Acoustic Trauma Augments the Cochlear Immune Response to Antigen

Abstract

Objectives/Hypothesis

To test whether noise-exposure, which activates a cochlear immune response with cytokine expression and infiltration of circulating leukocytes could augment the response to antigen (Ag).

Study Design

Randomized, prospective, mice.

Methods

We sensitized mice to an Ag, injected it intrathecally, and subsequently exposed the mice to noise (8–16 kHz, 90, 100, or 118 dB for 2 hours). Control mice received either noise exposure alone (100 or 118 dB), Ag challenge alone, intrathecal surgery and phosphate-buffered saline injection or no treatment. Four hours or 7 days later the mice were killed and cochlear sections were evaluated immunohistochemically for CD45, ICAM-1, and phospho-nuclear transcription factor-κB expression.

Results

Intrathecal Ag injection caused no hearing loss, but did result in a small immune response. Loud noise (118 dB) caused severe hearing loss and slight inflammation. The number of CD45-positive cells was significantly greater in the Ag plus-118 dB noise group relative to the Ag-alone group or 118 dB noise-exposure group. ICAM expression was seen in the lower part of the spiral ligament and small vessels within the normal cochlea. The amount of expression increased after Ag injection and acoustic trauma. Activated nuclear transcription factor-κB occurred in the nuclei of hair cells, supporting cells, spiral ligament fibrocytes, and neurons 4 hours after noise exposure.

Conclusions

It seems that noise exposure can activate a cochlear immune response, which in the presence of Ag, allows for greater recruitment of inflammatory cells than occurred in response to Ag alone.

INTRODUCTION

One putative cause for adult onset hearing loss (HL) is exposure of cochlear tissues to inflammation or immune activity.^1,2 At this point this is the only mechanism of HL that is amenable to therapeutic intervention, so to optimize this therapy, it is necessary to understand cochlear immune mechanisms. We investigated under what circumstances innate immune signals are triggered and whether this activation contributes to the magnitude of the adaptive response to antigen (Ag) in the cochlea as shown previously for systemic exposure to the Gram-negative, bacterial cell-wall component, lipopolysaccharide (LPS).³

Innate immunity is defined as the initial, immediate response to pathogens. It can also be activated by stress signals including tissue damage. Cellular components of the innate response that are relevant to our study are tissue macrophages and monocytes; signaling molecules include nuclear transcription factor κB (NF-κB). Adaptive immunity develops over time in response to foreign Ags that are most common from pathogens, but nonpathogenic molecules will also initiate an adaptive immune response. This response is mediated by macrophages presenting the Ags and lymphocytes (T and B cells) responding to them. We used the CD45 cell-surface marker to identify infiltrated T and B cells, but it also identifies leukocytes generally associated with the innate response (monocytes, macrophages, and neutrophils). An adaptive immune response in the cochlea can be initiated by systemically sensitizing an animal to an Ag, such as keyhole limpet hemocyanin (KLH), then challenging the cochlea with this Ag. This protocol induces a circulating pool of Ag-specific T and B cells and a serum titer of anti-KLH antibodies. When Ag is introduced into the cochlea, an adaptive immune response is initiated against it.^4–6

Noise-exposure is a well known cause of sensorineural HL, clinically and experimentally. Acoustic trauma seems to activate several pathways within the cochlea generally associated with innate immunity. For example, noise exposure causes a significant increase in the nuclear transport of NF-κB⁷ known for its role in regulation of inflammatory responses and apoptosis in response to insults in many cell types.⁸ Cytokines that show increased expression after noise exposure include interleukin (IL)-1β and IL-6.^9–11 The common leukocyte marker, CD45, a transmembrane glycoprotein present on all bone marrow-derived white blood cells, is expressed on cells in the spiral ligament and in scala tympani after noise exposure.^9,11 This leukocyte recruitment presumably is mediated by the observed chemokine upregulation. ICAM-1 is a member of the immunoglobulin superfamily and plays a critical role in mediating adhesion of leukocytes to vascular endothelial cells in preparation for parapedesis in a variety of acute and chronic inflammatory diseases.^12,13 ICAM-1 is present on the vascular endothelial cells of the normal cochlea¹⁴ and is upregulated after noise exposure, so that much of the spiral ligament becomes immunohistochemically labeled.¹¹ All of the above cochlear responses to noise exposure are signs of innate immune activation.

Previously we showed that systemic LPS stimulation markedly amplified the adaptive immune response to Ag in the inner ear.³ Because noise exposure is a common cochlear stress and can activate an innate immune response, it is important to know whether it can also serve as a stimulus to amplify an existing, adaptive immune response. We therefore, evaluated whether activation of innate cochlear immunity by noise exposure could increase the adaptive inner ear immune response to Ag. We hypothesized that noise exposure activates the cochlea and recruits inflammatory cells in a fashion similar to an innate immune response. We evaluated the number and distribution of CD45 expressing cells, cells with activated NF-κB and ICAM-1 expression patterns in cochlear tissues after exposure to exogenous Ag and subsequent noise exposure. To challenge the cochlea atraumatically, we injected Ag-sensitized mice, intrathecally with Ag that reaches the cochlea by diffusion through the cochlear aqueduct without any cochlear surgery.³ The following day, when Ag was in the cochlea, the mice were exposed to noise. HL was measured and histologic sections of the cochlea were used to assess pathology and protein expression patterns.

MATERIALS AND METHODS

Animals

Female albino Swiss-Webster or ND4 mice weighing 20 to 30 g were used. Before surgery and killing, each mouse was deeply anesthetized with a mixture of 50 mg/kg ketamine hydrochloride, 5 mg/kg xylazine hydrochloride, and 1 mg/kg acepromazine maleate intraperitoneally. All the animal work was performed in accordance with the Animal Use Committee of the Veterans Affairs Medical Center.

Induction of an Inner Ear Adaptive Immune Response

An adaptive inner ear immune response was created by sensitizing the mice to the Ag, KLH (Pacific Biomarine Supply Co. Port Hueneme, CA), a large immunogenic molecule. The KLH was mixed with phosphate-buffered saline (PBS) (pH 6.4), dialyzed aseptically against PBS (pH 7) and centrifuged for 30 minutes at 1,500g. The supernatant was centrifuged again for 2 hours at 37,000g and the precipitate adjusted to a concentration of 50 mg/mL in PBS (pH 6.4). This solution was used for systemic sensitization and intrathecal injection. Animals were systemically sensitized to the Ag by a subcutaneous injection of Ag in the back (0.2 mg KLH emulsified in Freund’s complete adjuvant), followed 2 weeks later, by a second injection (0.2 mg KLH in Freund’s incomplete adjuvant). This systemic sensitization protocol activates an adaptive immune response creating high serum titers of anti-KLH antibody, cochlear perilymph titers of the antibody at about 1/10 the concentration found in the serum⁴ and circulating B and T-cells reactive to the Ag, KLH.

To Ag-challenge the cochlea without any cochlear surgery which can itself induce an innate immune response, Ag was injected (5 mg in 50 µL) into the subarachnoid space over the cerebellum. This was done using an operating microscope and sterile technique. A skin incision was made near the top of the occipital bone and the muscle scraped from the bone. A hole was drilled through the occipital bone in the posterior fossa without brain injury. A scalpel was used to make an incision through the dura and Ag from a glass micropipette was slowly injected into the cerebrospinal fluid. The hole was sealed with bone wax and connective tissue and the wound closed. The animal was allowed to recover from anesthesia in a heated chamber before being returned to the animal facility. Ag-challenge by this procedure was done 7 days after completion of the systemic sensitization. The Ag diffuses rapidly from the injection site into the cochlea where a weak, adaptive immune response is generated.³

Auditory Brainstem Response Recording

Auditory thresholds were measured using click-evoked auditory brainstem responses (ABR). The ABR thresholds of both ears were determined before surgery (1 day before noise exposure) and just before killing. The animals were deeply anesthetized and placed on a heating pad in a single-walled acoustic booth (Industrial Acoustics Co, Bronx, NY). An earphone with a probe that fit into the external auditory canal was used to deliver the clicks. Subcutaneous electrodes (Astro-Med, Grass Instrument Division, West Warwick, RI) were inserted at the vertex as active electrode, at the mastoid as reference electrode and at the hind leg as ground electrode. Click stimuli (0.1milliseconds; 10/seconds) were computer generated and delivered to the earphone. The recorded ABR was amplified, filtered with a battery-operated amplifier and input to a Tucker-Davis Technologies ABR recording system (Alachua, FL). The stimuli were presented in 5 dB amplitude steps that decreased successively and the recorded stimulus-locked activity was averaged 512 times at each sound level. Threshold was defined as the stimulus level between the record with a clearly identifiable response and one with no visibly detectable response. A one-way analysis of variance (ANOVA and a Newman-Keuls post hoc test) was used to compare hearing losses among the groups.

Noise Exposure

Mice were exposed to noise 1 day after Ag challenge to the subarachnoid space. This was done by placing the mice in a custom-made chamber.¹⁵ They were exposed to an octave-band noise (8–16 kHz) at one of three intensities (90, 100 or 118 dB) SPL for 2 hours (JBL speaker, model 2446 H). The timing of the noise exposure was intended to activate the cochlear innate immune response when Ag to which the animal was sensitive was already in the cochlea.³

Experimental Design

The cochleas from each animal were randomly assigned to one of five groups as follows:

1. Ag and noise exposure, experimental group: Ag was injected intrathecally into Ag-sensitive mice. Noise exposure was done the following day. Mice were killed 4 hours (100 dB n = 4; 118 dB n = 10) or 7 days (90 dB n = 10; 100 dB n = 8; 118 dB n = 20) after noise exposure.

2. Ag control group: Ag was injected intrathecally into Ag-sensitive mice. Animals were killed 8 days after injection (n = 15).

3. Noise exposure control group: Mice were exposed to noise and survived 4 hours (118 dB n = 10) or 7 days (100 dB n = 8; 118 dB n = 15) after the exposure. They were not sensitized or challenged with Ag.

4. Surgical control group: PBS was injected intrathecally with-out noise exposure. Animals were killed 8 days after injection (n = 10).

5. Normal group: No experimental treatments were given to these animals (n = 10).

Histological Preparation

All animals were killed by intracardiac perfusion of periodate-lysine-paraformaldehyde fixative. The final concentration of paraformaldehyde was 2%. The cochleas were dissected and left in fixative overnight (4°C). They were decalcified with buffered 5% ethylenediaminetetraacetate (pH 7.2) for 14 days at 4°C. After decalcification, the cochleas were embedded in O.C.T. compound (Sakura Finetek, Torrance, CA), snap-frozen in liquid nitrogen and sectioned (6 µm). In two mice, the spleen was harvested to serve as a positive control for immunohistochemistry. The tissue was frozen and sectioned in the same manner as the cochleas.

Immunohistochemistry

All of the slides were washed with PBS, incubated in 3% hydrogen peroxide for 20 minutes, rinsed in PBS, and incubated in 9% normal horse serum and 1% bovine serum albumin for 1 hour at room temperature. Then, primary antibodies, purified rat antimouse CD45 antibody (Leukocyte Common Ag, Ly-5) (BD Pharmingen, San Jose, CA, catalogue 550539) or hamster anti-mouse CD54 (ICAM-1) antibody (BD Pharmingen, San Jose, CA, catalogue 550287) were added to the sections at concentrations of 5 µg/mL and 0.16 µg/mL in PBS, respectively. Sections were incubated at 4°C overnight. Control sections were incubated with rat IgG2b in place of primary antibody (Serotec Ltd, Oxford, UK). A phospho-specific antibody, phospho-NF-κB p65 (Ser 276) (3037; Cell Signaling Technology, Beverly, MA) was used to identify the active form of NF-κB. In addition to the protocol described above for immunohistochemical assays, the NF-κB assay included subjecting sections to an Ag retrieval step by steamer treatment for 10 minutes in sodium citrate buffer. The concentration of primary antibody was 2 µg/mL.

After incubation, the sections were washed with PBS for 15 minutes. Biotinylated mouse antirat IgG2b antibody (BD Pharmingen), biotinylated goat antihamster IgG, (Jackson Immunore-search Laboratories, West Grove, PA) or biotinylated antirabbit IgG (Vector Laboratories, Burlingame, CA) were used for 30 minutes as secondary antibodies. Antibody binding was detected with an avidin-biotin horseradish peroxidase complex (Elite ABC Kit, Vector Laboratories, Burlingame, CA) for 30 minutes and 3, 3-diaminobenzidine-H₂O₂ as the chromogen. The sections were washed in water, dehydrated, and cover-slipped.

From each cochlea, five randomly chosen sections were stained with anti-CD45 antibody. For each section, the number of immuno-stained cells was counted in cochlear structures using Nomarski-Differential Interference Contrast optics and expressed as the number of cells/section. Cells were identified within the scala vestibuli, scala media, scala tympani, spiral ligament, stria vascularis, modiolus, and limbus. ANOVA with a post hoc test was used on the individual-section data to test for differences in the number of CD45-positive cells/section among the experimental and control groups. Intervening cochlear sections were stained with Hematoxylin and Eosin (H & E) to serve as a reference for pathology for each cochlea.

RESULTS

Hearing Loss

Cochlear damage was examined functionally by determining the HL in each mouse at the end of the experimental period (Fig. 1). The small amount of inflammation induced by intrathecal Ag did not result in any HL (4 ± 9 dB). Noise exposure alone however, did cause HL that was minimal at 100 dB (12 ± 5 dB) and near the maximum detectable level by our ABR system after 118 dB exposure (63 ± 19 dB) (Fig. 1a). The combination of Ag plus noise exposure resulted in a slightly larger mean HL in each group with different exposure levels (90 dB; 7 ± 9 dB, 100 dB; 16 ± 10 dB, 118 dB; 66 ± 19 dB), but these differences were not statistically significant given the sample sizes and that in many animals the loss was at the maximal amount that could be detected.

Fig. 1.

(a) HL for the experimental and control groups 7 days after treatment. The ABR threshold was measured for each animal before and after treatment and the difference in these values were calculated and used as a measure of HL. The mean (SD) value for each group is shown in the figure. Intrathecal injection of Ag did not cause a statistically significant HL. Exposure to 100 dB noise resulted in a slight loss of hearing, but this level of noise exposure in combination with Ag challenge did not lead to a larger HL than noise exposure alone. Exposure to 118 dB noise caused a large HL (P < .001, ANOVA) with thresholds at the maximum level of measurement with our system. It was not possible to detect whether the loss was worse because of prior Ag challenge. (b) Mean (SD) HL measured 4 hours after noise exposure with and without prior intrathecal Ag challenge. There was no measurable difference between the noise (118 dB) and Ag plus noise (118 dB) groups.

In the animals used for the histochemical, NF-κB activation assay, the hearing was measured 4 hours after noise exposure and just before killing (Fig. 1b). The threshold shift at this time point was maximal in the mice exposed to 118 dB noise (73 ± 8 dB) and was not different in the animals that also had been previously challenged with Ag (69 ± 13 dB). The mice exposed to 100 dB noise after intrathecal Ag challenge also had a mild HL (28 ± 15 dB) which was greater than that measured in the 7-day survival animals indicative that this level induced a temporary threshold shift.

Pathology

Sections stained with H & E were used to evaluate cochlear inflammation and pathology in response to noise exposure. The cochlea of Swiss-Webster or ND4 mice is similar to other mouse strains (Fig. 2a). As reported previously, 4 hours after noise exposure (118 dB) the stria vascularis was edematous and thickened, with intercellular spaces present.¹⁶ The organ of Corti in the upper basal turn (50–60% distance from the apex) was often collapsed (Fig. 2b). By 7 days after noise exposure, the edema was mostly resolved and the thickening of the stria vascularis, therefore, was reduced (Fig. 2c). The spiral ligament still contained some large vacuoles and there seemed to be fewer cell nuclei, however, many nuclei were still present. The organ of Corti, on the other hand, had experienced a large loss of outer hair cells (Fig. 3) as described previously. ¹⁵ In the Ag challenged, surgical control, and normal cochleas, however, the organ of Corti looked normal.

Fig. 2.

Photomicrographs of the upper basal cochlear turn from (a) a normal mouse cochlea, (b) a cochlea 4 hours after 118 dB noise exposure (arrowhead, edematous thickened stria vascularis), and (c) a cochlea 7 days after 118 dB noise exposure (arrow, atrophic stria vascularis). H & E stained cryostat sections (6 µm). Scale bar = 50 µm. (SL, spiral ligament; SV, scala vestibuli; ST, scala tympani).

Fig. 3.

Photomicrograph of the upper basal cochlear turn 8 days after intrathecal injection of Ag and 7 days after noise exposure (118 dB) in a systemically sensitized mouse. Leukocytes are present in the scala tympani (arrows). The loss of outer hair cells is a result of the noise exposure (arrowheads). H & E staining of cryostat section (6 µm). Scale bar = 50 µm. (SL, spiral ligament; ST, scala tympani).

Leukocyte Infiltration

Intrathecal Ag challenge results in a small amount of cochlear inflammation indicated by the presence of leukocytes (monocytes, lymphocytes, and neutrophils) mainly in the scala tympani. Noise exposure also induced the influx of a small number of leukocytes (monocytes) into the scala tympani and expression of CD45 among the fibrocytes of the spiral ligament. Ag challenge, followed by the loudest noise exposure (118 dB) induced more inflammation than either Ag challenge or noise exposure alone (Fig. 3 and Fig. 4). The lower levels of noise exposure caused less inflammation and the combination with Ag did not induce a striking difference. Few leukocytes were seen in the cochlear scalae of normal cochleas or those mice that received intrathecal PBS injection.

Fig. 4.

Photomicrographs of the cochlear upper basal turn illustrating CD45-positive cells in the experimental conditions. (a) After intrathecal Ag challenge alone there are labeled cells in the scala tympani (arrows) 8 days post-treatment. (b) After noise exposure (118 dB) alone there are labeled cells in the spiral ligament (arrow) 7 days post-treatment, and (c) After intrathecal Ag challenge plus noise exposure (118 dB) labeled cells are found in both locations (arrows) in greater numbers than in either alone 8 days after Ag-challenge and 7 days after noise exposure. (SL, spiral ligament; SM, scala media; ST, scala tympani) Scale bar = 50 µm.

The number and location of CD45-positive cells/section was used to test whether there was a difference in the amount of inflammation among the experimental groups. Figure 5a shows the mean number of CD45-positive cells per section for all experimental conditions after 8 days. The number of CD45-positive cells was significantly greater in the Ag-plus-118 dB noise group (72 ± 50 cells/section) relative to the Ag-alone group (34 ± 20 cells/section) or the 118 dB-noise-exposure group (28 ± 16 cells/section) (P < .01, ANOVA with a Newman-Keuls post hoc test). Exposure to 90 or 100 dB noise levels after Ag challenge did not induce an increase of CD45-positive cells/section relative to that from Ag challenge alone (90 dB-14 ± 16 cells/section; 100 dB-28 ± 13 cells/section). In the surgical control group, there were fewer than 5 cells/section in any cochlea. Normal cochleas also showed only a few labeled cells. The Ag-alone group had a significantly greater number of CD45-positive cells/section than the surgical control group and normal cochlea (P < .01, ANOVA with post hoc test).

Fig. 5.

The mean (SD) number of panleukocyte marker, CD45-positive cells/section in each of the experimental and control groups. (a) Seven or 8 days after treatment. The location of the cells within the cochlea in each condition is also indicated with shading. Horizontal tie bars indicate statistical significance in the number of labeled cells between the indicated groups (P < .01, ANOVA with Newman-Keuls post hoc test). The cochleas from the combined Ag plus noise group contained more leukocytes than any of the other conditions. (b) Four hours after noise exposure. Ag was injected intrathecally 24 hours before the noise exposure. There is no statistical difference in the number of labeled cells among the groups at this early time point.

Ag-challenge in the absence of other experimental manipulations resulted in the majority of CD45-positive cells in the cochlear scalae, modiolus, and limbus. Noise exposure resulted in a majority of the CD45-positive cells in the spiral ligament and stria vascularis (Fig. 5a). They were not restricted to the location of the maximal noise induced trauma, but could also be found more basal or apical to this region. As seen by the amount of cells in Figure 5b, the number of leukocytes/section increased between 4 hours and 7 days (Fig. 5a, b).

ICAM-1 and NF-κB immunohistochemistry was performed to identify cochlear mechanisms that underlie the recruitment of leukocytes. The intracellular adhesion molecule, ICAM-1 is expressed on vascular endothelial cells and fibrocytes of the spiral ligament adjacent to the scala tympani in normal cochleas. Seven days after noise exposure (118 dB), the amount of ICAM-1 expression was slightly increased relative to normal cochleas. In contrast, in Ag-challenged mice subsequently exposed to noise, ICAM-1 expression increased a great deal, especially in the middle to inferior portion of the spiral ligament and in endosteal cells lining the scala tympani (Fig. 6). Intrathecal PBS injection did not show more labeled cells than a normal cochlea (not shown).

Fig. 6.

Photomicrographs of the spiral ligament from the cochlear basal turn illustrating ICAM-1 immuno-labeled cells (arrows) in (a) an untreated, normal mouse cochlea, (b) a cochlea obtained 7 days after noise exposure (118 dB), and (c) a cochlea obtained 7 days after noise exposure (118 dB) that was also exposed to Ag from intrathecal injection 24 hours before the noise. ICAM-1 expression is low in the normal cochlea and is increased and sustained for at least a week by noise exposure and by the combination of Ag challenge and noise. (SV, scala vestibuli; SL, spiral ligament; SM, scala media) Scale bar = 50 µm.

The activation pattern of NF-κB in response to noise exposure (118 dB) or Ag-challenge plus noise (118 dB) exposure was determined using immunohistochemistry on cochlear sections from mice killed at either 4 hours or 7 days after noise exposure. Activated NF-κB was identified in the nuclei of hair cells, supporting cells, neurons, and spiral ligament fibrocytes 4 hours after noise exposure (Fig. 7a). At this early time point there was no obvious difference between the labeling pattern in the “noise-alone” cochleas and “Ag-plus-noise exposed” cochleas. By 7 days however, only the cochleas that received Ag plus noise (118 dB) had labeled NF-κB cells in these same cell types (Fig. 7b). Neurons were labeled in all cochleas from all groups, so it seems this may represent background label.

Fig. 7.

Photomicrographs of the cochlear upper basal turn immunohistochemically labeled for activated NF-κB (a) Four hours after Ag challenge plus noise exposure (118 dB) the spiral ligament fibrocytes (arrows), hair cells, supporting cells and neurons have darkly stained nuclei indicative of NF-κB activation. (b) Eight days after Ag challenge plus noise exposure (118 dB) there are fewer labeled cells and the label is fainter. The arrows point to labeled cells in the organ of Corti that are still labeled. Spiral ligament cells are not labeled at this time point. (SL, spiral ligament; SG, spiral ganglion) Scale bar = 50 µm.

DISCUSSION

Immune-mediated HL is an otologic condition for which there is some amount of clinical restorative capability with the use of immunosuppressive agents.^1,2 For this reason alone, it is important that we understand the specific cochlear mechanisms related to immunity. It is clear that both adaptive and innate immune responses occur in the cochlea^5,17 and that innate immunity can exacerbate the adaptive immune response to Ag.³ The results presented here support the idea that loud noise exposure can also activate innate immunity and increase the adaptive response to Ag.

We are beginning to understand the normal immune mechanisms of the cochlea and how the cochlea interacts with systemic immunity. Although there are conflicting findings depending on which species and what antibodies were used to identify macrophages, it seems that the normal, immunologically “resting” cochlea does contain macrophages within the spiral ligament and the perivascular connective tissue of the modiolus.^5,9,11,18 In fact, it is possible that these cells constantly move in and out of the cochlea at a slow rate.¹⁹ The number of cells expressing macrophage, cell-surface markers within the spiral ligament increases in response to systemic immune activation by the bacterial Ag, LPS. Systemic LPS injection also results in the infiltration of a few leukocytes into scala tympani.³ It seems then, that the cochlea has an innate immune response that is activated by systemic events. This activation may be mediated by activation of NF-κB in spiral ligament fibrocytes²⁰ and the expression of the proinflammatory cytokine, IL-1β in these cells.³ It is also likely that the cochlea has connections to cervical lymph nodes via lymphatic channels so that extracellular proteins in the cochlea are sampled by systemic immune cells.²¹

Both noise exposure and foreign Ag, to which the mouse is immunologically sensitive, induced infiltration of circulating leukocytes into the cochlear scalae. Noise exposure that causes cellular damage presumably activates a wound healing response that involves upregulation of “innate-immune-like activity” in spiral ligament fibrocytes seen as CD45 expression followed by infiltration of circulating leukocytes. The relationship between the amount of cellular damage and the amount of CD45 expression was not determined in this study. Foreign Ag, on the other hand, presumably causes no direct cellular damage, no activation of innate immunity (CD45 expression in the spiral ligament) and must be recognized either by phagocytic cells in the cochlea or in lymph nodes²¹ to generate an adaptive immune response involving leukocyte infiltration to the cochlea. To explore the initial activation events, early time points need to be examined for NF-κB activation and IL-1β expression.²²

Innate immune activation in several tissues seems to exacerbate the adaptive immune response to Ag.²³ This also seems true for the cochlea. The presence of Ag in the cochlea, accompanied by systemic injection of LPS resulted in a vigorous adaptive immune response than when Ag was presented alone.³ The current study demonstrated that the cochlear stressor, loud noise, has a similar effect. The level of the noise exposure was important. Levels that cause minimal or only temporary threshold shifts did not induce an innate immune response or exacerbate the adaptive response to Ag, whereas levels that result in permanent threshold shift did.

Some components of the immune response to noise exposure have already been described. The most obvious is an influx of inflammatory cells from the vasculature within 2 to 4 days after acoustic trauma after which time the number decreased over the next 12 days.^9,11 In H & E stained sections and those immuno-labeled with antiCD45 and F4/80 antibodies, leukocytes, especially macrophages, were identified in the spiral ligament and stria vascularis and in scalae vestibuli and tympani.^9,11 In this study, we found that Ag alone induces the influx of circulating CD45-positive cells into the scalae and the addition of noise in the presence of Ag greatly increased the number of CD45-positive cells in the scalae (Fig. 5).

The mechanisms involved in the activation of a cochlear innate immune response are not fully known. NF-κB is associated with inflammatory and immune responses in many tissues, activating transcription of numerous genes involved in host defense and is likely to play a role in the cochlea as well. The large number of NF-κB-positive cells identified in cochlear sections 4 hours after noise exposure is consistent with previous studies^7,24 and identifies it as a possible mediator of innate-immune activation. The observation that the activation of NF-κB persisted for 1 week in the presence of Ag and noise, whereas its activation was limited after noise exposure alone makes it a potential regulator of the increased immune response to Ag in the presence of noise.

Increased expression of the adhesion molecule, ICAM-1, is also a component of the cochlear innate immune response and is seen 1 day after noise exposure. Maximal expression was seen at 2 and 4 days and the staining pattern returned to normal by 14 days after the noise exposure.¹¹ In the current study, ICAM-1 expression was greater in the cochleas exposed to Ag, and subsequently noise, and lasted as long as 7 days after the exposure. It seems then, that ICAM-1 participates throughout the immune response.

CONCLUSION

We suggest that very loud noise exposure activates a cochlear innate immune response, recruits inflammatory cells and, in the presence of Ag, allows for amplification of the adaptive response to that specific Ag. This response involves activation of NF-κB, increased expression of ICAM-1 and infiltration of leukocytes (CD45-positive cells). More moderate levels of noise do not have such a clear effect on innate immunity or the ability to exacerbate the adaptive response to Ag. The response may be part of a wound healing response activated by injured cochlear cells, but this was not directly tested. These results imply that individuals with autoreactive anticochlear antibodies and lymphocytes might be at risk of a cochlear attack after loud noise exposure.