- Email: [email protected]
Hair cell loss and regeneration after severe acoustic overstimulation in the adult pigeon
Hearing Research 120 (1998) 109^120
Hair cell loss and regeneration after severe acoustic overstimulation in the adult pigeon Danping Ding-Pfennigdor¡, Jean W.Th. Smolders *, Marcus Muëller, Rainer Klinke Physiologisches Institut III, Klinikum der J.W. Goethe Universitaët, Theodor-Stern-Kai 7, 60590 Frankfurt am Main, Germany Received 10 November 1997; revised 10 March 1998; accepted 11 March 1998
Abstract The extent of hair cell regeneration following acoustic overstimulation severe enough to destroy tall hair cells, was determined in adult pigeons. BrdU (5-bromo-2P-deoxyuridine) was used as a proliferation marker. Recovery of hearing thresholds in each individual animal was measured over a period of up to 16 weeks after trauma. In ears with loss of both short and tall hair cells, little or no functional recovery occurred. In ears with less damage, where significant functional recovery did occur, there were always a few rows of surviving hair cells left at the neural edge of the basilar papilla. In the region of hair cell loss, numerous BrdU labeled cells were found. However, only a small minority of these cells were regenerated hair cells, the majority being monolayer cells. Irrespective of the extent of the region of hair cell loss, regenerated hair cells were observed predominantly in a narrow strip at the transition from the abneural area of total hair cell loss and the neural area of hair cell survival. With increasing damage this strip moved progressively towards the neural edge of the papilla. No regeneration of hair cells was observed in the abneural region of total hair cell loss, even up to 16 weeks after trauma. The results indicate that there is a gradient in the destructive effect of loud sound across the width of the basilar papilla, from most detrimental at the abneural edge to least detrimental at the neural edge. Both tall and short hair cells can regenerate after sound trauma. Whether they do regenerate or not depends on the degree of damage to the area of the papilla where they normally reside. Regeneration of new hair cells occurs only in a narrow longitudinal band, which moves from abneural into the neural direction with increasing damage. In the area neural to this band, hair cells survive the overstimulation. In the area abneural to this band, sound damage is so severe, that no regeneration of hair cells occurs. As a consequence morphological and functional deficits persist. z 1998 Elsevier Science B.V. All rights reserved. Key words: Bird; Pigeon; Acoustic trauma; Regeneration; Auditory nerve; Hair cell; Basilar papilla; 5-Bromo-2P-deoxyuridine labeling
1. Introduction The hair cells in the avian basilar papilla are capable of regeneration after acoustic or ototoxic trauma (reviewed by Cotanche et al., 1994; Cotanche, 1997). Ototoxic trauma a¡ects both short and tall hair cells, across the entire width of the basilar papilla. The damage begins in the base and progresses towards the apex with increase in the amount or duration of ototoxic treat-
* Corresponding author. Tel.: +49 (69) 63016980; Fax: +49 (69) 63016987; E-mail: [email protected]
ment (Duckert and Rubel, 1990 ; Hashino et al., 1991). In contrast, damage resulting from sound overexposure begins at the abneural side of the papilla and mainly damages the short hair cells (Saunders and Tilney, 1982 ; Ryals and Rubel, 1985 ; Cotanche et al., 1987; Marsh et al., 1990) at a longitudinal position related to the frequency and pressure level of the sound (Rebillard et al., 1982, Ryals and Rubel, 1982; Saunders and Tilney, 1982 ; Muëller et al., 1996). The short hair cells receive mainly e¡erent innervation (Fischer, 1992, 1994) and their role in inner ear function is still unclear (Gleich, 1989; Patuzzi and Bull, 1991 ; Manley and Gleich, 1992 ; Smolders et al., 1992, 1995). The majority of the a¡erent auditory nerve ¢bers innervate tall hair
0378-5955 / 98 / $19.00 ß 1998 Elsevier Science B.V. All rights reserved PII: S 0 3 7 8 - 5 9 5 5 ( 9 8 ) 0 0 0 5 5 - 0
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
110
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
cells, which are not destroyed by most sound overexposure regimes up to 125 dB SPL, even with very long exposure times (Cotanche and Dopyera, 1990; Raphael, 1991 ; Saunders et al., 1993; Pugliano et al., 1993). Functional recovery after such acoustic trauma is very fast (a few days, McFadden and Saunders, 1989; Saunders et al., 1992; Niemiec et al., 1994; Muëller et al., 1996) and occurs largely before the newly generated short hair cells become mature. This type of recovery is believed to result from the re-establishment of function of the surviving hair cells and accessory inner ear structures (Cotanche et al., 1994; Poje et al., 1995; Ryals et al., 1995), rather than from the regeneration of new hair cells (reviewed by Saunders et al., 1996a). Damage to the tall hair cells over the neural limbus of the inner ear occurs only at very high sound pressure levels, well above 125 dB SPL (Muëller et al., 1996, 1997). After such damage, functional recovery takes much longer (3^4 weeks) and is incomplete (Muëller et al., 1996, 1997). This type of functional recovery is very similar to that observed after aminoglycoside intoxication of the inner ear (Tucci and Rubel, 1990). The purpose of the present experiments was to determine the proliferation pattern of new hair cells after the destruction of both short and tall hair cells after very severe acoustic trauma. We used the S-phase marker, 5-bromo-2P-deoxyuridine (BrdU), to label new cells. The experiments were performed on adult pigeons. Sound trauma was severe enough to produce damage to both short and tall hair cells. The time course of functional recovery after trauma was recorded in both ears of each individual animal. 2. Methods 2.1. Animals Adult pigeons, 0.5^2 yr old, were obtained from a local breeder. The weight of the animals ranged from 350 to 550 g. The care and the use of these animals were approved by the German authorities for animal welfare (Regierungspraësidium Darmstadt). 2.2. Sound trauma Animals were exposed under general anesthesia to a pure tone of 0.7 kHz presented to the left ear for 1 h at a sound pressure level of 142 þ 2 dB SPL. The sound was produced by a Beyer DT 48 transducer ¢tted to the ear canal with an ear speculum. Sound pressure level was measured throughout the exposure using a calibrated BpK probe microphone (Type 4170) placed 1^ 2 mm from the eardrum. With this stimulus paradigm it is possible to cause damage to both the short and the tall hair cells. Details are given in Muëller et al. (1996).
2.3. Compound action potential audiograms All animals were implanted a few weeks prior to the start of the experiments with gold-wire electrodes placed at the round window of both cochleae for recording of compound action potential (CAP)-audiograms (details in Muëller et al., 1996). CAP-audiograms were measured before and immediately after the sound exposure. To monitor the functional recovery, CAPaudiograms were measured once or twice a week during the survival period (1^16 weeks after the acoustic overstimulation). The CAP-audiograms, response thresholds to gaussian-shaped tone pips (2/3 octave bandwidth), were measured in a frequency range between 0.1 and 8 kHz at 3 logarithmically equidistant frequencies per octave. 2.4. BrdU injections and immunohistochemistry The BrdU-labeling method was adapted from Raphael (1992) and Stone and Cotanche (1994). BrdU (Sigma) was dissolved in sterile phosphate-bu¡ered saline (PBS). To detect proliferating cells, four groups of pigeons were injected (i.m.) with BrdU at a dose of 50 mg/kg after the acoustic trauma. Group 1: 11 pigeons received three injections over a period of 6 h at day 2, between 40^54 h post exposure (day 0 is the day of sound exposure). Group 2: 12 pigeons received seven injections over four days (twice a day, day 0 to day 3, 0^76 h post exposure), in order to detect as many new cells as possible. To investigate if some cells in the basilar papilla still enter S-phase later than three days after sound trauma, 5 additional animals were used. Three of these (Group 3) received six injections between day 4 and day 6 (92^148 h post exposure, twice a day). The other two (Group 4) received three injections over a period of 6 h, respectively at day 7 (165^171 h post exposure) and day 8 (188^194 h post exposure). Pigeons survived for 1^2, 3^4, 8^10 or 16 weeks after the sound exposure and were then decapitated under general anesthesia. Both cochleae were extirpated, exposed at their basal and apical ends and ¢xed in 4% paraformaldehyde in 0.1 M phosphate bu¡er overnight at 4³C. Following ¢xation, cochleae were rinsed in PBS. The bone around the cochlear duct was taken away and the tegmentum vasculosum was removed. Cochlear ducts were treated with 2 N HCl in PBS containing 0.5% Triton X-100 at 37³C for 30 min to denature DNA. Endogenous peroxidase was inactivated by treatment with 0.3% H2 O2 for 20 min. To inhibit nonspeci¢c binding of the antibodies, cochlear ducts were immersed in blocking solution (2% normal horse serum, 2% bovine serum albumin (BSA) and 0.5% Triton X-100 in PBS) for 20 min. Following the above pretreatment, cochlear ducts were incubated overnight in mouse monoclonal antibody to BrdU (Boehringer Mannheim,
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
111
1:100 diluted in PBS), followed by horse antimouse biotinylated secondary antibody (Vector Labs) for 1.5 h and avidin-biotin-peroxidase-complex (Vector Labs, Elite ABC Kit) for 45 min. Thorough rinses in PBS followed each antibody treatment. Cochlear ducts were then incubated in 0.5% 3,3P-diaminobenzidine (DAB) and 0.01% H2 O2 in PBS to visualize the HRPreaction product. After staining, the tectorial membrane and cochlear ganglion were removed in most cochleae. Specimens were mounted in Kaiser's glycerol gelatin and observed under the light microscope (Olympus BH 2). The surface views of the basilar papilla were photographed and camera lucida drawings were made. To determine the speci¢c cell type that was BrdU-labeled, some specimens were ¢nally embedded in EPON, transversely sectioned (7 Wm) and counterstained with toluidine blue. 2.5. Controls for BrdU labeling As a control for the speci¢city of antibody binding, two pigeons received neither sound exposure nor BrdU. As a control for non-regenerative proliferation, two pigeons received only BrdU, but no sound exposure. One control pigeon received both sound exposure and BrdU injections as in group 1. Pigeons were decapitated 20 h after the last BrdU injection. Normal BrdU immunohistochemistry was performed in all cochleae. However, as a control for unspeci¢c binding of the secondary antibody, the treatment of primary antibody was omitted in two cochleae from the pigeon which had received both sound exposure and BrdU. 3. Results Forty-two cochleae from 23 pigeons from group 1 and 2, in which both the CAP-audiograms during recovery and the morphological preparation were successful, were used for the analysis of the functional-morphological correlation. Both the ipsi- (exposed, left) and the contralateral (right) cochlea of each animal were used since the contralateral cochleae were damaged by the sound exposure as well, due to the small interaural attenuation (Rosowski and Saunders, 1980; Muëller et al., 1996). 3.1. CAP-audiograms after acoustic trauma The CAP-audiograms of an individual ear (C117L) before, immediately post exposure (1 h, 0.7 kHz, 142 dB SPL) and 8 weeks after sound exposure are given in Fig. 1a. Fig. 1b depicts these audiograms as hearing loss relative to the mean audiogram of 91 untreated pigeon ears. The time course of the recovery of CAPthreshold from acoustic trauma at ¢ve frequencies in
Fig. 1. a: CAP audiograms of an individual ear (C117L) before, immediately after severe acoustic trauma (0.7 kHz, 142 dB SPL, 1 h) and after 8 weeks survival. Mean control audiogram is averaged from 91 adult normal pigeons. b: The same data plotted as hearing loss relative to the mean control audiogram.
the same ear is shown in Fig. 2. Immediately after sound exposure, hearing loss ranged from 35^53 dB depending on frequency. The loss recovered to stable values in about 4 weeks. Only minor recovery occurred thereafter. After 8 weeks of survival, a frequency dependent residual hearing loss of about 20^35 dB persisted (Fig. 1b, Fig. 2). On the basis of the time course of functional recovery, the animals were divided into 3 groups (see Muëller et al., 1996): `fast' (Fig. 3a), `slow' (Fig. 3b) and `no recovery' (Fig. 3c). The mean hearing loss at 5 frequencies, before (pre-exposure), immediately after the sound exposure (acute) and after at least 8 weeks of survival (residual), are shown in Fig. 3d^f. `Fast recovery' (Fig. 3a,d) was seen in 14 out of 42 cochleae (33%, 4 ipsi- and 10 contralateral). The maximum hearing loss after sound exposure was 44 þ 10 dB at 1.0 kHz. CAP-thresholds recovered very fast, starting immediately after the trauma and reaching pre-exposure levels within two weeks. There was no residual hearing loss (t-test, n = 14, P s 0.1 at each of the 5 frequencies). `Slow recovery' (Fig. 3b,e) was seen in 15 out of 42
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
112
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
cochleae (36%, 9 ipsi- and 6 contralateral). The maximum acute hearing loss was 63 þ 9 dB at 1.0 kHz, 19 dB higher than in the `fast recovery' group. After trauma CAP-thresholds recovered over about 4 weeks and then remained unchanged. There was a signi¢cant residual hearing loss (Fig. 3e, minimum 18 dB at 0.25 kHz, maximum 33 dB at 1 and 2 kHz, t-test, n = 15, P 6 0.01 at each of the 5 frequencies). `No recovery' (Fig. 3c,f) occurred in 13 out of 42 cochleae (31%, 10 ipsi- and 3 contralateral). The acute hearing loss was similar to that in the `slow recovery' group. However, CAP-thresholds showed no or very little recovery after the trauma. After at least 8 weeks of survival, the mean residual hearing loss (Fig. 3f, minimum 37 dB at 0.25 kHz, maximum 52 dB at 2 kHz) was about 20 dB higher than that in the `slow recovery' group. 3.2. Morphological observations after acoustic trauma 3.2.1. Controls No BrdU labeling was seen in the basilar papillae without acoustic trauma and BrdU (n = 4). Similarly, no BrdU labeling was seen in the papillae of animals that were given BrdU, but had not received acoustic trauma (n = 4). So there was no cell proliferation in the normal, undamaged basilar papilla of adult pigeons. No BrdU labeling was seen either, when the treatment of primary antibody was omitted (n = 2). This shows that no unspeci¢c binding of antibodies occurred. 3.2.2. Hair cell loss and BrdU labeling after more than 2 months of survival The extent of hair cell loss (observed after at least 2 months of recovery) was correlated with the type of functional recovery after trauma. In 10 out of 14 cochleae with `fast recovery', no hair cell loss and no BrdU labeled cells were observed (Fig. 4a). In the 4 other `fast recovery' cochleae, minor hair cell losses were observed, at the very abneural edge of the basilar papilla. The longitudinal location of the damaged region was at a distance of 30^60% of the papillar length, measured from the apex (Fig. 4b). Only very few BrdU labeled cells were found in the damaged region in these papillae. In all 15 cochleae with `slow recovery', large abneural areas were devoid of hair cells, but there were always surviving hair cells on the neural side of the papilla (Fig. 4c,d). In 4 out of 15 `slow recovery' cochleae, two crescent-shaped areas of hair cell loss occurred, located at the abneural side (arrows in Fig. 4c). In 8 `slow recovery' cochleae, the region of hair cell loss extended over nearly the whole area of the basilar membrane extending between the neural and abneural limbus (free basilar membrane), leaving surviving hair cells only over the neural limbus. The borderline of the area
Fig. 2. Recovery from severe acoustic trauma. Hearing loss of the same ear as in Fig. 1 (C117L) at 5 frequencies as a function of survival time. Values at each frequency are the means of three frequencies equally spaced on a logarithmic scale in a 0.66-octave band centered on the given frequency.
where hair cells survived was irregular. Fig. 4d gives an example. In 3 further `slow recovery' cochleae, hair cells over the neural limbus were also lost. The area of surviving hair cells was a strip along the neural edge of the papilla. In all cases the regions of hair cell loss were occupied by BrdU-labeled cells. In 13 cochleae, there was `no recovery'. Hair cells remained only in the apical area of the basilar papilla (up to 15^40% distance from the apex) at the very neural edge over the superior ¢brocartilaginous plate (Fig. 4e). Again, numerous BrdU labeled cells occupied the whole region of hair cell loss. 3.2.3. Type of labeled cells Despite the fact that all regions where hair cell loss was observed were covered by numerous labeled cells, the number of labeled hair cells was surprisingly low, as revealed from the serial cross-sections. The vast majority of labeled cells were monolayer cells. In some cochleae no labeled hair cells were found at all, although the damaged region was full of labeled cells. The highest number of labeled hair cells found in a single cochlea was about 40. Fig. 5a shows a cross-section through a cochlea from the `slow recovery' group, at 30% distance from the apex, 8 weeks after acoustic overstimulation. Over the neural limbus, unlabelled supporting cells and hair cells, covered by tectorial membrane are present. Above the free basilar membrane, hair cells and supporting cells are lost and replaced by a monolayer of £at cells, most of them labeled with BrdU. No labeled hair cells were observed. Also, no labeling was seen in the hyaline cell region. The above pattern of damage and recovery was seen in the majority of the cochleae with more than three
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
113
Fig. 3. Mean recovery from acoustic trauma. Hearing loss of ears from the `fast' (a), `slow' (b) and `no recovery' group (c) at 5 frequencies as a function of the survival time. d^f: Mean hearing losses of the three groups before (pre-exposure), immediately after severe acoustic trauma (acute) and after at least 8 weeks survival time (residual), plotted at 5 frequencies. Mean values beyond 8 weeks are combined together as `residual'. Vertical bars are one standard deviation.
weeks survival. Cochleae with `fast', `slow' and `no recovery' di¡ered mainly in the extent of the area of total hair cell loss and the location of the few labeled hair cells (Figs. 5 and 6). Some hair cells and supporting cells, not labeled with BrdU, over the neural limbus had survived. The abneural area without hair cells and supporting cells was covered with numerous labeled monolayer cells. In the narrow transition-region between the damaged and the undamaged area a few labeled hair cells and supporting cells were observed (Fig. 5b, Fig. 6). In cochleae with `fast recovery' (Fig. 5b) or in the apical part of the lesion in other cochleae (where
damage was minor), some labeled hair cells were located above the free basilar membrane at the abneural edge of the undamaged region. In the cochleae with `slow recovery', labeled hair cells were found just at the linkage between the neural limbus and the free basilar membrane (Fig. 6a). Fig. 6b^d shows cross sections from cochleae with `no recovery' after 8 weeks of survival. Some labeled tall hair cells were seen over the neural limbus. No or few unlabeled tall hair cells have survived. In very rare cases a pair of labeled hair cells was seen as an island on the free basilar membrane (Fig. 6d, Fig. 8).
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
114
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
Fig. 4. Camera lucida drawings of ¢ve basilar papillae with increasing damage. As indicated by the frequency axis (calculated from Smolders et al., 1995) above the drawings, the basilar papilla is oriented with the apical end to the left and the basal end to the right of the drawings. The neural and abneural side of the basilar papilla is on the top and the bottom, respectively. Dashed line represents the edge of the neural limbus. The area abneural to this line is referred to as free basilar membrane. Dark grey: area of surviving hair cells. Light grey: area of labeled cells with BrdU. White: damaged area without labeled cells. a: C97L, fast functional recovery, 16 weeks survival. The basilar papilla appears intact. No labeled cells are seen. b: C98L, fast functional recovery, 16 weeks survival. Hair cell loss is limited to the abneural edge of the basilar papilla, extending 30^60% of distance from apex. c: C112L, slow functional recovery, 8 weeks survival. Two crescent-shaped regions of hair cell loss extended across the abneural side, extending 7^45% and 45^73% of distance from apex. d: C99L, slow functional recovery, 16 weeks survival. The region of hair cell loss expanded to the whole free basilar membrane. e: C96L, no functional recovery, 10 weeks survival. Hair cells remained only in the very apical-neural region.
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
115
Fig. 5. Cross sections from two basilar papillae with BrdU labeling (nuclei labeled brown), counterstained with toluidine blue. a: C95R (`slow recovery', BrdU injected at day 2, 8 weeks survival, shown as mirror image). b: C87L (`fast recovery', BrdU injected at day 7, 3 weeks survival). Hair cells and supporting cells in the neural part of the basilar papilla are unlabeled. In the abneural part of the basilar papilla, neither hair cells nor supporting cells exist. In a, almost all of the monolayer cells were labeled in the damaged region (bracket), but no labeled hair cells are seen. In b, only ¢ve labeled cells were seen in and near the damaged region (arrows), one of which is a hair cell (arrowhead). The inset shows the higher magni¢cation of the basilar papilla. Bar: 100 Wm (31.8 Wm in inset).
3.3. Temporal pattern of DNA replication and cell proliferation To clarify if some cells in the basilar papilla still
enter the S-phase later than three days after the acoustic overstimulation, the BrdU labeling in group 2, injected with BrdU at day 0^3, was compared to that in group 3, injected with BrdU at day 4^6 (Fig. 7). Both cochleae
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
116
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
Fig. 6. Labeled hair cells (arrowheads) in the cross sections from cochleae with di¡erent functional recovery pattern. a: Labeled hair cells from cochleae with functional `slow recovery' are seen adjacent to the undamaged region, at the linkage between the neural limbus and the free basilar membrane. b, c, d: Labeled hair cells from cochleae with functional `no recovery' are seen over the neural limbus. A pair of `island' cells are seen over the free basilar membrane in d. Bar: 50 Wm.
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
in Fig. 7 are from the `slow recovery' group, and both animals survived 8 weeks after the acoustic trauma. In group 2 (Fig. 7a), numerous cells with BrdU labeling were seen in the damaged abneural half of the basilar papilla. In group 3 (Fig. 7b), there were less labeled cells than in group 2. In particular, no labeled cells were seen at the very abneural edge of the basilar papilla. In the basilar papillae of the two pigeons, which received BrdU at day 7 and day 8 after sound trauma (group 4), some labeled monolayer cells and a labeled hair cell were observed (Fig. 5b). To answer the question, if cochleae with longer survival time have more regenerated hair cells, cochleae with functional `slow recovery' were examined after a survival time of 1^16 weeks. In cochleae with one week survival time (C84L), there were numerous labeled monolayer cells in the hair cell loss region, a few labeled supporting cells, but no labeled hair cells. A pair of labeled cells that may have been regenerated but still immature hair cells, was found at the surface of the sensory epithelium (Fig. 8). However, typical features of the hair cell, cuticular plate and stereocilia were not
117
Fig. 8. Cross section from a cochlea (C86L) with one week survival time. A pair of BrdU-labeled cells, probably immature hair cells (arrowheads), are seen at the surface of the basilar papilla. Bar: 10 Wm.
seen in these two cells. In cochleae with three weeks survival time, clearly labeled hair cells were already observed. There was no evidence of an increase in the number of labeled hair cells with survival time in the range of 3^16 weeks. 4. Discussion 4.1. Cell proliferation after trauma
Fig. 7. Comparison of the BrdU labeling on whole mounts of the basilar papilla. Top: neural; bottom: abneural. a: C133L (injected with BrdU at day 0^3 after acoustic trauma), labeled cells are seen in the abneural half of the basilar papilla. b: C131L (injected with BrdU at day 4^6 after acoustic trauma), the amount of the labeled cells is less than that in a. At the very abneural edge of the basilar papilla there are no labeled cells.
In the present study, BrdU labeling was observed neither in the control basilar papillae, nor in the undamaged regions of sound-damaged basilar papillae of the adult pigeon. This result supports the ¢nding that hair cells in the normal avian basilar papilla do not divide during postembryonic life (Corwin and Cotanche, 1988 ; Girod et al., 1989). However, postembryonic proliferation was observed in adult quail (Ryals and Westbrook, 1990). Also in young chicks, a very low level of proliferation of supporting cells was found in the apical region of the normal basilar papilla (Oesterle and Rubel, 1993 ; Raphael, 1993). A possible reason for the di¡erence may be that we have not given pigeons BrdU continuously, so that not all of the dividing cells could be labeled. It is also conceivable that the rate of spontaneous proliferation di¡ers between species and with age. BrdU labeled cells were observed only in animals that su¡ered from sound trauma, and only in damaged areas of the basilar papilla. This indicates that cell pro-
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
118
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
liferation was triggered by the trauma. The fact that it made no di¡erence in the amount of labeled cells in the present experiments whether BrdU was injected during 3 days post trauma, or only in a time window 40^54 h post trauma, suggests that the peak of cell proliferation is reached around this time after trauma. The same observation was made after sound trauma in the chick (Hashino and Salvi, 1993; Stone and Cotanche, 1994). In the present experiments, a low level of proliferation of non-sensory cells and of hair cells was observed up to 8 days after trauma. It could well be that these cells result from more than one round of cell division, as suggested by Stone and Cotanche (1994) and that the BrdU was passed on to the daughter cells. This possibility also indicates that the time of birth of a labeled cell may be after the time of pulse labeling by BrdU, because the cell can result from further division. Some labeled hair cells appeared in pairs (Figs. 6 and 8). This observation which was also made by Raphael (1992, 1993), Stone and Cotanche (1994) supports the hypothesis, that both symmetric and asymmetric di¡erentiation of new dividing supporting cells into new hair cells occurs. 4.1.1. Hair cells Only a very small number of the labeled cells were hair cells. One reason for this is that hair cell regeneration after the severe sound trauma was incomplete. Large abneural areas of the basilar papilla remained without new hair cells, as observed up to 16 weeks after trauma. Similar observations made in the pigeon with even longer recovery times indicate that these de¢cits are permanent (Muëller et al., 1996, 1997). New hair cells were only generated in a small area of the basilar papilla, bordering the neural side of the abneural region of irreversible hair cell loss. These ¢ndings, together with published results on sound trauma at lower exposure levels (Rubel and Ryals, 1982 ; Cousillas and Rebillard, 1985; Cotanche, 1987a ; Cotanche et al., 1987 ; Corwin and Cotanche, 1988; Girod et al., 1989 ; Adler et al., 1992; Hashino and Salvi, 1993 ; Chen et al., 1996a) suggest that there is a gradient across the width of the papilla in the susceptibility of the sensory epithelium to trauma. With increasing exposure level and time, sound trauma increases from the abneural edge of the papilla towards the neural edge (Cotanche et al., 1987; Cotanche and Dopyera, 1990 ; Adler et al., 1992). At levels high enough to eradicate also the neural tall hair cells, as observed in the present study, damage to the abneural portion of the basilar papilla becomes so severe that regeneration of hair cells does not occur any more in these abneural areas. This is probably due to destruction of the supporting cells (Cotanche et al., 1995). In the narrow transition area between the damaged and undamaged area a number of supporting cells survive, proliferate and give rise to new
hair cells. With increasing sound trauma this transitional zone moves progressively from abneural into the neural direction. The lower susceptibility of the neural side of the papilla may be related to the mechanics of the inner ear. The neural hair cells are located over the rigid neural limbus which vibrates at much lower amplitude than the free basilar membrane (Gummer et al., 1987). The neural hair cells are thought to receive their input via the motion of the tectorial membrane (Raphael, 1991 ; Smolders et al., 1995; Salvi et al., 1996). During sound damage the tectorial membrane decouples from the hair cells (Cotanche, 1987b), which will reduce the vibration input to the neural hair cells and protect them from sound damage (Raphael, 1991). The fact that the area of the papilla over the neural limbus and the area over the free basilar membrane roughly correspond to the areas occupied by `tall' and `short' hair cells, respectively, may be purely coincidental (Smolders et al., 1995). At very high sound pressure levels, the vibration amplitude of the free basilar membrane may become so large as to destroy even the supporting cells over the free basilar membrane. The vibration amplitude of the neural limbus may become large enough to damage or even destroy the overlying tall hair cells, but leave the supporting cells and thus the potential of hair cell regeneration in the transitional zone between neural limbus and free basilar membrane. 4.1.2. Non-sensory cells The majority of the labeled progeny of the dividing cells were non-sensory monolayer cells. However, labeling in the hyaline cell area was present only in the cochleae after one week survival, not after more than three weeks survival. This ¢nding, together with that from Girod et al. (1989), suggests that new hyaline cells are present shortly after sound exposure, but then may di¡erentiate, migrate away or die during a longer survival period (Girod et al., 1989). Almost all of the monolayer cells that occupied irreversibly damaged areas were labeled. This indicates that the proliferation of the progenitors of the monolayer cells occurred at the same time as that of the hair-cell progenitors. However, as there is evidence from the chicken, that the precursors of the hair cells (supporting cells, Raphael, 1992, 1993; Tsue et al., 1994) are di¡erent from those of the monolayer cells (hyaline cells, Girod et al., 1989; Cotanche et al., 1995), this may just mean that the proliferation of both cell types is triggered by the same event, the sound trauma which causes the loss of the hair cells and the supporting cells, respectively. The monolayer cells then rapidly proliferate and cover the whole damaged region where both hair cells and supporting cells are missing. Our ¢nding that the areas with monolayer cells are not repopulated with hair cells 8 weeks or more (up to 16 weeks) after trauma supports the hypothesis that the monolayer cells ac-
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
tually prevent this repopulation (Cotanche et al., 1995, 1996). 4.1.3. Functional recovery It appears that the three groups of functional recovery, `fast', `slow' and `no' that were distinguished in the present study represent three instances of a continuum of increasing damage across the basilar papilla. Fast recovery without residual hearing loss represents the situation where no or negligible hair cell loss occurs. In this case, hair cell loss and regeneration play an insigni¢cant role in the functional impairment and recovery. Rather, hearing loss after sound exposure results from reversible functional impairment, possibly of hair cell transduction, as well as from decoupling of the tectorial membrane (Cotanche, 1992; Marsh et al., 1990 ; Saunders et al., 1996a) and impairment of the function of the tegmentum vasculosum (Ryals et al., 1995), involving loss of the endocochlear potential in young (Poje et al., 1995), but not in adult (Chen et al., 1995) chickens. These impaired functions recover rapidly after trauma, leading to restoration of function within days (McFadden and Saunders, 1989; Adler et al., 1992 ; Niemiec et al., 1994; Muëller et al., 1996 ; Saunders et al., 1996a). When the damage expands towards the neural edge of the basilar papilla, irreversible damage in the abneural area occurs. Not only hair cells but also their precursor cells are lost and no regeneration of hair cells in these areas is possible. Instead this area is populated by monolayer cells, which may prevent its reoccupation by hair cell precursor cells (Cotanche et al., 1995). Only a limited number of new hair cells regenerates in the small transitional zone between the abneural damaged area and the neural undamaged area. The abneural part of the tectorial membrane is destroyed as well. Because of the loss of the supporting cells in the abneural damaged region, no new tectorial membrane can be regenerated (Girod et al., 1995). As a consequence, the sound induced vibration of the free basilar membrane can no longer be transferred to the tall hair cells over the neural limbus, resulting in threshold elevation, reduction of tuning sharpness and abnormal rate intensity functions of the appertaining auditory nerve ¢bers (Chen et al., 1996b, Saunders et al., 1996b; Muëller et al., 1996, 1997). This is the situation designated `slow recovery'. Recovery takes longer because there is more damage to repair and because regeneration of hair cells is involved. Because the structural regeneration is incomplete, there is incomplete functional recovery and a permanent residual hearing loss remains. It appears, as one would expect, that an intact structure of the basilar papilla is necessary for normal sensitivity. It is not su¤cient that the tall hair cells remain, which are innervated by the majority of the a¡erent cochlear nerve ¢bers (von
119
Duëring et al., 1985 ; Fischer, 1992, 1994 ; Smolders et al., 1995). `No recovery' represents the other extreme, where damage has extended to the neural tall hair cells and their precursor cells have been destroyed over large parts of the papilla. The resulting permanent hearing loss represents cochleae with only a few rows of neural hair cells remaining, a destroyed tectorial membrane and probably signi¢cant loss of ganglion cells (Ryals et al., 1989). Acknowledgments Supported by the Deutsche Forschungsgemeinschaft (SFB 269, B1). We thank S. Hoidis and H. Schalk for expert technical assistance. References Adler, H.J., Kenealy, J.F.X., Dedio, R.M., Saunders, J.C., 1992. Threshold shift, hair cell loss, and hair bundle sti¡ness following exposure to 120 and 125 dB pure tones in the neonatal chick. Acta Otolaryngol. (Stockh.) 112, 444^454. Chen, L., Trautwein, P.G., Miller, K., Salvi, R.J., 1995. E¡ects of kanamycin ototoxicity and hair cell regeneration on the DC endocochlear potential in adult chickens. Hear. Res. 89, 28^34. Chen, L., Trautwein, P.G., Shero, M., Salvi, R., 1996a. Correlation of hair cell regeneration with physiology and psychophysics in adult chickens following acoustic trauma. In: R.J. Salvi, D. Henderson, V. Colletti, F. Fiorino (Eds.), Auditory System Plasticity and Regeneration. Thieme Medical Publ., New York, pp. 43^61. Chen, L., Trautwein, P.G., Shero, M., Salvi, R., 1996b. Tuning, spontaneous activity and tonotopic map in chicken cochlear ganglion neurons following sound-induced hair cell loss and regeneration. Hear. Res. 00, 152^164. Corwin, J.T., Cotanche, D.A., 1988. Regeneration of sensory hair cells after acoustic trauma. Science 240, 1772^1774. Cotanche, D.A., 1987a. Regeneration of hair cell stereociliary bundles in the chick cochlea following severe acoustic trauma. Hear. Res. 30, 181^196. Cotanche, D.A., 1987b. Regeneration of the tectorial membrane in the chick cochlea following severe acoustic trauma. Hear. Res. 30, 197^206. Cotanche, D.A., Saunders, J.C., Tilney, L.G., 1987. Hair cell damage produced by acoustic trauma in the chick cochlea. Hear. Res. 25, 267^286. Cotanche, D.A., Dopyera, C.E.J., 1990. Hair cell and supporting cell response to acoustic trauma in the chick cochlea. Hear. Res. 46, 29^40. Cotanche, D.A., 1992. Video-enhanced DIC images of the noise-damaged and regenerated chick tectorial membrane. Exp. Neurol. 115, 23^26. Cotanche, D.A., Lee, K.H., Stone, J.S., Picard, D.A., 1994. Hair cell regeneration in the bird cochlea following noise damage or ototoxic drug damage. Anat. Embryol. 189, 1^18. Cotanche, D.A., Messana, E.P., Ofsie, M.S., 1995. Migration of hyaline cells into the chick basilar papilla during severe noise damage. Hear. Res. 91, 148^159. Cotanche, D.A., Messana, E.P., Riedl, 1996. Hyaline cell migration into the basilar papilla involves the secretion of a new layer of extracellular matrix. Assoc. Res. Otolaryngol. (Abstr.) 19, 199.
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
120
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
Cotanche, D.A., 1997. Hair cell regeneration in the avian cochlea. Ann. Otol. Rhinol. Laryngol. 106, 9^15. Cousillas, H., Rebillard, G., 1985. Age-dependent e¡ects of a pure tone trauma in the chick basilar papilla: Evidence for a development of the tonotopic organization. Hear. Res. 19, 217^226. Duckert, L.G., Rubel, E.W., 1990. Ultrastructural observations on regenerating hair cells in the chick basilar papilla. Hear. Res. 48, 161^182. Fischer, F.P., 1992. Quantitative analysis of the innervation of the chicken basilar papilla. Hear. Res. 61, 167^178. Fischer, F.P., 1994. General pattern and morphological specializations of the avian cochlea. Scanning Microsc. 8, 351^364. Girod, D.A., Duckert, L.G., Rubel, E.W., 1989. Possible precusors of regenerated hair cells in the avian cochlea following acoustic trauma. Hear. Res. 42, 175^194. Girod, D.A., Ryals, B.M., Fankhauser, C.E., Westbrook, E.W., 1995. Long term structural changes in the chick tectorial membrane following severe acoustic damage. Assoc. Res. Otolaryngol. (Abstr.) 18, 787. Gleich, O., 1989. Auditory primary a¡erents in the starling: Correlation of function and morphology. Hear. Res. 37, 255^268. Gummer, A.W., Smolders, J.W.T., Klinke, R., 1987. Basilar membrane motion in the pigeon measured with the Moëssbauer technique. Hear. Res. 29, 63^92. Hashino, E., Tanak, Y., Sokabe, M., 1991. Hair cell damage and recovery following chronic application of kanamycin in the chick cochlea. Hear. Res. 52, 356^368. Hashino, E., Salvi, R.J., 1993. Changing spatial patterns of DNA replication in the noise-damaged chick cochlea. J. Cell Science 105, 23^31. Manley, G.A., Gleich, O., 1992. The evolutionary biology of hearing. In: D. Webster, R.R. Fay, A.N. Popper (Eds.), Evolution and Specialization of Function in the Avian Auditory Periphery. Springer Verlag, New York, pp. 561^580. Marsh, R.R., Xu, L.R., Moy, J.P., Saunders, J.C., 1990. Recovery of the basilar papilla following intense sound exposure in the chick. Hear. Res. 46, 229^238. McFadden, E.A., Saunders, J.C., 1989. Recovery of auditory function following intense sound exposure in the neonatal chick. Hear. Res. 41, 205^216. Muëller, M., Smolders, J.W.T., Ding-Pfennigdor¡, D., Klinke, R., 1996. Regeneration after tall hair cell damage following severe acoustic trauma in adult pigeons: correlation between cochlear morphology, compound action potential responses and single ¢ber properties in single animals. Hear. Res. 102, 133^154. Muëller, M., Smolders, J.W.T., Ding-Pfennigdor¡, D., Klinke, R., 1997. Discharge properties of pigeon single auditory nerve ¢bers after recovery from severe acoustic trauma. Int. J. Dev. Neurosci. 15, 401^416. Niemiec, A.J., Raphael, Y., Moody, D.B., 1994. Return of auditory function following structural regeneration after acoustic trauma: behavioral measures from quail. Hear. Res. 79, 1^16. Oesterle, E.C., Rubel, E.W., 1993. Postnatal production of supporting cells in the chick cochlea. Hear. Res. 66, 213^224. Patuzzi, R.B., Bull, C.L., 1991. Electrical responses from the chicken basilar papilla. Hear. Res. 53, 57^77. Poje, C.P., Sewell, D.A., Saunders, J.C., 1995. The e¡ects of exposure to intense sound on the DC endocochlear potential in the chick. Hear. Res. 82, 197^204. Pugliano, F.A., Wilcox, T.O., Rossiter, J., Saunders, J.C., 1993. Recovery of auditory structure and function in neonatal chicks exposed to intense sound for 8 days. Neurosci. Lett. 151, 214^ 218. Raphael, Y., 1991. Pure-tone overstimulation protects surviving avian hair cells from acoustic trauma. Hear. Res. 53, 173^184.
Raphael, Y., 1992. Evidence for supporting cell mitosis in responce to acoustic trauma in the avian inner ear. J. Neurocytol. 21, 663^671. Raphael, Y., 1993. Reorganization of the chick basilar papilla after acoustic trauma. J. Comp. Neurol. 330, 521^532. Rebillard, G., Ryals, B.M., Rubel, E.W., 1982. Relationship between hair cell loss on the chick basilar papilla and threshold shift after acoustic overstimulation. Hear. Res. 8, 77^82. Rosowski, J.J., Saunders, J.C., 1980. Sound transmission through the avian interaural pathways. J. Comp. Physiol. A 136, 183^190. Rubel, E.W., Ryals, B.M., 1982. Patterns of hair cell loss in chick basilar papilla after intense auditory stimulation. Exposure duration and survival time. Acta Otolaryngol. (Stockh.) 93, 31^41. Ryals, B.M., Rubel, E.W., 1982. Patterns of hair cell loss in chick basilar papilla after intense auditory stimulation. Frequency organisation. Acta Otolaryngol. (Stockh.) 93, 205^210. Ryals, B.M., Rubel, E.W., 1985. Di¡erential susceptibility of avian hair cells to acoustic trauma. Hear. Res. 19, 73^84. Ryals, B.M., Teneyck, B., Westbrook, E.W., 1989. Ganglion cell loss continues during hair cell regeneration. Hear. Res. 43, 81^90. Ryals, B.M., Westbrook, E.W., 1990. Hair cell regeneration in senescent quail. Hear. Res. 50, 87^96. Ryals, B.M., Stalford, M.D., Lambert, P.R., Westbrook, E.W., 1995. Recovery of noise-induced changes in the dark cells of the quail tegmentum vasculosum. Hear. Res. 83, 51^61. Salvi, R., Henderson, D., Fiorino, F., Colletti, V., 1996. Auditory System Plasticity and Regeneration. Thieme Medical Publishers, New York. Saunders, J.C., Tilney, L.G., 1982. Species di¡erences in susceptibility to noise exposure. In: R.P. Hamernik, D. Henderson, R. Salvi (Eds.), New Perspectives on Noise-Induced Hearing Loss. Raven, New York. Saunders, J.C., Adler, H.J., Pugliano, F.A., 1992. The structural and functional aspects of hair cell regeneration in the chick as a result of exposure to intense sound. Exp. Neurol. 115, 13^17. Saunders, J.C., Torsiglieri, A.J., Dedio, R.M., 1993. The growth of hearing loss in neonatal chicks exposed to intense pure tones. Hear. Res. 69, 25^34. Saunders, J.C., Doan, D.E., Cohen, Y.E., Adler, H.J., Poje, C.P., 1996a. Recent observations on the recovery of structure and function in the sound-damaged chick ear. In: R.J. Salvi, D. Henderson, F. Fiorino, V. Colletti (Eds.), Auditory System Plasticity and Regeneration. Thieme Medical Publishers, New York, pp. 62^83. Saunders, J.C., Doan, D.E., Poje, C.P., Fisher, K.A., 1996b. Cochlear nerve activity after intense sound exposure in neonatal chicks. J. Neurophysiol. 76, 770^787. Smolders, J.W.T., Ding, D., Klinke, R., 1992. Normal tuning curves from primary a¡erent ¢bres innervating short and intermediate hair cells in the pigeon ear. Adv. Biosci. 83, 197^204. Smolders, J.W.T., Ding-Pfennigdor¡, D., Klinke, R., 1995. A functional map of the pigeon basilar papilla: Correlation of the properties of single auditory nerve ¢bres and their peripheral origin. Hear. Res. 92, 151^169. Stone, J.S., Cotanche, D.A., 1994. Identi¢cation of the timing of S phase and the patterns of cell proliferation during hair cell regeneration in the chick cochlea. J. Comp. Neurol. 341, 50^67. Tsue, T.T., Walting, D.L., Weisleder, P., Coltrera, M.D., Rubel, E.W., 1994. Identi¢cation of hair cell progenitors and intermetotic migration of their nuclei in the normal and regenerating avian inner ear. J. Neurosci. 14, 140^152. Tucci, D.L., Rubel, E.W., 1990. Physiological status of regenerated hair cells in the avian inner ear following aminoglycoside ototoxicity. Otolaryngol. Head Neck Surg. 103, 443^445. von Duëring, M., Andres, K.H., Simon, K., 1985. The comparative anatomy of the basilar papillae in birds. Fortschr. Zool. 30, 681^ 685.
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
Hair cell loss and regeneration after severe acoustic overstimulation in the adult pigeon Danping Ding-Pfennigdor¡, Jean W.Th. Smolders *, Marcus Muëller, Rainer Klinke Physiologisches Institut III, Klinikum der J.W. Goethe Universitaët, Theodor-Stern-Kai 7, 60590 Frankfurt am Main, Germany Received 10 November 1997; revised 10 March 1998; accepted 11 March 1998
Abstract The extent of hair cell regeneration following acoustic overstimulation severe enough to destroy tall hair cells, was determined in adult pigeons. BrdU (5-bromo-2P-deoxyuridine) was used as a proliferation marker. Recovery of hearing thresholds in each individual animal was measured over a period of up to 16 weeks after trauma. In ears with loss of both short and tall hair cells, little or no functional recovery occurred. In ears with less damage, where significant functional recovery did occur, there were always a few rows of surviving hair cells left at the neural edge of the basilar papilla. In the region of hair cell loss, numerous BrdU labeled cells were found. However, only a small minority of these cells were regenerated hair cells, the majority being monolayer cells. Irrespective of the extent of the region of hair cell loss, regenerated hair cells were observed predominantly in a narrow strip at the transition from the abneural area of total hair cell loss and the neural area of hair cell survival. With increasing damage this strip moved progressively towards the neural edge of the papilla. No regeneration of hair cells was observed in the abneural region of total hair cell loss, even up to 16 weeks after trauma. The results indicate that there is a gradient in the destructive effect of loud sound across the width of the basilar papilla, from most detrimental at the abneural edge to least detrimental at the neural edge. Both tall and short hair cells can regenerate after sound trauma. Whether they do regenerate or not depends on the degree of damage to the area of the papilla where they normally reside. Regeneration of new hair cells occurs only in a narrow longitudinal band, which moves from abneural into the neural direction with increasing damage. In the area neural to this band, hair cells survive the overstimulation. In the area abneural to this band, sound damage is so severe, that no regeneration of hair cells occurs. As a consequence morphological and functional deficits persist. z 1998 Elsevier Science B.V. All rights reserved. Key words: Bird; Pigeon; Acoustic trauma; Regeneration; Auditory nerve; Hair cell; Basilar papilla; 5-Bromo-2P-deoxyuridine labeling
1. Introduction The hair cells in the avian basilar papilla are capable of regeneration after acoustic or ototoxic trauma (reviewed by Cotanche et al., 1994; Cotanche, 1997). Ototoxic trauma a¡ects both short and tall hair cells, across the entire width of the basilar papilla. The damage begins in the base and progresses towards the apex with increase in the amount or duration of ototoxic treat-
* Corresponding author. Tel.: +49 (69) 63016980; Fax: +49 (69) 63016987; E-mail: [email protected]
ment (Duckert and Rubel, 1990 ; Hashino et al., 1991). In contrast, damage resulting from sound overexposure begins at the abneural side of the papilla and mainly damages the short hair cells (Saunders and Tilney, 1982 ; Ryals and Rubel, 1985 ; Cotanche et al., 1987; Marsh et al., 1990) at a longitudinal position related to the frequency and pressure level of the sound (Rebillard et al., 1982, Ryals and Rubel, 1982; Saunders and Tilney, 1982 ; Muëller et al., 1996). The short hair cells receive mainly e¡erent innervation (Fischer, 1992, 1994) and their role in inner ear function is still unclear (Gleich, 1989; Patuzzi and Bull, 1991 ; Manley and Gleich, 1992 ; Smolders et al., 1992, 1995). The majority of the a¡erent auditory nerve ¢bers innervate tall hair
0378-5955 / 98 / $19.00 ß 1998 Elsevier Science B.V. All rights reserved PII: S 0 3 7 8 - 5 9 5 5 ( 9 8 ) 0 0 0 5 5 - 0
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
110
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
cells, which are not destroyed by most sound overexposure regimes up to 125 dB SPL, even with very long exposure times (Cotanche and Dopyera, 1990; Raphael, 1991 ; Saunders et al., 1993; Pugliano et al., 1993). Functional recovery after such acoustic trauma is very fast (a few days, McFadden and Saunders, 1989; Saunders et al., 1992; Niemiec et al., 1994; Muëller et al., 1996) and occurs largely before the newly generated short hair cells become mature. This type of recovery is believed to result from the re-establishment of function of the surviving hair cells and accessory inner ear structures (Cotanche et al., 1994; Poje et al., 1995; Ryals et al., 1995), rather than from the regeneration of new hair cells (reviewed by Saunders et al., 1996a). Damage to the tall hair cells over the neural limbus of the inner ear occurs only at very high sound pressure levels, well above 125 dB SPL (Muëller et al., 1996, 1997). After such damage, functional recovery takes much longer (3^4 weeks) and is incomplete (Muëller et al., 1996, 1997). This type of functional recovery is very similar to that observed after aminoglycoside intoxication of the inner ear (Tucci and Rubel, 1990). The purpose of the present experiments was to determine the proliferation pattern of new hair cells after the destruction of both short and tall hair cells after very severe acoustic trauma. We used the S-phase marker, 5-bromo-2P-deoxyuridine (BrdU), to label new cells. The experiments were performed on adult pigeons. Sound trauma was severe enough to produce damage to both short and tall hair cells. The time course of functional recovery after trauma was recorded in both ears of each individual animal. 2. Methods 2.1. Animals Adult pigeons, 0.5^2 yr old, were obtained from a local breeder. The weight of the animals ranged from 350 to 550 g. The care and the use of these animals were approved by the German authorities for animal welfare (Regierungspraësidium Darmstadt). 2.2. Sound trauma Animals were exposed under general anesthesia to a pure tone of 0.7 kHz presented to the left ear for 1 h at a sound pressure level of 142 þ 2 dB SPL. The sound was produced by a Beyer DT 48 transducer ¢tted to the ear canal with an ear speculum. Sound pressure level was measured throughout the exposure using a calibrated BpK probe microphone (Type 4170) placed 1^ 2 mm from the eardrum. With this stimulus paradigm it is possible to cause damage to both the short and the tall hair cells. Details are given in Muëller et al. (1996).
2.3. Compound action potential audiograms All animals were implanted a few weeks prior to the start of the experiments with gold-wire electrodes placed at the round window of both cochleae for recording of compound action potential (CAP)-audiograms (details in Muëller et al., 1996). CAP-audiograms were measured before and immediately after the sound exposure. To monitor the functional recovery, CAPaudiograms were measured once or twice a week during the survival period (1^16 weeks after the acoustic overstimulation). The CAP-audiograms, response thresholds to gaussian-shaped tone pips (2/3 octave bandwidth), were measured in a frequency range between 0.1 and 8 kHz at 3 logarithmically equidistant frequencies per octave. 2.4. BrdU injections and immunohistochemistry The BrdU-labeling method was adapted from Raphael (1992) and Stone and Cotanche (1994). BrdU (Sigma) was dissolved in sterile phosphate-bu¡ered saline (PBS). To detect proliferating cells, four groups of pigeons were injected (i.m.) with BrdU at a dose of 50 mg/kg after the acoustic trauma. Group 1: 11 pigeons received three injections over a period of 6 h at day 2, between 40^54 h post exposure (day 0 is the day of sound exposure). Group 2: 12 pigeons received seven injections over four days (twice a day, day 0 to day 3, 0^76 h post exposure), in order to detect as many new cells as possible. To investigate if some cells in the basilar papilla still enter S-phase later than three days after sound trauma, 5 additional animals were used. Three of these (Group 3) received six injections between day 4 and day 6 (92^148 h post exposure, twice a day). The other two (Group 4) received three injections over a period of 6 h, respectively at day 7 (165^171 h post exposure) and day 8 (188^194 h post exposure). Pigeons survived for 1^2, 3^4, 8^10 or 16 weeks after the sound exposure and were then decapitated under general anesthesia. Both cochleae were extirpated, exposed at their basal and apical ends and ¢xed in 4% paraformaldehyde in 0.1 M phosphate bu¡er overnight at 4³C. Following ¢xation, cochleae were rinsed in PBS. The bone around the cochlear duct was taken away and the tegmentum vasculosum was removed. Cochlear ducts were treated with 2 N HCl in PBS containing 0.5% Triton X-100 at 37³C for 30 min to denature DNA. Endogenous peroxidase was inactivated by treatment with 0.3% H2 O2 for 20 min. To inhibit nonspeci¢c binding of the antibodies, cochlear ducts were immersed in blocking solution (2% normal horse serum, 2% bovine serum albumin (BSA) and 0.5% Triton X-100 in PBS) for 20 min. Following the above pretreatment, cochlear ducts were incubated overnight in mouse monoclonal antibody to BrdU (Boehringer Mannheim,
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
111
1:100 diluted in PBS), followed by horse antimouse biotinylated secondary antibody (Vector Labs) for 1.5 h and avidin-biotin-peroxidase-complex (Vector Labs, Elite ABC Kit) for 45 min. Thorough rinses in PBS followed each antibody treatment. Cochlear ducts were then incubated in 0.5% 3,3P-diaminobenzidine (DAB) and 0.01% H2 O2 in PBS to visualize the HRPreaction product. After staining, the tectorial membrane and cochlear ganglion were removed in most cochleae. Specimens were mounted in Kaiser's glycerol gelatin and observed under the light microscope (Olympus BH 2). The surface views of the basilar papilla were photographed and camera lucida drawings were made. To determine the speci¢c cell type that was BrdU-labeled, some specimens were ¢nally embedded in EPON, transversely sectioned (7 Wm) and counterstained with toluidine blue. 2.5. Controls for BrdU labeling As a control for the speci¢city of antibody binding, two pigeons received neither sound exposure nor BrdU. As a control for non-regenerative proliferation, two pigeons received only BrdU, but no sound exposure. One control pigeon received both sound exposure and BrdU injections as in group 1. Pigeons were decapitated 20 h after the last BrdU injection. Normal BrdU immunohistochemistry was performed in all cochleae. However, as a control for unspeci¢c binding of the secondary antibody, the treatment of primary antibody was omitted in two cochleae from the pigeon which had received both sound exposure and BrdU. 3. Results Forty-two cochleae from 23 pigeons from group 1 and 2, in which both the CAP-audiograms during recovery and the morphological preparation were successful, were used for the analysis of the functional-morphological correlation. Both the ipsi- (exposed, left) and the contralateral (right) cochlea of each animal were used since the contralateral cochleae were damaged by the sound exposure as well, due to the small interaural attenuation (Rosowski and Saunders, 1980; Muëller et al., 1996). 3.1. CAP-audiograms after acoustic trauma The CAP-audiograms of an individual ear (C117L) before, immediately post exposure (1 h, 0.7 kHz, 142 dB SPL) and 8 weeks after sound exposure are given in Fig. 1a. Fig. 1b depicts these audiograms as hearing loss relative to the mean audiogram of 91 untreated pigeon ears. The time course of the recovery of CAPthreshold from acoustic trauma at ¢ve frequencies in
Fig. 1. a: CAP audiograms of an individual ear (C117L) before, immediately after severe acoustic trauma (0.7 kHz, 142 dB SPL, 1 h) and after 8 weeks survival. Mean control audiogram is averaged from 91 adult normal pigeons. b: The same data plotted as hearing loss relative to the mean control audiogram.
the same ear is shown in Fig. 2. Immediately after sound exposure, hearing loss ranged from 35^53 dB depending on frequency. The loss recovered to stable values in about 4 weeks. Only minor recovery occurred thereafter. After 8 weeks of survival, a frequency dependent residual hearing loss of about 20^35 dB persisted (Fig. 1b, Fig. 2). On the basis of the time course of functional recovery, the animals were divided into 3 groups (see Muëller et al., 1996): `fast' (Fig. 3a), `slow' (Fig. 3b) and `no recovery' (Fig. 3c). The mean hearing loss at 5 frequencies, before (pre-exposure), immediately after the sound exposure (acute) and after at least 8 weeks of survival (residual), are shown in Fig. 3d^f. `Fast recovery' (Fig. 3a,d) was seen in 14 out of 42 cochleae (33%, 4 ipsi- and 10 contralateral). The maximum hearing loss after sound exposure was 44 þ 10 dB at 1.0 kHz. CAP-thresholds recovered very fast, starting immediately after the trauma and reaching pre-exposure levels within two weeks. There was no residual hearing loss (t-test, n = 14, P s 0.1 at each of the 5 frequencies). `Slow recovery' (Fig. 3b,e) was seen in 15 out of 42
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
112
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
cochleae (36%, 9 ipsi- and 6 contralateral). The maximum acute hearing loss was 63 þ 9 dB at 1.0 kHz, 19 dB higher than in the `fast recovery' group. After trauma CAP-thresholds recovered over about 4 weeks and then remained unchanged. There was a signi¢cant residual hearing loss (Fig. 3e, minimum 18 dB at 0.25 kHz, maximum 33 dB at 1 and 2 kHz, t-test, n = 15, P 6 0.01 at each of the 5 frequencies). `No recovery' (Fig. 3c,f) occurred in 13 out of 42 cochleae (31%, 10 ipsi- and 3 contralateral). The acute hearing loss was similar to that in the `slow recovery' group. However, CAP-thresholds showed no or very little recovery after the trauma. After at least 8 weeks of survival, the mean residual hearing loss (Fig. 3f, minimum 37 dB at 0.25 kHz, maximum 52 dB at 2 kHz) was about 20 dB higher than that in the `slow recovery' group. 3.2. Morphological observations after acoustic trauma 3.2.1. Controls No BrdU labeling was seen in the basilar papillae without acoustic trauma and BrdU (n = 4). Similarly, no BrdU labeling was seen in the papillae of animals that were given BrdU, but had not received acoustic trauma (n = 4). So there was no cell proliferation in the normal, undamaged basilar papilla of adult pigeons. No BrdU labeling was seen either, when the treatment of primary antibody was omitted (n = 2). This shows that no unspeci¢c binding of antibodies occurred. 3.2.2. Hair cell loss and BrdU labeling after more than 2 months of survival The extent of hair cell loss (observed after at least 2 months of recovery) was correlated with the type of functional recovery after trauma. In 10 out of 14 cochleae with `fast recovery', no hair cell loss and no BrdU labeled cells were observed (Fig. 4a). In the 4 other `fast recovery' cochleae, minor hair cell losses were observed, at the very abneural edge of the basilar papilla. The longitudinal location of the damaged region was at a distance of 30^60% of the papillar length, measured from the apex (Fig. 4b). Only very few BrdU labeled cells were found in the damaged region in these papillae. In all 15 cochleae with `slow recovery', large abneural areas were devoid of hair cells, but there were always surviving hair cells on the neural side of the papilla (Fig. 4c,d). In 4 out of 15 `slow recovery' cochleae, two crescent-shaped areas of hair cell loss occurred, located at the abneural side (arrows in Fig. 4c). In 8 `slow recovery' cochleae, the region of hair cell loss extended over nearly the whole area of the basilar membrane extending between the neural and abneural limbus (free basilar membrane), leaving surviving hair cells only over the neural limbus. The borderline of the area
Fig. 2. Recovery from severe acoustic trauma. Hearing loss of the same ear as in Fig. 1 (C117L) at 5 frequencies as a function of survival time. Values at each frequency are the means of three frequencies equally spaced on a logarithmic scale in a 0.66-octave band centered on the given frequency.
where hair cells survived was irregular. Fig. 4d gives an example. In 3 further `slow recovery' cochleae, hair cells over the neural limbus were also lost. The area of surviving hair cells was a strip along the neural edge of the papilla. In all cases the regions of hair cell loss were occupied by BrdU-labeled cells. In 13 cochleae, there was `no recovery'. Hair cells remained only in the apical area of the basilar papilla (up to 15^40% distance from the apex) at the very neural edge over the superior ¢brocartilaginous plate (Fig. 4e). Again, numerous BrdU labeled cells occupied the whole region of hair cell loss. 3.2.3. Type of labeled cells Despite the fact that all regions where hair cell loss was observed were covered by numerous labeled cells, the number of labeled hair cells was surprisingly low, as revealed from the serial cross-sections. The vast majority of labeled cells were monolayer cells. In some cochleae no labeled hair cells were found at all, although the damaged region was full of labeled cells. The highest number of labeled hair cells found in a single cochlea was about 40. Fig. 5a shows a cross-section through a cochlea from the `slow recovery' group, at 30% distance from the apex, 8 weeks after acoustic overstimulation. Over the neural limbus, unlabelled supporting cells and hair cells, covered by tectorial membrane are present. Above the free basilar membrane, hair cells and supporting cells are lost and replaced by a monolayer of £at cells, most of them labeled with BrdU. No labeled hair cells were observed. Also, no labeling was seen in the hyaline cell region. The above pattern of damage and recovery was seen in the majority of the cochleae with more than three
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
113
Fig. 3. Mean recovery from acoustic trauma. Hearing loss of ears from the `fast' (a), `slow' (b) and `no recovery' group (c) at 5 frequencies as a function of the survival time. d^f: Mean hearing losses of the three groups before (pre-exposure), immediately after severe acoustic trauma (acute) and after at least 8 weeks survival time (residual), plotted at 5 frequencies. Mean values beyond 8 weeks are combined together as `residual'. Vertical bars are one standard deviation.
weeks survival. Cochleae with `fast', `slow' and `no recovery' di¡ered mainly in the extent of the area of total hair cell loss and the location of the few labeled hair cells (Figs. 5 and 6). Some hair cells and supporting cells, not labeled with BrdU, over the neural limbus had survived. The abneural area without hair cells and supporting cells was covered with numerous labeled monolayer cells. In the narrow transition-region between the damaged and the undamaged area a few labeled hair cells and supporting cells were observed (Fig. 5b, Fig. 6). In cochleae with `fast recovery' (Fig. 5b) or in the apical part of the lesion in other cochleae (where
damage was minor), some labeled hair cells were located above the free basilar membrane at the abneural edge of the undamaged region. In the cochleae with `slow recovery', labeled hair cells were found just at the linkage between the neural limbus and the free basilar membrane (Fig. 6a). Fig. 6b^d shows cross sections from cochleae with `no recovery' after 8 weeks of survival. Some labeled tall hair cells were seen over the neural limbus. No or few unlabeled tall hair cells have survived. In very rare cases a pair of labeled hair cells was seen as an island on the free basilar membrane (Fig. 6d, Fig. 8).
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
114
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
Fig. 4. Camera lucida drawings of ¢ve basilar papillae with increasing damage. As indicated by the frequency axis (calculated from Smolders et al., 1995) above the drawings, the basilar papilla is oriented with the apical end to the left and the basal end to the right of the drawings. The neural and abneural side of the basilar papilla is on the top and the bottom, respectively. Dashed line represents the edge of the neural limbus. The area abneural to this line is referred to as free basilar membrane. Dark grey: area of surviving hair cells. Light grey: area of labeled cells with BrdU. White: damaged area without labeled cells. a: C97L, fast functional recovery, 16 weeks survival. The basilar papilla appears intact. No labeled cells are seen. b: C98L, fast functional recovery, 16 weeks survival. Hair cell loss is limited to the abneural edge of the basilar papilla, extending 30^60% of distance from apex. c: C112L, slow functional recovery, 8 weeks survival. Two crescent-shaped regions of hair cell loss extended across the abneural side, extending 7^45% and 45^73% of distance from apex. d: C99L, slow functional recovery, 16 weeks survival. The region of hair cell loss expanded to the whole free basilar membrane. e: C96L, no functional recovery, 10 weeks survival. Hair cells remained only in the very apical-neural region.
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
115
Fig. 5. Cross sections from two basilar papillae with BrdU labeling (nuclei labeled brown), counterstained with toluidine blue. a: C95R (`slow recovery', BrdU injected at day 2, 8 weeks survival, shown as mirror image). b: C87L (`fast recovery', BrdU injected at day 7, 3 weeks survival). Hair cells and supporting cells in the neural part of the basilar papilla are unlabeled. In the abneural part of the basilar papilla, neither hair cells nor supporting cells exist. In a, almost all of the monolayer cells were labeled in the damaged region (bracket), but no labeled hair cells are seen. In b, only ¢ve labeled cells were seen in and near the damaged region (arrows), one of which is a hair cell (arrowhead). The inset shows the higher magni¢cation of the basilar papilla. Bar: 100 Wm (31.8 Wm in inset).
3.3. Temporal pattern of DNA replication and cell proliferation To clarify if some cells in the basilar papilla still
enter the S-phase later than three days after the acoustic overstimulation, the BrdU labeling in group 2, injected with BrdU at day 0^3, was compared to that in group 3, injected with BrdU at day 4^6 (Fig. 7). Both cochleae
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
116
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
Fig. 6. Labeled hair cells (arrowheads) in the cross sections from cochleae with di¡erent functional recovery pattern. a: Labeled hair cells from cochleae with functional `slow recovery' are seen adjacent to the undamaged region, at the linkage between the neural limbus and the free basilar membrane. b, c, d: Labeled hair cells from cochleae with functional `no recovery' are seen over the neural limbus. A pair of `island' cells are seen over the free basilar membrane in d. Bar: 50 Wm.
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
in Fig. 7 are from the `slow recovery' group, and both animals survived 8 weeks after the acoustic trauma. In group 2 (Fig. 7a), numerous cells with BrdU labeling were seen in the damaged abneural half of the basilar papilla. In group 3 (Fig. 7b), there were less labeled cells than in group 2. In particular, no labeled cells were seen at the very abneural edge of the basilar papilla. In the basilar papillae of the two pigeons, which received BrdU at day 7 and day 8 after sound trauma (group 4), some labeled monolayer cells and a labeled hair cell were observed (Fig. 5b). To answer the question, if cochleae with longer survival time have more regenerated hair cells, cochleae with functional `slow recovery' were examined after a survival time of 1^16 weeks. In cochleae with one week survival time (C84L), there were numerous labeled monolayer cells in the hair cell loss region, a few labeled supporting cells, but no labeled hair cells. A pair of labeled cells that may have been regenerated but still immature hair cells, was found at the surface of the sensory epithelium (Fig. 8). However, typical features of the hair cell, cuticular plate and stereocilia were not
117
Fig. 8. Cross section from a cochlea (C86L) with one week survival time. A pair of BrdU-labeled cells, probably immature hair cells (arrowheads), are seen at the surface of the basilar papilla. Bar: 10 Wm.
seen in these two cells. In cochleae with three weeks survival time, clearly labeled hair cells were already observed. There was no evidence of an increase in the number of labeled hair cells with survival time in the range of 3^16 weeks. 4. Discussion 4.1. Cell proliferation after trauma
Fig. 7. Comparison of the BrdU labeling on whole mounts of the basilar papilla. Top: neural; bottom: abneural. a: C133L (injected with BrdU at day 0^3 after acoustic trauma), labeled cells are seen in the abneural half of the basilar papilla. b: C131L (injected with BrdU at day 4^6 after acoustic trauma), the amount of the labeled cells is less than that in a. At the very abneural edge of the basilar papilla there are no labeled cells.
In the present study, BrdU labeling was observed neither in the control basilar papillae, nor in the undamaged regions of sound-damaged basilar papillae of the adult pigeon. This result supports the ¢nding that hair cells in the normal avian basilar papilla do not divide during postembryonic life (Corwin and Cotanche, 1988 ; Girod et al., 1989). However, postembryonic proliferation was observed in adult quail (Ryals and Westbrook, 1990). Also in young chicks, a very low level of proliferation of supporting cells was found in the apical region of the normal basilar papilla (Oesterle and Rubel, 1993 ; Raphael, 1993). A possible reason for the di¡erence may be that we have not given pigeons BrdU continuously, so that not all of the dividing cells could be labeled. It is also conceivable that the rate of spontaneous proliferation di¡ers between species and with age. BrdU labeled cells were observed only in animals that su¡ered from sound trauma, and only in damaged areas of the basilar papilla. This indicates that cell pro-
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
118
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
liferation was triggered by the trauma. The fact that it made no di¡erence in the amount of labeled cells in the present experiments whether BrdU was injected during 3 days post trauma, or only in a time window 40^54 h post trauma, suggests that the peak of cell proliferation is reached around this time after trauma. The same observation was made after sound trauma in the chick (Hashino and Salvi, 1993; Stone and Cotanche, 1994). In the present experiments, a low level of proliferation of non-sensory cells and of hair cells was observed up to 8 days after trauma. It could well be that these cells result from more than one round of cell division, as suggested by Stone and Cotanche (1994) and that the BrdU was passed on to the daughter cells. This possibility also indicates that the time of birth of a labeled cell may be after the time of pulse labeling by BrdU, because the cell can result from further division. Some labeled hair cells appeared in pairs (Figs. 6 and 8). This observation which was also made by Raphael (1992, 1993), Stone and Cotanche (1994) supports the hypothesis, that both symmetric and asymmetric di¡erentiation of new dividing supporting cells into new hair cells occurs. 4.1.1. Hair cells Only a very small number of the labeled cells were hair cells. One reason for this is that hair cell regeneration after the severe sound trauma was incomplete. Large abneural areas of the basilar papilla remained without new hair cells, as observed up to 16 weeks after trauma. Similar observations made in the pigeon with even longer recovery times indicate that these de¢cits are permanent (Muëller et al., 1996, 1997). New hair cells were only generated in a small area of the basilar papilla, bordering the neural side of the abneural region of irreversible hair cell loss. These ¢ndings, together with published results on sound trauma at lower exposure levels (Rubel and Ryals, 1982 ; Cousillas and Rebillard, 1985; Cotanche, 1987a ; Cotanche et al., 1987 ; Corwin and Cotanche, 1988; Girod et al., 1989 ; Adler et al., 1992; Hashino and Salvi, 1993 ; Chen et al., 1996a) suggest that there is a gradient across the width of the papilla in the susceptibility of the sensory epithelium to trauma. With increasing exposure level and time, sound trauma increases from the abneural edge of the papilla towards the neural edge (Cotanche et al., 1987; Cotanche and Dopyera, 1990 ; Adler et al., 1992). At levels high enough to eradicate also the neural tall hair cells, as observed in the present study, damage to the abneural portion of the basilar papilla becomes so severe that regeneration of hair cells does not occur any more in these abneural areas. This is probably due to destruction of the supporting cells (Cotanche et al., 1995). In the narrow transition area between the damaged and undamaged area a number of supporting cells survive, proliferate and give rise to new
hair cells. With increasing sound trauma this transitional zone moves progressively from abneural into the neural direction. The lower susceptibility of the neural side of the papilla may be related to the mechanics of the inner ear. The neural hair cells are located over the rigid neural limbus which vibrates at much lower amplitude than the free basilar membrane (Gummer et al., 1987). The neural hair cells are thought to receive their input via the motion of the tectorial membrane (Raphael, 1991 ; Smolders et al., 1995; Salvi et al., 1996). During sound damage the tectorial membrane decouples from the hair cells (Cotanche, 1987b), which will reduce the vibration input to the neural hair cells and protect them from sound damage (Raphael, 1991). The fact that the area of the papilla over the neural limbus and the area over the free basilar membrane roughly correspond to the areas occupied by `tall' and `short' hair cells, respectively, may be purely coincidental (Smolders et al., 1995). At very high sound pressure levels, the vibration amplitude of the free basilar membrane may become so large as to destroy even the supporting cells over the free basilar membrane. The vibration amplitude of the neural limbus may become large enough to damage or even destroy the overlying tall hair cells, but leave the supporting cells and thus the potential of hair cell regeneration in the transitional zone between neural limbus and free basilar membrane. 4.1.2. Non-sensory cells The majority of the labeled progeny of the dividing cells were non-sensory monolayer cells. However, labeling in the hyaline cell area was present only in the cochleae after one week survival, not after more than three weeks survival. This ¢nding, together with that from Girod et al. (1989), suggests that new hyaline cells are present shortly after sound exposure, but then may di¡erentiate, migrate away or die during a longer survival period (Girod et al., 1989). Almost all of the monolayer cells that occupied irreversibly damaged areas were labeled. This indicates that the proliferation of the progenitors of the monolayer cells occurred at the same time as that of the hair-cell progenitors. However, as there is evidence from the chicken, that the precursors of the hair cells (supporting cells, Raphael, 1992, 1993; Tsue et al., 1994) are di¡erent from those of the monolayer cells (hyaline cells, Girod et al., 1989; Cotanche et al., 1995), this may just mean that the proliferation of both cell types is triggered by the same event, the sound trauma which causes the loss of the hair cells and the supporting cells, respectively. The monolayer cells then rapidly proliferate and cover the whole damaged region where both hair cells and supporting cells are missing. Our ¢nding that the areas with monolayer cells are not repopulated with hair cells 8 weeks or more (up to 16 weeks) after trauma supports the hypothesis that the monolayer cells ac-
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
tually prevent this repopulation (Cotanche et al., 1995, 1996). 4.1.3. Functional recovery It appears that the three groups of functional recovery, `fast', `slow' and `no' that were distinguished in the present study represent three instances of a continuum of increasing damage across the basilar papilla. Fast recovery without residual hearing loss represents the situation where no or negligible hair cell loss occurs. In this case, hair cell loss and regeneration play an insigni¢cant role in the functional impairment and recovery. Rather, hearing loss after sound exposure results from reversible functional impairment, possibly of hair cell transduction, as well as from decoupling of the tectorial membrane (Cotanche, 1992; Marsh et al., 1990 ; Saunders et al., 1996a) and impairment of the function of the tegmentum vasculosum (Ryals et al., 1995), involving loss of the endocochlear potential in young (Poje et al., 1995), but not in adult (Chen et al., 1995) chickens. These impaired functions recover rapidly after trauma, leading to restoration of function within days (McFadden and Saunders, 1989; Adler et al., 1992 ; Niemiec et al., 1994; Muëller et al., 1996 ; Saunders et al., 1996a). When the damage expands towards the neural edge of the basilar papilla, irreversible damage in the abneural area occurs. Not only hair cells but also their precursor cells are lost and no regeneration of hair cells in these areas is possible. Instead this area is populated by monolayer cells, which may prevent its reoccupation by hair cell precursor cells (Cotanche et al., 1995). Only a limited number of new hair cells regenerates in the small transitional zone between the abneural damaged area and the neural undamaged area. The abneural part of the tectorial membrane is destroyed as well. Because of the loss of the supporting cells in the abneural damaged region, no new tectorial membrane can be regenerated (Girod et al., 1995). As a consequence, the sound induced vibration of the free basilar membrane can no longer be transferred to the tall hair cells over the neural limbus, resulting in threshold elevation, reduction of tuning sharpness and abnormal rate intensity functions of the appertaining auditory nerve ¢bers (Chen et al., 1996b, Saunders et al., 1996b; Muëller et al., 1996, 1997). This is the situation designated `slow recovery'. Recovery takes longer because there is more damage to repair and because regeneration of hair cells is involved. Because the structural regeneration is incomplete, there is incomplete functional recovery and a permanent residual hearing loss remains. It appears, as one would expect, that an intact structure of the basilar papilla is necessary for normal sensitivity. It is not su¤cient that the tall hair cells remain, which are innervated by the majority of the a¡erent cochlear nerve ¢bers (von
119
Duëring et al., 1985 ; Fischer, 1992, 1994 ; Smolders et al., 1995). `No recovery' represents the other extreme, where damage has extended to the neural tall hair cells and their precursor cells have been destroyed over large parts of the papilla. The resulting permanent hearing loss represents cochleae with only a few rows of neural hair cells remaining, a destroyed tectorial membrane and probably signi¢cant loss of ganglion cells (Ryals et al., 1989). Acknowledgments Supported by the Deutsche Forschungsgemeinschaft (SFB 269, B1). We thank S. Hoidis and H. Schalk for expert technical assistance. References Adler, H.J., Kenealy, J.F.X., Dedio, R.M., Saunders, J.C., 1992. Threshold shift, hair cell loss, and hair bundle sti¡ness following exposure to 120 and 125 dB pure tones in the neonatal chick. Acta Otolaryngol. (Stockh.) 112, 444^454. Chen, L., Trautwein, P.G., Miller, K., Salvi, R.J., 1995. E¡ects of kanamycin ototoxicity and hair cell regeneration on the DC endocochlear potential in adult chickens. Hear. Res. 89, 28^34. Chen, L., Trautwein, P.G., Shero, M., Salvi, R., 1996a. Correlation of hair cell regeneration with physiology and psychophysics in adult chickens following acoustic trauma. In: R.J. Salvi, D. Henderson, V. Colletti, F. Fiorino (Eds.), Auditory System Plasticity and Regeneration. Thieme Medical Publ., New York, pp. 43^61. Chen, L., Trautwein, P.G., Shero, M., Salvi, R., 1996b. Tuning, spontaneous activity and tonotopic map in chicken cochlear ganglion neurons following sound-induced hair cell loss and regeneration. Hear. Res. 00, 152^164. Corwin, J.T., Cotanche, D.A., 1988. Regeneration of sensory hair cells after acoustic trauma. Science 240, 1772^1774. Cotanche, D.A., 1987a. Regeneration of hair cell stereociliary bundles in the chick cochlea following severe acoustic trauma. Hear. Res. 30, 181^196. Cotanche, D.A., 1987b. Regeneration of the tectorial membrane in the chick cochlea following severe acoustic trauma. Hear. Res. 30, 197^206. Cotanche, D.A., Saunders, J.C., Tilney, L.G., 1987. Hair cell damage produced by acoustic trauma in the chick cochlea. Hear. Res. 25, 267^286. Cotanche, D.A., Dopyera, C.E.J., 1990. Hair cell and supporting cell response to acoustic trauma in the chick cochlea. Hear. Res. 46, 29^40. Cotanche, D.A., 1992. Video-enhanced DIC images of the noise-damaged and regenerated chick tectorial membrane. Exp. Neurol. 115, 23^26. Cotanche, D.A., Lee, K.H., Stone, J.S., Picard, D.A., 1994. Hair cell regeneration in the bird cochlea following noise damage or ototoxic drug damage. Anat. Embryol. 189, 1^18. Cotanche, D.A., Messana, E.P., Ofsie, M.S., 1995. Migration of hyaline cells into the chick basilar papilla during severe noise damage. Hear. Res. 91, 148^159. Cotanche, D.A., Messana, E.P., Riedl, 1996. Hyaline cell migration into the basilar papilla involves the secretion of a new layer of extracellular matrix. Assoc. Res. Otolaryngol. (Abstr.) 19, 199.
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart
120
D. Ding-Pfennigdor¡ et al. / Hearing Research 120 (1998) 109^120
Cotanche, D.A., 1997. Hair cell regeneration in the avian cochlea. Ann. Otol. Rhinol. Laryngol. 106, 9^15. Cousillas, H., Rebillard, G., 1985. Age-dependent e¡ects of a pure tone trauma in the chick basilar papilla: Evidence for a development of the tonotopic organization. Hear. Res. 19, 217^226. Duckert, L.G., Rubel, E.W., 1990. Ultrastructural observations on regenerating hair cells in the chick basilar papilla. Hear. Res. 48, 161^182. Fischer, F.P., 1992. Quantitative analysis of the innervation of the chicken basilar papilla. Hear. Res. 61, 167^178. Fischer, F.P., 1994. General pattern and morphological specializations of the avian cochlea. Scanning Microsc. 8, 351^364. Girod, D.A., Duckert, L.G., Rubel, E.W., 1989. Possible precusors of regenerated hair cells in the avian cochlea following acoustic trauma. Hear. Res. 42, 175^194. Girod, D.A., Ryals, B.M., Fankhauser, C.E., Westbrook, E.W., 1995. Long term structural changes in the chick tectorial membrane following severe acoustic damage. Assoc. Res. Otolaryngol. (Abstr.) 18, 787. Gleich, O., 1989. Auditory primary a¡erents in the starling: Correlation of function and morphology. Hear. Res. 37, 255^268. Gummer, A.W., Smolders, J.W.T., Klinke, R., 1987. Basilar membrane motion in the pigeon measured with the Moëssbauer technique. Hear. Res. 29, 63^92. Hashino, E., Tanak, Y., Sokabe, M., 1991. Hair cell damage and recovery following chronic application of kanamycin in the chick cochlea. Hear. Res. 52, 356^368. Hashino, E., Salvi, R.J., 1993. Changing spatial patterns of DNA replication in the noise-damaged chick cochlea. J. Cell Science 105, 23^31. Manley, G.A., Gleich, O., 1992. The evolutionary biology of hearing. In: D. Webster, R.R. Fay, A.N. Popper (Eds.), Evolution and Specialization of Function in the Avian Auditory Periphery. Springer Verlag, New York, pp. 561^580. Marsh, R.R., Xu, L.R., Moy, J.P., Saunders, J.C., 1990. Recovery of the basilar papilla following intense sound exposure in the chick. Hear. Res. 46, 229^238. McFadden, E.A., Saunders, J.C., 1989. Recovery of auditory function following intense sound exposure in the neonatal chick. Hear. Res. 41, 205^216. Muëller, M., Smolders, J.W.T., Ding-Pfennigdor¡, D., Klinke, R., 1996. Regeneration after tall hair cell damage following severe acoustic trauma in adult pigeons: correlation between cochlear morphology, compound action potential responses and single ¢ber properties in single animals. Hear. Res. 102, 133^154. Muëller, M., Smolders, J.W.T., Ding-Pfennigdor¡, D., Klinke, R., 1997. Discharge properties of pigeon single auditory nerve ¢bers after recovery from severe acoustic trauma. Int. J. Dev. Neurosci. 15, 401^416. Niemiec, A.J., Raphael, Y., Moody, D.B., 1994. Return of auditory function following structural regeneration after acoustic trauma: behavioral measures from quail. Hear. Res. 79, 1^16. Oesterle, E.C., Rubel, E.W., 1993. Postnatal production of supporting cells in the chick cochlea. Hear. Res. 66, 213^224. Patuzzi, R.B., Bull, C.L., 1991. Electrical responses from the chicken basilar papilla. Hear. Res. 53, 57^77. Poje, C.P., Sewell, D.A., Saunders, J.C., 1995. The e¡ects of exposure to intense sound on the DC endocochlear potential in the chick. Hear. Res. 82, 197^204. Pugliano, F.A., Wilcox, T.O., Rossiter, J., Saunders, J.C., 1993. Recovery of auditory structure and function in neonatal chicks exposed to intense sound for 8 days. Neurosci. Lett. 151, 214^ 218. Raphael, Y., 1991. Pure-tone overstimulation protects surviving avian hair cells from acoustic trauma. Hear. Res. 53, 173^184.
Raphael, Y., 1992. Evidence for supporting cell mitosis in responce to acoustic trauma in the avian inner ear. J. Neurocytol. 21, 663^671. Raphael, Y., 1993. Reorganization of the chick basilar papilla after acoustic trauma. J. Comp. Neurol. 330, 521^532. Rebillard, G., Ryals, B.M., Rubel, E.W., 1982. Relationship between hair cell loss on the chick basilar papilla and threshold shift after acoustic overstimulation. Hear. Res. 8, 77^82. Rosowski, J.J., Saunders, J.C., 1980. Sound transmission through the avian interaural pathways. J. Comp. Physiol. A 136, 183^190. Rubel, E.W., Ryals, B.M., 1982. Patterns of hair cell loss in chick basilar papilla after intense auditory stimulation. Exposure duration and survival time. Acta Otolaryngol. (Stockh.) 93, 31^41. Ryals, B.M., Rubel, E.W., 1982. Patterns of hair cell loss in chick basilar papilla after intense auditory stimulation. Frequency organisation. Acta Otolaryngol. (Stockh.) 93, 205^210. Ryals, B.M., Rubel, E.W., 1985. Di¡erential susceptibility of avian hair cells to acoustic trauma. Hear. Res. 19, 73^84. Ryals, B.M., Teneyck, B., Westbrook, E.W., 1989. Ganglion cell loss continues during hair cell regeneration. Hear. Res. 43, 81^90. Ryals, B.M., Westbrook, E.W., 1990. Hair cell regeneration in senescent quail. Hear. Res. 50, 87^96. Ryals, B.M., Stalford, M.D., Lambert, P.R., Westbrook, E.W., 1995. Recovery of noise-induced changes in the dark cells of the quail tegmentum vasculosum. Hear. Res. 83, 51^61. Salvi, R., Henderson, D., Fiorino, F., Colletti, V., 1996. Auditory System Plasticity and Regeneration. Thieme Medical Publishers, New York. Saunders, J.C., Tilney, L.G., 1982. Species di¡erences in susceptibility to noise exposure. In: R.P. Hamernik, D. Henderson, R. Salvi (Eds.), New Perspectives on Noise-Induced Hearing Loss. Raven, New York. Saunders, J.C., Adler, H.J., Pugliano, F.A., 1992. The structural and functional aspects of hair cell regeneration in the chick as a result of exposure to intense sound. Exp. Neurol. 115, 13^17. Saunders, J.C., Torsiglieri, A.J., Dedio, R.M., 1993. The growth of hearing loss in neonatal chicks exposed to intense pure tones. Hear. Res. 69, 25^34. Saunders, J.C., Doan, D.E., Cohen, Y.E., Adler, H.J., Poje, C.P., 1996a. Recent observations on the recovery of structure and function in the sound-damaged chick ear. In: R.J. Salvi, D. Henderson, F. Fiorino, V. Colletti (Eds.), Auditory System Plasticity and Regeneration. Thieme Medical Publishers, New York, pp. 62^83. Saunders, J.C., Doan, D.E., Poje, C.P., Fisher, K.A., 1996b. Cochlear nerve activity after intense sound exposure in neonatal chicks. J. Neurophysiol. 76, 770^787. Smolders, J.W.T., Ding, D., Klinke, R., 1992. Normal tuning curves from primary a¡erent ¢bres innervating short and intermediate hair cells in the pigeon ear. Adv. Biosci. 83, 197^204. Smolders, J.W.T., Ding-Pfennigdor¡, D., Klinke, R., 1995. A functional map of the pigeon basilar papilla: Correlation of the properties of single auditory nerve ¢bres and their peripheral origin. Hear. Res. 92, 151^169. Stone, J.S., Cotanche, D.A., 1994. Identi¢cation of the timing of S phase and the patterns of cell proliferation during hair cell regeneration in the chick cochlea. J. Comp. Neurol. 341, 50^67. Tsue, T.T., Walting, D.L., Weisleder, P., Coltrera, M.D., Rubel, E.W., 1994. Identi¢cation of hair cell progenitors and intermetotic migration of their nuclei in the normal and regenerating avian inner ear. J. Neurosci. 14, 140^152. Tucci, D.L., Rubel, E.W., 1990. Physiological status of regenerated hair cells in the avian inner ear following aminoglycoside ototoxicity. Otolaryngol. Head Neck Surg. 103, 443^445. von Duëring, M., Andres, K.H., Simon, K., 1985. The comparative anatomy of the basilar papillae in birds. Fortschr. Zool. 30, 681^ 685.
HEARES 3021 27-5-98 Cyaan Magenta Geel
Zwart