Nitric oxide in glutamate-induced compound action potential threshold shifts

Results: The control group showed no CAP threshold elevations ( p < 0.05) when APS was perfused after systemic pre-treatment with 7-NI. GA perfusion alone caused significant elevation ( p < 0.05) of the mean cochlear CAP threshold (25 dB SPL ± 5.8 dB to 78 dB SPL ± 19.5 dB). The CAP threshold elevation was prevented ( p < 0.05) when the animals were pretreated with 7-NI before GA perfusion (24 dB SPL ± 4.2 dB to 27 dB SPL ± 6.7 dB). Conclusion: NO mediates excitotoxicity when the cochlea is perfused with l -glutamate. Abbreviations EAA excitatory amino acid Ca +2 calcium ions nNOS neuronal nitric oxide synthase NO nitric oxide CAP compound action potential NMDA N -methyl- d -aspartate AMPA quisqualate/alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid NOS nitric oxide synthase 7-NI 7-nitroindazole CNS central nervous system i.p. intraperitoneal MAP mean arterial blood pressure APS artificial perilymph solution GA glutamic acid/glutamate/ l -glutamate L-NAME l -N G -nitroarginine methyl ester cGMP cyclic guanosine monophosphate ONOO − peroxynitrate anion cAMP cyclic adenosine monophosphate Keywords Cochlea Excitatory amino acids Glutamate Nitric oxide 7-Nitroindazole Excitotoxicity 1 Introduction The excitatory amino acid (EAA) l -glutamate is thought to be the primary neurotransmitter acting at the synapse between cochlear hair cell and the afferent auditory nerve dendrites ( Eybalin, 1993 ). Over-stimulation of glutamate receptors can lead to neuronal damage and disruption of cell electrophysiology ( Garthwaite, 1991 ), a process referred to as excitotoxicity ( Rothman and Olney, 1995 ). Part of the excitotoxic pathway involves massive intracellular influxes of calcium (Ca +2 ) ions, resulting in the formation of calcium–calmodulin complexes and activation of neuronal nitric oxide synthase (nNOS), which catalyzes the formation of nitric oxide (NO) and peroxynitrate, ultimately leading to cell death. In the cochlea, excessive glutamate and aspartate have been shown to cause ototoxicity, as demonstrated by elevated cochlear compound action potential (CAP) thresholds ( Bobbin and Thompson, 1978 ). The EAAs in the cochlea function primarily via three types of glutamate receptors: (1) N -methyl- d -aspartate (NMDA) ( [Eybalin, 1993 ) which binds l -glutamate (2) kainate ( Bledsoe et al., 1981; Jenison et al., 1986 ) (3) quisqualate/alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptor subtypes ( Jenison et al., 1986 ). Activation of NMDA and kainate receptors are most commonly associated with excitotoxicity ( Lefebvre et al., 1991 ). Nitric oxide is enzymatically produced from arginine via nitric oxide synthase (NOS) and participates in numerous functions as a mediator of cellular responses within the central nervous system, including excitotoxicity ( Dawson et al., 1991 ). NO specifically mediates EAA-induced cyclic GMP formation in post-synaptic structures ( Garthwaite, 1991 ). Studies have shown that NOS is an active enzyme in cochlear spiral ganglion cells ( Zdanski et al., 1994 ). Kainic acid, a conformationally restricted glutamate analog, is excitotoxic in the cochlea ( Pujol et al., 1985 ) and elevates CAP thresholds ( Bledsoe et al., 1981 ). Furthermore, 7-nitroindazole (7-NI), a competitive inhibitor specific for nNOS ( Moore et al., 1993 ) significantly prevents CAP threshold elevations when administered systemically prior to kainic acid cochlear perfusion ( Johnson et al., 1998 ). Based on these studies, we postulate that the function of NO in cochlea neuronal tissue is analogous to its function in the central nervous system (CNS), and that NO participates in EAA-induced excitotoxicity. This study investigates the role of NO as a neurotransmitter in the gerbil cochlea and the effects of (7-NI) on CAP threshold elevations induced by l -glutamate, an agonist at the NMDA glutamate receptor subtype. The goal of this study was to further elucidate the role of NO in cochlear excitotoxicity. 2 Materials and methods 2.1 Surgical procedure Mongolian gerbils (50–80 g) were anesthetized with an intraperitoneal (i.p.) injection of α-chloralose (100 mg/kg) and urethane (425 mg/kg) to provide surgical anesthesia, allow non-ventilator dependent oxygenation, and maintain mean arterial blood pressure (MAP) in the range of 80 ± 10 mmHg. A maintenance dose, 25% of the initial dose, was injected i.p. every 60 min to maintain constant blood levels. Each animal was placed supine in a head holder apparatus and body temperature was maintained between 37° and 38 °C with a thermo-coupled, controlled heating pad. Surgery was performed only on animals in deep anesthesia, with regular breathing and adequate MAP. Anesthetics and non-specific inhibitors of NOS potentially affect blood pressure; therefore, blood pressure was monitored throughout the experiments using an intra-arterial cannula, which was placed in the carotid artery and connected to a blood pressure monitor (Cardiovascular Analyzer CVA-1, Buxco Electronics Inc., Sharon, CT, USA). The intra-arterial cannula was attached to a micro-infusion pump (Model A-99, Razel Scientific Instruments, Inc., Stamford, Connecticut) for continuous infusion of Krebs Ringer solution at a rate of 80 μl/min/kg. A ventral midline incision was made in the neck and a tracheostomy was performed. The right pinna was removed to expose the auditory meatus. The tympanic bulla, contralateral to the intra-carotid cannula, was opened carefully to expose the middle ear space while preserving the tympanic membrane and middle ear ossicles. 2.2 Cochlear perfusion A circular hole, 50 μm in diameter, was drilled in the scala tympani at the first, basal-most turn of the cochlea for infusion, and a second hole, 40 μm in diameter, was similarly drilled at the apex to allow escape of the perfusate. This allowed perfusion of the entire scala tympani, from base to apex ( Johnson et al., 1998 ). The position of the apical hole is critical, as the scala tympani and scala media are in closer approximation at the apex than at the base of the cochlea. A breach into the scala media would cause mixing of endolymph and perilymph, with subsequent disappearance of cochlear action potentials ( Nuttall et al., 1977 ). The holes were created by carefully rotating a hand-held drill bit. Glass micropipettes, to deliver the perfusate, were pulled on a vertical pipette puller (Model 700C, David Kopf Instruments). A 40 μm orifice was created by breaking the tip, and a small, silicone ball was fashioned a few micrometers above the tip (1–2577 conformal coating, Dow Corning, Midland, Michigan) to afford a tight seal. The pipette was connected via polyurethane tubing to a syringe and micro-infusion pump (Model #943, Harvard Apparatus Co., Millis, MA, USA). After evacuation of air bubbles, the pipette was lowered into the hole at the basal turn such that the silicone ball provided a tight seal between the pipette and the cochlea. All solutions were perfused at a rate 1.4 μl/min, a rate sufficiently slow enough to minimize mechanical interference on the cochlea. Perfusates were all 22 °C when infused to prevent confounding effects of cochlear cooling. Excess fluid was suctioned from the middle ear cavity with plastic tubing with care taken to avoid direct suction at and around both perfusion holes. A small amount of cotton was placed near, but not directly touching, the apical outflow hole to serve as a wick for excess perfusate. Excess perfusate was suctioned periodically from the cotton. 2.3 Solutions The artificial perilymph solution (APS) was composed of 130 mM NaCl, 3.00 mM KCl, 1.25 mM KH 2 PO 4 , 10.0 mM glucose, 20.0 mM NaHCO 3 , 1.20 mM CaCl 2 · H 2 O, and 1.30 mM MgCl 2 · 6H 2 O dissolved in deionized water. Prior to each experiment, APS was filtered, maintained at 300 mOsm, and the pH was adjusted to 7.2–7.4 by bubbling 5% CO 2 through the solutions. Glutamic acid (GA) (Sigma) was freshly dissolved in APS prior to each experimental perfusion. Test solutions of 10 μM, 25 μM, 50 μM, and 100 μM were perfused, each for 15 min, to determine the concentration that yielded reliable and significant CAP threshold shifts compared to APS perfusion. A dose-response curve was established using this data. All dose-response challenges were performed in an isolated group of 7 gerbils ( n = 7), none of which were used in experiments involving 7-NI. A total dose of 50 mg/kg of 7-nitroindazole (7-NI) (Sigma) was freshly sonicated and suspended in 0.8 ml peanut oil (Sigma) prior to each experiment requiring i.p. 7-NI injection. Previous studies involving i.p. 7-NI have successfully used similar dosages ( Moore et al., 1993; Johnson et al., 1998 ). 2.4 Experimental groups Anesthetized Mongolian gerbils were divided into three groups. Group 1, ( N = 5), the APS/7-NI control group, underwent cochlear perfusion with APS for 10 min, received an i.p. injection of 7-NI (50 mg/kg in 0.8 ml peanut oil), followed 60 min later by cochlear perfusion with APS for 60 min. Group 2, ( N = 7), the GA group, underwent cochlear perfusion with APS for 10 min, then perfusion of 50 μM l -glutamate for 60 min. Group 3, ( N = 7), the 7-NI/GA group, underwent cochlear perfusion with APS for 10 min, received an i.p. injection of 7-NI, followed later by cochlear perfusion with 50 μM l -glutamate for 60 min. In each animal from each group, CAP thresholds were measured before and after drilling perfusion holes, before and after initial 10-min APS perfusion, and at five 15-min intervals following the 60-min perfusion of either APS or GA. The CAP threshold measurements before drilling and after the initial 10-min APS perfusion were compared to establish the integrity of the scala tympani after drilling. Animals were used only if they exhibited threshold shifts less than 10 dB after drilling and after the initial 10 min of APS perfusion. The 7-NI was administered i.p. prior to glutamate perfusion in an effort to block nNOS and thus block the toxic effect of glutamate. Cochlear hypo-perfusion may lead to threshold shifts; therefore, blood pressure levels were continuously monitored and recorded at 15-min intervals for all animals. 2.5 Cochlear compound action potential threshold recording Cochlear CAPs were recorded in response to pure tone-pip stimuli. The CAP threshold was used as an index of cochlear function in response to presented stimuli. CAPs were recorded through differential silver electrodes placed on the round window and apex of the exposed cochlea. Tone-pip stimuli were calibrated using a microphone probe (ER& C probe system, Etymotic Research, Elk Grove, Illinois). Stimuli were 5 μs 8 kHz tones with 0.5 μs rise/fall time, generated at a rate of 3 per second with a sampling rate of 0.5 μs (TDT System II hardware and software, Tucker-Davis Technologies, Gainesville, FL, USA). The intensity of the each tone-pip was adjusted such that they were presented initially at 85 dB SPL, and decreased to 5 dB SPL in 5-dB SPL decrements, through a speaker (Motorola Piezoelectric Tweeter) fitted to an earphone tube sealed to the gerbil’s right external auditory meatus. The action potentials were amplified (Model P15D, Grass Instruments, Quincy, MA, USA and Model AM502, Tektronicks Inc., Beaverton, OR, USA) and averaged using a PC-based system (TDT System II). At each stimulus intensity level, responses to 100 individual stimuli were averaged with 180° alternation of phase at each individual presentation to cancel cochlear microphonics. The CAP threshold was defined as the intensity level at which a response was discernable above the noise floor. For each animal, CAP thresholds recorded after initial APS perfusion were compared to CAP threshold recorded after subsequent experimental perfusions of experimental of control solutions. 2.6 Analysis CAP threshold and mean arterial blood pressure data were analyzed by one-way repeated measures ANOVA (for comparison of time-point values within each group) and one-way ANOVA (for comparison of mean time-point values between groups). A p value less than 0.05 was considered statistically significant. Graphs were constructed from the means with vertical bars representing the standard errors of the means. 2.7 Statement This study was conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals, and the Animal Welfare Act. The animal use protocol was approved by the Institutional Animal Care and Use Committee (IACUC) of the University of North Carolina. 3 Results The effect of various glutamate (GA) concentrations on CAP thresholds demonstrate significant threshold shifts were observed with 15-min cochlear perfusions of 25 μM, 50 μM, and 100 μM glutamate. A concentration of 50 μM GA consistently produced a significant threshold shift and was chosen for subsequent experiments ( Fig. 1 ). Neither the drilling of cochlear perfusion holes nor perfusion with APS alone caused significant changes in CAP thresholds in any experimental group ( Fig. 2 ). In Group 1 (7-NI/APS control group), 7-NI given 60 min prior to a second APS perfusion showed no significant CAP threshold shifts above its mean APS baseline. The mean threshold shifts of Group 1 for the 5 separate time intervals were calculated to be 3, 2, 1, 4 and 4 dB measured at 60, 75, 90, 105 and 120 min post-perfusion, respectively. In Group 2 (GA without 7-NI), perfusion of l -glutamate alone caused significant shifts ( p < 0.05) in the mean CAP thresholds. The threshold shifts of the 7 animals were averaged for each of the five 15-min time intervals. Thus, at 60, 75, 90, 105, and 120 min post-perfusion the mean threshold shifts of Group 2 were 51, 51, 51, 50 and 50 dB, respectively, above its mean APS baseline. In comparison, Group 3 (the 7-NI/GA group) showed no significant mean CAP threshold shifts compared with its mean APS baseline. The mean threshold shifts of Group 3 for each of the 5 time intervals were calculated to be 5, 4, 6, 6 and 7 dB at 60, 75, 90, 105, and 120 min post-perfusion, respectively. In addition to showing CAP threshold elevations significantly higher than its own baseline at each time interval, Group 2 (GA without 7-NI) demonstrated CAP thresholds significantly higher ( p < 0.05) than CAP thresholds of control Group 1 (7-NI/APS) at each 15-min interval. In comparison, animals pre-treated with 7-NI (Group 3) exhibited threshold shifts that were insignificant with regard to the control Group 1 thresholds at each 15-min interval ( Fig. 2 ). Averaged mean arterial pressure measurements among all groups ranged from 74.5 mmHg to 99.3 mmHg. Mean arterial pressure was not significantly affected by administration of either 7-NI or l -glutamate ( Fig. 3 ). 4 Discussion The results of this study demonstrate that elevation of cochlear compound action potential thresholds in gerbils due to cochlear perfusion of l -glutamate can be significantly blocked by intraperitoneal pre-treatment with 7-NI, a competitive inhibitor of nNOS, thus implicating NO as a mediator of glutamate-induced elevation of CAP thresholds. Because of its specificity for nNOS and lack of systemic side effects in animal models, 7-NI, not l -N G -nitroarginine methyl ester (L-NAME), was used as the NOS inhibitor in this series of experiments ( Martins-Pinge et al., 2007 ). This was evident in our experimental model, the results of which suggest that 7-NI is sufficiently specific for nNOS, precluding significant systemic side effects ( Moore et al., 1993 ) such as anoxia or hypoxia, and evident by insignificant alterations in mean arterial blood pressures throughout all experiments, including those involving i.p. 7-NI injection. Previous studies in this lab have shown that constitutive NOS is an active enzyme in spiral ganglion cells in the rat cochlea ( Zdanski et al., 1994 ). It has also been demonstrated that perfusing the cochlea with sodium nitroprusside (an NO donor) causes CAP threshold elevations, and that these elevations may be significantly blocked by inhibiting guanylate cyclase ( Dais et al., 1996 ). We have also shown that the threshold shifts caused by kainate perfusion are prevented by 7-NI ( Johnson et al., 1998 ). However in the present study the use of glutamate in vivo, rather than kainate, suggests physiologically relevant results. Kainate is a naturally occurring substance derived from algae and is a well established neurotoxin to the cochlea in vivo that causes a loss of spiral ganglion neurons under inner hair cells, a phenomenon observed in both ischemia- and noise-induced swelling ( Puel et al., 1994 ). Kainate induced damage to the spiral ganglion is prevented by pre-treatment of the cochlea with intra-cochlear perfusion of the AMPA receptor antagonists GYKI 53784 ( Ruel et al., 2007 ). Glutamate is the primary agonist of AMPA receptors, which studies demonstrate is the only glutamate receptor subtype responsible for fast excitatory synaptic transmission between inner hair cells and auditory nerve fibers ( Ruel et al., 2007 ); thus binding of the AMPA receptor in vivo may be the means of cochlea excitotoxicity. Although EAAs act on various receptor subtypes in the cochlea, including the NMDA ( Eybalin, 1993 ), the kainate ( Bledsoe et al., 1981; Jenison et al., 1986 ), and the quisqualate/AMPA subtypes ( Jenison et al., 1986 ), glutamate is considered the best candidate for mediating neurotransmission between cochlear inner hair cells and the afferent auditory nerve dendrites ( Eybalin, 1993 ). l -glutamate activates all glutamate receptor subtypes; however, most of the neuronal cyclic guanosine monophosphate (cGMP) response to glutamate is initiated by Ca +2 flux through the NMDA receptor ( Garthwaite, 1991 ). It has been shown that cochlear excitotoxicity is most commonly associated with over stimulation of NMDA and kainate receptors ( Lefebvre et al., 1991 ). The binding of EAAs to NMDA receptors leads to an intracellular influx of calcium ( Moncada et al., 1991 ), which may bind with calmodulin to form a complex that activates the neuronal isoform of NOS, thus generating NO from l -arginine ( Snyder and Bredt, 1992 ). What follows is a cascade of intracellular events, all induced by NO. One biochemical pathway for NO is its binding with the superoxide radical O 2 - to form the highly toxic peroxynitrate (ONOO − ) anion, leading to cell damage ( Beckman et al., 1990; Hogg et al., 1992 ). Another effect of NO is its activation of soluble guanylate cyclase, which stimulates the production of cGMP ( Dais et al., 1996; Schuman and Madison, 1994 ). Excess cGMP has been shown to cause multiple cellular alterations affecting neurotransmission, such as increasing spontaneous neuronal firing rates, stimulating further EAA release, and altering cyclic adenosine monophosphate (cAMP) levels ( Garthwaite, 1991 ). Excess cGMP precipitating these events may occur from either EAA over-stimulation and/or excess NO. Studies have shown that EAA over-stimulation in the CNS causes injury that is mediated by NO ( Dawson et al., 1991 ) and that injury caused by NO-induced EAA over-stimulation is associated with NO-induced elevation of cGMP ( Snyder and Bredt, 1992 ). Glutamate receptor agonists dose-dependently increase NO levels, whereas the NOS inhibitors, including 7-NI, inhibit NMDA-induced increases in NO, but not AMPA-induced increases in NO ( Yamada et al., 1996 ). Accordingly, our results suggest that excitotoxicity mediated events are via NMDA, rather than AMPA, activation by glutamate and subsequent release of NO via NOS. However, kainate is not an established agonist of NMDA receptors and results from previous studies, including our own, demonstrate kainate induced excitotoxicity in vivo. The mechanism by which kainate induces NO excitoxicity may be secondary to kainate receptor activation that ultimately results in glutamate release at the synaptic junction, leading to glutamate receptor activation, and NMDA mediated NO release. In addition to its role in neurotransmission, NO is also cytotoxic ( Garthwaite, 1991; Moncada et al., 1991 ), and has been shown to mediate toxic EAA over-stimulation in the CNS ( Dawson et al., 1991 ). Nitric oxide plays a significant role in ischemic cell death and selective neuronal NOS inhibitors have been found to protect against ischemic brain damage ( O’Neill et al., 1996 ). Because NO is clearly toxic, nitric oxide synthase inhibitors have been examined as possible neuroprotective agents ( Buisson et al., 1993 ). Given that EAA receptor over-stimulation causes neurotoxic changes in the cochlea ( Eybalin, 1993 ), and that CAP threshold elevations caused by kainic acid perfusion are significantly blocked by the NOS inhibitor 7-NI ( Johnson et al., 1998 ), it follows that NO may mediate excitotoxicity in the cochlea as it does in the CNS. Acknowledgements The research presented in this paper was funded and made possible by the Deafness Research Foundation Medical Student Otologic Fellowship and the Southern Medical Association. The technical assistance of Stephen Pulver, Scott Lin and John Grosse, PhD is greatly appreciated, as is the assistance of Dr. Paul Manis. References Beckman et al., 1990 J.S. Beckman T.W. Beckman J. Chen P.A. Marshall B.A. Freeman Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide Proc. Natl. Acad. Sci. USA 87 4 1990 1620 1624 Bledsoe et al., 1981 S.C. Bledsoe R.P. Bobbin D.M. Chihal Kainic acid, an evaluation of its action on cochlear potentials Hear. Res. 4 1 1981 109 120 Bobbin and Thompson, 1978 R.P. Bobbin M.H. Thompson Effects of putative transmitters on afferent cochlear transmission Ann. Otol. Rhinol. Laryngol. 87 2 Pt 1 1978 185 190 Buisson et al., 1993 A. Buisson I. Margaill J. Callebert M. Plotkine R.G. Boulu Mechanisms involved in the neuroprotective activity of a nitric oxide synthase inhibitor during focal cerebral ischemia J. Neurochem. 61 2 1993 690 696 Dais et al., 1996 C.G. Dais J. Prazma S.S. Ball H.C. Pillsbury Effect of sodium nitroprusside on compound action potential thresholds in the gerbil cochlea Hear. Res. 99 1–2 1996 1 6 Dawson et al., 1991 V.L. Dawson T.M. Dawson E.D. London D.S. Bredt S.H. Snyder Nitric oxide mediates glutamate toxicity in primary cortical cultures Proc. Natl. Acad. Sci. USA 88 14 1991 6368 6371 Eybalin, 1993 M. Eybalin Neurotransmitters and neuromodulators in the mammalian cochlea Physiol. Rev. 73 Apr 1993 309 373 Review Garthwaite, 1991 J. Garthwaite Glutamate, nitric oxide, and cell–cell signaling in the nervous system Trend Neurosci. 14 Feb 1991 60 67 Review Hogg et al., 1992 N. Hogg V.M. Darley-Usmar M.T. Wilson S. Moncada Production of hydroxyl radicals from the simultaneous generation of superoxide and nitric oxide Biochem. J. 281 Pt2 1992 419 424 Jenison et al., 1986 G.L. Jenison S. Winbery R.P. Bobbin Comparative actions of quisqualate and N -methyl- d -aspartate, excitatory amino acid agonists, on guinea-pig cochlear potentials Comp. Biochem. Physiol. C 84 2 1986 385 389 Johnson et al., 1998 K.L. Johnson V. Carrasco J. Prazma C.J. Zdanski W.F. Durland H.C. Pillsbury Role of nitric oxide in kainic acid-induced elevation of cochlear compound action potential thresholds Acta Otolaryngol. 118 5 1998 660 665 Lefebvre et al., 1991 P.P. Lefebvre T. Weber P. Leprince J.M. Rigo B. Rogister G. Moonen Kainate and NMDA toxicity for cultured developing and adult rat spiral ganglion neurons: further evidence for a glutamatergic excitatory neurotransmission at the inner hair cell synapse Brain Res. 555 1 1991 75 83 Martins-Pinge et al., 2007 M.C. Martins-Pinge M.R. Garcia D.B. Zoccal C.C. Crestani P. Pinge-Filho Differential influence of iNOS and nNOS inhibitors on rostral ventrolateral medullary mediated cardiovascular control in conscious rats Auton. Neurosci. 131 1–2 2007 65 69 Moncada et al., 1991 S. Moncada R.M. Palmer E.A. Higgs Nitric oxide: physiology, pathophysiology, and pharmacology Pharmacol. Rev. 43 2 1991 109 142 Moore et al., 1993 P.K. Moore R.C. Babbedge P. Wallace Z.A. Gaffen S.L. Hart 7-Nitro indazole, an inhibitor of nitric oxide synthase, exhibits anti-nociceptive activity in the mouse without increasing blood pressure Brit. J. Pharmocol. 108 2 1993 296 297 Nuttall et al., 1977 A.L. Nuttall D.M. Marques M. Lawrence Effects of perilymphatic perfusion with neomycin on the cochlear microphonic potential in the guinea pig Acta Otolaryngol. 83 5–6 1977 393 400 O’Neill et al., 1996 M.J. O’Neill C. Hicks M. Ward Neuroprotective effects of 7-nitroindazole in the gerbil model of global cerebral ischaemia Eur. J. Pharmacol. 310 2–3 1996 115 122 Puel et al., 1994 J.L. Puel R. Pujol F. Tribillac S. Ladrech M. Eybalin Excitatory amino acid antagonists protect cochlear auditory neurons from excitotoxicity J. Comp. Neurol. 341 2 1994 241 256 Pujol et al., 1985 R. Pujol M. Lenoir D. Robertson M. Eybalin B.M. Johnstone Kainic acid selectively alters auditory dendrites connected with cochlear inner hair cells Hear. Res. 18 2 1985 145 151 Rothman and Olney, 1995 S.M. Rothman J.W. Olney Excitotoxicity and the NMDA receptor—still lethal after eight years Trend Neurosci. 18 2 1995 57 58 Feb Ruel et al., 2007 J. Ruel J. Wang G. Rebillard M. Eybalin R. Lloyd R. Pujol J.L. Puel Physiology, pharmacology and plasticity at the inner hair cell synaptic complex Hear. Res. 227 1–2 2007 19 27 Schuman and Madison, 1994 E.M. Schuman D.V. Madison Nitric oxide and synaptic function Annu. Rev. Neurosci. 17 1994 153 183 Snyder and Bredt, 1992 S.H. Snyder D.S. Bredt Biologic role of nitric oxide Sci. Am. 266 5 1992 68 77 Yamada et al., 1996 M. Yamada K. Momose E. Richelson M. Yamada Sodium nitroprusside–induced apoptotic cellular death via production of hydrogen peroxide in murine neuroblastoma N1E-115 cells J. Pharmacol. Toxicol. Methods 35 1 1996 11 17 Zdanski et al., 1994 C.J. Zdanski J. Prazma P. Petrusz G. Grossman E. Raynor T.L. Smith H.C. Pillsbury Nitric oxide synthase is an active enzyme in the spiral ganglion cells of the rat cochlea Hear. Res. 79 1–2 1994 39 47