The distribution and the modulation of tyrosine hydroxylase immunoreactivity in the lateral olivocochlear system of the guinea-pig

Neuroscience 125 (2004) 725–733

THE DISTRIBUTION AND THE MODULATION OF TYROSINE HYDROXYLASE IMMUNOREACTIVITY IN THE LATERAL OLIVOCOCHLEAR SYSTEM OF THE GUINEA-PIG X. NIU,a N. BOGDANOVICb AND B. CANLONa*

role for tyrosine hydroxylase (TH), a key enzyme for dopamine synthesis, in protecting against acoustic trauma by sound conditioning has been demonstrated (Niu and Canlon, 2002). In that study, it was found that acoustic trauma decreased TH within the lateral efferents under the inner hair cells, while sound conditioning up-regulated TH. Moreover, pre-treatment with 6-hydroxydopamine, a toxin that disrupts TH synthesis, blocked the protective effect of sound conditioning and thereby suggests a role for the lateral efferents in sound conditioning (Niu and Canlon, 2002). The lateral efferents make axo-dendritic synapses with the afferent dendrites under the inner hair cells, and arise from neurons located in the brainstem (Warr and Guinan, 1979). One functional role of the lateral efferents is to inhibit the activity of the inner hair cell afferent dendrites, and protect against excitotoxicity (Gil-Loyzaga et al., 1994; d’Aldin et al., 1995a,b). The main purpose of the present study was to determine whether protection against acoustic trauma by sound conditioning is governed by central mechanisms, or if the lateral efferent synapses under the inner hair cells in the cochlea are acting locally. The lateral olivocochlear system (LOC) system originates from the lateral superior olive (LSO) and the surrounding periolivary nuclei (Warr et al., 2002). The LOC neurons originating in the LSO have relatively small cell bodies and give rise to small unmyelinated axons, while the LOC neurons from the periolivary regions are larger and myelinated (Aschoff and Ostwald, 1988). There are at least five different neuronal types in the LSO and the differences in immunocytochemical profiles suggest unique functional responses from these neurons (Helfert and Schwartz, 1986, 1987). In addition, there appear to be two different types of fibers innervating the base of the inner hair cells as described in the guinea-pig and the rat (Brown, 1987; Warr et al., 1997). One fiber type originates in the LSO proper and terminates as dense branches under the inner hair cells, while the other fiber type originates immediately adjacent to the LSO proper (periolivary regions) and terminates as sparse branches under the inner hair cells. In order to characterize the morphology and location of TH positive neurons within the LOC system we have employed retrograde dextran tracings combined with TH immunocytochemistry. Comparisons were made among different sound stimulation models including acoustic trauma, sound conditioning, and the combination of sound conditioning and acoustic trauma in guinea-pigs.

a

Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, Sweden b Geriatric Section, NEUROTEC, Karolinska Institutet, Stockholm, Sweden

Abstract—It was previously shown that tyrosine hydroxylase (TH) immunoreactivity in the terminals of the lateral efferents of the cochlea is decreased by acoustic trauma and that sound preconditioning counteracted this decrease [Hear Res 174 (2002) 124]. Here we identify those neurons in the lateral olivocochlear system (LOC) in the brainstem that regulates the peripheral expression of TH in the cochlea. By employing retrograde tracing techniques, dextran-labeled neurons were found predominantly in the ipsilateral LOC system including lateral superior olive (LSO), and the surrounding periolivary regions (dorsal periolivary nucleus [DPO], dorsolateral periolivary nucleus [DLPO], lateral nucleus of trapezoid body [LNTB]). Employing immunocytochemistry, it was found that a control group had 35% of the ipsilateral LOC neurons positively stained with TH. Of the total population of TH neurons, 77% were double-stained (TH and dextran) in the LOC system. Acoustic trauma decreased the number of TH positive neurons in the LSO and the surrounding DLPO, and caused a reduction of TH fiber immunolabeling in these regions. Changes were not found in the DPO or the LNTB after acoustic trauma. Sound conditioning protected against the decrease of TH immunolabeling by acoustic trauma and increased the fiber staining for TH in the LSO and DLPO, but not in the DPO or the LNTB. These results provide evidence that TH positive neurons are present in the LOC system in the guinea-pig. It is now demonstrated that protection against acoustic trauma by sound conditioning has a central component that is governed by TH in the LSO and the surrounding periolivary DLPO region. © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: dopamine, acoustic trauma, cochlea, dextran, sound conditioning, stereology.

Sound conditioning, a low-level acoustic stimulus delivered prior to a traumatic stimulus, is an established means of protecting the cochlea against the morphological, functional, and molecular consequences induced by acoustic trauma (Canlon et al., 1988; Subramaniam et al., 1992, 1993; Ryan et al., 1994; Kujawa and Liberman, 1997). A *Corresponding author. Tel: ⫹46-8-728-7248; fax: ⫹46-8-327-026. E-mail address: [email protected] (B. Canlon). Abbreviations: DAB, 3,3⬘-diaminobenzidine tetrahydrochloride; DLPO, dorsolateral periolivary nucleus; DPO, dorsal periolivary nucleus; LNTB, lateral nucleus of trapezoid body; LOC, lateral olivocochlear system; LSO, lateral superior olive; PBS, phosphate buffer; TH, tyrosine hydroxylase.

0306-4522/04$30.00⫹0.00 © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.02.023

725

726

X. Niu et al. / Neuroscience 125 (2004) 725–733

EXPERIMENTAL PROCEDURES A total of 25 pigmented guinea-pigs (250 – 400 g body weight) without any evidence of middle ear pathology were used in this study. They were divided into the following groups: dextran retrograde tracing (fluorescein, n⫽4 and biotinylated, n⫽3); TH– dextran double labeling, n⫽3; and NeuN neuronal counting, n⫽3. In addition, a total of 12 guinea-pigs were used for the different sound stimulation paradigms. The Ethical Committee at the Karolinska Institutet approved the care and use of the animals. All experiments conformed to local guidelines on the ethical use of animals. The number of animals used in this study and their suffering were minimized as much as possible.

Sound conditioning and noise trauma exposure paradigms The guinea-pigs were divided into four groups: (1) acoustic trauma (2.7 kHz tone at 103 dB, SPL for 30 min; n⫽3) where the stimulus was monaurally delivered through a closed system into the ear canals of anesthetized (50 mg/kg ketamine and 10 mg/kg xylazine i.m.) guinea-pigs. The tone used for the acoustic trauma was generated by a synthesizer (8904A Hewlett-Packard Company, Palo Alto, CA, USA) and delivered through a Beyer DT880 headphone; (2) sound conditioning (1.0 kHz tone at 81 dB SPL for 24 h, n⫽3) where the sound conditioning stimulus was generated by a Wavetek signal generator (model 187), amplified by a Quad 303 (Acoustic Manufacturing Co., England) in an open field acoustic chamber (225⫻120⫻100 cm); (3) sound conditioning followed by acoustic trauma (n⫽3). The animals were given a 1-h recovery before being exposed to the acoustic trauma, and (4) control animals were maintained in an ambient sound environment (n⫽3).

Neuronal labeling Two hours after each sound exposure, guinea-pigs were deeply anesthetized with sodium pentobarbital (600 mg/kg, i.p.) and transcardially perfused with 300 ml 0.1 M phosphate buffer, pH 7.4 (PBS) containing heparin 10 iu/ml, followed by 400 ml 4% paraformaldehyde in PBS. The brains were removed and immersed for 5 h in the same fixative at 4 °C, and cryoprotected first in 10% sucrose/PBS for 24 h and then in 20% sucrose for 48 h. The brains were frozen in a mixture of dry ice and isopentane (⫺40 °C) and were coronally sectioned at the level of superior olivary complex on a cryostat (HM 500M, Zeiss) at ⫺24 °C and mounted on slides (Super Frost Plus; Menzel Glaser, Germany). Care was taken to acquire sections that were symmetric. Brainstem sections were processed for immunohistochemical staining using the avidin– biotin–peroxidase method (Vectastain, ABC kit; Vector Laboratories, Burlingame, CA, USA). The sections were incubated in 0.3% hydrogen peroxide and 0.3% Triton X-100 for 30 min and then blocked with 1% bovine serum albumin containing 0.15% Triton X-100 for 30 min in a humid chamber. Sections were then incubated in 1:2000 anti-TH (rabbit, polyclonal; Calbiochem, La Jolla, CA, USA) or in 1:600 anti-NeuN (mouse, monoclonal; Chemicon, Temecula, CA, USA) at 4 °C for 24 –30 h. Sections were then incubated in 1:2000 biotinylated secondary antibody (anti-rabbit for TH and anti-mouse for NeuN; Vector Laboratories) for 1 h and were processed in an avidin– biotin–peroxidase complex solution for 30 min. Visualization of the immunoreactivity was carried out using 3,3⬘-diaminobenzidine tetrahydrochloride (DAB; for NeuN immunostaining) mixed with nickel ammonium sulfate (DAB-Ni, for TH immunostaining) for 5 min. The slides were rinsed and air-dried. Tests were performed to adjust the dilutions of the primary antibody, and of the secondary and tertiary system for selecting the optimal condition. For standardization of the immunohistochemical procedure we used a dilution of the primary antibody and a concentration of the DAB-Ni far from saturation

and an incubation time adjusted so that the darkest elements in the brain section were below saturation. Glial cells in the LSO were not immunoreactive. Negative controls were obtained by replacing the primary antibody with non-immune rabbit IgG or mouse IgG. After the sections were immunostained for either TH or NeuN, they were then stained for Nissl. The slides were placed in xylene to remove the cover glass and then were successively immersed in 99%, 95%, 70% ethanol for 2 min, then rinsed in distilled water for 5 min, and finally in 2.5% Cresyl Violet solution until adequate staining was achieved. The slides were placed in PBS and then through a graded series of alcohol solutions 70%, 95%, and 99% ethanol and finally immersed in xylene for 5 min, and coverslipped using Accu-Mount 280 (Baxter, McGaw Park, IL, USA).

Dextran tracing and double labeling with TH Guinea-pigs were anesthetized with an i.m. injection of 50 mg ketamine and 10 mg xylazine per kg body weight. The tympanic bulla was opened and a tracer of 20% dextran fluorescein, or biotinylated dextran (10,000 MW; Molecular Probes, Eugene, OR, USA) at a volume of 10 ␮l was injected into the scala tympani through the round window with a Hamilton syringe. After a survival time of 6 – 8 days, fixation and sectioning of the brainstems were performed as described above. BDA labeling was visualized by the Vectorstain, ABC kit, using DAB-Ni as the chromogen. For the TH immunofluorescence staining, sections were processed in TH primary antibody (rabbit; Calbiochem) as described above. The sections were incubated in TRITC conjugated to anti-rabbit secondary antibody (1:200; Sigma) for 1 h and then treated with Citi Fluor and then coverslipped.

Anatomical boundaries The LOC system consists of the LSO and its periolivary regions. Anatomical boundaries were determined from sections that were stained with a neuron-specific antibody, anti-NeuN in conjunction with different anatomical maps found in the literature (Tsuchitani, 1977; Schofield and Cant, 1991; Warr et al., 2002). The LSO is a convoluted structure with a clearly defined border due to the fibers that encircle this nucleus. The LSO is divided into a medial and a lateral limb by the hilus. There are several periolivary nuclei found immediately adjacent to the LSO. The dorsal periolivary regions include the dorsal periolivary nucleus (DPO) and the dorsolateral periolivary nucleus (DLPO). The DLPO is located lateral to the DPO and is relatively easy to identify due to its large and tightly clustered neurons. The DPO is a region dorsalmedial to the MSO and was a region with a relatively small number of neurons. In this study, the lateral nucleus of trapezoid body (LNTB) region covered the LNTB and nearby regions including the anterolateral periolivary and ventrolateral periolivary nuclei. The LNTB region extended the ventrolateral area of the LSO to the ventral area of the DPO, along the full rostrocaudal length of the LSO.

Stereological analysis The stereological analysis was performed using an Olympus video stereological analysis system (BICO, Copenhagen) and the GRID 1.2 version (Interactivision ApS; Silkeborg, Denmark) was used for the generation of the counting frames of the required areas. The systematically sampled sections were first viewed on the screen at low magnification (2.5⫻ objective). The entire LSO was delineated on the screen with a cursor. Subsequently, an automated systematic sampling procedure within the area of interest was performed using a 100⫻ oil immersion objective with a high numerical aperture (1.4). A neuron was classified as being immunopositive when a brown deposit from the peroxidase staining in the nucleus was darker than background staining. Nissl staining served as a counter-stain for the NeuN-immunolabeled sections

X. Niu et al. / Neuroscience 125 (2004) 725–733

727

Fig. 1. Neurons immunostained with NeuN in the LSO regions. The neurons present in the LSO include: (a) principal neurons; (b) small neurons; (c) marginal neurons; (d) multiplanar neurons; and (e) slender neurons. The neurons present in the DLPO include: (a) bipolar neurons; and (b) large multipolar neurons. The neurons present in the DPO include (a) multipolar neurons; and (b) bipolar neurons. The neurons present in the LNTB include (a) bipolar neurons; and (b) multipolar neurons. Top right is a schematic drawing showing the anatomical location and arrangement of the LSO, DLPO, DPO and LNTB.

allowing the neurons to be clearly visualized and the borders of the LSO and the periolivary regions to be easily discerned. Nissl stain also helped for identifying the cytoplasm and nucleolus of the neurons. The stereological probe was superimposed on the microscopic image, and Nissl/NeuN-immunoreactive neurons were sampled if their nucleoli were inside the sampling frame and did not touch either one of the two (left and bottom) forbidden lines. The optical disector (Gundersen et al., 1988) was used to obtain unbiased estimates of total numbers of neurons. The coefficient of error was calculated according to West et al. (1991) and was well below biological variability.

Quantifications of dextran- and TH-labeled neurons The relationship between dextran-labeled and TH-labeled neuronal profiles was quantified in sections that were double-labeled (dextran and TH). For this analysis counting was performed on all sections (30 –32 sections/guinea-pig, 30 ␮m/section) using FITC and TRITC filters at a magnification of 40⫻. To quantify TH positive neurons in the four groups of animals, every third section was analyzed (64 – 68 sections/guinea-pig, 14 ␮m/section). The total number of TH neurons was obtained by calculating the neuronal number obtained from every third section multiplied by the total number of sections. Neurons with ambiguous labeling or an unidentifiable nucleolus (via Nissl staining) was excluded from the analysis. To determine soma size, a 100⫻ oil immersion objective lens was used.

Quantifications of TH fibers To quantify the densities of TH fibers, a multiple z-plane image was scanned with the aid of a motor controller (Prior, England; model H128) that compiled these images into one picture (Image Pro Plus; Media Cybernetics, CA, USA). From this composite picture, the LSO, DLPO, DPO, and LNTB were silhouetted and the mean pixel density of TH immunoreactivity was obtained by randomly setting the same sampling box within these regions. The

background pixel density was obtained from a neutral region in proximity to the LSO. The values of the mean pixel density for the regions of interest were then background-subtracted. Measurements were determined on every third section for each animal. The values from all animals in each group were averaged and then normalized to the mean value of the pixel density from the control animals.

Statistical analysis Data values were presented as mean⫾S.D. Comparisons were performed using Student’s t-test or one-way ANOVA followed by Tukey test. Differences were considered statistically significant when the P value was ⬍0.05.

RESULTS Morphology of neurons in the LSO and periolivary regions The total number of NeuN immunopositive neurons in the LSO was 3572⫾414. The mean depth of the entire guineapig lateral superior olivary complex was 94⫾3 ␮m. NeuN immunostaining allowed for the neurons in the LSO to be classified according to their distinct morphology (Fig. 1). A schematic illustration of the LSO and the periolivary regions are shown (Fig. 1, top right). In the LSO proper, the majority of neurons were the principal neurons. This neuron was bipolar and possessed a fusiform cell body with at least one, and often two dendrites, measuring around 10 –25 ␮m in diameter (Fig. 1, LSO a). Small neurons in the LSO possessed an oval or round cell body, measuring less than 7–9 ␮m in diameter and often seen with a single dendrite (Fig. 1, LSO b). Marginal neurons appeared similar to princi-

728

X. Niu et al. / Neuroscience 125 (2004) 725–733

Fig. 2. Representative examples of dextran–fluorescein-labeled neurons in the LOC (top left and right). Higher magnification micrographs showing examples of retrogradely dextran-labeled neurons in the LOC, LSO, DLPO, DPO, and LNTB (middle panel). Quantification of dextran-labeled neurons (mean⫾S.D.) in the LOC including the LSO, DLPO, DPO, and the LNTB. The majority of LOC neurons were found within LSO (bottom).

pal neurons in morphology, but were identified along the contours of the LSO and oriented orthogonal to principal neurons. The cell body measured 12–32 ␮m in length (Fig. 1, LSO c). A few of multiplanar neurons were found with dendrites randomly distributed over the surface of the cell and measuring 10 –25 ␮m in diameter (Fig. 1, LSO d). A few neurons were found located in the matrix of the LSO, as previously identified as class 5 (Helfert et al., 1986, 1987; Fig. 1, LSO e). The DLPO neurons were generally larger than the neurons in the LSO. The majority of DLPO neurons were bipolar or multipolar measuring 15–25 ␮m in length (Fig. 1, DLPO a), while others were polygonal with multipolar extensions and measured 26 – 45 ␮m in length (Fig. 1, DLPO b). In the region of the DPO, neurons were basically bipolar and measured 12–25 ␮m in length (Fig. 1, DPO a– b). In the LNTB neurons were middle-sized (12–25 ␮m) and were typically round or oval and displayed bipolar or multipolar morphology (Fig. 1, LNTB a– b). Identification of LOC neurons Examples of the retrograde tracer, dextran–fluorescein throughout the LSO is shown (Fig. 2 top left and right). The results from each tracer, dextran–fluorescein and dextranbiotinylated (not shown) resulted in a similar number and location of the neurons. The only difference between the two methods was that the biotinylated labeling stained the fibers more intensely than the fluorescent label. The majority of dextran labeled neurons was found in the ipsilateral LOC system. Somata and long stretches of dendrites were labeled throughout the LOC. Examples of dextran labeled neurons in the areas of the LSO, DLPO, DPO and LNTB are shown at higher magnification (Fig. 2). The total number of dextran labeled neurons was counted bilaterally

in all regions of the LOC system (Fig. 2). The majority of the neurons were found in the ipsilateral LSO proper (282⫾35), and significantly fewer in the periolivary regions (DLPO⫽92⫾10; DPO⫽43⫾9; LNTB⫽31⫾7). The average number of labeled neurons on the ipsilateral LOC system was 448⫾31, while the number on the contralateral LOC system was 44⫾11. Distribution and morphology of TH positive neurons in the LOC TH immunoreactivity showed a strong positive reaction and was localized to the cell body and their processes. The most common morphology of TH positive neurons within the LSO was bipolar cells measuring 10 –25 ␮m in length (Fig. 3a). These cells constituted more than 74% of the total population of TH positive neurons in the LSO. Multiplanar neurons were polygonal or round and composed roughly 8% of the TH neurons in the LSO. Their dendrites were distributed over the surface of cell bodies and not restricted to any single plane of section (Fig. 3b). Another LSO type of TH positive neuron was small sized and unipolar (12%) with an oval or round cell body, often with a single dendrite (Fig. 3c). The less frequently observed marginal TH positive neurons (⬍6%) were distinguished along the contours of the LSO. These neurons were seen as bipolar in transverse sections (Fig. 3d). In addition, it was noted that TH neurons above described belonged to the classification of principal neurons, small neurons and marginal neurons. Furthermore, a moderately dense network of TH positive fibers was observed in this nucleus. Most of fibers were thin, beaded and randomly orientated. In the region of the DLPO, a variety of shapes and sizes of TH positive neurons were found. In general, these neurons

X. Niu et al. / Neuroscience 125 (2004) 725–733

729

Fig. 3. Representative micrographs illustrating the different morphological classes of TH positive neurons in LSO and its periolivary regions. LSO, a– d; DLPO, e– h; DPO, i;. and LNTB, j.

have larger soma than the neurons in the LSO. The intensity of the staining differed in these neurons and was considered as being either moderate or strongly labeled. Morphologically, the most common type of TH positive neuron in the DLPO was the multipolar cell, which constituted 56% of the total population of neurons analyzed (Fig. 3e, f), followed by the bipolar oblong type (24%; Fig. 3g). Small neurons (20%) were bipolar (Fig. 3h). Dendrites in the DLPO were strongly stained. Most of fibers ran randomly and some of fibers formed small accumulations parallel around the LSO. In the region of the DPO, TH positive neurons were middle size and multipolar (Fig. 3i). In the region of the LNTB, TH positive neurons were middle size. They were typically round or oval and displayed bipolar or tripolar morphology (Fig. 3j). A series of projection drawings obtained from eight levels showing the location and number of TH immunopo-

sitive neurons (Fig. 4). The abbreviations for the different regions are indicated on the top left drawing. These drawings illustrate that in the LSO there are between two and five neurons in each section and a similar number in the DLPO. Both the LSO and the DLPO contained the more TH immunopositive neurons compared with the DPO and the LNTB. Double labeling of TH and dextran The direct identification of LOC neurons containing TH was determined after double-staining the sections (dextran and TH). The double-stained neurons were found in the LSO proper as well as in the periolivary regions. Examples of these double-stained neurons from the different regions of the LOC are shown (Fig. 5). The quantification of dextranand TH-labeled neurons in the LOC is shown (Fig. 5). On

Fig. 4. Camera lucida drawings showing the distribution of TH immunopositive neurons at eight levels throughout the LSO regions (LSO, DLPO, DPO, LNTB).

730

X. Niu et al. / Neuroscience 125 (2004) 725–733

Fig. 5. Top: Examples of double-labeled neurons showing TH (red), dextran-fluorescein (green) and double-labeling (yellow). Bottom: The quantification of dextran (white bars), TH (gray bars), and double-stained (black) neurons in the ipsilateral LOC system. Values are mean and standard deviations.

the ipsilateral side to the dextran injection, the total number of double-stained neurons in the LOC (LSO⫹DLPO⫹DPO⫹LNTB) with dextran and TH was 155⫾38. This corresponded to 35% of the total dextranlabeled neurons (448⫾31) and 77% TH-labeled neurons (202⫾21). Within the LSO, double-labeled neurons (75⫾14) were composed of 27% dextran-labeled neurons (282⫾35) and 85% TH-labeled neurons (88⫾22). In the DLPO, double-labeled neurons (53⫾11) were composed of 58% dextran-labeled neurons (92⫾10) and 65% THlabeled neurons (82⫾3). In the DPO, double-labeled neurons (13⫾7) were composed of 42% dextran-labeled neurons (31⫾7) and 81% TH-labeled neurons (16⫾3). In the LNTB, double-labeled neurons (14⫾7) were composed of 33% dextran-labeled neurons (43⫾9) and 88% TH-labeled neurons (16⫾1). The majority of double-stained neurons were found in the LSO and the DLPO, while fewer were found in DPO and LNTB. The contralateral LOC did not have any double-stained neurons. Within the LSO, dextran-labeled neurons corresponded to 8% (282⫾35) and TH positive neurons to 2.5% (88⫾22) in relation to the total number of neurons (3572⫾414). The effect of acoustic stimulation on TH immunoreactivity The effect of acoustic stimulation on the immunoreactivity of TH was region specific and dependent on the intensity of stimulation. Representative micrographs illustrating TH immunopositive neurons from the DLPO region from the four

different experimental groups are shown (Fig. 6 top). An increased number of positively stained neurons are found after sound conditioning, or by the combined treatment of sound conditioning and acoustic trauma compared with the acoustic trauma group. Quantification of the number of neurons immunopositive for TH is shown (Fig. 6, bottom). The control animals showed 83⫾3 TH positive neurons in the LSO. After acoustic trauma, the number of TH neurons was decreased to 30⫾5 and was statistically significant compared with the control group. When animals were pretreated with sound conditioning and acoustic trauma the number of TH neurons was not affected. In the DLPO the total number of TH neurons in control animals was 80, while acoustic trauma reduced the number of TH neuron to 38⫾5 (P⬍0.001). No significant changes were found after sound conditioning or the combined treatment of sound conditioning and acoustic trauma. In the control group, TH positive neurons in LNTB and DPO were 16⫾5 and 17⫾5 respectively. After acoustic trauma, there was 12⫾4 and 14⫾5 in the LNTB and DPO, respectively and these numbers were not different from the control group. After sound conditioning the numbers of neurons in LNTB and in DPO were not altered compared with the control values. Fiber staining for TH in the LOC The densities of TH immunoreactive fibers in the regions of the LSO and the DLPO exhibited changes after acoustic trauma while the DPO and LNTB did not illustrate any

X. Niu et al. / Neuroscience 125 (2004) 725–733

731

Fig. 6. Representative micrographs illustrating TH immunopositive neurons in the DLPO from the four different experimental groups. The positively stained neurons were fewer in acoustic trauma group. Bottom: The quantification of TH-immunolabeled neurons in the ipsilateral LSO regions. Control (open bars), sound-conditioned (gray bars), sound-conditioned followed by acoustic trauma (hatched bars), and acoustic trauma (black bars). In the LSO and DLPO the number of TH-labeled neurons was reduced after acoustic trauma but was not changed in sound conditioning or after sound conditioning followed by acoustic trauma. Compared with the control group the acoustic trauma group showed a statistically significant decrease (*** P⬍0.001). Statistical results were obtained with one-way ANOVA followed by post hoc Tukey test.

changes. Representative micrographs of the fiber density for the DLPO region from the different groups of animals are shown (Fig. 7 top). The control animal shown has a moderate to low density of TH fiber staining. A slight decrease from the control staining was illustrated for the acoustic trauma group. The sound-conditioned animals and the animals that received the combined treatment of sound conditioning followed by acoustic trauma show significantly more dense TH immunostaining in the fibers of the DLPO compared with the control group or the acoustic trauma group. Quantification of TH immunoreactivity revealed a relative decrease in the density of staining after acoustic trauma compared with the control group (Fig. 7 bottom). This corresponds to approximately a 68% decrease in the LSO and 45% decrease in the DLPO from control (unexposed) values. After either sound conditioning or the combined treatment of sound conditioning and trauma, an increase in TH density was found relative to the control group. The trauma alone group showed a statistically significant decrease from the control group while the sound-conditioned group as well as the soundconditioned⫹trauma group showed statistically significant increases in immunoreactivity compared with the trauma alone group (ANOVA post hoc Tukey test, *** P⬍0.001, ** P⬍0.01). These findings are in agreement with the immunoreactive TH changes measured by immunohistochemistry and Western blotting methods in the cochlea as previously reported (Niu and Canlon, 2002).

DISCUSSION The present study demonstrates a specific immunostaining of TH neurons and describes their distribution and morphology in the LOC system. The fact that TH immuno-

reactivity shows region-specific plastic responses after sound stimulation with varying intensities indicates a key role for this enzyme in modulating auditory sensitivity in the LOC system. In a previous study from this laboratory, it was shown that TH levels in the lateral efferent terminals under the inner hair cells were decreased after acoustic trauma (Niu and Canlon, 2002). In contrast, sound conditioning or the combination of sound conditioning and acoustic trauma, resulted in a significant up-regulation of TH at the level of the lateral efferents under the inner hair cells (Niu and Canlon, 2002). These findings taken together with the results of the present study shed additional light on the importance of the LOC system in modulating the peripheral hearing organ against acoustic damage. TH is the initial and rate-limiting enzyme in the biosynthesis of dopamine. Dopamine is known to influence a variety of physiological and behavioral processes, and alterations in its function have been related to numerous disorders. Dopamine, released from the lateral efferents, is thought to exert a tonic inhibition of auditory nerve activity whereby removal of this tonic inhibition results in excitotoxicity (Gil-Loyzaga, 1995; Ga´borja´n et al., 1999; Sun and Salvi, 2001). Thus, the role of the lateral efferents is to protect the afferent dendrites from excitotoxicity and thereby preserve auditory sensitivity after glutamate overstimulation. Direct evidence for this comes from physiological studies where the application of either dopamine or agonists to dopamine results in a decrease in eighth nerve activity (Eybalin et al., 1993; Puel, 1995; Gil-Loyzaga et al., 1997; Oestreicher et al., 1997; Ga´borja´n et al., 1999; Ruel et al., 2001). Further evidence for a modulatory role of the lateral efferents comes from direct injections of a toxin into the LSO (Prell et al., 2003). When the LSO was chemically lesioned, the amplitude of compound action potential of the

732

X. Niu et al. / Neuroscience 125 (2004) 725–733

Fig. 7. Top: Representative micrographs showing TH immunopositive fibers in the DLPO. Control (upper left); acoustic trauma resulted in a reduction in TH labeling (upper right); sound conditioning (SC; lower left) or the combined treatment with SC and acoustic trauma (lower right) caused an intense TH labeling compared with the trauma alone or the control. Bottom: Relative density (% control) of fibers positive for TH in the region of LSO (left) and DLPO (right). The control group (open bars), acoustic trauma (black bars), SC (gray bars), and the combined treatment of SC and trauma (hatched bars). Acoustic trauma significantly decreased the densities of TH, while SC or the combined treatment of SC and trauma (SC⫹Trauma) increased the densities of TH. (*** P⬍0.001; ** P⬍0.01). Statistical results were obtained with one-way ANOVA followed by Tukey test.

eighth nerve was depressed, and distortion product otoacoustic emissions, a measure of outer hair cell activity, remained normal. A variety of stimuli are known to increase TH activity via phosphorylation of pre-existing enzymes (Onali and Olianas, 1987; Zigmond, 1988; Zigmond et al., 1989). Activation of the lateral efferent system by 24-h sound conditioning stimulus may lead to a facilitation of gene expression in the cell nucleus and trigger an increased synthesis of TH. In this way, the lateral efferents

would have the capacity to release more dopamine and thereby protect the afferent dendrites from the excitotoxic actions of glutamate overstimulation from the inner hair cells during acoustic overstimulation. It is well known that glutamate excitotoxicity, leading to the swelling and degeneration of afferent dendrites, is one mechanism underlying hearing loss induced by acoustic trauma (Spoendlin, 1971; Goulios and Robertson, 1983; Puel et al., 1998). TH positive neurons were found in all regions of the LOC (LSO, DLPO, DPO and LNTB). The number of immunopositive staining was more prominent within the LSO and the DLPO compared with the DPO and the LNTB. There were a total of 155 positive labeled TH neurons in the ipsilateral LOC corresponding to only 35% of the total number of neurons in the LOC, including the periolivary regions. This is a relatively low percentage considering the dense network of TH fibers that are present at the base of the inner hair cell in the cochlea (Niu and Canlon, 2002). One possible explanation for the difference between the number of TH positive neurons in the LOC and the dense network of lateral efferent fibers under the inner hair cells may be due to the dual lateral efferent system that has been describe in the guinea-pig and the rat (Brown, 1987; Warr et al., 1997). These two different systems have their origins either in the intrinsic or extrinsic regions of the LSO, and differ primarily in their degree of bifurcation and number of terminal swellings. The fibers from the intrinsic region have both numerous bifurcations and many terminal swellings, a pattern that is reminiscent of the TH staining in the cochlea (Niu and Canlon, 2002). Physiological differences between these different terminals beneath the inner hair cells may be expected since the different fiber groups have distinct immunoreactive profiles. The neurons in the LSO proper (intrinsic neurons) are immunoreactive to calcitonin gene-related peptide and choline acetyltransferase, while the neurons in the extrinsic region appear to contain only choline acetyltransferase (Vetter et al., 1991). However, at the level of the lateral efferents beneath the inner hair cells additional neurotransmitters and peptides have been demonstrated including dopamine, and GABA, serotonin, acetylcholine, calcitonin gene-related peptide and opioid peptide (Eybalin et al., 1993; Puel, 1995; GilLoyzaga et al., 1997; Safieddine et al., 1997; Ga´borja´n et al., 1999; Maison et al., 2003). Safieddine et al. (1997) demonstrated a high degree of colocalization in neurons in the LSO from both rats and guinea-pigs for choline acetyltransferase, glutamate decarboxylase, and TH. It remains to be determined how neurotransmitter release is controlled during sound stimulation in these terminals with co-existing transmitters. If sound stimulation affects only TH, or if the co-localized transmitters are simultaneously release during sound conditioning needs to be established. A more complete understanding of the neurochemical balance and regulation between the LOC neurons and the lateral efferent terminals is needed to understand the physiological role of the LOC system. In summary, the results of the present study demonstrate TH immunoreactivity in the LSO and its periolivary regions. Quantitative analysis revealed a decrease in the

X. Niu et al. / Neuroscience 125 (2004) 725–733

number of immunoreactive neurons after acoustic trauma and protection against this decrease by sound conditioning. The results specifically identify the LSO and the DLPO as controlling the plastic response of TH immunoreactivity in the peripheral hearing organ in response to acoustic trauma. Acknowledgements—This study was supported by grants from the Swedish Council for Work Life Research (98-0300), Swedish Medical Research Council (09476), The Royal Institute for Deaf People, AMF trygghetsfo¨rsa¨kring, and Stiftelsen Tysta Skolan. The technical assistance of Agneta Viberg is gratefully acknowledged.

REFERENCES Aschoff A, Ostwald J (1988) Distribution of cochlear efferents and olivo-collicular neurons in the brainstem of rat and guinea pig. Exp Brain Res 71:241–251. Brown MC (1987) Morphology of labeled efferent fibers in the guinea pig cochlea. J Comp Neurol 260:605–618. Canlon B, Borg E, Flock Å (1988) Protection against noise trauma by pre-exposure to a low level acoustic stimulus. Hear Res 34:197–200. d’Aldin C, Eybalin M, Puel JL, Charachon G, Ladrech S, Renard N, Pujol R (1995a) Synaptic connections and putative functions of the dopaminergic innervation of the guinea pig cochlea. Eur Arch Otorhinolaryngol 252:270 –274. d’Aldin C, Puel JL, Ledquc R, Crambes O, Eybalin M, Pujol R (1995b) Effects of a dopaminergic agonist in the guinea pig cochlea. Hear Res 90:202–211. Eybalin M, Charachon G, Renard N (1993) Dopaminergic lateral efferent innervation of the guinea-pig cochlea: immunoelectron microscopy of catecholamine-synthesizing enzymes and effect of 6-hydroxydopamine. Neuroscience 54:133–142. Ga´borja´n A, Lendvai B, Vizi ES (1999) Neurochemical evidence of dopamine release by lateral olivocochlear efferents and its presynaptic modulation in guinea-pig cochlea. Neuroscience 90:131–138. Gil-Loyzaga P, Vicente-Torres MA, Fernandez-Mateos P, Arce A, Esquifino A (1994) Piribedil affects dopamine turnover in cochleas stimulated by white noise. Hear Res 79:178 –182. Gil-Loyzaga P, Bartolome MV, Vicente-Torres MA (1997) Serotonergic innervation of the organ of Corti of the cat cochlea. Neuroreport 8:3519 –3522. Gil-Loyzaga PE (1995) Neurotransmitters of the olivocochlear lateral efferent system: with an emphasis on dopamine. Acta Otolaryngol 115:222–226. Goulios H, Robertson D (1983) Noise-induced cochlear damage assessed using electrophysiological and morphological criteria: an examination of the equal energy principle. Hear Res 11:327–341. Gundersen HJ, Bagger P, Bendtsen TF, Evans SM, Korbo L, Marcussen N, Møller A, Nielsen K, Nyengaard JR, Pakkenberg B, Sørensen FB, Versterby A, West MJ (1988) The new stereological tools: disector, fractionator, nucleator and point sampled intercepts and their use in pathological research and diagnosis. Acta Pathol Microbiol Immunol Scand 96:857–881. Helfert RH, Schwartz IR (1986) Morphological evidence for the existence of multiple neuronal classes in the cat lateral superior olivary nucleus. J Comp Neurol 244:533–549. Helfert RH, Schwartz IR (1987) Morphological features of five neuronal classes in the gerbil lateral superior olive. Am J Anat 179:55–69. Kujawa SG, Liberman MC (1997) Conditioning-related protection from acoustic injury: effects of chronic deefferentation and sham surgery. J Neurophysiol 78:3095–3106.

733

Maison SF, Adams JC, Liberman MC (2003) Olivocochlear innervation in the mouse: immunocytochemical maps, crossed versus uncrossed contributions, and transmitter colocalization. J Comp Neurol 455:406 –416. Niu X, Canlon B (2002) Activation of tyrosine hydroxylase in the lateral efferent terminals by sound conditioning. Hear Res 174:124 –132. Oestreicher E, Arnold W, Ehrenberger K, Felix D (1997) Dopamine regulates the glutamatergic inner hair cell activity in guinea pigs. Hear Res 107:46 –52. Onali P, Olianas MC (1987) Stimulation of dopamine synthesis and activation of tyrosine hydroxylase by phorbol diesters in rat striatum. Life Sci 40:1219 –1228. Prell CG, Shore SE, Hughes LF, Bledsoe SC (2003) Disruption of lateral efferent pathways: functional changes in auditory evoked responses. J Assoc Res Otolaryngol 4:276 –290. Puel JL (1995) Chemical synaptic transmission in the cochlea. Prog Neurobiol 47:449 –476. Puel JL, Ruel J, Gervais d’Aldin C, Pujol R (1998) Excitotoxicity and repair of cochlea synapses after noise-trauma induced hearing loss. Neuroreport 22:2109 –2114. Ruel J, Nouvian R, Gervais d’Aldin C, Pujol R, Eybalin M, Puel JL (2001) Dopamine inhibition of auditory nerve activity in the adult mammalian cochlea. Eur J Neurosci 14:977–986. Ryan AF, Bennett TM, Woolf NK, Axelsson A (1994) Protection from noise-induced hearing loss by prior exposure to a nontraumatic stimulus: role of the middle ear muscles. Hear Res 72:23–28. Safieddine S, Prior AM, Eybalin M (1997) Choline acetyltransferase, glutamate decarboxylase, tyrosine hydroxylase, calcitonin generelated peptide and opioid peptides coexist in lateral efferent neurons of rat and guinea-pig. Eur J Neurosci 9:356 –367. Schofield BR, Cant NB (1991) Organization of the superior olivary complex in the guinea pig: I. Cytoarchitecture, cytochrome oxidase histochemistry, and dendritic morphology. J Comp Neurol 314:645–670. Spoendlin H (1971) Primary structural changes in the organ of Corti after acoustic overstimulation. Acta Otolaryngol 71:159 –165. Subramaniam M, Henderson D, Campo P, Spongr V (1992) The effect of ‘conditioning’ on hearing loss from a high frequency traumatic exposure. Hear Res 58:57–62. Subramaniam M, Henderson D, Spongr V (1993) Effect of lowfrequency ‘conditioning’ on hearing loss from high-frequency exposure. J Acoust Soc Am 93:952–956. Sun W, Salvi RJ (2001) Dopamine modulates sodium currents in cochlear spiral ganglion neurons. Neuroreport 12:803–807. Tsuchitani C (1977) Functional organization of lateral cell groups of cat superior olivary complex. J Neurophysiol 40:296 –318. Vetter DE, Adams JC, Mugnaini E (1991) Chemically distinct rat olivocochlear neurons. Synapse 7:21–43. Warr WB, Boche BJ, Neely ST (1997) Efferent innervation of the inner hair cell region: origins and termination of two lateral olivocochlear systems. Hear Res 108:89 –111. Warr WB, Guinan JJ Jr (1979) Efferent innervation of the organ of Corti: two separate systems. Brain Res 173:152–155. Warr WB, Beck Boche JE, Ye Y, Kim DO (2002) Organization of olivocochlear neurons in the cat studied with the retrograde tracer cholera toxin-B. J Assoc Res Otolaryngol 3:457–478. West MJ, Slomianka L, Gundersen HJ (1991) Unbiased stereological estimation of the total number of neurons in the subdivisions of the rat hippocampus using the optical fractionator. Anat Rec 231:482–497. Zigmond RE (1988) A comparison of the long-term and short-term regulations of tyrosine hydroxylase activity. J Physiol 83:267–271. Zigmond RE, Schwarzschild MA, Rittenhouse AR (1989) Acute regulation of tyrosine hydroxylase by nerve activity and by neurotransmitters via phosphorylation. Annu Rev Neurosci 12:415–461.

(Accepted 11 February 2004)