�������� �� �������� ��� ��������

�������� � ����� ��� ���� ��� ����

From the Universite´ Montpellier I and Inserm U 583, Physiopathologie et The´rapie des De´ficits Sensoriels et Moteurs, Hoˆpital SaintElop, 34295 Montpellier cedex 5, France

Guitton MJ, Puel J-L. Cochlear NMDA Receptors and Tinnitus.Audiological Medicine 2004; 2: 3–7.

Large dosage of aspirin produces reversible hearing loss and tinnitus. These effects have been attributedto the salicylate ion, the active component of aspirin. Salicylate acts as a competitive antagonist at theanion-binding site of prestin, the motor protein of sensory outer hair cells. This provides an explanationfor the hearing loss induced by aspirin. However, the molecular mechanism of salicylate-induced tinnitusremains obscure. One physiological basis of salicylate ototoxicity probably originates from alteredarachidonic acid metabolism. Arachidonic acid potentiates NMDA receptor currents. We therefore testedthe involvement of cochlear NMDA receptors in the occurrence of tinnitus. Tinnitus was assessed with abehavioural test based on an active avoidance paradigm. Results showed that the occurrence of tinnitusinduced by salicylate can be suppressed by application of NMDA antagonists into the cochlear fluids. Todetermine if the activation of NMDA receptors was linked to cyclooxygenase inhibition, we investigatedthe effect of mefenamate (a potent cyclooxygenase inhibitor). Since NMDA antagonists also blockedmefenamate-induced tinnitus, we propose that salicylate-induced tinnitus is mediated by cochlear NMDAreceptors through an inhibition of cyclooxygenase activity. Thus target cochlear NMDA receptors maypresent a therapeutic strategy for treatment of tinnitus.Key words:salicylate, tinnitus, NMDA receptor,arachidonic acid, cyclooxygenase, cochlea.

����������

In industrialised nations, 8–10% of the adult populationcurrently experiences tinnitus (ringing in the ears), i.e.approximately 20 million people in the USA. Tinnitus is asubjective perception of a sound in the absence of acorresponding external source. This corresponds to anabnormal sound which is clearly perceived in quietsurroundings and which very often becomes highlydebilitating. Although some palliative treatments exist(e.g. maskers, behavioural therapy), no convincing andspecific treatment is available. This is due to the limitedevidence of the physiological basis of tinnitus both fromclinical observations and from animal experiments.

Salicylate, which belongs to a family of non-steroidalanti-inflammatory drugs (NSAIDS), is widely recognisedfor its antipyretic, analgesic and anti-inflammatoryproperties. It has been well known for more than acentury that a large dose of salicylate produces hearingloss and tinnitus in humans (1). Both hearing loss andtinnitus develop over the initial days of treatment but maythen level off, fluctuate or decrease, and are reversiblewithin a few days of cessation of treatment. These effectsof salicylate were also demonstrated to occur in animals(2–4). In humans, the pitch of salicylate-induced tinnitusappears to be for almost all cases in the high frequencyrange (5). In rats, salicylate induces tinnitus which has afrequency close to 10 kHz (2, 4). Salicylate-inducedhearing loss results from the blocking of prestin, themolecular motor of outer cell electromotility (6, 7). In

contrast, the site and the mechanism of generation of thetinnitus induced by salicylate remain unclear.

Electrophysiological studies reported that injection ofsalicylate increased spontaneous activity of single units ofthe auditory nerve (8, 9), and modified the averagespectrum activity recorded from the round window,which is a gross measure of spontaneous activity of theauditory nerve (10). The characteristics of these changesappear to be similar to the characteristics of salicylate-induced tinnitus in animals (10). This suggests that, atleast in part, tinnitus induced by salicylate is associatedwith dysfunction of neurotransmission within the cochlea.

�� ��� ���������� ����� �������������������

Numerous biochemical processes underlying the effectsof salicylate have been identified: inhibition of prosta-glandins synthesis through the inhibition of cylooxy-genase, inhibition of numerous metabolic enzymes(phospholipase C, cholesterol ester synthase, NADPHoxydase, ATPase), inhibition of free radicals, insertioninto membranes and interference with ion transport,uncoupling of oxidative phosphorylation, or activation ofheat shock transcription factor (11–13). Among them, theinhibition of cyclooxygenase activity is the best knownpharmacological effect of salicylate (11, 14, 15). Cyclo-oxygenase 1 (COX-1) and its inducible isoform COX-2convert arachidonic acid to prostaglandin H2 (13, 16).

In the mammalian cochlea, the stria vascularis and the

������������ ����� � �� �� ���

2004 Taylor & Francis.ISSN 1651-386XDOI 10.1080/16513860410027394 AUDIOLOGICAL MEDICINE 2004

Aud

iol M

ed D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/26/

14Fo

r pe

rson

al u

se o

nly.

spiral ligament of guinea pigs produce PGI2, PGF2a andPGE2 from arachidonic acid (17). Systemic administra-tion of aspirin or indomethacin has been demonstrated tosignificantly decrease prostaglandin levels measured inthe perilymph by radioimmunoassay (18). Similarly, ithas been shown using radiolabelled arachidonic acid thatPGF2a and PGE2 levels were decreased at 30 minutes and3 hours post-aspirin injection (17). In normal conditions,arachidonic acid is competitively metabolised in thecochlea by cyclooxygenase to prostaglandins and trom-boxanes, and by lipooxygenase to leukotrienes (19).Sodium salicylate application on the round window of thecochlea induces a reversible decrease of prostaglandinsand tromboxane B2, with an increase of leukotrienes B4

and C4 (19). Electrophysiological studies have demon-strated that arachidonic acid increases the channelopening probability of NMDA receptor in varioussystems (Figure 1), including cerebellar granule cells,dissociated pyramidal cells, cortical neurons and adulthippocampal slices (20–22). The potentiation of NMDAresponses by arachidonic acid is observed both in nativeand recombinant receptors (23), and occurs even withsaturating levels of agonists at the glutamate-binding siteand glycine-binding site (20). Therefore, one possibilitywould be that salicylate induces tinnitus through the

activation of cochlear NMDA receptors. Testing thishypothesis requires the development of an adaptedbehavioural model of tinnitus in an animal.

�� ��� �������� �� �������� �� ���������

Validating a behavioural procedure to assess the presenceof tinnitus in animals is an unusual and difficult task. Thefirst behavioural model of tinnitus was developed in ratsby Jastreboff et al. The protocol was a conditionedsuppression paradigm based on water deprivation (2, 3).Animals were trained to stop drinking whenever thebroadband noise was turned off by combining its absencewith shocks to the foot. One major feature of Jastreboff’sprotocol is that animals have to be kept thirsty, leading toa loss of body weight up to 20%. When the loss of bodyweight was reduced, salicylate-induced modifications ofbehaviour were drastically reduced. This led to theproposal that changes in physiological state stronglyaffect the motivational level of the animals, making thetest difficult to interpret in terms of tinnitus (2).

To avoid changes in the physiological state of theanimal, we recently designed a new behavioural modelwhere tinnitus was assessed on an active avoidance

Fig. 1. Salicylate potentiates NMDA responses A. Cyclooxygenase converts arachidonic acid to prostaglandin H2 (PGH2). In normalconditions, the composition of the lipid bilayer is regulated in arachidonic acid by cyclooxygenase. B. The inhibition of cyclo-oxygenase following salicylate treatment leads to an increase of arachidonic acid concentration in lipid bilayers. Change ofcomposition of lipid bilayers results in an increase in open probability of NMDA receptors.

AUDIOLOGICAL MEDICINE 2004

� MJ Guitton and J-L Puel

Aud

iol M

ed D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/26/

14Fo

r pe

rson

al u

se o

nly.

paradigm (4). Animals had to display a motor task (e.g. tojump on a climbing pole) when hearing a sound matchingsalicylate-induced tinnitus (10 kHz). Salicylate treatmentprovoked a progressive decrease in the score and aconcomitant development of hearing loss as demonstratedby CAP threshold recordings (4). When the intensity ofsound eliciting behavioural responses was adjusted as afunction of CAP threshold shift, no significant decrease inthe score was observed. Taken together with the striking

similarity between time pattern of CAP threshold shifts,DPOAEs recordings, and score measurements, theseresults demonstrate that the score was linked to hearingperformance. In contrast, salicylate treatment drasticallyincreased the number of false positive responses. Becausethey have a sound hallucination (i.e. tinnitus), salicylate-treated animals are more likely to execute the motor taskduring the silent periods. Thus, animals treated withsalicylate would behave as if they heard a sound when no

Fig. 2. Molecular mechanisms of salicylate-induced tinnitus Animals were conditioned to jump on a climbing pole in response to asound stimulation. The false positive responses (i.e. jumping during silent periods) are a good indicator of the occurrence of tinnitus.A. Intraperitoneal saline injections did not provoke the occurrence of false positives. B. In contrast, intraperitoneal injections ofsalicylate (300 mg/kg/day for 4 days, grey area) significantly increased the number of false positives. C. Mefenamate treatment(35 mg/kg/day for 4 days, grey area) significantly increased the number of false positive responses (n = 10). D. Shown are the numberof false positive responses measured at day 4 in animals injected with saline solution (saline), and in animals injected withsalicylate� gelfoams bathed with artificial perilymph alone (AP,n = 10) or MK-801 (10�M, n = 10), 7-chlorokynurenate (7-CK,50�M, n = 10), or gacyclidine (50�M, n = 10). When compared with AP alone, local application of MK-801, 7-CK, or gacyclidinesignificantly (p� 0.001) reduced the occurrence of the false positive responses. When compared with animals injected withmefenamate alone (control), application of gelfoam containing 50�M 7-CK significantly reduced the occurrence of the false positiveresponses. NMDA antagonists were applied into the perilymphatic fluids using gelfoam placed on the round window of both ears.Figure adaptated from Guitton et al., 2003.

AUDIOLOGICAL MEDICINE 2004

Cochlear NMDA receptors and tinnitus 5

Aud

iol M

ed D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/26/

14Fo

r pe

rson

al u

se o

nly.

external sound was presented. Accordingly, resultsdemonstrated that animals treated with salicylate signi-ficantly increased the number of false positive responses(jump on the climbing pole) during silent periods (4). Inother words, the number of false positive responses is apowerful indicator of the occurrence of tinnitus.

�������� ���������� � ���������� ������� ��������

To determine if changes of behavioural responses inducedby salicylate were linked to cyclooxygenase inhibition,we investigated the effect of mefenamate (a potentcyclooxygenase inhibitor). We found that mefenamatetreatment significantly increased the number of falsepositive responses (Figure 2), attesting that inhibition ofthe cyclooxygenase pathway is one of the mechanisms bywhich salicylate and mefenamate induce tinnitus (4).

Local application of three different NMDA antagonists(7-chlorokynurenate, gacyclidine, and MK-801) intoperilymphatic fluids via a gelfoam placed on the roundwindow of each cochlea blocked the occurrence of bothsalicylate- and mefenamate-induced tinnitus (Figure 2).Thus, salicylate-induced tinnitus is mediated by cochlearNMDA receptors, through an inhibition of cyclooxy-genase activity (4). The results discussed here provideevidence for a new pharmacological effect of salicylate ininner ear physiology. In addition to reducing OHCselectromotility, salicylate may act on cochlear fastsynaptic transmission via the activation of NMDAreceptors, accounting for the occurrence of tinnitus.

��� ���������� ���������� �� ��������� ���� ���������

In the cochlea, the normal synaptic transmission betweeninner hair cells and primary auditory neurons is mediatedby the AMPA receptor (24). However, analysis with geneexpression, immunocytochemistry and in situ hybridisa-tion indicates that the cochlea express NR1 and NR2A-Dsubunits of NMDA receptors (25, 26). Although theseNMDA receptors are not involved in cochlear synaptictransmission, they are implicated in synaptic repair afterexcitotoxicity (27) and NMDA antagonists protectsensory hair cells from aminoglycoside ototoxicity (28),and prevent excitotoxicity induced by cochlear ischaemiaand acoustic trauma (29–31). Although experiments areneeded to confirm the implication of cochlear NMDAreceptors in other models of tinnitus (noise trauma,ischaemia, aminoglycosides or cisplatin ototoxicity),NMDA antagonists may thus constitute an attractivecandidate for the treatment of tinnitus in humans (4, 30).

�������������

The authors thank Dr. C. Nicolas-Puel for helpful comments onthe manuscript. This work was supported by Acouphe`nesLanguedoc-Roussillon.

����������

1. See G. Etudes sur l’acide salicylique et les salicylates;traitement du rhumatisme aigu et chronique et de la goutte,et de diverses affections du syste`me nerveux sensitif par lessalicylates. Bulletin de l’Acade´mie Nationale de Me´decine(Paris) 1877; 26: 689–706.

2. Jastreboff PJ, Brennan JF, Coleman JK, Sasaki CT. Phantomauditory sensation in rats: an animal model for tinnitus.Behav Neurosci 1988; 102: 811–22.

3. Jastreboff PJ, Sasaki CT. An animal model of tinnitus: adecade of development. Am J Otol 1988; 15: 19–27.

4. Guitton MJ, Caston J, Ruel J, Johnson RM, Pujol R, Puel JL.Salicylate induces tinnitus through activation of cochlearNMDA receptors. J Neurosci 2003; 23: 3944–52.

5. Cazals Y. Auditory sensorineural alterations induced bysalicylate. Prog Neurobiol 2000; 62: 583–631.

6. Kakehata S, Santos-Sacchi J. Effects of salicylate andlanthanides on outer hair cell motility and associated gatingcharge. J Neurosci 1996; 16: 4881–9.

7. Zheng J, Shen W, He DZ, Long KB, Madison LD, Dallos P.Prestin is the motor protein of cochlear outer hair cells.Nature 2000; 405: 149–55.

8. Evans EF, Borerwe TA. Ototoxic effects of salicylates onthe responses of single cochlear nerve fibres and on cochlearpotentials. Br J Audiol 1982; 16: 101–8.

9. Stypulkowski PH. Mechanisms of salicylate ototoxicity.Hear Res 1990; 46: 113–46.

10. Cazals Y, Horner KC, Huang ZW. Alterations in averagespectrum of cochleoneural activity by long-term salicylatetreatment in the guinea pig: a plausible index of tinnitus. JNeurophysiol 1998; 80: 2113–20.

11. Vane JR. Inhibition of prostaglandin synthesis as amechanism of action for aspirin-like drugs. Nature 1971;231: 232–5.

12. Levy GN. Prostaglandin H synthases, non-steroidal anti-inflammatory drugs, and colon cancer. FASEB J 1997; 11:234–47.

13. Vane JR, Bakhle YS, Botting RM. Cyclooxygenases 1 and2. Annu Rev Pharmacol Toxicol 1998; 38: 97–120.

14. Mitchell JA, Akarasereenont P, Thiemermann C, Flower RJ,Vane JR. Selectivity of non-steroidal anti-inflammatorydrugs as inhibitors of constitutive and inducible cycloox-ygenase. Proc Natl Acad Sci USA 1993; 90: 11693–7.

15. Vane JR, Botting RM. Mechanism of action of non-steroidalanti-inflammatory drugs. Am J Med 1998; 104: 2S–8S.

16. Thiemermann C. Biosynthesis and interaction of endothe-lium-derived vasoactive mediators. Eicosanoids 1991; 4:187–202.

17. Escoubet B, Amsallen P, Ferrary E, Tran Ba Huy P.Prostaglandin synthesis by the cochlea of the guinea pig:influence of aspirin, gentamicin, and acoustic stimulation.Prostaglandins 1985; 29: 589–99.

18. Jung TTK, Juhn SK. Prostaglandins in the perilymph. AssRes Otolaryngology Abs 1984; 107.

19. Jung TT, Miller SK, Rozehnal S, Woo HY, Park YM, BaerW. Effect of round window membrane application of sali-cylate and indomethacin on hearing and levels of arachi-

AUDIOLOGICAL MEDICINE 2004

� MJ Guitton and J-L Puel

Aud

iol M

ed D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/26/

14Fo

r pe

rson

al u

se o

nly.

donic acid metabolites in perilymph. Acta OtolaryngolSuppl 1992; 493: 81–7.

20. Miller B, Sarantis M, Traynelis SF, Attwell D. Potentiationof NMDA receptor currents by arachidonic acid. Nature1992; 355: 722–5.

21. Horimoto N, Nabekura J, Ogawa T. Developmental changesin arachidonic acid potentiation of NMDA currents incortical neurons. NeuroReport 1996; 7: 2463–7.

22. Yamakura T, Shimoji K. Subunit- and site-specific pharma-cology of the NMDA receptor channel. Prog Neurobiol1999; 59: 279–98.

23. Casado M, Ascher P. Opposite modulation of NMDAreceptors by lysophospolipids and arachidonic acid: com-mon features with mechanosensitivity. J Physiol (Lond)1998; 513: 317–30.

24. Ruel J, Chen C, Pujol R, Bobbin RP, Puel J-L. Ampa-preferring glutamate receptors in cochlear physiology ofadult guinea pig. J Physiol (Lond) 1999; 518: 667–80.

25. Usami S, Matsubara A, Fujita S, Shinkawa H, Hayashi M.NMDA (NMDAR1) and AMPA-type (GluR2/3) receptorsubunits are expressed in the inner ear. NeuroReport 1995;6: 1161–4.

26. Niedzielski AS, Wenthold RJ. Expression of AMPA,kainate, and NMDA receptor subunits in cochlear andvestibular ganglia. J Neurosci 1995; 15: 2338–53.

27. d’Aldin CG, Ruel J, Assie R, Pujol R, Puel JL. Implicationof NMDA type glutamate receptors in neural regeneration

and neoformation of synapses after excitotoxic injury in theguinea pig cochlea. Int J Dev Neurosci 1997; 15: 619–29.

28. Basile AS, Huang JM, Xie C, Webster D, Berlin C, SkolnickP. N-methyl-D-aspartate antagonists limit aminoglycosideantibiotic-induced hearing loss. Nat Med 1996; 2: 1338–43.

29. Puel J-L, Pujol R, Tribillac F, Ladrech S, Eybalin M.Excitatory amino acid antagonists protect cochlear auditoryneurons from excitotoxicity. J Comp Neurol 1994; 341:241–56.

30. Puel J-L. Chemical synaptic transmission in the cochlea.Prog Neurobiol 1995; 47: 449–76.

31. Duan M, Agerman K, Ernfirs P, Canlon B. Complementaryroles of neurotrophin 3 and a N-methyl-D-aspartateantagonist in the protection of noise and aminoglycoside-induced ototoxicity. Proc Natl Acad Sci USA 2000; 97:7597–602.

Address for correspondence:

Jean-Luc PuelUniversiteMontpellier I and Inserm U 583Physiopathologie et The´rapie des De´ficits Sensoriels et MoteursHopital Saint Eloi34295 Montpellier cedex 5FranceTel: �33 499 636 009Fax:�33 499 636 020E-mail: puel@montp.inserm.fr

AUDIOLOGICAL MEDICINE 2004

Cochlear NMDA receptors and tinnitus 7

Aud

iol M

ed D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/26/

14Fo

r pe

rson

al u

se o

nly.