Research 121 (1998) 71^76

Benzodiazepines alter cochleo-cochlear loop in humans

N. Morand a, E. Veuillet a;*, M.C. Gagnieu b, P. Lemoine c, L. Collet a

a Laboratoire `Neurosciences et Systeémes Sensoriels', 3 place d'Arsonval, Pavillon U, Hoêpital Edouard Herriot, 69003 Lyon, France b Laboratoire de Pharmacologie, Hoêpital Edouard Herriot, Lyon, France c Uniteè Clinique de Psychiatrie Biologique, Le Vinatier, Bron, France

Received 26 November 1997; revised 26 March 1998; accepted 7 April 1998

Abstract

By using otoacoustic emission, we looked for change in outer (OHC) motile activity and medial olivocochlear (MOC) system inhibition due to benzodiazepine administration, a drug that is known to produce a pharmacological effect by interacting with GABAergic inhibitory neurotransmission. No effect was observed on OHC motile activity, in contrast benzodiazepines decreased MOC system effectiveness suggesting the existence of GABAergic fibers projecting onto the MOC system. z 1998 Elsevier Science B.V. All rights reserved.

Key words: Gamma-aminobutyric acid; Outer hair cell; Olivocochlear e¡erent system; Otoacoustic emission

1. Introduction OHCs, principally acetylcholine (Ach) and the inhibi- tory neurotransmitter Q-aminobutyric acid (GABA) (for The organ of Corti is innervated by the olivocochlear review see Eybalin, 1993). Concerning GABA, several e¡erent system (Rasmussen, 1946), in which two sub- studies seem to prove the presence of GABAergic syn- components have been clearly distinguished (Warr, apses in the . The presence of GABA and glu- 1992; Warr and Guinan, 1979). The lateral olivo- tamic acid decarboxylase in ¢bers and e¡erent endings cochlear (LOC) system originates from lateral superior synapsing onto OHC bodies has been revealed by im- olivary neurons making synaptic contact with the a¡er- munocytochemistry data (Fex and Altschuler, 1984; ent inner radial ¢bers of the inner hair cell system. The Fex et al., 1986; Thompson et al., 1986; Eybalin et medial e¡erent system originates from the medial al., 1990). Furthermore, the presence of GABAA recep- (SOC) or , tors in the outer cell membrane of isolated OHCs has synapsing with the basal pole of outer hair cells been demonstrated using monoclonal antibodies against (OHCs). the K- and L-subunits of GABAA receptors (Plinkert et Various neurotransmitters seem to be involved be- al., 1989). GABAA receptors are at the basal pole of tween the medial olivocochlear (MOC) system and the OHCs. GABA uptake by e¡erent endings has also been demonstrated (Schwartz and Ryan, 1983; Ryan et al., 1992). The action of GABA or agonists of GABA on OHC * Corresponding author. Tel.: +33 04 72 11 05 02; mechanisms has been studied, leading to contradictory Fax: +33 04 72 11 05 04; E-mail: [email protected] conclusions. In vitro, OHC hyperpolarization was ob- served after GABA application; this hyperpolarization Abbreviations: Ach, acetylcholine; CAP, compound action potential; DP, distortion product; EOAE, evoked otoacoustic emission; GABA, was increased by an application of benzodiazepines Q-aminobutyric acid; IC, ; LOC, lateral olivococh- (Plinkert et al., 1993). GABA also decreased the electro- zlear; MOC, medial olivocochlear; OAE, otoacoustic emission; SL, motile response (Sziklai et al., 1996). In contrast, Evans sensation level; SPL, sound pressure level; SOAE, spontaneous otoacoustic emission; SOC, superior olivary complex et al. (1996) observed no OHC response during whole-

0378-5955 / 98 / $19.00 ß 1998 Elsevier Science B.V. All rights reserved. PII: S0378-5955(98)00068-9

HEARES 3034 29-6-98 72 N. Morand et al. / Hearing Research 121 (1998) 71^76 cell recording after GABA application. In vivo, OHC GABA is an inhibitory transmitter which is abun- function can be explored by otoacoustic emission dant in the central auditory pathways (Wynne et al., (OAE) recordings. OAEs are sounds recorded in the 1995; Vater et al., 1992; Fubara et al., 1996). It has external ear canal in the presence of acoustic stimula- been shown that benzodiazepines increase the CAP and tion (evoked OAEs (EOAEs) and distortion products decrease cochlear microphonics in guinea pigs, in con- (DP)) or without acoustic stimulation (spontaneous trast to electrical stimulation of the crossed olivo- emissions (SOAEs)). OAEs are generated in the coch- cochlear bundle. These benzodiazepine e¡ects have lea, probably by OHC motile activity. The in£uence of been thought to re£ect a central action of GABA on benzodiazepines on EOAE amplitude has been studied the e¡erent system (Velluti and Pedemonte, 1986). in humans: Hauser et al. (1992) found no signi¢cant Thus GABA may have a peripheral action on outer di¡erence in EOAE amplitude under premedication hair cell mechanisms, and a central action on the medial with midazolam; likewise diazepam and £unitrazepam e¡erent system. The aim of the present study was to had no e¡ect on EOAE amplitude (Delb et al., 1994). It investigate these two hypotheses in humans by studying has been shown that contralateral acoustic stimulation benzodiazepines known to produce their pharmacolog- decreases the EOAE amplitude in humans (Collet et al., ical e¡ect by interacting with GABAergic inhibitory 1990) and acoustic DPs in guinea pig (Puel and Rebil- neurotransmission and their e¡ect on EOAE amplitude lard, 1990). The medial e¡erent system is involved in and on the MOC system, by comparing EOAE ampli- this contralateral suppression e¡ect, which is greatly tude with and without contralateral acoustic stimula- reduced in humans after sectioning of the vestibular tion (Collet et al., 1990; Veuillet et al., 1991). nerve, which severs olivocochlear e¡erents (Giraud et al., 1995). Thus this suppression e¡ect observed on EOAEs in the presence of contralateral acoustic stim- 2. Methods ulation is a good means of investigating the MOC sys- tem in vivo. In guinea pigs, the perfusion of the cochlea 2.1. Subjects with bicuculline (GABA antagonist) abolished the f23f1 DP reduction normally observed after contralateral The study was performed on 10 healthy men ranging stimulation in guinea pigs (Kirk and Johnstone, 1993), in age from 20 to 34 years (mean 24 years, S.E. 1.23). thus con¢rming a role for GABA in the OHC mecha- All the subjects were volunteers and this study was nism. So the terminal elements involved in the contra- carried out with their written consent and with the lateral acoustical suppression of f23f1 can be identi¢ed agreement of the local institutional Ethics Committee. as being part of a GABAergic innervation, whether The subjects were selected according to the following LOC or MOC system, of the apical OHCs. acoustic criteria: normal auditory thresholds ( 6 20 dB Thus a majority of studies con¢rm the presence of loss between 250 and 8000 Hz by octave) measured by a GABAergic e¡erents on OHCs and presence of Madsen DAIII audiometer, normal tympanometric re- GABAA receptors on the basal pole of OHC, but there cordings done using an Amplaid 702, presence of is no evidence of a real action of GABA on outer hair EOAEs (amplitude better than 5 dB SPL at 80 þ 3 dB cell mechanisms. Furthermore the presence of GA- SPL), no spontaneous otoacoustic emissions. The vol- BAergic synapses in human cochlea has not been unteers were all right-handed (Ol¢eld, 1971). studied. A questionnaire and a medical examination further On the other hand, the olivocochlear e¡erent system enabled selection of subjects without prior acoustic could be under the control of central auditory path- pathology, taking no medical treatment, using no hyp- ways. The inferior colliculus (IC) receives descending notics or benzodiazepines. On the day of experimenta- input from the , and sends ipsilateral tion a urine assay for benzodiazepine was made to con- and contralateral projections onto the medial olivo- ¢rm the last of these conditions. cochlear system (Hu¡man and Henson, 1990). So it is anatomically possible that the IC plays a role in the 2.2. Otoacoustic emissions activity of the olivocochlear e¡erent system. Further- more, physiological studies show an IC contribution EOAE recordings were made with subjects reclining to the olivocochlear e¡erent system. Stimulating the in a sound-proof room, using an Otodynamics0 ILO IC induces a decrease in compound action potential 88 apparatus. A probe placed in the external ear canal (CAP) (Dolan and Nuttal, 1988a) that is similar to delivered acoustic stimulation through a loudspeaker the change seen after electrical stimulation of the (Knowles 1712 transducer) and recorded responses crossed olivocochlear bundle (Dolan and Nuttal, with a microphone (Knowles 1843). To be sure that 1988b). Furthermore, stimulating the IC protects the the probe was always placed in the same position cochlea from noise, as does stimulation of the crossed over successive recordings, a silicone mold of the probe olivocochlear bundle (Rajan, 1990). and of the external ear canal was made.

HEARES 3034 29-6-98 N. Morand et al. / Hearing Research 121 (1998) 71^76 73

Stimuli were non-¢ltered clicks with a duration of 80 Ws and a rate of 50/s. Linear and non-linear di¡er- ential cochlear echo methods were used (Kemp et al., 1990).

2.3. Procedure

During this study a double-blind administration of oxazepam versus placebo was made (the kind of molec- ular administered was not known either to subjects or to experimenters). So the data were collected in a blinded manner. The subjects were tested before and 1, 3, 7 and 24 h after oxazepam (20 mg) or placebo administration. The order of testing (oxazepam versus placebo) was randomized. We employed a crossover design: all 10 subjects were tested twice with a 2-week interval (once after oxazepam intake, and once after placebo intake). EOAEs were recorded in right ears. Fig. 1. Mean plasma concentration 1, 3, 7 and 24 h after oxazepam A ¢rst recording was made with a non-linear click at intake. The standard error is represented. 80 þ 3 dB SPL. The amplitude of the response was cal- culated from 2.5 to 20 ms. Then the e¡erent system was investigated according to the protocol proposed by Collet et al. (1992): 3. Results EOAEs were recorded with linear clicks for ¢ve ipsi- lateral intensities levels ranging between 60 and 72 dB 3.1. Plasma oxazepam concentration SPL. Presentation of these ¢ve intensities with and without contralateral acoustic stimulation delivered by Fig. 1 shows the mean plasma concentration of ox- the ILO 88 apparatus was randomized. Contralateral azepam. The highest plasma concentration was found acoustic stimulation was a 30 dB SL broadband noise 3 h after intake. Individual variations in maximum plas- (bandwidth 500^8000 Hz). The response window was ma concentration can be noted, values ranging between set at 3.2^20 ms. 0.076 and 0.4 mg/l (mean 0.185, S.E. 0.028). Oxazepam versus placebo was administrated at the Like data were collected in a double-blind manner, same hour of the day to all the subjects. Before each plasma oxazepam concentration was measured after recording, a sample of blood was taken, to measure placebo administration: no oxazepam was found in plasma oxazepam concentration. any blood samples taken after placebo administration and this for all the subjects. 2.4. Data processing and statistical analysis 3.2. EOAE amplitudes EOAE amplitudes were studied after non-linear clicks at 80 þ 3 dB SPL. In order to eliminate Fig. 2 shows non-linear EOAE amplitude as a func- artifacts, the amplitude of the response was calculated tion of time. There was no signi¢cant variation in am- from 2.5 to 20 ms. The overall click-evoked response plitude between the various recordings made on the day level, corresponding to the sum of echo power in the of placebo administration (paired t-test, P s 0.05). whole frequency range of the power spectrum, was Response amplitude was also studied after oxazepam noted. intake. EOAE amplitudes after oxazepam intake were The e¡ectiveness of contralateral acoustic stimula- compared with themselves and with EOAE amplitudes tion was expressed as an equivalent attenuation which before oxazepam intake (paired t-test). These results is the decrease in ipsilateral stimulus inducing the same were further compared with results obtained after a EOAE amplitude e¡ect as the contralateral broadband 2-week interval (paired t-test). We note that the noise (Collet et al., 1992; Veuillet et al., 1991). EOAE amplitude obtained at the same time at two Statistical analysis was performed using Sigmastat0 intervals was statistically comparable before oxazepam software and included parametric tests (normality test intake (11.7 dB, S.E. 1.17) and before placebo intake being passed). ANOVAs on repeated measures, paired (11.2 dB, S.E. 1.17), suggesting no damage and no ex- t-test and Pearson product moment correlation were posure to noise trauma during the course of the 2-week conducted. Probabilities below 0.05 were considered study period. No signi¢cant e¡ect of oxazepam was signi¢cant. observed.

HEARES 3034 29-6-98 74 N. Morand et al. / Hearing Research 121 (1998) 71^76

(one-way ANOVA). We compared the results obtained after oxazepam intake with the result obtained before oxazepam intake (paired t-test). A decrease in equiva- lent attenuation, signi¢cant 3 h after oxazepam intake, was observed (paired t-test, t = 32.569, P = 0.030). A strong correlation was observed between mean plasma oxazepam level and mean equivalent attenuation (Pear- son product moment correlation, r = 0.976, P = 0.005). At 3 h (maximum plasma concentration) a signi¢cant correlation between oxazepam concentration and the change in equivalent attenuation between the recording made without oxazepam and the recording 3 h after oxazepam intake was observed (Pearson product mo- ment correlation, r = 0.684, P = 0.03): the higher the oxazepam concentration, the more the equivalent at- tenuation decreases (Fig. 4). Perception threshold for contralateral stimulation Fig. 2. Amplitudes of EOAEs before and after oxazepam and place- (broadband noise) in the left ear did not change be- bo administration (non-linear click at 80 þ 3 dB SPL, amplitude cal- culated from 2.5 to 20 ms). The standard error is represented. tween successive recordings (ANOVA, paired t-test).

3.3. E¡ectiveness of the MOC system 4. Discussion

Fig. 3 shows the mean equivalent attenuation ob- The ¢rst aim of this study was to test the e¡ect of tained before and after administration of placebo. No benzodiazepines on OHC mechanisms in humans. No signi¢cant di¡erence between the recordings was ob- statistically signi¢cant changes could be detected before served under placebo (one-way ANOVA, paired t- and after oxazepam administration with respect to non- test). The e¡ectiveness of MOC is constant across the linear EOAE amplitude. This result con¢rms previous hours of day. studies (Hauser et al., 1992; Delb et al., 1994): benzo- The mean results obtained before and after oxaze- diazepines have no direct e¡ect on OHC motility in pam intake are shown in Fig. 3. The mean equivalent humans. attenuation obtained at the same time at 2-week inter- Benzodiazepines are lipophilic substances and so vals being statistically comparable before oxazepam in- easily cross the blood-brain barrier to penetrate brain take (31.75 dB, S.E. 0.32) and before placebo intake tissue. Therefore benzodiazepines should be present in (31.88 dB, S.E. 0.36), it was possible to compare equiv- alent attenuation calculated at the same time but at a 2- week interval. So we note that the e¡ectiveness of the MOC system is the same after the 2-week interval. To evaluate treatment e¡ects (oxazepam and placebo) on equivalent attenuation, a two-way ANOVA for re- peated measures was conducted and showed a signi¢- cant treatment e¡ect (F = 14.83, df = 1, P = 0.004). First, we decided to observe if the equivalent attenu- ation is di¡erent under placebo and oxazepam. By paired t-test, there was a signi¢cant e¡ect of the treat- ment 1 h after oxazepam intake (paired t-test, t = 4.9, P = 0.02). Oxazepam intake induced a decrease in equivalent attenuation. A tendency to decrease 3 h after oxazepam intake was, however, not statistically signi¢- cant (paired t-test, t = 2.01, P = 0.076). So there is a decrease of MOC system e¡ectiveness after oxazepam intake compared to placebo intake. This e¡ect appears 1 h after oxazepam intake. Fig. 3. Mean equivalent attenuation and standard error. The treat- ment has a signi¢cant e¡ect on medial e¡erent system performance Second, we decided to observe the e¡ect of oxaze- (P = 0.04, F = 14.8). The asterisk indicates a signi¢cant change of pam over time. Fig. 3 shows a variation in the e¡ect of equivalent attenuation 3 h after oxazepam intake compared to basal oxazepam with time which is not statistically signi¢cant equivalent attenuation (before oxazepam intake).

HEARES 3034 29-6-98 N. Morand et al. / Hearing Research 121 (1998) 71^76 75

tory neurons; GABAA receptors have been found by autoradiography in the of the guinea pig (Juiz et al., 1994). Thus there may be reduced audi- tory a¡erent ¢ring after oxazepam intake. Nevertheless the contralateral acoustic threshold was measured be- fore all the tests: a 30 dB broadband noise was always used to elicit the contralateral e¡ect. The decrease in MOC performance could be ex- plained by inhibition. Benzodiazepines may act directly on the synapse between a¡erent pathway and e¡erent system. GABA would be liberated in the synapse by the a¡erent pathway; benzodiazepines by acting on GABAA receptors would enhance the GABA e¡ect and the inhibition of the MOC. These hypotheses are supported by the fact that GAD and GABA positive cells have been found in the SOC (Vater et al., 1992) and that the medial nucleus of the trapezoid body con- Fig. 4. Correlation between oxazepam concentration and the change tains GABAergic endings (Suneja et al., 1995); never- in equivalent attenuation between the recording made without oxa- theless the nature of the neurotransmitters involved be- zepam and the recording 3 h after oxazepam intake (r = 0.684, P = 0.03). tween the a¡erent pathways and MOC system in the SOC is unknown. The decrease in MOC system per- formance could also re£ect some central control of the perilymph. Benzodiazepines act as coreceptors in- the MOC, suggesting the existence of GABAergic de- creasing GABA-induced chloride ion in£ux into the scending ¢bers projecting onto the MOC system. cell. Benzodiazepines thus increase the e¡ect of This study shows that benzodiazepines have an ac- GABA. Although benzodiazepines act on OHCs, tion on the MOC system in humans; some ascending or GABA must be released in the synaptic gap. Thus if descending GABAergic projections act on the MOC the medial e¡erent system has no tonic action on OHC, which controls active OHC mechanisms. no variation in EOAE amplitudes can be observed after oxazepam intake. It may also be that the GABAA re- ceptors at the basal pole of the OHCs are non-func- tional as in vitro studies on guinea pig have shown References (Evans et al., 1996). The presence of GABAergic syn- Collet, L., Kemp, D.T., Veuillet, E., Duclaux, R., Moulin, A., Mor- apses in human cochlea could also be discussed. Fi- gon, A., 1990. E¡ect of contralateral auditory stimuli on active nally, the dose of oxazepam used (20 mg) was perhaps cochlear micro-mechanical properties in human subjects. Hear. not high enough to in£uence EOAEs signi¢cantly. Res. 43, 251^262. Thus it can be concluded that, with the doses of Collet, L., Veuillet, E., Bene, J., Morgon, A., 1992. E¡ects of con- oxazepam used in our study, no e¡ect on OHC active tralateral white noise on click-evoked emissions in normal and sensorineural ears: Towards an exploration of the medial olivo- micromechanisms was obtained. cochlear system. 31, 1^7. The second aim of this study was to test, for the ¢rst Delb, W., Merl, E., Andes, Ch., Koch, A., 1994. The in£uence of time in humans, the e¡ect of benzodiazepines on the sedation on otoacoustic emissions. Laryngol. Rhinol. Otol. 73, MOC system. Our ¢ndings clearly demonstrate a de- 315^319. crease in contralateral suppression e¡ect, the contra- Dolan, D.F., Nuttal, A.L., 1988a. Cochlear whole nerve responses altered by electrical stimulation of the inferior colliculus. Soc. lateral acoustic stimulation having less e¡ect on OHC Neurosci. Abstr. 14, 650. active micromechanisms under oxazepam. This decrease Dolan, D.F., Nuttal, A.L., 1988b. Masked cochlear whole nerve re- in e¡ectiveness seems to be a true consequence of ox- sponse intensity functions altered by electrical stimulation of the azepam administration, being strongly correlated with crossed olivocochlear bundle. J. Acoust. Soc. Am. 83, 1081^ plasma oxazepam concentration. Since benzodiazepines 1086. Evans, M.G., Kiln, J., Pinch, D., 1996. No evidence for functional did not modify OHC responses, the decrease in suppres- GABA receptors in outer hair cells isolated from the apical half of sion can be attributed to a change in MOC system the guinea-pig cochlea. Hear. Res. 101, 1^6. performance rather than to a modi¢cation in the Eybalin, M., 1993. Neurotransmitters and neuromodulators of the OHCs. Such a change could be due to a reduction in mammalian cochlea. Physiol. Rev. 73, 309^373. MOC stimulation. The message carried by the auditory Eybalin, M., Merchan-Perez, A., Gil-Loyzagaya, P., 1990. Immuno- cytochemical detection of glutamate decarboxylase in the postna- a¡erents may be modi¢ed by benzodiazepines: GABA tal developing rat organ of Corti. Int. J. Dev. Neurosci. 8, 613^ is thought to be a LOC neurotransmitter (for review see 620. Eybalin, 1993), presumably linked to the ¢ring of audi- Fex, J., Altschuler, R.A., 1984. Glutamic acid carboxylase immuno-

HEARES 3034 29-6-98 76 N. Morand et al. / Hearing Research 121 (1998) 71^76

reactivity of olivocochlear neurons in the organ of Corti of guinea Rasmussen, G., 1946. The olivary peduncle and other ¢brous projec- pig and rat. Hear. Res. 15, 123^131. tions of the superior olivary complex. J. Comp. Neurol. 84, 141^ Fex, J., Altschuler, R.A., Kachar, B., 1986. et al. GABA visualized by 219. immunocytochemistry in the guinea pig cochlea in axons and end- Ryan, A.F., Schwartz, I.R., Keithley, E.M., Wang, Z.X., 1992. Selec- ings of e¡erent neurons. Brain Res. 366, 106^117. tive retrograde transport of nicopetic acid, a GABA analog, labels Fubara, B.M., Casseday, J.H., Corvey, E., Schwartz-Bloom, R.D., a subpopulation of gerbil olivocochlear neurons. J. Comp. Neurol. 1996. Distribution of GABA/A, GABA/B and glycine receptors 326, 337^346. in the central of the big brown bat, Eptisicus Schwartz, I.R., Ryan, A.F., 1983. Di¡erential labeling of sensory cell fuscus. J. Comp. Neurol. 369, 83^92. and neural populations in the organ of Corti following amino acid Giraud, A.L., Collet, L., Cheèry-Croze, S., Magnan, J., Chays, A., incubations. Hear. Res. 9, 185^200. 1995. Evidence of medial olivocochlear involvement in contra- Suneja, S.K., Benson, C.G., Gross, J., Potashner, S.J., 1995. Uptake lateral suppression of otoacoustic emissions in humans. Brain and release of D-aspartate, GABA, glycine in guinea pig brainstem Res. 705, 15^23. auditory nuclei. J. Neurochem. 64, 147^160. Hauser, R., Harris, F.P., Probst, R., Frei, F., 1992. In£uence of gen- Sziklai, I., He, D.Z.Z., Dallos, P., 1996. E¡ect of acetylcholine and eral anesthesia on transiently evoked otoacoustic emissions in hu- GABA on the transfer function of electromotility in isolated outer mans. Ann. Oto-Rhino-Laryngol. 101, 994^999. hair cells. Hear. Res. 95, 87^99. Hu¡man, R.F., Henson, O.W., Jr., 1990. The descending auditory Thompson, G.C., Cortez, A.M., Igarashi, M., 1986. GABA like im- pathway and the acousticomotor systems: connections with the munoreactivity in the squirrel monkey organ of Corti. Brain Res. inferior colliculus. Brain Res. Rev. 15, 295^323. 372, 72^79. Juiz, J.M., Albin, R.L., Helfert, R.M., Altschuler, R.A., 1994. Distri- Vater, M., Kossl, H., Horn, A.K., 1992. GAD and GABA immuno- bution of GABA/A and GABA/B binding sites in the cochlear reactivity in the ascending auditory of horseshoe and moustached nucleus of the guinea pig. Brain Res. 639, 193^201. bats. J. Comp. Neurol. 325, 183^206. Kemp, D.T., Ryan, S., Bray, P., 1990. A guide to the e¡ective use of Velluti, R., Pedemonte, M., 1986. Di¡erential e¡ects of benzodiaze- otoacoustic emissions. Ear Hear. 11, 93^105. pines on cochlear and auditory nerve responses. Electroencepha-

Kirk, D.L., Johnstone, B.M., 1993. Modulation of f23f1 : Evidence logr. Clin. Neurophysiol. 64, 556^562. for a GABAergic e¡erent system in apical cochlea of the guinea Veuillet, E., Collet, L., Duclaux, R., 1991. E¡ect of contralateral pig. Hear. Res. 67, 20^34. acoustic stimulation on active micromechanical properties in hu- Ol¢eld, R.C., 1971. The assessment and analysis of handedness: The man subjects: dependence on stimulation variables. J. Neuro- Edinburgh Inventory. Neuropsychologia 9, 97^111. physiol. 65, 724^735. Plinkert, P.K., Mohlher, H., Zenner, H.P., 1989. A subpopulation of Warr, W.B., Guinan, J.J., 1979. E¡erent innervation of the organ of outer hair cells possessing GABA receptors with tonotopic organ- Corti: two separate systems. Brain Res. 173, 152^155. isation. Arch. Oto-Rhino-Laryngol. 246, 417^422. Warr, B. (1992) Organisation of olivocochlear e¡erent systems in Plinkert, P.K., Gitter, A.H., Mohler, H., Zenner, H.P., 1993. Struc- mammals. In: Wester, D.B. (Ed.), Mammalian Auditory Path- ture, pharmacology and function of GABA/A receptors in coch- ways: Neuroanatomy, Handbook of Auditory Research, Spring- lear outer hair cells. Eur. Arch. Oto-Rhino-Laryngol. 250, 351^ er-Verlag, New York, pp. 410^448. 357. Wynne, B., Harvey, A.R., Robertson, D., Sirinathsinghzi, D.J., 1995. Puel, J.L., Rebillard, G., 1990. E¡ect of contralateral sound-stimula- Neurotransmitter and neuromodulation systems of the rat inferior tion on distortion product 2f13f2 : Evidence that medial e¡erent colliculus and auditory brainstem studied by in situ hybridization. system is involved. J. Acoust. Soc. Am. 87, 1630^1635. J. Chem. Neuroanat. 9, 289^300. Rajan, R., 1990. Electrical stimulation of the inferior colliculus at low rates protects the cochlea from auditory desensitization. Brain Res. 506, 192^204.

HEARES 3034 29-6-98