INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cochlear centrifugal pathways terminate on afferent dendrites carrying cochlear outflow, or receptor outer hair cells (OHCs) critical for transduction sensitivity and frequency selectivity. The former constitutes the lateral olivocochlear system (LOCS). The latter constitutes the medial olivocochlear system (MOCS) with uncrossed and crossed subsystems (UMOCS and CMOCS) originating in ipsilateral or contralateral brain stem, respectively (Warr et al. 1986). Cochlear effects of MOCS subsystems are not differentiated since both are cholinergic, and cholinergic effects on OHCs appear exercised through a nicotinic-type receptor (e.g., Guth and Norris 1996). However, differences are suggested by recent detailed studies of contributions of the two subsystems to protection of hearing sensitivity from sound-induced temporary threshold shifts (TTSs). In these studies, TTSs were induced by loud pure tones by themselves or in combination with atraumatic noise backgrounds. In essence, CMOCS only appeared to reduce TTSs (Rajan 1995, 2000, 2001b). However, UMOCS could, under different conditions, prevent noise exacerbation of tone-induced TTSs without affecting tone-alone TTSs (Rajan 2000) or exacerbate tone-induced TTSs (Rajan 2001a,b). Thus MOCS actions at cholinergic synapses on OHCs appear likely to be more diverse than currently appreciated. I examine this hypothesis by studying how the two OHC efferent subsystems modulated TTSs created by broadband trauma that I predicted would activate both systems to act on TTSs. This would also extend study of the utility of MOCS protection since normally occurring traumata are more likely to be complex-sound trauma.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

General procedures are detailed elsewhere (Rajan 2001b). Adult cats (3-6 kg) were anesthetized (60 mg/kg ip) and maintained with continuous intravenous pentobarbital sodium (Nembutal, 2-3 mg · kg¹ · h¹), with monitoring of parameters including areflexia and output from electrocardiographic/electromyographic electrodes. Body temperature was maintained at 37.5 ± 0.5°C by a warming blanket, regulated by rectal-probe feedback. Cats were tracheostomized and ventilated on room air, at 20-30 breaths/min depending on weight, with tidal volume set from normogram data. Stainless steel electrodes were implanted against round windows, and VIIIth nerve compound action potential (CAP) thresholds were measured at frequencies from 1 to 40 kHz. Only animals with bilaterally normal hearing sensitivity from 1 to 40 kHz were used (Rajan et al. 1991).

Tone and noise stimuli to each ear were generated independently digitally by fast Fourier transform and "brick-wall" filtered. Noise output was flat at frequencies within the noise band. Output at frequencies outside the band, due to intermodulation distortion in the sound delivery system, was > 50 dB from within-band output. Stimuli were gated under computer control, fed through separate attenuators into one of four channels of an electronic mixer, and manually switched to a Sennheiser HD 535 speaker connected to a tube in the external meatus (Rajan 2000). Animals were exposed to binaural sound trauma with efferent pathways intact bilaterally or after lesions at the fourth ventricle floor (Rajan 2000) to cut all efferent pathways to only one cochlea, or only crossed pathways (bilaterally) leaving uncrossed pathways intact, or to totally de-efferent one cochlea and only crossed pathways to the other ear. Ears were grouped according to efferent status: all pathways intact (OC+ ears), all pathways cut (OC ears), or only crossed pathways cut (COC). Postmortem histology was used to confirm location of cuts.

Traumatic stimuli were 100 dB SPL binaural noise bands at 1-6 kHz for 40 min or at 8-13 kHz for 15 min. CAP thresholds were re-measured 5-min postnoise, from 1 to 13 kHz for 1- to 6-kHz noise and 6-32 kHz for 8- to 13-kHz noise. Frequency-specific TTSs were differences between pre- and posttrauma thresholds. Only frequencies with significant TTSs in OC ears (group mean TTSs  5dB; 2- to 12-kHz TTSs for 1- to 6-kHz noise and 7- to 30-kHz TTSs for 8- to 13-kHz noise) were analyzed in all groups. Two-way repeated measures ANOVAs compared effects between groups; if a significant condition difference or condition × frequency interaction existed, unpaired Student's t-tests were used to compare TTSs at corresponding frequencies in the groups.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Generally similar patterns of effects were seen for both noise bands. Hence, effects for high-frequency noise (8-13 kHz) are first detailed before presenting data for 1-to 6-kHz noise. To aid in reading, crossed OC pathway effects are attributed to the CMOCS because the crossed pathway is almost exclusively the crossed MOCS to OHCs. Uncrossed OC effects are attributed to UMOCS effects on OHCs rather than effects of the uncrossed LOCS on afferent dendrites, for reasons detailed in DISCUSSION.

In de-efferented (OC) ears, 8- to 13-kHz noise caused TTSs from 7 to 30 kHz, with a broad desensitization peak from 11 to 17 kHz (Fig. 1A). When olivocochlear pathways were intact (OC+), TTS reductions by >5 dB occurred from 7 to 16 kHz (Fig. 1B), a range containing all noise frequencies ("within-band" frequencies) and extending just outside the band but not at much higher frequencies. Peak protection of 20 dB was at 10 kHz with protection 15 dB from 9 to 14 kHz. When only uncrossed pathways were intact (COC ears), TTSs at <12 kHz were similar to or even greater than in de-efferented ears but at >15 kHz decreased below those even in efferent-intact ears (Fig. 1A, Table 1). Thus in frequency ranges on either side of a crossover frequency, it appeared that uncrossed and crossed OC pathways (specifically, UMOCS and CMOCS) exerted complex, opposing effects.

View larger version (36K): [in this window] [in a new window]   Fig. 1. Role of outer hair cell (OHC) efferents on compound action potential (CAP) threshold desensitization, after binaural exposure to (A-C) 8- to 13-kHz noise or (D-F) 1- to 6-kHz noise. In each row, the black bar indicates the noise band. A and D: "raw" temporary threshold shifts (TTSs) with all efferent pathways intact (OC+), all cut (OC), or crossed pathways cut (COC) leaving intact uncrossed pathways. Data are mean TTSs (error bars, standard error), 5-min postexposure. B and E: TTS modulation by all OC pathways (see text). C and F: TTS modulation separately by crossed and uncrossed medial olivocochlear system (MOCS; see text). In B, C, E, and F, positive differences indicate increased TTSs, negative differences TTS reductions, closed symbols indicate frequencies with significant differences (Student's t-test; P < 0.05) between raw TTSs in groups (in A and D) used to derive difference data in B, C, E, and F, and open symbols are frequencies without significant differences.

To examine this phenomenon, the separate effects of the two MOCS subsystems were quantified (Fig. 1C). Differences in mean frequency-specific TTSs in COC and OC+ groups (i.e., only uncrossed pathways intact or all OC pathways intact) were used to quantify CMOCS-alone effects. UMOCS-alone effects were quantified as differences in mean TTSs in COC and OC groups (i.e., only crossed pathways absent or all OC pathways absent).

The CMOCS caused two effects. At "within-band" frequencies, it protected. This peaked with 26 dB protection at 10 kHz, decreasing on either side of this peak. At frequencies >15 kHz ("high-side" frequencies, outside the noise), it exacerbated TTSs by as much as 21 dB (at 19 kHz). The UMOCS produced almost reverse effects: protection occurred at high-side frequencies 14 kHz and slight TTS exacerbation occurred at within-band frequencies. Peak protection was 15 dB from 17 to 19 kHz, and exacerbation was maximally 7 dB at 8-9 kHz. The conjoint effect of both pathwaysthe total OC effect in OC+ ears compared with OC ears (Fig. 1B)was significant protection from 7 to 16 kHz.

With 1- to 6-kHz noise (Fig. 1, D-F), TTSs from 2 to 12 kHz in efferent-intact ears were lower than in de-efferented ears. Protection >5 dB in efferent-intact ears occurred from 3 to 11 kHz (Fig. 1E; Table 1). With only CMOCS lesioned, TTSs at within-band frequencies of 3-4 kHz were now greater than in efferent-intact ears and similar to those in de-efferented ears, but at high-side frequencies (7-12 kHz) were even lower than in efferent-intact ears. The separate effects of the two MOCS pathways, calculated as in the preceding text, are illustrated in Fig. 1F. The CMOCS had significant protective effects at within-band frequencies of 3 and 4 kHz (peak 11 dB at 4 kHz). At 5-6 kHz, it had no effects, and at high-side frequencies of 7-10 kHz, it caused a broad desensitization of 12 dB. Conversely there were no UMOCS effects at within-band frequencies <5 kHz, but from 5 to 12 kHz, there was strong protection, peaking at almost 30 dB at 9 kHz. The conjoint total OC effect (Fig. 1E) was protection at frequencies from 3 to 12 kHz in OC+ ears compared with OC ears.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

As noted, crossed OC effects must be due to the crossed MOCS to OHCs. Uncrossed OC effects are likely UMOCS effects on OHCs than effects of the uncrossed LOCS to afferent dendrites: TTSs 5 min postexposure (as here) appear to be only OHC effects (Patuzzi et al. 1989) and, also, actions at afferent dendrites are unlikely to counter CMOCS effects on OHCs as observed here.

The two OHC efferent subsystems, acting through nicotinic cholinergic receptors, exerted complex effects. For both, noise bands crossed MOCS exerted two effects: protection at within-band frequencies, but at high-side frequencies outside the noise, it contributed to "TTSs." Uncrossed MOCS exercised, at most, small effects at within-band frequencies but protected at high-side frequencies. Differences in cross-cochlear innervation densities (Warr et al. 1986) or in slow OC effects (Sridhar et al. 1997) may account for some differences in effects with the two noise bands. Both sets of CMOCS effects were stronger for the high-frequency noise. Second, UMOCS exercised no within-band effects for low-frequency noise but a small depressive effect for high-frequency noise and caused stronger high-side protection for low-frequency noise.

Crossed MOCS protection at within-band frequencies is consistent with the fact that with traumatic pure tones, it only protectsi.e., it acts in stimulus frequency-appropriate cochlear regions to protect from trauma. Novel findings are that at frequencies higher than trauma spectral content, it exerted depressive effects and that these were opposed by the uncrossed MOCS. It is unlikely the two crossed MOCS effects are causally related because crossed MOCS protection from traumatic pure tones occurs without MOCS-induced TTS exacerbation at frequencies higher than the tone or than the range suffering TTSs (Rajan 1995). The presence of two crossed MOCS effects likely relates to the fact traumatic sound here was spectrally broadband and may be high-level extensions of lower-level roles with such sounds.

One such role may be contrast enhancement for narrowband sounds: crossed MOCS enhancement of within-band sensitivity, while depressing high-side sensitivity could result in strong edge-enhancement. (Because of cochlear mechanics, for contrast enhancement, suppression would primarily be required at frequencies higher than noise content.) At atraumatic levels both CMOCS effects may be small but sufficient for contrast enhancement. At traumatic levels, stronger CMOCS drive may, incidentally, protect within-band frequencies but worsen high-side TTSs. To prevent the latter, the uncrossed MOCS appears to act in a state-restoration role to selectively counteract only high-side CMOCS effects. (This could account for the fact that with loud tones, where CMOCS protection occurs without high-side depression of sensitivity, there is no UMOCS effect.)

This proposed contrast-enhancement role for the CMOCS likely complements another postulated MOCS role, of "unmasking" to improve detection of tonal stimuli in noise (Winslow and Sachs 1988). Different mechanisms appear involved. Unmasking appears to involve both UMOCS and CMOCS, causing similar cochlear effects (Kawase et al. 1993), whereas the proposed contrast-enhancement role appears to involve a remarkable diversity of CMOCS effects. Finally, with loud broadband sound, there is also a UMOCS effect that opposes one CMOCS effect and may restore cochlear state.