigs were then exposed to a single shot impulse noise from a modified 2-mm caliber pistol placed 24 cm from a speculum inserted into the ear canal. The recorded level was 187 dB pSPL. The low-Mg²⁺ group developed significantly higher CTS measured two hours after the exposure. However, by the measurement one week after the exposure, there were no differences between treatment groups' threshold shifts. This may be attributable to a floor effect, in that the animals had mean threshold shifts of 0–5 dB at that test point (Scheibe et al., 2000). In the same study, 150 and 167 dB pSPL impulses were generated digitally and delivered via a speaker connected to a speculum in the external ear canal. The 150 dB pSPL impulses were delivered at 1 impulse/s for 1000 impulses. Once again, CTS was detected at two hours, but by one week, mean threshold shifts were ∼0–7 dB, and no differences were detected between the low- Mg²⁺ animals and those with supplementation. The 167 dB pSPL impulses were also delivered 1/s, but a total of 2280 impulses were delivered. Even though this was a unilateral exposure paradigm, the noise dose was strong enough to evoke ∼10 dB CTS in the contralateral ears at the two-hour test time. At one week, both the low-Mg²⁺ and the supplemented-Mg²⁺ animals had threshold shifts greater than 10 dB. However, the low-Mg²⁺ animals' threshold shifts were ∼15–20 dB higher than the supplemented-Mg²⁺ animals, indicating the Mg²⁺ deficiency can increase susceptibility to injury from impulse noise (Scheibe et al., 2000). The report lists the one-week test time as “PTS,” but it is unclear if any additional threshold recovery might have been measured at later test times. Similarly, in the damage phase time window four days after exposure to 15 (1/s) gunshots of 176 dB pSPL, lipid peroxidation products in the guinea pig cochlea and physiological auditory brainstem response (ABR) thresholds were increased, both of which negatively correlated with cochlear levels of Mg²⁺ as assessed with energy dispersive X-ray analysis (Xiong et al., 2013).

Mg²⁺ supplementation has also been explored as a rescue agent after impulse noise exposure. Guinea pigs were lightly anesthetized and then exposed to blank gunfire from a FAMAS (French acronym that translates to “Assault Rifle from the Saint-Étienne Weapon Factory”) F1 assault rifle. This impulse exposure was particularly relevant to the French military population because the FAMAS F1 rifle was the service-issued rifle for the French military for several years. The guinea pigs were placed 60 cm from the rifle to create a 170 dB pSPL impulse or 30 cm away to create a 176 dB pSPL impulse. Animals were exposed to three shots of either the 170 or 176 dB pSPL impulses. The rescue-treated animals received daily sub-cutaneous injections of 3.5 mg/kg MgSO₄ for three days after the noise, as well as supplemental MgCl₂ in their drinking water. The rescue treatments began one hour after the noise exposures. Interestingly, the animals exposed to the 170 dB pSPL impulses showed a reduced CTS from the Mg²⁺ supplementation at 48-h and 7-day test times, but no differences in PTS measured at 14 days post noise. The Mg²⁺ supplementation did not induce any rescue effect from threshold shift from the 176 dB pSPL exposure. This may be attributable to a lack of TTS injury. The animals' threshold shifts were ∼25–40 dB at 24 h post noise and stayed fairly stable out to the final test at two weeks (Sendowski et al., 2006). The lack of recovery indicates that the animals essentially presented with PTS at the 24-h test point. Since Mg²⁺ has shown to have a significant impact on TTS pathology, possibly through the prevention of glutamate excitotoxicity (Sendowski, 2006), there may not have been an opportunity for Mg²⁺ to make an impact since there was little TTS pathology to rescue. It is also possible that the lack of TTS indicated a predominant mechanical injury to the cochlea that manifested very quickly after the noise and offered no avenue for pharmaceutical protection.

To further explore the rescue effect of Mg²⁺ after the series of 170 dB pSPL FAMAS F1 rifle impulses, the Sendowski research group repeated the experiment in 2009 using a longer rescue period of Mg²⁺ treatment. Again, beginning one hour after the noise exposure, the sub-cutaneous injections of 3.5 mg/kg MgSO₄ for three days were repeated, but the supplemental MgCl₂ in the drinking water was given for either 7 or 30 days to test extending the rescue treatment period. Hearing thresholds from the impulse noise showed a significant CTS at 20 min post noise. The thresholds recovered substantially in all groups by Day 7 post noise but continued to recover between Day 7 and Day 30 in the animals given the supplemental MgCl₂ in the drinking water for 30 days. Intriguingly, the distortion product optoacoustic emissions (DPOAE) data indicated a different damage pattern in which amplitudes were depressed at 20 min post noise, but instead of recovering, worsened by Day 7 and beyond. The animals with Mg²⁺ supplementation showed a greater preservation of DPOAE amplitudes in the higher frequencies, indicating retention of OHC function. This corresponded with lower OHC losses in the group with 1-month Mg²⁺ supplementation, compared to controls (Abaamrane et al., 2009).

Overall, Mg²⁺ supplementation exerted a protective effect against impulse noise, as demonstrated by higher susceptibility to impulse NIHL in animals with low Mg²⁺ diets. The heightened susceptibility was offset by supplementation of Mg²⁺ in those animals. Further, in a rescue paradigm, Mg²⁺ supplementation promoted better recovery of ABR thresholds and hair cell survival than control animals.

Noise, both impulse and continuous, has been shown to increase reactive oxygen species (ROS) in the cochlea (Yamane et al., 1995; Ohlemiller et al., 1999; Ohinata et al., 2000; Yamashita et al., 2004; Shi et al., 2007). Antioxidants can be defined as, “any substance that delays, prevents, or removes oxidative damage to a target molecule” (Halliwell and Gutteridge, 2015). Antioxidant compounds have a lengthy history of exploration as ototoprotective compounds against impulse NIHL. Donald Henderson's research team studied antioxidants and impulse NIHL in chinchillas, using an impulse noise that simulated the U.S. Army's M-16A1 rifle utilizing a 5.52 caliber round (Henselman et al., 1994) that was delivered to awake animals in a restraint tube. A time domain representation of the impulses, as a pair, are presented in Fig. 1. Hight et al. (2003) exposed the chinchillas to 50 pairs of the impulses (50 ms between the first and second impulse in each pair) at a rate of 1 pair/s at a level of 145 dB pSPL with the speaker positioned at the animals' ear level at 0° azimuth. One ear in each animal was pre-treated with glutathione monoethyl ester, surgically delivered onto the round window 40 min before the noise, with the goal of increasing glutathione (GSH) levels in the cochlea, and thus rendering the ears more resistant to NIHL. GSH is a potent antioxidant that scavenges some of the most potentially toxic ROS, hydrogen peroxide, the hydroxyl radical, and peroxynitrite. The control ears sustained ∼35–50 dB PTS across the 0.5–16 kHz tested range. The glutathione monoethyl ester compound at a concentration of 50 mM significantly reduced PTS and OHC loss from the impulse noise, although the 100 mM concentration exerted no protective effect, and the 150 mM concentration induced OHC protection, but no significant protection from PTS (Hight et al., 2003).

The same impulses were also used in chinchilla experiments of NIHL protection with n-acetyl, l-cysteine (NAC). NAC acts as a free radical scavenger and increases tissue levels of GSH (Zafarullah et al., 2003). In the experiments, the impulses were presented in 75 pairs (1 pair/s) at a level of 155 dB pSPL (Kopke et al., 2005; Kopke et al., 2007; Bielefeld et al., 2007). With those exposures, the PTS did not increase over the 100 pairs of 145 dB pSPL impulses used in Hight et al. (2003), and were in fact slightly lower in the low frequencies. In all experiments, NAC was given twice daily by intra-peritoneal injection starting two days before the noise and finishing two days after. The drug effectively reduced PTS and OHC losses when given at 325 mg/kg (Kopke et al., 2005) and also reduced PTS when given at lower doses of 50 or 100 mg/kg (Bielefeld et al., 2007). The protective effect of NAC against impulse NIHL has been demonstrated in the rat as well as the chinchilla. Sprague-Dawley rats were exposed to 50 impulses at 160 dB pSPL. NAC-treated rats were given 350 mg/kg NAC in saline for a total of either five (one day before noise, one hour before, immediately after, three hours after, and one day after) or three (one hour before, immediately after, three hours after) doses. The impulse noise exposure caused ∼45–65 dB CTS that recovered to ∼15–25 dB PTS in the control animals. The rats treated with the three doses of NAC showed essentially complete recovery of hearing with no PTS, nor any inner hair cell (IHC) or OHC loss. However, the rats given the five-dose regimen of NAC injections showed less recovery from the CTS than the controls, and finished with larger PTS, which was potentially attributable to much higher IHC loss (20%–50% across the length of the basilar membrane) than the controls (<15%) (Duan et al., 2004). The investigators did assess NAC toxicity to the cochlea without noise and found no evidence of ototoxicity. Thus, the results create a complex picture of how NAC affects recovery of threshold after impulse noise, with the particularly surprising target of IHC mortality as the determining factor.

Astragaloside IV, a traditional Chinese herb with antioxidant properties, has been tested in guinea pigs as an otoprotectant against impulse noise. Astragaloside IV has unspecified mechanisms of action, but is believed to exert antioxidant effects. Guinea pigs were noise exposed to 15 shots from a 7.62 mm Chinese Army 81–1 assault rifle at a distance of 35 cm, creating an exposure level of 176 dB pSPL. Threshold shift measurements were taken 72 h after the noise, and therefore likely reflect a CTS with both TTS and PTS components, although the relative distribution of the two components cannot be determined. Threshold shifts at 72 h in the control animals were 20–30 dB for clicks and 8–32 kHz tone bursts. The animals treated with astragaloside IV (200 mg/kg/day for three days by intragastric administration) showed lower threshold shifts of 10–15 dB compared to control animals treated with saline (Xiong et al., 2011). The threshold shift findings were replicated in a later experiment that explored mechanisms and demonstrated reduced lipid peroxidation activity, as indexed by lower 4-HNE staining, in the cochleae that had been treated with astragaloside IV (Xiong et al., 2012a). A similar protective effect of 10–15 dB against the same impulse noise was detected with the herbal form of astragaloside IV, radix astragali, delivered intragastrically at 6000 mg/kg/day (Xiong et al., 2012b).

Antioxidants have been shown to protect against NIHL in a number of different delivery protocols. What is the most striking about glutathione monotethyl ester and the NAC studies is the tendency for the compounds to exert otoprotective effects at certain doses or schedules, but then those effects are not present at higher doses or more frequent delivery schedules. Ultimately, there is clear proof of concept that antioxidants can improve recovery from impulse noise exposure, but further research is still needed to elucidate the optimal delivery system and dosing schedule.

Immune cells invade into cochlear tissue that has been exposed intense noise (Tornabene et al., 2006; Frye et al., 2018). An increased presence of circulating leukocytes in the cochlea has been documented in the period of several days after a continuous noise that induces significant PTS and hair cell loss (Tornabene et al., 2006). Further, leukocyte response in the cochlea to intrathecal injection of keyhole limpet hemocyanin, a large antigen molecule, was sensitized by previous exposure to high-level continuous noise (Miyao et al., 2008). The supporting cells in the organ of Corti themselves express immune-related genes in response to damaging continuous noise (Cai et al., 2014). Macrophages have been detected near the scala tympani side of the basilar membrane, and after intense noise exposure, monocytes migrate into the region and transform into macrophages. The macrophages at the base of the cochlea, where the bulk of cell damage occurred, presented an increased level of antigen protein expression (Yang et al., 2015). It remains unclear what role inflammation and the immune cells play in the progression of cochlear injury after impulse noise, as they may have beneficial or deleterious effects on the cells of the organ of Corti. This has been addressed indirectly with impulse noise otoprotection using glucocorticoid compounds. Glucocorticoids are used across a variety of applications due to their anti-inflammatory and immunosuppressive actions. They act to inhibit pro-inflammatory molecules, including cytokines, chemokines, arachidonic acid metabolites, and adhesion molecules (reviewed by van der Velden, 1998). Glucocorticoids have been used extensively to treat idiopathic sudden sensorineural hearing losses (Chandrasekhar, 2001; Kopke et al., 2001; Gouveris et al., 2005) and tested to minimize cochlear implant electrode insertion trauma (James et al., 2008; Eastwood et al., 2010; Van de Water et al., 2010; Connolly et al., 2011).

The Sendowski research group explored the glucocorticoid compound, methylprednisolone (MP), against impulse noise (Sendowski et al., 2006). They once again used the 170 dB pSPL impulse exposure of three blank shots delivered from FAMAS F1 rifle. MP was delivered to the cochlea after noise. Surgically-implanted osmotic pumps were used to infuse MP in artificial perilymph into scala tympani of guinea pigs at a concentration of 300 μM over a seven-day period. The treated ears showed no differences in threshold shift at 20 min after the noise but did show significantly lower threshold shifts at 48 h post noise. However, at seven days after noise, the treated ears only had significantly lower threshold shifts at one tested frequency. By 14 days after noise, the differences were small and not statistically significant. Interestingly, the treated ears showed much lower OHC and IHC losses that were spread over a much smaller cochlear region than the untreated ears, this despite the lack of difference in PTS at Day 14. So overall, the MP treatment decreased CTS but not PTS, and decreased hair cell loss (Sendowski et al., 2006). This presents some intriguing possible explanations. One is that the hair cells that survived in the ears with MP treatment were not fully functional, and therefore the hair cell cochleograms did not provide a good index of the otoprotective effectiveness of the compound. Further, the threshold shift detected in both sets of ears may have been influenced by injury to other structures, such as the lateral wall or the IHC-spiral ganglion cell afferent pathway.

The rescue effects of both intra-muscular and intra-tympanic delivery of MP were compared in guinea pigs exposed to 60 impulses of 0.5 ms duration at level of 165 dB pSPL, presented with an inter-impulse interval of 2 s (Zhou et al., 2009). The intra-muscular course of MP consisted of four doses of 40 mg/kg delivered over the 48-h period after the noise, beginning at one hour post noise. One group received a single injection at one hour post noise. Another group received four injections over 48 h, beginning at one hour post noise. A third group received the four injections over 48 h but beginning at seven days after noise. Within one hour after the noise, mean threshold shifts were 60–65 dB in the 2–16 kHz range. PTS at four weeks post noise settled at 25–50 dB, depending on the frequency and condition. The ABR thresholds are displayed in Fig. 2. The animals treated with MP systemically by intra-muscular injection had significantly lower thresholds at four weeks than those treated with saline. Similarly, the animals treated with MP by intra-tympanic injection starting one hour after the noise both had lower thresholds than the saline-treated animals. However, giving the four intra-tympanic doses did not offer any additional benefit over the single dose given at one hour post noise. Also, beginning the treatment at seven days after noise eliminated any otoprotective effect, as those animals had thresholds equal to the controls. The PTS results were consistent with hair cell loss data, with the MP-treated groups that were treated starting at one hour post noise retained more hair cells after noise than the control group (Zhou et al., 2009). Overall, the results are consistent with the general finding of otoprotection from MP. However, Zhou et al. (2009) found otoprotection for both hair cell loss and PTS, unlike the similar experiment that used fewer impulses at a higher level (Sendowski et al., 2006) and found that MP protects against hair cell loss, but not PTS.

Rescue from impulse noise with local infusion of MP or prednisolone across the round window after impulse noise was also tested in guinea pigs (Müller et al., 2017). In the experiment, the impulses were created from 500 ms white noise bandpass filtered from 0.25 to 4 kHz and then multiplied by a sin² ramp. The measured dB pSPL was 165, with rms values of 140–144 dB SPL. Animals were exposed to either 15, 30, 45, 60, or 120 impulses. The authors found that 15- and 30-impulse exposures induced significant variability with complete recovery in some animals and large PTS in others. The 60- and 120-impulse exposures induced ceiling-level threshold shifts with no recordable compound action potential thresholds and massive loss of hair cells. Therefore, the investigators focused their otoprotection experiments on the 45-impulse condition, as it induced a CTS. Immediately after the exposure, mean threshold shifts were 50–60 dB. There was recovery of 10–20 dB by 24 h after noise in the frequencies below 12 kHz. An additional recovery of ∼5–15 dB occurred between 24 h and two weeks post noise. PTS, as calculated by threshold shift at two weeks, was approximately 30 dB across the range of 0.5–12 kHz. Cochlear infusion of 2.5 mg/ml prednisolone significantly reduced the PTS at 1 kHz and below, with a reduction of ∼20–30 dB. With a higher dose of 25 mg/ml, PTS was significantly reduced at 2 kHz and below (∼15–30 dB protection) and above 10 kHz (∼35–50 dB protection). However, it was noted that the ears treated with the prednisolone had a hair cell lesion in the hook region of the extreme base of the cochlea that did not occur in other noise-exposed groups. With MP, a low-frequency protective effect (2 kHz and below) was detected, with reductions of PTS ∼15–30 dB, at an MP dose of 12.5 mg/ml. The higher tested dose, 50 mg/ml, provided no protection (Müller et al., 2017). The results indicate that prednisolone may offer a more effective pharmaceutical otoprotection strategy than MP, even though MP has been studied more extensively against impulse noise.

Further investigation of the optimal route for delivery of the glucocorticoid compounds is warranted, since it has been shown across multiple studies to offer effective otoprotection in a rescue paradigm. Mechanisms of protection also need to be explored. Glucocorticoids' anti-inflammatory and immunosuppressive properties are well-known, but possible anti-apoptotic mechanisms have also been offered as an explanation for their otoprotective effects against noise.

Apoptosis is one of the two main forms of cell death in the cochlea, along with necrosis. Other cell death mechanisms in the noise-exposed cochlea have been identified (Bohne et al., 2007), but a large portion of the hair cell lesion that appears after an intense noise exposure is a combination of necrosis and apoptosis (Hu et al., 2000; Hu et al., 2002; Yang et al., 2004). Apoptosis is an active form of cell death in which a series of intracellular signals are engaged that result in disassembly of the cell without significant spillage of the intracellular contents into the extra-cellular space. Necrosis is a passive process that results from gross injury to the cell and can result in spillage of the intracellular contents out through the plasma membrane of the cell. Apoptosis is visibly distinct from necrosis in terms of the cell death process and its inflammatory consequences (Kerr et al., 1972). While it was initially unclear if apoptosis was occurring as a primary form of cell death in the noise-exposed cochlea or if it was occurring secondarily as a consequence of necrosis, it has become clear that impulse noise can induce primary apoptosis within minutes of an impulse exposure (Hu et al., 2006; Hu and Zheng, 2008). Because hair cell apoptosis from metabolic injuries occurs over a period of several hours or days, the rapid induction of apoptosis after impulse exposure can be attributed to the mechanical injuries created by the high level of the input sound (Hu et al., 2006).

With the primacy of impulse noise-induced apoptosis as a mechanism of cochlear injury and hearing loss, several compounds have been tested to block apoptosis signaling and promote cell survival after impulse noise. Harris et al. (2005) tested a series of Src-protein tyrosine kinase (PTK) inhibitors against NIHL in the chinchilla. Src was targeted because of its potential involvement in several mechanisms that lead to noise-induced OHC death, including mechanical injury-induced apoptosis and generation of ROS through the NADPH oxidase pathway. An NADPH oxidase inhibitor in the cochlea has previously been shown to reduce impulse NIHL in the chinchilla (Bielefeld, 2013). Harris et al. (2005) determined that the most effective compound against continuous noise was KX1–004, which is a non-ATP competitive Src-PTK inhibitor that was hypothesized to prevent apoptosis and the generation of ROS in the organ of Corti. The investigators delivered the compound surgically to the round window in chinchillas, and then exposed them to 75 pairs of the impulses described above from chinchilla experiments (Henselman et al., 1994; Hight et al., 2003; Kopke et al., 2005; Bielefeld et al., 2007; Bielefeld, 2013). The control subjects sustained 40–50 dB threshold shifts (as measured with implanted inferior colliculus electrodes) from 0.5–8 kHz at 24 h after noise. They recovered to 15–25 dB PTS. The loss was associated with 20%–40% OHC losses at the region of the basilar membrane 40%–90% from the apex. This lesion size was considerably smaller than that which was detected in a previous experiment using the 75 pairs 155 dB pSPL impulses (Kopke et al., 2005), which found 70%–90% OHC losses in the base of the cochlea. This discrepancy speaks to the inter-subject variability in impulse noise-induced cochlear injury. The animals pre-treated with KX1–004 on the round window sustained 15–25 dB threshold shifts at 24 h, and ∼5 dB PTS, with <20% OHC loss in the 40%–80% region from the apex.

The impulse noise exposure and local delivery of KX1–004 was repeated by Fetoni et al. (2014) to help determine the mechanisms through the Src inhibitor might be acting, and to compare KX1–004's otoprotective effect to that of pifithrin-α. Pifithrin-α is an inhibitor of p53, which is a tumor suppressor that is involved in apoptosis of neural cells during development and in cases of acquired injury (reviewed in Miller et al., 2000). Intense noise exposure increases expression of genes related to p53 either upstream or downstream (Hu et al., 2009), and inhibition of p53 had been shown to provide otoprotection against cisplatin ototoxicity (Zhang et al., 2003). Both KX1–004 and pifithrin-α were delivered locally onto the round window. At 24-h after the noise, the KX1–004 afforded ∼15–25 dB of protection from 2 to 8 kHz, and the pifithrin-α provided ∼10–20 dB of protection. In order to capture the active mechanisms of the noise-induced cell death, the cochleae were sampled immediately after the 24-h test, and no further hearing tests were performed. At 24 h post noise, a large lesion of damaged OHCs was detected, with ∼40%–70% of the OHCs showing damage in the region 40%–90% from the apex. That finding was supported by observations of a large number of shrunken, condensed OHC nuclei at 4 and 24 h after the noise, consistent with evidence of active apoptosis. There was also a high level of p53 activity at 4 h after noise in OHCs whose nuclei showed early evidence of apoptosis. The Src inhibitor and p53 inhibitor both reduced the OHC damage lesion to ∼10%–30%, and reduced the level of the activated form of phospho-p53 (Ser) in the OHCs (representative images shown in Fig. 3) (Fetoni et al., 2014).

The otoprotective effect of an alternative Src-PTK inhibitor, KX2–329, was also tested via round window delivery in chinchillas before the 75 pairs of 155 dB pSPL impulses (Bielefeld et al., 2011). The impulse noise exposure induced 40–50 dB threshold shifts (again measured with implanted inferior colliculus electrodes) at 24 h after noise, and 18–30 dB PTS at 28 days in the control ears. Pre-treatment with KX2-329 30 min before the noise reduced the PTS to 5–15 dB. KX2-329 is a biaryl-based non-ATP competitive Src-PTK inhibitor that inhibits Src-driven cell growth. KX1-004 is also non-ATP competitive, but is an indole-based compound rather than biaryl-based. KX2-329 offered the additional effect of acting as a tubulin polymerization inhibitor. OHCs have high levels of tyrosinated tubulin, which is unusual in that tyrosinated tubulin is typically found in cell populations that undergo frequent cell division (Slepecky et al., 1995). The tyrosinated tubulin may be involved in the OHCs' electromotility (Slepecky and Ulfendahl, 1992). Noise alters the spatial distribution of tubulin in the OHCs (Liberman, 1987; Raphael et al., 1993; Spongr et al., 1998), but the impact of inhibition of tubulin polymerization on NIHL is unknown. Therefore, it is unclear if KX2-329 acted any differently from KX1–004 in offering impulse noise otoprotection.

Beyond Src, c-Jun N-terminal kinase (JNK) has also been a target of intervention in the apoptosis signaling pathway to prevent impulse NIHL. JNK is part of the stress-activated protein kinase pathways, and is a key molecule in multiple apoptosis pathways (Basu and Kolesnick, 1998). JNK inhibitors have been shown to exert otoprotective effects against continuous noise exposures (Pirvola et al., 2000; Wang et al., 2003; Wang et al., 2007), and JNK activation has been detected in cochlear cells three hours after an acoustic blast exposure of 194 kPa (Murai et al., 2008) generated by detonating plastic explosive at one end of a 1-meter shock tube (Chan et al., 1998). A JNK inhibitor, AM-111 (also known as D-JNKI), was used in a rescue paradigm in chinchillas after the aforementioned 75 pairs of 155 dB pSPL impulses. The compound was delivered: (1) locally into the cochlea through an implanted osmotic pump that was placed either one or four hours after the noise (100 μM concentration), (2) locally through the round window with hyaluronic acid gel formulation (100 μM concentration), or (3) systemically through conventional intra-peritoneal injections (3.3 mg/kg) at one or four hours post noise. The noise created 25–40 dB PTS in the range 4–8 kHz. Each group treated with AM-111 showed lower PTSs, with ∼15 dB of protection with the osmotic pump, ∼20 dB of protection from the gel formulation, and ∼10 dB of protection from the systemic injections. There was not a drastic difference in any of the three delivery systems between beginning the treatments at one hour versus four hours (Coleman et al., 2007).

Like Mg²⁺ supplementation, antioxidant intervention, and glucocorticoid treatment, intervening in the apoptosis pathway has demonstrated otoprotection from impulse NIHL. Ultimately, the exact mechanisms through which Src or JNK inhibitors exert their protective effects is unclear because the apoptosis pathways are so complex and interconnected, and because impulse noise causes a cascade of mechanical and metabolic effects that themselves are interconnected. The challenge with these compounds going forward will be to maximize the otoprotective effects without inducing side effects in systemic delivery. Further understanding of the mechanisms through which the compounds exert their effects is a crucial step toward that goal.

Short-duration, high-level sound impulses are a prevalent source of hazardous noise. They present unique patterns of cochlear damage, and a unique time course for injury and hearing loss. Because of the mechanical component of the damage, there are limits on the amount of protection that can be expected from pharmaceutical intervention. Indeed, complete protection from PTS was not seen in the experiments reviewed above, nor was complete protection of hair cells. Inter-subject variability tended to be high in these experiments, which is to be expected with the short duration of the impulses and the combinations of mechanical and metabolic cochlear injuries that can result. Also noteworthy was the relatively small number of impulse noise experiments that have been conducted with pharmaceutical otoprotection, compared to the number of experiments that have used continuous exposures at various sound pressure levels and durations. With the ongoing public health concerns about noise injury, and the potential with impulse noise to promote significant threshold and cochlear recovery in a rescue paradigm, further exploration of potential otoprotection strategies is warranted.