| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratoire d'Audiologie Expérimentale et Clinique, Unité Propre de Recherche-Enseijnement Supérieur Université de Bordeaux II, Hôpital Pellegrin, 33076 Bordeaux, France
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Lima da Costa, Deise, Anne Chibois, Jean-Paul Erre, Christophe Blanchet, Renaud Charlet de Sauvage, and Jean-Marie Aran. Fast, slow, and steady-state effects of contralateral acoustic activation of the medial olivocochlear efferent system in awake guinea pigs: action of gentamicin. J. Neurophysiol. 78: 1826-1836, 1997. The function of the medial olivocochlear efferent system was observed in awake guinea pigs by recording, in the absence of ipsilateral external acoustic stimulation, the ensemble background activity (EBA) of the VIIIth nerve from an electrode chronically implanted on the round window of one ear. The EBA was measured by calculating the power value of the round window signal in the 0.5- to 2.5-kHz band after digital or analog (active) filtering. This EBA was compared with and without the addition of a low-level broadband noise to the opposite ear. The contralateral broadband noise (CLBN, 55 dB SPL) induced, via the efferent system, a decrease (suppression) of this EBA. With the use of noise bursts of different durations, two components in this suppression could be observed. After the onset of a 1-s CLBN, the power value of the EBA decreased rapidly by 38.0 ± 4.2% (mean ± SD, n = 3), with a latency of <10 ms and a decay time constant of 13.1 ± 1.0 ms (fast effect). At the offset of the 1-s CLBN, EBA came back to prestimulation values with a similar latency and a time constant of 15.5 ± 2.9 ms. During longer CLBN stimulation (
1 min), EBA presented, after the fast decrease, an additional, slower decrease of 15.6 ± 3.1%, with a delay of 9.8 ± 1.3 s and a decay time constant of 16.1 ± 5.0 s (n = 12, slow effect), and then remained remarkably constant for as long as observed, i.e., >2 h (steady state). The average global suppression was thus up to 47.8 ± 5.8% of the basal, pre-CLBN-stimulation EBA value. At the offset of the CLBN, EBA returned to pre-CLBN level with fast and slow phases, with, for the slow phase, no delay and a time constant of 32.1 ± 8.1 s. Fast and slow changes in EBA power values were observed after a single injection of gentamicin (GM) at different doses (150, 200, and 250 mg/kg). At 150 and 200 mg/kg, GM progressively and reversibly blocked the rapid effect, but the slow component of the efferent medial suppression remained remarkably unchanged. However, at higher doses both the fast and slow suppressions were totally yet still reversibly blocked. These observations indicate that the medial olivocochlear efferent system exerts sustained influences on outer hair cells and that this effect develops in two different steps that may have different basic cellular mechanisms.
The medial efferent nerve fibers seem to regulate the contractile state of the outer hair cells (OHCs) and thus modulate the mechanical properties of the cochlea (Brownell et al. 1985 Experimental animals were female pigmented GPs chronically implanted with an indwelling platinum ball electrode on the RW membrane of the left ear and reference screw electrodes on the skull. Implantation was realized under general anesthesia with an intramuscular injection of xylazine (8 mg/kg) and ketamine (33 mg/kg). The wires from the electrodes terminated in a five-pin electronic connector fixed with the screws and methacrylate dental cement onto the skull. Details of the implantation technique can be found in Aran and Erre (1979) CLBN stimulation
Electrophysiological measurements were performed in the GPs in an acoustically attenuated chamber (IAC model AC-2, New York, NY) placed in a quiet laboratory room. The signal-generating and recording equipment was situated in an adjacent room, except for a preamplifier that was placed inside the soundproof room. Each animal was placed in a standing-yoke tube with the animal's head fixed by a sliding ring placed over the nose.
Recording and analysis of the EBA
The signal from the RW/vertex electrodes was amplified 5,000 times with a custom-made, low-noise preamplifier (band pass 3 Hz to 10.0 kHz, 6 dB/octave, 0.5-µV RMS noise in the 3-Hz to 10.0-kHz band). After further amplification (3 or 10 times), the signal was recorded and analyzed through two different procedures, depending on the duration of the recorded epoch.
Recordings during intermittent CLBN stimulation
SHORT-DURATION CLBN (1 S).
For short-duration recordings, the signals were recorded in a PC-based CED data acquisition system (Cambridge Electronic Design, Cambridge, UK, model 1401-plus, 12-bit A/D conversion) and sampled at 12,000 Hz. Single recordings consisted of a 2-s epoch made with a 1-s CLBN stimulation digitally triggered 0.5 s after the start of the recording. Successive recordings were separated by intervals close to 20 s. Recordings that included muscular movement or electrical artifacts were rejected. The raw signals were analyzed off-line (LabView software, National Instruments, Austin, TX, Fig. 1). Namely, the signal was digitally filtered (Butterworth, order 2) between 0.5 and 2.5 kHz. Then the filtered signal was squared and the mean power was calculated every millisecond by averaging over 12 successive points. The resulting power signal was visualized on a video screen and stored for further analysis (Fig. 1).
LONG-DURATION CLBN ( Recordings during continuous CLBN stimulation
Measurements of EBA power values were also performed during continuous CLBN (up to 2-3 h after an intramuscular injection of NaCl 0.9% or GM) in three GPs with the use of series of three to five 2-s-average recordings. With these short-duration recordings, it was not necessary to sedate the animals; signals with movement artifacts were manually rejected.
Experiments with GM
Most of the GPs used in the above experiments subsequently received a single intramuscular injection of GM (Gentamicin,Dakota Pharm, Créteil, France). In five of the GPs used in the 1-min CLBN normative study, similar recordings (2-min recordings with 1-min CLBN) were performed before and after a 150-mg/kg GM injection. The 12 GPs used in the other experiment (5-min recordings with 1.5-min CLBN) were divided into three groups that received an intramuscular injection of GM at doses of 150, 200, and 250 mg/kg (6, 3, and 3 GPs, respectively) and were similarly tested. In both cases, between the series of recordings (10-15 min intervals) there was no acoustic stimulation. Changes in CL suppression were followed during the 2-8 h after the injection. Measures were repeated 24 and/or 48 h afterward. The three GPs submitted to the >2-h continuous CLBN, received, a few days after the serum injection, a 150-mg/kg GM injection and were subsequently followed as after the serum injection and 24 and 48 h afterward.
1-s CLBN
Within a few milliseconds after the onset of the 55-dB SPL CLBN, the EBA amplitude exhibited a reduction of 38.0 ± 4.2% (mean ± SD, n = 3, Figs. 2 and 3). The latencies of the fast onset and offset of suppression following the CL stimulation were typically between 5 and 10 ms (Fig. 3). The decay and recovery of the EBA power amplitude, obtained after averaging 100 recordings, could be approximated by nonlinear regressions to exponential functions, with time constant of 13.1 ± 1.0 ms (onset) and 15.5 ± 2.9 ms (offset).
CLBN > 1 min
When the CLBN was presented for
Continuous CLBN
Three GPs were submitted to different set of measurements before and after an intramuscular injection of NaCl 0.9% (see METHODS and Fig. 8) used as a control for further experiments in which the effects of a single injection of GM in the same GPs were investigated (see below). The mean EBA power value without CLBN (EBA0) was found to be 63.7 ± 7.4 µV2 (Table 2). After the fast effect (EBA1), measured with 1-s CLBN, the mean EBA power value fell to 40.5 ± 9.6 µV2. Then, the CLBN was presented for 1 min and the mean EBA power value decreased further to a stable, low value (EBA2) of 29.7 ± 7.1 µV2. The suppression coefficients, relative to the power level without CLBN, were S = 53.7 ± 7.3% (global suppression), S1 = 37.0 ± 8.7% (fast component), and S2 = 16.7 ± 4.2 (slow component). Then the continuous CLBN was applied for >2 h. As shown in Fig. 8, the power level (EBA2) during the >2-h continuous CLBN remained constant. The global suppression (S) was relatively unchanged from the beginning (1 min after the onset) and throughout the continuous >2-h stimulation period. After the end of the continuous CLBN, although EBA0 was slightly smaller than before, both fast (S1) and slow (S2) suppression levels were also unchanged (Table 2, Fig. 8).
Effects of GM
DURING CONTINUOUS CLBN.
The same three GPs (Table 2) were further submitted to the same set of measurements, except that a few days later the intramuscular injection of NaCl 0.9% was replaced by an intramuscular injection of GM (150 mg/kg). During continuous CLBN stimulation, we observed an increase of the EBA values (EBA2, i.e., a decrease in CL suppression), which started <30 min after the injection and continued progressively until the end of the 2- to 3-h CLBN (Fig. 8). After termination of the continuous CLBN, EBA0 (without CLBN) was slightly larger than before, the fast effect (S1) was almost completely suppressed (it completely disappeared in 1 GP, Fig. 8); the slow effect (S2), however, was unchanged. In other words, GM seemed to affect only the fast component, but not the slow one. Such observations were consistent in the three GPs studied (Table 2). For each GP, fast and slow ON effects were examined 24 and 48 h after GM injection and a complete recovery was observed (see Fig. 8).
DURING INTERMITTENT CLBN.
The fast, slow, and global components of the CL suppression were also observed in five GPs with the use of intermittent 1-min CLBN stimulations (during the 2-min recording epochs) before and at different times (10-min intervals) after a single intramuscular injection of GM (150 mg/kg). Initially, the global suppression was S = 51.6 ± 8.3%, and the fast and slow suppression effects were S1 = 36.8 ± 9.3% and S2 = 14.8 ± 4.8%, respectively. Between 30 min and 1 h after the GM injection, the EBA amplitude during the CLBN started to increase (i.e., the suppression effect decreased) and maximal decrease in suppression was observed within 1.5-3 h postinjection; then suppression started to slowly increase and was completely back to preinjection value 24 or 48 h later. At time of maximal effect of GM, S was reduced to 25.2 ± 5.9%. The fast component of the suppression (S1) was significantly reduced to 11 ± 3.5%, whereas the slow component (S2) was unchanged (14.2 ± 3.3%). However, at this time, as well as at any other time, particularly at 24 and 48 h, EBA amplitude without CLBN (EBA0) was not significantly different from its value before the injection, and CAP thresholds did not change significantly.
In the present study we demonstrate, in the awake animal, that CL acoustic stimulation triggers fast and slow suppressions of basal ipsilateral cochlear activity followed by fast and slow return to prestimulation levels after the end of the CLBN. We characterized these different effects in terms of amplitudes, latencies, and time constants. Interestingly, we showed that the fast and slow ON effects led to a steady-state level that remained stable for hours. Furthermore, we showed that after a single injection of GM the fast effect only was suppressed at 150 mg/kg, whereas both the fast and slow effects were completely suppressed at higher doses (250 mg/kg). These effects were all completely reversible within 24-48 h, and the afferent function, as assessed by CAP thresholds and EBA amplitude without CLBN (EBA0), was never significantly affected.
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Fex 1967; Kemp 1978; Mountain 1980; Siegel and Kim 1982; Weiderhold 1986). Activation of this medial olivocochlear (OC) efferent system can be achieved by electrical stimulation of the OC bundle (OCB) on the floor of the fourth ventricle (Galambos 1956) or of the round window (RW) of the contralateral (CL) ear (Rajan and Johnstone 1983). It is also more physiologically activated by acoustic stimulation of the CL ear. During or briefly after CL acoustic stimulation, a suppression of the ipsilateral cochlear activity has been observed in single afferent nerve fiber responses to ipsilateral sounds (Buño 1978), in tone-pip-evoked VIIIth nerve compound action potentials (CAPs) (Liberman 1989), in otoacoustic emissions (Guinan 1986; Puel and Rebillard 1990), and in the ensemble background activity (EBA) of the VIIIth nerve (Popelar et al. 1994, 1997). The first described OC effects occurred within time constants for onset and offset of suppression in the 50- to 100-ms range (Wiederhold and Kiang 1970). More recently, an additional slow suppression of cochlear response has been described with a decay time constant of 25-50 s; this was termed the "slow" effect (Sridhar et al. 1995), in comparison with the "fast" effect originally described. These different fast and slow effects were observed on cochlear potentials after repeated electrical stimulation of the OCB at the floor of the IVth ventricle in anesthetized, tracheotomized, and artificially ventilated guinea pigs (GPs).
, 1997). The spectrum of this background bioelectrical noise, in the absence of any ipsilateral acoustic stimulation, presents a peak at ~1.0 kHz that is supposed to derive mainly from partly synchronized ongoing activities of auditory nerve fibers (particularly fibers with high characteristic frequencies), either spontaneous or induced by endogenous internal noise (Cazals and Huang 1996; Dolan et al. 1990; Lima da Costa et al. 1997c). A low-level CL broadband noise (CLBN) significantly reduces the EBA, similar to what is observed for tone-pip-evoked CAPs and otoacoustic emissions. These suppressive effects are commonly accepted as being mediated by the CL activation of the medial efferent system innervating the OHCs of the ipsilateral cochlea.
; Smith et al. 1994).
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
. The animals were allowed to recover for 1 wk before normal cochlear function was assessed by measuring the VIIIth nerve CAP thresholds to ipsilateral tone pips (rise-fall time 2 ms, no plateau) from 0.5 to 25 kHz (CAP audiogram). At various times before and after GM treatment, CAP thresholds were also determined. Animal handling throughout these experiments was performed with authorization of the French Ministery of Agriculture in agreement with EEC regulations.
). The level and the frequency spectra varied little between GPs or in the same GP with several repeated placements of the tube in the ear canal. The primary components of the broadband noise were in the frequency range of 1.5-13.0 kHz; the maximum difference between GPs did not exceed 5 dB for frequencies <9.0 kHz and did not exceed 10 dB for frequencies >9.0 kHz. The CL noise level used was typically 55 dB SPL, which is in the range of maximum suppression effect, supposedly without any risk of cross talk from one ear to the other (Avan et al. 1996; Popelar et al. 1997; Smith et al. 1994). The left ear was occluded with cotton to further attenuate any outside sound. The most commonly studied durations were 1 s, 1 and 1.5 min, and ~2 h. Each CLBN had 2.5-ms rise-fall times.
View larger version (90K):
[in a new window]
FIG. 1.
Front panel LabView of 2-s recordings signal analysis, with the 1-s contralateral broadband noise (CLBN) presented 0.5 s after the beginning of the record (black horizontal bar). Top trace: raw signal after ×15,000 amplification with 3-Hz to 10-kHz bandwidth (6 dB/octave) and A/D conversion (12,000 Hz). Middle trace: signal after digital filtering in the 0.5- to 2.5-kHz band (Butterworth order 2). Bottom trace: mean power of the filtered signal calculated over each ms. This is the signal that is averaged, after 100 or 500 acquisitions, to display the time constants of the suppression (see Figs. 2 and 3). Numbers at bottom: mean power calculated over the periods without (µV2 WO) and with (µV2 W) CLBN (after suppression of the 1st 50 ms after the onset and offset of the CLBN) and corresponding suppression coefficient.
2. Next we calculated, as before, the power of this sum signal per millisecond and averaged successive individual 2-s pairs. The resulting power signal, thus averaged over half the number of sweeps, presented without the peak at the onset of the CLBN, but with a fluctuation 2 times larger than without this AEP cancellation procedure.
View larger version (30K):
[in a new window]
FIG. 2.
Top: graphs displaying average of the power signal (see Fig. 1) before, during, and after the presentation of the 55-dB SPL CLBN, without (left) and with (right) the auditory evoked potential (AEP) cancellation procedure, in 1 guinea pig (GP, n = 500 sweeps). Bottom: this AEP cancellation procedure is explained by showing the averaged recordings of the unfiltered round window (RW) signal over 30 ms at time of the onset (left, 
) and of the offset (right, ) of the 1-s CLBN, without (a) and with (b) alternating polarity of the RW signal (n = 500 sweeps); this shows the suppression of the evoked response in the later condition (b).
View larger version (21K):
[in a new window]
FIG. 3.
Details of the time course of the power values of the RW power signal from Fig. 2 (with AEP cancellation) after the onset (left) and the offset (right) of the 1-s CLBN. Nonlinear regression analysis (exponential functions) after the onset and offset are represented by solid curves and the corresponding equations (time in ms).
1 MIN).
Because the awake animals would not stay quiet enough over longer duration periods (in the range of min), for recording they had to be sedated with one intramuscular injection of xylazine (
12 mg/kg) and their body temperature had to be maintained at ~38°C. In a previous study we demonstrated that in such conditions the EBA power value and the CL suppression were not affected (Lima da Costa et al. 1997c).
1 min. Because the duration of the CLBN was determined by an analog pulse generator, its duration could not be accurately fixed; its onset triggering, however, was precisely determined by the data acquisition computer.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
1 min, the EBA first showed a fast decrease followed by a slower decrease for ~30 s, and then the EBA amplitude remained constant until the end of the CLBN. At the offset of the CLBN, EBA recovery also displayed fast and slow components. However, the complete return of the EBA to the pre-CLBN level did not occur before 30 s and thus was measurable only on the 5-min recordings, which lasted >2 min after the end of the CLBN. A schematic representation of the time courses and amplitudes of EBA changes during and after the CLBN is represented in Fig. 4, and individual recordings taken from different animals before the injection of GM (0 h) can be seen in Figs. 5-7. Several levels of EBA amplitude are distinguished: before CLBN (EBA0), just after the fast effect at the onset of the CLBN (EBA1), and after the following slow effect (EBA2). Most of the time a delay (d) of several seconds can be clearly observed between the fast effect and the beginning of the slow effect. Then the offset of the CLBN is followed by a partial fast recovery of EBA amplitude to a third level (EBA3), usually different (larger) than EBA1, and a slow progressive return to a post-CLBN level EBA4 identical to the pre-CLBN level (EBA0). These different steps can be expressed as different suppression effect coefficients in relation with the CLBN presentation: a fast and slow ON suppression (S1 and S2, respectively) and similarly a fast and slow OFF recovery (S3 and S4, respectively). The global suppression (S) is thus the sum of S1 + S2 (SON) or S3 + S4 (SOFF, Fig. 4). The slow ON and OFF effects could be fitted with exponential functions providing time constants 
ON and OFF. Mean values calculated from measurements in the 12 GPs with the 5-min recording epochs and 1.5-min CLBN are summarized in Table 1 and Fig. 4. Although there are no statistically significant differences between the ON and the OFF amplitudes of the fast and slow effects, the ON and OFF slow time constants are significantly different (P < 0.01): the slow OFF time constant (OFF = 32.1 ± 8.1 s) is about twice the slow ON time constant (ON = 16.1 ms ± 5.0 s).
View larger version (23K):
[in a new window]
FIG. 4.
Schematic representation, with the use of mean values from Table 1, of the time course of changes in ensemble background activity (EBA) power before, during, and after the 1.5-min CLBN (black horizontal bar) over the 5-min recording epoch. Mean EBA amplitude is measured in % of its value before the CLBN (EBA0, 100%). After the CLBN onset, at time 0, there is a fast decrease from EBA0 to EBA1 levels corresponding to a fast ON suppression coefficient (S1). This fast effect is followed, with a delay (d), by an additional slow decrease from EBA1 to EBA2 corresponding to a slow ON suppression coefficient (S2), with a slow ON time constant ON indicated in the figure. Then EBA remains constant until the end of the CLBN. At the offset, there is a fast and a slow recovery from levels EBA2 to EBA3 and EBA3 to EBA4 with corresponding reversed suppression coefficients S3 and S4 and a slow OFF time constant OFF. Global effect (S) is thus measured from EBA0 to EBA2 (SON) or from EBA2 to EBA4 (SOFF).
View larger version (17K):
[in a new window]
FIG. 5.
Recordings of EBA before, during, and after the 1.5-min CLBN (black horizontal bar) at different times (from 0 to 48 h) after intramuscular injection of 150 mg/kg gentamicin (GM) in GP 264.
View larger version (17K):
[in a new window]
FIG. 6.
Same as in Fig. 5 for another GP (GP 301) after injection of 200 mg/kg GM.
View this table:
TABLE 1.
Mean parameter values
View larger version (16K):
[in a new window]
FIG. 7.
Same as in Figs. 5 and 6 for another GP (GP 304) after injection of 250 mg/kg GM.
ON were observed in the seven GPs studied with the 1-min CLBN and 2-min recording, respectively (S = 49.7 ± 9.5%, S1 = 33.9 ± 9.4%, S2 = 15.8 ± 5.2%, ON = 18.4 ± 7.4 s).
View larger version (14K):
[in a new window]
FIG. 8.
EBA power values measured before and after the intramuscular injection ( ) of serum (left) and GM (150 mg/kg, right) in 1 GP. EBA is measured in silence (
), during the 1-s CLBN (
) just before the end of 1-min CLBN, or during continuous (>2 h) CLBN (
). Measurements of the fast and slow effects were performed again after the end of the continuous CLBN (i.e., ~2 h after the injection) and, for GM, 24 and 48 h afterward.
View this table:
TABLE 2.
EBA power values and coefficient suppressions before and 2-3 h after the intramuscular injection of NaCl 0.9% and GM (after 2-3 h of continuous CLBN stimulation)
View larger version (13K):
[in a new window]
FIG. 9.
Means ± SD of the global (S), fast (S1), and slow (S2) suppression coefficients before and at time of maximum change after GM at doses of 150, 200, and 250 mg/kg (n = 6, 3, and 3, respectively). Measurements are based on 5-min recordings with 1.5-min CLBN (Figs. 4-7).
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
). As for the slow effect, Sridhar et al. (1995), on the basis of the action of numerous cholinergic antagonists, tended to demonstrate that this is also related to activation of the medial efferent system. Indeed, with their CAP recordings as well as with our observations on the EBA of the VIIIth nerve, there is no direct evidence that the lateral system controlling single auditory nerve fiber activities is not involved. However, in the monitoring of the EBA of the VIIIth nerve during CL acoustic stimulation, without any ipsilateral acoustic stimulation, there is no involvement of the numerous ipsilateral lateral efferents; and only the crossed efferents from the CL side, thus most exclusively medial efferents (Warr and Guinan 1979), are activated by the CLBN.
) or 50-250 ms (Brown and Nuttall 1984) for the rapid effects. The 25- to 50-s time constant for the slow effect reported by Sridhar et al. (1995) is the slow decay time constant after the end of the OCB stimulation; it compares well also with our finding of a slow OFF time constant (OFF) of 32.1 s. The fact that our measurements provide generally shorter time constants might be attributed to differences in the general condition of the experimental animals (awake vs. anesthetized). Indeed, anesthetic agents are known to alter the physiology of the peripheral auditory system (Dodd and Capranica 1992) and, more specifically, to slow down the sound-evoked efferent activity (Brown 1989; Liberman and Brown 1986; Lima da Costa et al. 1997c). This should lead to an overestimation of time constants. Another source of overestimation, especially for the fast effect, might arise from the examination of offset suppression instead of onset. Also, as suggested by our values, fast and slow ON suppression effects (13.1 ms and 16.1 s) appear to be shorter than fast and slow OFF recovery effects (15.5 ms and 32.1 s). Moreover the delay we observed in the development of the ON slow effect after the fast (9.8 ± 1.3 s) is similar to that already reported by Sridhar et al. (1995) and also remarquably identical to the delay recently observed by Dallos et al. (1997) in the change in the electromotile response of isolated OHCs after the application of acetylcholine (ACh).
reported fast components in CAP suppression that were twice as large as the suppression coefficients we found in this study. This might be explained on the basis of the two fundamental differences in the methods employed. First, the measures are different: Sridhar et al. measured synchronized responses to ipsilateral stimulation, whereas we measured spontaneous and/or only partly synchronized activities (Cazals and Huang 1996; Lima da Costa et al. 1997c), i.e., supposedly without any ipsilateral stimulation. Along these lines, our results are similar to those of Guinan and Gifford (1988), who measured efferent suppression of single-unit spontaneous activity. They found that, during the 1st s of crossed medial OC electrical stimulation, the spontaneous firing rate was depressed by up to 35%. Second, the electrical stimulation is expected to produce a larger release of ACh at the efferent-OHC synapses than that induced by the low-level activation of the efferent bundle by the 55-dB SPL CLBN.
and Sridhar et al. (1995), who reported that the effect could be maintained only for 1-2 min. Besides the different physiological condition of the animal (anesthetized vs. awake), it is likely that the differences might arise from the difference in the activation, and its consequences at the synaptic level, of the medial OC fibers.
), which may lead to desensitization of OHCs' ACh receptors. Accordingly, in the 300-s continuous shock paradigm presented by Reiter and Liberman (1995, their Fig. 4c), and in the 250-s shock train displayed by Sridhar et al. (1995, their Fig. 5c), there was a net reduction of the total suppression after 50-100 s, but it included a decrease of both the rapid and slow components. Otherwise, electrical stimulation rates commonly used to stimulate the OCB are ~300-400 shocks per second, whereas monaural stimulation with tones induces efferent fiber activities that rarely exceed 60 spikes/s (Liberman and Brown 1986; Robertson and Gummer 1985). Even high-level sound noise stimulation does not induce efferent activity at rates >140 spikes/s (Liberman 1988). Thus it is possible that electrical stimulation of the OCB for a few minutes causes an exhaustion of ACh contents at the efferent terminal level, preventing the adequate refill of ACh stores in the face of the sustained electrical stimulation. However, no efferent fatigue was seen under continuous acoustic stimulation up to several hours in awake or lightly sedated GPs. In other words, long-duration CL acoustic stimulation did not alter the effectiveness of the medial efferent system. We presume that no "depletion" is imposed on the efferents, and that, consequently, the cellular homeostasis is sustained and all the further molecular steps are preserved, comprising a continuous and regular release of ACh from presynaptic sites. Thus low-level CL acoustic stimulation appears to be a more physiological method of activating the medial efferent system. Such long-lasting effects are also consistent with the assumption that OC efferents might protect the ear against acoustic overstimulation (for review see Rajan 1992), notably when efferent stimulation parallels overstimulation exposure (Rajan 1988).
. Those researchers sectioned the OCB immediately after a train of shocks and observed that the magnitude of the lingering suppression of the CAP and the time required for recovery were similar to control responses with an intact OCB. Our results support this idea, because GM at 150 mg/kg seems to produce a progressive pharmacological inhibition of the medial efferent fast effect, whereas the magnitude of the slow effect (S2) remained unchanged. This effect of GM also suggests that the slow effect is not due, in our experimental conditions, to mobilization of additional efferent fibers or to any increase in ACh concentration at the efferent synapses over time.
also suggest that fast and slow effects are mediated by the same ACh receptor, because both are similarly blocked by a variety of cholinergic antagonists. Because of the widely varying time scales between the two effects, the authors proposed that fast and slow effects are produced by different intracellular mechanisms, possibly mediated by second-messenger systems. First, binding of ACh to the receptor would induce a small calcium influx that activates nearby calcium-activated potassium channels (Blanchet et al. 1996; Eróstegui et al. 1994; Evans 1996; Fuchs and Murrow 1992; Housley and Ashmore 1991). The hyperpolarizing potassium current generated might be responsible for the fast effect. Sridhar et al. (1995) suggested that the calcium entry associated with the fast effect is also responsible for generating the slow effect by activating distant calcium-activated potassium channels either directly or indirectly. Recent reports (Dallos et al. 1997; Sridhar et al. 1997) argue in favor of calcium as the second messenger involved in the slow effect. It is even suggested that a calcium-induced calcium release from intracellular stores amplifies the signal. We agree that calcium influx through ACh receptor channels may be responsible for the slow effect. We do not believe, however, that this slow effect is mediated by additional calcium-activated potassium channels, because in isolated OHCs, the ACh-induced current does not increase, on a time scale of a few minutes, but desensitizes (Nenov et al. 1996). We agree that calcium, possibly after amplification, triggers the slow calcium-dependent lengthening of OHCs (Dulon et al. 1990) and the decrease of their longitudinal stiffness (Dallos et al. 1997). These phenomena can be slow and would take time to develop. They could afford the delay of ~10 s and the low time constant of the ON slow effect. The absence of delay and the longer time constant of the slow OFF effect that we report here would correspond to the slow clearing out of free calcium and a progressive decrease of its effects on the lateral wall membrane of the OHC after the end of the CLBN, with a slow relaxation of the contractile proteins likely involved in this muscular-like micromechanical mechanism. However, according to Murugasu and Russel (1996), these changes in length and stiffness do not seem to modify the tuning of the basilar membrane, but they decrease its sensitivity and shift downward the CF. Basilar membrane velocity has also been shown recently to increase during OCB stimulation (Dolan et al. 1997). However, these effects might not appear in our global recordings. Still, this hypothesis may explain the protection by the efferent slow effect against temporary threshold shifts induced by high-level tones (Sridhar et al. 1997).
). Indeed, no suppression effect was observed and a similar slight increase of ipsilateral EBA could be recorded, particularly with higher levels of CLBN (60 dB SPL), whereas usually with 50 dB SPL no such EBA increase could be evoked. Moreover, when we applied the same broadband noise to the ipsilateral ear, a similar small EBA increase could be obtained with ipsilateral broadband noise levels as low as a few dB SPL. Thus it appears that the55-dB SPL level chosen for the CL stimulation was slightly over the interaural attenuation. A 45- or 50-dB CLBN would have been more appropriate, because the suppression effect would still be quite evident: it can be observed with levels of CL stimulation as low as 20 dB SPL (Popelar et al. 1997), and there would be practically no chance of interaural cross talk. Thus in these experiments the presumed absence of ipsilateral sound was not exactly true. However, we do not believe that this slight cross talk significantly altered the results: 1) the EBA before and after the CLBN (EBA0) was indeed typically without any ipsilateral acoustic stimulation; 2) the CL suppression in the normal condition was so large that it completely masked the eventual slight excitatory response; 3) in any case the suppression measured in our experimental conditions was thus only underestimated; and 4) it demonstrates the sensitivity of the EBA recordings!
), the distortion products acoustic emissions (Avan et al. 1996), and more recently the EBA of the auditory nerve (Popelar et al. 1997). Looking at the experimental parameters used in these different studies, particularly the duration of the CL stimulation, it appears that most of them observed the fast effect only, whereas the study on distortion product otoacoustic emissions, because of the continuous CL pure tone acoustic stimulation, must have investigated the global fast and slow effects (Avan et al. 1996). For the first time to our knowledge, we describe the evolution of the fast and slow components of medial OC effects over time following a GM injection. Moreover, using different doses, we demonstrate that the effect is dose dependent. In another study (Lima da Costa et al. 1997a), we investigated the effects of lower GM doses (60, 90, 120, and 150 mg/kg) on the fast suppression effect. No changes were observed with 60 and 90 mg/kg; a slight but significant effect was observed at 120 mg/kg, although smaller than with 150 mg/kg. The results with the dose of 150 mg/kg are interesting in that the fast effect only appears affected, whereas the slow effect is apparently unchanged. However, at the higher dose the slow effect also disappears.
). In this case, small and decreasing amounts of calcium would enter into OHCs, leading to a smaller hyperpolarization and so to a reduced fast effect. This calcium entry, however, could be sufficient to ensure slow OHC movements dependent on calcium if an amplifier mechanism, such as a calcium-induced calcium release from intracellular stores, is involved. Indeed at higher doses GM would completely inhibit the entry of calcium and no effect whatsoever could develop. This blockade of calcium entry could be due to the repulsion at the ACh receptor site of Ca2+ ions by the positively charged GM molecules.
| |
ACKNOWLEDGEMENTS |
|---|
The authors are greatly indebted to D. W. Smith for critically reading the manuscript and editing the English.
This work was supported by the Institut National de la Santé et de la Recherche Médicale and the University of Bordeaux II (UPR-ES). D. Lima da Costa was supported by a Doctoral scholarship from Conselho Nacional de Desenvolvimento Cientifico e Technologico, Brazil.
| |
FOOTNOTES |
|---|
Address reprint requests to J.-M. Aran.
Received 24 February 1997; accepted in final form 13 June 1997.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Blanchet, C. Erostegui, M. Sugasawa, and D. Dulon Gentamicin blocks ACh-evoked K+ current in guinea-pig outer hair cells by impairing Ca2+ entry at the cholinergic receptor J. Physiol., June 15, 2000; 525(3): 641 - 654. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Yoshida, M. C. Liberman, M. C. Brown, and W. F. Sewell Gentamicin Blocks Both Fast and Slow Effects of Olivocochlear Activation in Anesthetized Guinea Pigs J Neurophysiol, December 1, 1999; 82(6): 3168 - 3174. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. van den Abbeele, J. Teulon, and P. T. B. Huy Two types of voltage-dependent potassium channels in outer hair cells from the guinea pig cochlea Am J Physiol Cell Physiol, November 1, 1999; 277(5): C913 - C925. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
