A Novel Organ of Corti Explant Model for the Study of Cochlear Implantation Trauma

Abstract

This study presents a novel in vitro model of electrode insertion trauma-induced hair cell (HC) damage and loss and its application for testing the efficacy of otoprotective drugs. In the cochlear implant (CI) procedure as a treatment for profound deafness, an electrode array is surgically inserted to provide electrical stimulation to the auditory nerve. Mechanical trauma from insertion of a CI electrode into the scala tympani can lead to inflammation and a high level of oxidative stress, which can initiate the apoptosis of auditory HCs and intracochlear fibrosis. HC apoptosis and intracochlear fibrosis are thought to be causes of poor CI functional outcomes. In order to gain insight into the molecular mechanisms that initiate HC apoptosis and scala tympani fibrosis following electrode insertion trauma (EIT), and the otoprotective effects of dexamethasone (DXM) observed in previous studies, an in vitro model of EIT was designed. Here we present and characterize a novel, reproducible in vitro model for the study of cellular and molecular events that occur following an EIT procedure. Cochleae from 3-day-old rats were subjected to a cochleostomy and were then divided into three groups: (1) control, (2) EIT, and (3) EIT + DXM (20 μg/mL). In Groups 2 and 3, a 0.28-mm diameter monofilament fishing line was introduced through the small cochleostomy located next to the round window area, allowing for an insertion of between 110° and 150°. HC counts, gene expression for pro-inflammatory cytokines (i.e., TNFα and IL-1β), pro-inflammatory inducible enzymes (i.e., iNOS and COX-2) and growth factors (i.e., TGFβ1, TGFβ3 and CTGF), oxidative stress (i.e., CellROX), and analyses of apoptosis pathways (i.e., caspase-3, apoptosis induced factor and Endonuclease G) were carried out on all explants at different time points. The results of this EIT in vitro model show the initiation of wound healing in which an inflammatory response is followed by a proliferative-fibrosis phase. Moreover, DXM treatment of EIT explants inhibited the inflammatory response and promoted a nonscarring wound healing process. The novel in vitro model described here will improve our understanding of mechanisms underlying CI insertion trauma and protective strategies such as DXM treatment. Anat Rec, 2012. © 2012 Wiley Periodicals, Inc.

Profound to total sensorineural hearing loss (SNHL) has been successfully treated with cochlear implants (CI) over the past 2 decades and this therapeutic approach has allowed many individuals to enjoy a level of function comparable with that of less hearing-impaired individuals. Insertion of a CI electrode array can be associated with an inflammatory response that can produce a high level of oxidative stress which can induce the programmed cell death of the auditory hair cells (HCs) and the growth of fibrotic tissue around the electrode array. This can result in both a decrease in a patient's residual hearing as well as have a negative impact on the functioning of the patient's implant. Thus, current research efforts are focused on understanding the biochemical mechanisms underlying insertion trauma, and the development of otoprotective strategies to prevent the sequelae of this trauma.

Literature review reveals the widespread use of intact adult animal models, that is, mice, rats, gerbils and guinea pigs, for auditory function studies (Lorito et al., 2011; Makishima et al., 2011; Nowotny et al., 2011; Sly et al., 2011). There are limitations on the experiments that can be done in the specimens obtained from the animals due to the small size and number of cells that populate the organ of Corti (OC) (Cunningham and Tan, 2011); morphology studies and staining for the presence and distribution of enzymes of interest are described. Obtaining more detailed quantitative and statistically significant biochemical data, for example, gene and protein expression, micro RNA analysis, to understand the molecular mechanisms involved in a particular trauma model or as a result of a drug treatment would likely require a prohibitive number of adult animals if using current in vivo models.

Here we propose a novel, feasible, and reproducible in vitro model to study the cell and molecular biological processes that occur in response to electrode insertion trauma (EIT). Morphology studies, gene expression analyses of pro-inflammatory mediators as well as growth factors involved in fibrosis and scar formation, and immunofluorescence studies for mediators of programmed cell death are presented here, demonstrating the utility of this model system. Previously, our group has reported the otoprotective effects of dexamethasone (DXM) in tumor necrosis factor alpha (TNFα) challenged OC explants (Dinh et al., 2008, 2011; Haake et al., 2009; Hoang et al., 2009). Here, we further analyze the efficacy of this glucocorticoid in our newly described EIT in vitro model.

MATERIAL AND METHODS

Organ of Corti Explants

Three day-old (P-3) Wistar strain rats (Charles River Laboratories, Wilmington, MA) were anesthetized with ice for 30 min. Housing conditions and experimental procedures used in this study were in accordance with the Guidelines for the Care and Use of Laboratory Animals of the National Institutes of Health (NIH Publications No. 80-23, revised 1996) and in accordance with the University of Miami, Internal Animal Care and Use Committee, protocol No. 11-086.

Three OC explant groups were used in the study: (1) cochleostomy-only (control), (2) cochleostomy with monofilament line insertion (EIT), and (3) cochleostomy with monofilament line insertion and dexamethasone (DXM, 20 μg/mL, i.e., a 50 μM concentration, Sigma Aldrich, St. Louis, MO) treatment (EIT + DXM). In Groups 2 and 3, a 0.28-mm diameter monofilament fishing line (Cajun Line; W.C. Bradley, OK) was introduced three times through a small 0.35 mm diameter cochleostomy previously created with sharpened No. 5 Dumont forceps next to the round window (see Fig. 1 and Supporting Information online video clip) in order to achieve a high angle and depth of insertion into the scala tympani, which varied between 110° and 150°. All cochleae were then incubated for 10 min in phosphate saline buffer (PBS) followed by excision of full length OC explants, and were cultured in serum-free media consisting of Dulbecco's modified Eagle's medium (DMEM, Invitrogen, Carlsbad, CA) supplemented with glucose (final concentration at 6 g/L), 1% of N-1 supplement (Sigma Aldrich, St. Louis, MO), and penicillin G (30 U/mL). The images were captured with the aid of a Wild-Heerbrugg M400 stereomicroscope coupled to a Wild photoautomat MPS 55 (Wild-Heerbrugg, Switzerland) and to a SPOT color camera (Diagnostic Instruments, MI). The images were processed with ImageJ 1.45h (http://imagej.nih.gov/ij) software.

Figure 1. Photographs that depict the electrode insertion trauma procedure for the 3-day-old (P-3) rat cochlea. (A) P-3 cochlea with intact cartilaginous capsule, the location of the round and oval windows are identified by dashed lines and the coiling of the internal cochlear coils are also identified with dashed lines. (B) The site of a cochleostomy that is ∼0.35 mm diameter is indicated by dashed lines and is located next to the round window which is also denoted by dashed lines. (C) A 0.28-mm diameter monofilament fishing line is shown as it is introduced into the scala tympani through the cochleostomy. (D,E) Lateral views of the cochlea showing the monofilament fishing line's trajectory into the scala tympani of the cochlea with the cartilaginous otic capsule intact. (F) Top view of a P-3 cochlea with a logarithmic spiral scale superposed as indicated by the dashed and solid lines. Based on this scale, the insertion of the monofilament fishing line is estimated to be between 110° and 150°. (G) A portion of the cartilaginous otic capsule has been removed to show the internal structure of the cochlea with the site of the round window indicated by a dashed line and the direction of insertion of the monofilament line into the scala tympani indicated by an arrow. The site of the scala vestibuli is also indicated by an arrow. (H) With a portion of the cartilaginous otic capsule removed (see dashed lines), the monofilament fishing line introduced into the scala tympani is seen through its trajectory into this scala and underneath the scala media. (I) An OC explant dissected from a P-3 cochlea, 10 min post-cochleostomy plus EIT (i.e., EIT or EIT + DXM) or 10 min post-cochleostomy without EIT (i.e., control). The area where the full length OC explants are divided into apex and middle + base segments is indicated by a dashed line. Bars in A and I equal to 1 mm in all photographs seen in A–I.

Total RNA Extraction and Real Time RT-PCR

Twelve OC explants for each real time RT-PCR experiment were cultured for 24 or 96 hr, under the following conditions: (1) control, (2) EIT, and (3) EIT + DXM (20 μg/mL). Three independent experiments were carried out (N = 4 explants/condition). Full length OC explants were divided into apex and middle + base specimens (Fig. 1) prior to extraction of RNA. Total RNA was extracted with TRIzol reagent (Invitrogen, Carlsbad, CA) following the manufacturer's protocol. RNA purity and concentration were determined by the absorbance at 260 and 280 nm using Nano Drop ND-1000 (Thermo Fisher Scientific, Waltham, MA). cDNA was synthesized using an iScript kit (Bio-Rad, Hercules, CA). Quantitative real-time PCR was performed in duplicate by using iQ SYBR Green Supermix (Bio-Rad) on an iCycler Real-Time CFX96 detection system (Bio-Rad, Hercules, CA). The mRNA level was normalized by using the housekeeping gene β-actin. The primers were designed based on the cDNA sequences obtained from Ensembl Genome Browser (http://www.ensembl.org) and NCBI nucleotide database (http://www.ncbi.nlm.nih.gov/nuccore). The primer sequences used in this study are summarized in Table 1. Real-time PCR was carried out to 40 cycles at 95°C, 61°C, and 72°C, for 50 sec each. Melting curves were also performed to ensure primer specificity and to evaluate for any contamination. Relative changes in mRNA levels of genes were assessed using the 2^(-ΔΔCT) method (Livak and Schmittgen, 2001) and normalized to the house-keeping gene β-actin and then to the expression levels obtained from the control OC explants.

Table 1. List of primers used for the gene expression study

Target mRNA	Forward (5′ to 3′)	Reverse (3′ to 5′)	Reference sequence
The primers were designed based on the cDNA sequences obtained from Ensembl Genome Browser (http://www.ensembl.org) and NCBI nucleotide database (http://www.ncbi.nlm.nih.gov/nuccore).
COX-2	CGTGTTGACGTCCAGATCACA	GATTTAAGTCCACTCCATGGCC	ENSRNOT00000003567
TNF-α	AACTCGAGTGACAAGCCCGTAG-3′	GTACCACCAGTTGGTTGTCTTTGA	ENSRNOT00000001110
Interleukin (IL)-1β	ACCCAAGCACCTTCTTTTCC-3′	AGACAGCACGAGGCATTTTT	ENSRNOT00000006308
inducible Nitric oxide synthase (iNOS or NOS2)	CTGGGACTGCACAGAATGTTCC-3′	TTTGCCTCTTTGAAGGAGCC	ENSRNOT00000016133
Transforming Growth Factor (TGF)-β1	CGGACTACTACGCCAAAGAA	TCAAAAGACAGCCACTCAGG	NM_021578.2
TGFβ3	AGCACAATGAACTGGCAGT	AAGTTGGACTCTCTCCGCAA	ENSRNOT00000013516
Connective tissue growth factor (CTGF)	TGTGCACTGCCAAAGATGGT	TTCCAGTCGGTAGGCAGCTA	ENSRNOT00000020528
β-actin	CGTTGACATCCGTAAAGACC	AGCCACCAATCCACACAGAG	NM-031144.2

Morphology Studies

Twelve OC explants (N = 3 explants/group) were cultured for 6 days and were either: (1) control, (2) EIT, or (3) EIT + DXM groups. Culture medium was completely exchanged every 3 days in vitro. After 6 days, specimens were harvested and washed three times with PBS, then fixed in 4% paraformaldehyde in 0.1 M PBS and kept at 4°C for 48 hr. The fixed explants were washed three times in PBS and subsequently incubated in 5% normal goat serum (Sigma Aldrich, MO), 1% Triton X-100 (Sigma Aldrich, MO) in PBS for 30 min at 25°C. OC specimens were then incubated with FITC-labeled phalloidin for 45 min at 25°C. After washing, these OC explants were transferred to a glass slide with mounting medium, cover slipped, and viewed under a florescence Zeiss Axiovert 200 microscope. Stereocilia bundles of HCs stained with phalloidin-FITC were recognized and used for HC counts. A HC was counted if it possessed an intact cuticular plate with an intact stereociliary bundle. Total HCs were counted for the apex and middle + basal 150 μm regions of the OC explants and expressed as percentage of HCs lost. N = 3 counts/group for the apex specimens and N = 6 counts/group for the middle + base specimens. The formula used to calculate the percentage of protection of DXM is [(x – y)/x] × 100. The % THC loss corresponding to control was subtracted from the EIT and EIT + DXM groups. “x” is the % THC loss in the EIT group and “y” is the % THC loss in the EIT + DXM group.

Total Reactive Oxygen Species Detection Studies

Nine OC explants (N = 3 explants/group) were cultured for 96 hr and were either: (1) control, (2) EIT, or (3) EIT + DXM group explants. Three independent experiments were carried out. CellROX Deep Red (5 μM, Molecular Probes, OR) was added to each culture well and incubated at 37°C for 30 min. The OC explant samples were then washed three times with PBS fixed in 4% paraformaldehyde in 0.1 M PBS for 20 min. These fixed explants were washed three times in PBS and subsequently incubated in 5% normal goat serum (Sigma Aldrich, MO), 1% Triton X-100 (Sigma Aldrich, MO) in PBS for 30 min at 25°C. Specimens were next incubated with FITC-labeled phalloidin for 45 min at 25°C. After another washing, these OC explants were incubated in 600 nM 4′,6-diamidino-2-phenylindole (DAPI) solution (Sigma Aldrich, MO) for 5 min at 25°C; following this, there were three additional washings with PBS and then these triple stained OC explants were transferred to a glass slide with mounting medium, cover slipped, and viewed under a confocal Zeiss Axiovert 700 microscope. ImageJ 1.45h (http://imagej.nih.gov/ij) software was used for processing and analyzing the images. Red signal intensity (CellROX) was measured and normalized with the blue signal intensity (DAPI), which corresponds to stained cell nuclei. The size of the region of interest (ROI) was the same for all images and the brightness was in the range of 0 to 255 arbitrary units.

Cleaved Caspase-3, Apoptosis Induced Factor and Endonuclease G Immunofluorescent Labeling Studies

Nine OC explants for each staining (N = 3 explants/condition) were cultured in the same three conditions described in the section above. In the case of apoptosis induced factor (AIF) and endonuclease G (Endo G) immunostaining, MitoTracker® Orange CM-H₂TMRos (500 nM, Molecular Probes, OR) was added to each well and incubated at 37°C for 30 min. The tissues were then fixed in 4% paraformaldehyde, in 0.1 M PBS for 48 hr at 4°C. The explants were washed three times in PBS and subsequently incubated in 5% normal goat serum, 1% Triton X-100 in PBS for 30 min at 25°C. The OC explants were then separated into the three staining groups and incubated with either (1) anti-cleaved caspase-3 (Asp¹⁷⁵) rabbit polyclonal antibody or (2) AIF rabbit polyclonal antibody (Cell Signaling Technology, MA), or (3) Endo G rabbit polyclonal antibody (Abcam, MA) overnight at 4°C. After washing with PBS, the tissues from each of these three staining groups were incubated with secondary antibody Alexa 633-labelled goat anti-rabbit IgG (Invitrogen, NY) for 90 min at 25°C. The tissues from all three staining groups were then washed in PBS three times and incubated with FITC-labeled phalloidin for 45 min at 25°C. After washing, the OC explants were incubated in 600 nM 4′,6-diamidino-2-phenylindole (DAPI, Sigma) solution for 5 min at 25°C following three additional washings with PBS and transferred to a glass slide with mounting medium, cover slipped, and viewed under a confocal Zeiss Axiovert 700 microscope. ImageJ 1.45h (http://imagej.nih.gov/ij) software was used for processing and analyzing the images. In the case of cleaved caspase-3, the red signal intensity was measured and normalized with the blue signal intensity (DAPI), which corresponds to the nuclei. In the case of nuclear translocation of Endo G, the red signal intensity was measured for each cell nucleus within a defined area (ROI). The size and location of the region of the ROI was the same for all images and the brightness was in the range of 0 to 255 arbitrary units.

Statistical Analysis

Two-way analysis of variance (ANOVA) test was used for gene expression analysis, morphology studies, and mean fluorescent intensity comparisons, followed by a Bonferroni post test for multiple comparisons. The data are expressed as mean ± S.E.M., control and EIT + DXM groups were compared versus EIT group and a P value of <0.05 was considered significant. All calculations were performed on a computer equipped with GraphPad Prism v 5.00c software for Mac OS X®.

RESULTS

Gene Expression Studies

Pro-inflammatory cytokines (TNFα, IL-1β).

At 24-hr post-EIT, there was an increase in mRNA levels for TNFα in both the apical (1.83 mfu) and the middle + basal turn segments (1.50 mfu) of OC explants compared with Control explant group levels where these explants were not exposed to the EIT trauma caused by the insertion of the monofilament line (apex: P < 0.01; base: ns). Treatment of EIT exposed explants with DXM significantly inhibited the expression of this pro-inflammatory cytokine in both the apex specimens (0.64 mfu, P < 0.001) and the middle + basal turn specimens (0.77 mfu, P < 0.05) (Fig. 2A), when compared with EIT group mRNA levels. Also, the mRNA levels of IL-1β were up-regulated after EIT in the apex (5.64 mfu) and in the middle + base turn specimens (3.02 mfu) compared to these same regions of the control OC explants (apex: P < 0.001; base: 0.01), while treatment of EIT exposed explants with DXM reduce these IL-1β mRNA levels to a control explant level in the apex (1.32 mfu, P < 0.001) and also in the middle + base turn regions of these DXM treated EIT explants (0.68 mfu, P < 0.001) (Fig. 2C). However, at a later time point, that is, 96 hr in vitro, there was a change in the pattern of the mRNA levels of the pro-inflammatory cytokines. At 96 h, the levels of TNFα mRNA for the EIT and EIT + DXM OC explants dramatically decreased in the apex (0.14 and 0.09 mfu, P < 0.001, respectively), while this decrease in TNFα levels was not significant in the middle + base turn segment of these OC specimens (0.78 and 0.67 mfu, respectively) (Fig. 2B). Similarly, the levels of IL-1β were less in EIT and EIT + DXM OC explants in the apex (0.07 and 0.06 mfu, P < 0.001, respectively), but in the case of this cytokine, the mRNA levels in EIT and EIT + DXM OC explants now were also decreased in the middle + basal turn segment (0.42 and 0.33 mfu, P < 0.001, respectively) (Fig. 2D).

Figure 2. Results from the gene expression studies performed in the P-3 OC explants at 24 and 96 hr post-EIT and control. The first row histograms (A,C,E,G) corresponds to gene expression results of the indicated genes (i.e., TNF-α, IL-1β, iNOS, and COX-2) obtained after at 24 hr of culture, the second row (B,D,F,H) corresponds to these same genes but after 96 hr in culture, and the third row correspond to growth factor genes (i.e., TGFβ1, CTGF, and TGFβ3) also studied after 96 hr in culture. (A,B) show TNF-α mRNA levels; (C,D) show IL-1β mRNA levels; (E,F) show iNOS mRNA levels; (G,H) show COX-2 mRNA levels; (I) shows TGFβ1 mRNA levels; (J) shows CTGF mRNA levels; and (K) shows TGFβ3 mRNA levels. Two-way ANOVA test followed by a Bonferroni post test was used for multiple comparisons. The data are expressed as mean values ± SEM, control and EIT + DXM groups were compared versus EIT group. N = 6/group. ***P < 0.001, **P < 0.01, *P < 0.05, not significant if P > 0.05.

Pro-inflammatory enzymes (iNOS, COX-2).

At 24 hr, the levels of mRNA for the inducible form of nitric oxide synthase (iNOS), an enzyme involved in the inflammatory process and that produces the cytotoxic compound peroxynitrite, increased in the apical segment (11.26 msu) and in the middle + basal turn segment of the EIT OC explant group (3.65 mfu) compared to the levels observed in the control OC explants (P < 0.001). DXM treatment of EIT explants reduced the iNOS mRNA levels in the apical segment (1.07 mfu, P < 0.001) and in the middle + basal segment of these OC explants (2.27 msu, P < 0.01) (Fig. 2E). Cyclooxygenase 2 (COX-2) is an enzyme that transforms arachidonic acid into prostaglandin, resulting in localized pain and inflammation. COX-2 was upregulated in the EIT explant group mainly in the apex (1.94 mfu) compared to control OC explant results (P < 0.001), and these elevated mRNA levels were reduced in EIT explants when treated by DXM (0.26 mfu, P < 0.001); however, no difference was observed in the middle + base segments for all three groups of OC explans (Fig. 2G). At 96 hr, the mRNA levels of iNOS in the apex segments remained high in the EIT OC explants (19.37 mfu) when compared with control group values for this same area (P < 0.001), while in the EIT + DXM treatment group the iNOS mRNA levels remained low (5.18 mfu, P < 0.001). No significant differences in mRNA levels for iNOS between the three groups of OC explants were observed in the middle + basal turn segments (Fig. 2F). In the case of COX-2, at 96 hr, the mRNA levels in the EIT group in the apex segment (0.88 mfu) fell to control group levels, but DXM treatment of the EIT explants still inhibited COX-2 and the mRNA levels remained low both in the apex (0.22 mfu, P < 0.001) and at this stage in time also in the middle + base segments (0.39 mfu, P < 0.001) (Fig. 2H).

Wound healing associated growth factors (TGFβ-1, TGFβ-3, CTGF).

Due to the wound healing tendency in live animals and the possible complications with tissue scarring associated with this process, three important growth factors (transforming growth factor beta (TGFβ) -1 and -3 and connective tissue growth factor (CTGF)) were studied at the later time period in vitro, that is 96 hr. mRNA levels of TGFβ1 were increased in the apex segments after the EIT trauma induced by the insertion of the monofilament line (8.71 mfu) compared to the values obtained for the control OC explants (not significant, ns), these levels, however, increased significantly when the EIT OC explants were treated with DXM (28.46 mfu, P < 0.01). Interestingly, no differences were observed in the middle + basal turn segments within the three explant groups (Fig. 2I). The same pattern of changes in TGFβ1 mRNA levels was observed in the case of CTGF mRNA (Fig. 2J), while EIT increased the expression of transcripts for this growth factor in the apex segments (2.72 mfu), DXM treatment of EIT explants induced a higher expression level of CTGF mRNA in apex specimens (7.70 mfu, P < 0.01). However, there were no changes in mRNA levels for this growth factor in the middle + basal turn specimens from either EIT or EIT + DXM OC explants. An increase of TGFβ3 mRNA levels after EIT was observed both in apex (2.00 mfu) and in base (3.42 mfu). DXM treatment of EIT explants did not alter these TGFβ3 mRNA levels in the apex segments; however, there was a significant increase in TGFβ3 mRNA in the middle + basal turn specimens (5.64 mfu, P < 0.05) compared to the levels of transcripts for TGFβ3 in the EIT group of explants (Fig. 2K).

Morphology Studies

There was an increase of total hair cell (THC) loss in the EIT OC explants when compared with control group HC counts (Fig. 3). This damage to HCs was not significant in the apex segment of the EIT explants (i.e., 1% THC loss) when compared with the THC counts from the control group explants; however, a 56% loss of THCs (P < 0.001) was observed in the middle + basal turn specimens. Treatment of EIT explants with DXM had a highly significant otoprotective effect reducing the EIT induced loss of THCs to 11.3% (P < 0.001), which is a reduction in the level of EIT induced THC loss by 94% in response to DXM treatment.

Figure 3. Total hair cell (THC) counts corresponding to 150 μm of explant basilar membrane in apex and middle-base sections of OC explants fixed and stained with phalloidin-FITC after 6 days in vitro. Two-way ANOVA test followed by a Bonferroni post test was used for multiple comparisons. The data are expressed as mean values ± SEM, control and EIT + DXM groups were compared versus the EIT group. N = 3 counts/group for the apex specimens and N = 6 counts/group for the middle + base specimens. ***P < 0.001, not significant if P > 0.05.

Total Reactive Oxygen Species (ROS-CellROX Detection) Studies

The insertion of the monofilament line through the cochleostomy resulted in an increased production of the total reactive oxygen species (ROS) in both the HCs and the supporting cells (SC), mainly in the middle + basal turn specimens from the EIT explants (Fig. 4A). As is shown in Fig. 4B, the normalized mean fluorescence intensity (NMFI) values for the control group explants remained low in both HCs (0.008 NMFI, P < 0.001) and SCs (0.002 NMFI, P < 0.001), while in the EIT explant group these NMFI values (i.e., reflecting the level of ROS) are significantly increased for both the HCs (0.384 NMFI) and SCs (0.416 NMFI). Treatment of EIT OC explants with DXM significantly decreased the levels of ROS (Fig. 4A,B) in both the HCs (0.028 NMFI, P < 0.001) and SCs (0.048 NMFI, P < 0.001) located in the middle + base segments of the treated explants.

Figure 4. Confocal images of OC, after 4 days in vitro, middle + base area of explant labeled for total ROS with CellROX Deep Red dye and nuclei stained with DAPI (blue). (A) The first column shows the explant's stereocilia bundles stained with phalloidin FITC, the second column shows total ROS (CellROX Deep Red) at the HC nuclei level (nuclei stained with DAPI). Third column shows total ROS at the supporting cell (SC) nuclei level. Bars = 20 μm for all nine microphotographs. (B) Graph showing mean fluorescent (red) signal intensity levels (CellROX Deep Red) normalized with the blue signal intensity (DAPI), which corresponds to stained HC and SC nuclei. The data presented is only for the middle + base segments of the OC explants from control, EIT, and EIT + DXM groups of explants, since no difference with control group was observed. Two-way ANOVA test followed by a Bonferroni post test was used for multiple comparisons. The data are expressed as mean values ± SEM, control and EIT + DXM groups were compared versus EIT group. N = 9 areas measured/group. ***P < 0.001, not significant if P > 0.05.

Immunofluorescence Assay Results for Apoptosis Related Factors (Cleaved Caspase-3, Apoptosis Inducing Factor and Endonuclease G

At 96-hr post-EIT, the activation and location of three important apoptosis-related factors described in the literature were studied in the three different groups of OC explants.

While minimum activation of caspase-3 was observed in the control explants in HCs (0.010 NMFI, P < 0.001) and SCs (0.046 NMFI, ns), a significant increase of the activated form of this enzyme was observed only in the HCs of the EIT group (0.350 NMFI) with SCs in these explants not showing a significant increase in response to the trauma paradigm (0.099 NMFI in EIT). Treatment of EIT OC explants with DXM had a highly significant effect in reducing the levels of activated caspase-3 detected in the explant HCs (0.064 NMFI, P < 0.001) (Fig. 5A,B).

Figure 5. Confocal images of OC, after 4 days in vitro from the middle + base area of the explants labeled for the presence of activated caspase-3 and with DAPI. (A) First column shows the explant's stereocilia bundles stained with phalloidin FITC, the second column shows caspase-3 activation (anti-cleaved-caspase-3-Alexa 633) at the HC nuclei level (nuclei stained with DAPI). Third column shows caspase-3 activation at the SC nuclei level. Bars = 20 μm for all nine microphotographs. (B) Graph showing mean fluorescent (red) signal intensity (anti-cleaved-caspase-3-Alexa 633) normalized with the blue signal intensity (DAPI), which corresponds to stained cell nuclei. The data presented is only for the middle + base segments of the OC explants from control, EIT, and EIT + DXM groups of explants, since no difference with control group was observed. Two-way ANOVA test followed by a Bonferroni post test was used for multiple comparisons. The data are expressed as mean values ± SEM, control and EIT + DXM groups were compared versus EIT group. N = 9 areas measured/group. ***P < 0.001, not significant if P > 0.05.

AIF is the first mitochondrial protein shown to mediate cell death independent of caspase activation. As depicted in Fig. 6, AIF did not appear to play a role in the induction of either HC or SC apoptosis in the EIT explant group, since the immunolabeled AIF remained associated with the mitochondria (i.e., mitotracker orange stain), and immunostaining for AIF was not observed in either the peri-nuclear area or to be translocated into the nuclei (i.e., DAPI stain) of the HCs or the SCs.

Figure 6. Confocal images of OC, after 4 days in vitro, immunostained for apoptosis-inducing factor (red) and cell nuclei (DAPI, blue). First column shows the stereocilia bundles stained with phalloidin FITC, the second column shows AIF (anti-AIF-Alexa 633, red) either in the mitochondria (Mitotracker Orange) or in the nuclei (DAPI) at the level of the HC nuclei. Third column shows AIF either in the mitochondria (orange) or in the nuclei (blue) at the level of the SC nuclei. Bars = 20 μm for all nine photomicrographs. Because no translocation of AIF from the mitochondria to either the HC or the SC nuclei was observed in the middle + base segments of any of the three OC groups of explants quantification of immunostaining was not done.

Endo G is a recently discovered mitochondrial enzyme that is also known to be involved in the caspase-independent nuclear DNA degradation during apoptosis. Our results indicate that after the mechanical trauma to the cochlea, that is EIT, Endo G (i.e., immunolabeled with anti-Endo G-Alexa 633) is released from the inter-membrane space of the mitochondria (i.e., mitotracker orange) and translocated into the nuclei (DAPI stain) of affected cells (Fig. 7A). Interestingly, this phenomena of Endo G translocation was observed in both the HCs and the SCs of the EIT explants. There was also an apparent reduction in the size of the HC and SC nuclei in the EIT group compared with the DAPI stained nuclei in these cell types of both the control and EIT + DXM groups of explants (Fig. 7A). As seen in Fig. 7A,B, the level of nuclear translocation of Endo G is at a minimum in the control explant group (HCs: 19.77 MFI, P < 0.001; SCs: 5.88 MFI, P < 0.001), while this level has significantly increased in the nuclei of both the HCs and SCs in EIT explants (HCs: 53.57 MFI; SCs: 49.20 MFI). DXM treatment of EIT explants protected the SCs from Endo G release (6.51 MFI, P < 0.001), but did not have this effect on the HCs of the treated explants (42.81 MFI, ns).

Figure 7. Confocal images of OC middle + base segments of OC explants after 4 days in vitro, for Endo G nuclear translocation immunolabeling. (A) First column shows the stereocilia bundles stained with phalloidin FITC, the second column shows Endo G (anti-Endo G-Alexa-633, red) either in the mitochondria (Mitotracker Orange) or in the nuclei (DAPI) at the level of the HC nuclei. Third column shows Endo G either in the mitochondria (orange) or in the nuclei (blue) at the level of thr SC nuclei. Bars = 20 μm for all nine photomicrographs. (B) Graph showing mean fluorescent (red) signal intensity for each cell nucleus within a defined area (ROI). The size and location of the region of the ROI was the same for all images and the brightness was in the range of 0 to 255 arbitrary units. Two-way ANOVA test followed by a Bonferroni post test was used for multiple comparisons. The data are expressed as mean ± SEM, control and EIT + DXM groups were compared versus EIT group. N = 9 defined areas measured/group. ***P < 0.001, not significant if P > 0.05.

DISCUSSION

Inflammation is the first response of the body to an injury and is considered as a nonspecific immune response aimed at the neutralization of aggressor agents, and the repair of damaged tissues, in an attempt to assure the survival of the organism. The symptoms of inflammation are usually characterized by redness, warmth, swelling, pain, and loss of function (Gomes et al., 2008). The sequence of events in wound healing process begins immediately after injury, with the activation of the coagulation cascade and the initiation of the inflammatory phase. Injury usually causes disruption of blood vessels and leakage of blood constituents into the wound space. Activated platelets degranulate and release substances that attract and activate macrophages and fibroblasts into the wound site. Circulating monocytes enter the wound area and mature into tissue macrophages and dendritic cells. As described by Tsirogianni et al. (2006) in their review about the immunological aspects of the wound healing process, injury leads to the activation of nuclear factor (NF)-kB through the Toll-like receptors or cytokine receptors. NF-kB mediates gene transcription and the production of inflammatory mediators, such as chemokines, adhesion molecules, growth factors, and pro-inflammatory cytokines, especially TNFα and IL-1β.

Li et al. (2007) reviewed wound healing as a complex process that can be roughly divided into three overlapping phases of inflammatory reaction, proliferation, and remodeling. We observed in the in vitro model of EIT reported here that there are different patterns of pro-inflammatory mediators depending on the time that has elapsed post-EIT that corresponds to an early stage, when the inflammatory response picks up and a later stage, when the inflammatory response “deflates” and the wound healing process takes place and appears to predominate (Fig. 8).

Figure 8. Graphic schematic presentation of the trauma and wound healing process that takes place in the in vitro electrode induced trauma (EIT) model which is based on the three overlapping phases described by Li et al. (2007). After inducing injury in the cochlea there is an innate inflammatory response characterized by an increase of pro-inflammatory cytokines such as TNF-α and IL-1β and induction of the pro-inflammatory enzymes iNOS and COX-2. Under the high level of oxidative stress that is initiated by a mechanical trauma, many cells are conducted to a programmed cell death (apoptosis), which is mainly executed by activated caspase-3. In a later stage post-trauma, the inflammatory process is reduced and is overtaken by a proliferative response which drives the expression of growth factors and the process of wound healing. An inflammatory exacerbated response within the cochlea can result in the elimination of cochlear sensory cells causing a resultant loss of hearing. In the same side, if the wound healing process gets out of control, it can cause extensive fibrosis within the cochlea which can in turn impair the optimal functioning of the CI's electrode array.

Early (Inflammatory) Stage

Within 24 hr after inducing a mechanical trauma to the P-3 rat cochlea, we observed an increase of the pro-inflammatory cytokines TNFα and IL-1β both in the apex and middle + base turn segments of the OC in the EIT group compared to the control group explants. A well known anti-inflammatory glucocorticoid, dexamethasone (DXM), was used in this in vitro model of EIT because of its extensive use in otology clinics to treat idiopathic sudden SNHL (Gouveris et al., 2011). Recently Dinh et al. (2011) reported the protective effect of DXM against TNFα-induced apoptosis of rat auditory HCs in vitro through activation of the NF-kB signaling pathway which regulates the expression of inflammatory and apoptosis-related genes.

During the inflammatory process, a localized area with a high level of oxidative stress can occur as a result of the activation of macrophages and fibroblast to migrate into the area of damage. Generation of ROS has been shown to regulate signal transduction and also to play an essential role in growth factor and cytokine stimulation (Sundaresan et al., 1996). By means of CellROX fluorescence imaging, we have observed a highly significant increase in the level of total ROS in HCs and SCs of the EIT OC explants that were injured by the monofilament line insertion. In this model, DXM treatment of EIT explants caused a highly significant reduction in the levels of total ROS thereby effectively blocking the oxidative stress pathway.

Reactive nitrogen species are also produced during an inflammatory response to a trauma event. Nitric oxide (NO) is produced by iNOS and is involved primarily in the mediation of cellular immune response. Besides its pro-inflammatory effects, NO can exert toxicity by generating more destructive reactive nitrogen species, such as the peroxynitrite anion, by reacting with the superoxide anion (Gomes et al., 2008). In our model, we show an upregulation of iNOS in the EIT group of explants in the early stage, that is 24 hr, of the inflammatory process. Even though iNOS was up-regulated through all the OC, the levels were surprisingly higher in the apex segment of the explants. As expected, the treatment of the EIT explants with a glucocorticoid, that is, DXM, inhibited the expression of iNOS in all turns of the EIT OC explants.

Arachidonic acid is a polyunsaturated fatty acid which is a second-messenger molecule released by phospholipase A2 in stimulated cells. Cyclooxygenase (COX) and lipoxygenase further metabolize arachidonic acid into different eicosanoids ncluding prostaglandins and leukotrienes in diverse cells and into thromboxane A in platelets (Gomes et al., 2008). While COX-1 is a constitutive enzyme found in several tissues, such as the stomach, and is responsible for the basal and beneficial levels of prostaglandins, COX-2 is induced in the inflammatory process and it highly increases the levels of prostaglandins, resulting in pain and fever. We found that after inducing a mechanical trauma, there was an increase of COX-2 mRNA levels in the EIT OC explant apex specimens. DXM treatment of these EIT explants reduced the levels of COX-2 in the activated cells (apex segment), as expected since glucocorticoids have been demonstrated to inhibit both phospholipase A2 and COX (Goppelt-Struebe et al., 1989).

In summary, after inducing an insertion trauma injury in the cochlea by introducing a monofilament fishing line, which is the tool that we used to mimic an electrode array, there is an innate inflammatory response, characterized by an increase of pro-inflammatory cytokines such as TNFα and IL-1β and induction of the pro-inflammatory enzymes iNOS and COX-2. Interestingly, there was a difference in the enzyme expression between the apex and middle + base specimens. This may be a consequence of the difference in the stage of cell development in these two different areas of the P-3 OC explants, or possibly a direct relationship to the insertion length, and more studies will be needed to better understand the activity of these enzymes in this model.

Late (Wound Healing-Proliferation) Stage

In gene expression study at 96-hr post-EIT, we observed a change in the pattern of the pro-inflammatory cytokines and in COX-2 compared to the levels observed at 24-hr post-EIT. While the levels of iNOS in the apex remained high for the EIT group of OC explants, in the same apex specimens the levels of COX-2 mRNA decreased to control explant levels. Treatment of EIT explants with DXM maintained the mRNA levels low for these two enzymes in the apex, and at this 96-hr time point also reduced the levels of COX-2 in the segments from the base.

We also observed differences in the mRNA levels for the pro-inflammatory cytokines TNFα and IL-1β that were related to time post-EIT. At 96 hr, these cytokines were dramatically down-regulated in the apical turns of OC explants subjected to the mechanical trauma paradigm of this study. Also, in the middle + base turns the mRNA levels of IL-1β were significantly lower than control and in the case of TNFα the levels of transcripts for this cytokine were reduced to nearly the same as the levels observed in control explants. Treatment of EIT explants with DXM did not make a significant difference in the expression levels of the pro-inflammatory genes at this post-EIT time point. This result suggests an earlier effect of this glucocorticoid on the transcription of these inflammatory process related genes.

The TGF-β family is multifunctional and plays important roles in cell proliferation, differentiation, migration, and survival that affect multiple biological processes, including development, carcinogenesis, fibrosis, wound healing, and immune responses. TGF-β1 is the predominant isoform expressed in the immune system, but all three isoforms have similar properties in vitro (Li et al., 2006). In this study, the observed increases in TGF-β1 mRNA expression levels correlate with increases in CTGF mRNA in the apical turns of EIT group explants, that is, levels that were even significantly higher after DXM treatment of the EIT explants. No differences in TGF-β1 and CTGF levels were seen between the EIT and control explant groups in the middle + base specimens. TGF-β1 and its downstream effector CTGF are activators of fibrogenesis. Haydont et al. (2005) suggested that TGF-β1-induced CTGF transactivation mainly depends on the Smad signal pathway. However, in addition to its role in fibrosis, TGF-β modulates immune responses by promoting inflammation, as well as exerting an immunosuppressive activity (Blom et al., 2001). Bissonnette et al. (1997) reported that TGF-β1 inhibits the release of histamine and TNFα by mast cells, but not the release of NO from these cells. Moreover, the regulation of TGFβ1 over mast cells on the release of pro-inflammatory mediators was in an autocrine manner. On the other hand, TNFα has been reported to suppress CTGF gene expression in bovine aortic endothelial cells and also DXM-mediated induction of CTGF mRNA in fibroblasts (Abraham et al., 2000). Batuman et al. (1991) showed that DXM significantly increases the expression of the TGFβ1 gene and its protein product in mitogen-activated human T cells. On the other hand, TGFβ3 has been shown to play a key role in embryonic development and scarless wound healing (Tandon et al., 2010). In our experimental EIT model, there was an increase of TGFβ3 mainly in the middle + base specimens of the OC explants subjected to an EIT mechanical trauma, and again DXM treatment of EIT explants promoted the up-regulation of TGFβ3 over the values of the EIT only group of explants. Namazi et al. (2011) reviewing current strategies for prevention of scars described how wounds treated with TGFβ3 showed a marked reduction in the immunoreactivity for TGFβ1 and TGFβ2 post-wounding compared with the expression levels of these two growth factors in untreated, control wounds. This result suggests that TGFβ3 can down-regulate TGFβ1 and TGFβ2 expression. This could explain why in our model the levels of TGFβ3 were up-regulated in the middle + base turn specimens of the OC, while there was no difference in TGFβ1 and therefore in CTGF between groups for these middle + base specimens.

Regarding the immunosuppressive effects of TGFβ, Park et al. (2000) reported that TGF-β1 as well as DXM inhibited the cytokine (TNFα and IL-1β)- induced expression of VCAM-1 in capillary and post-capillary venular endothelial cells, resulting in a reduction of the recruitment of leukocytes to the locus of inflammation.

Vodovotz et al. (1996) reported that TGF-β1 is a central negative regulator of inducible NO production in macrophages, helping to maintain homeostasis. Based on TGFβ1–/– knock out mice, these authors concluded that TGF-β1 is the primary negative regulator of the expression of iNOS in vivo. This is consistent with our results, where an increase of TGF-β1 induced by DXM treatment of EIT explants correlates with an increase in the iNOS mRNA levels.

During the inflammatory and wound healing processes many cytokines, products from the activity of enzymes such as COX-2 and iNOS, and growth factors are released. These mediators can cause inflammation and oxidative stress in the structures of the cochlea, leading to the death of auditory HCs. The results of this study demonstrate that apoptosis of the auditory HCs in the EIT explants involved the activated caspase-3 pathway, and that activation of this enzyme was not present in the SCs of these EIT explants. DXM treatment of the EIT explants inhibited the apoptosis of HCs via caspase-3 activation. We also investigated other caspase-independent proteins involved in apoptosis, that is, AIF and Endo G. AIF does not seem to be involved in this process of EIT-initiated HC death because it was not translocated into the nucleus or localized to the peri-nuclear area of the HCs or the SCs in the EIT group explants. In contrast, Endo G was effectively translocated into the nuclei of the HCs in both the EIT and the EIT + DXM groups of explants where this enzyme can potentially cause degradation of the DNA. Interestingly, this phenomenon was observed in the HCs and SCs of the EIT explants as noted by the observable reduction in nuclear size, which is typical in cells undergoing apoptosis. However, when the EIT group was treated with DXM, Endo G was no longer translocated into the nucleus of the SCs; interestingly, DXM did not prevent the nuclear translocation of this enzyme in the HCs. Unlike HC nuclei in the EIT only explants, the HC nuclei in DXM treated EIT explants maintained the same size as the HC nuclei of the control group and were qualitatively healthy-appearing, which suggests that an auto-repair mechanism may be activated by the steroid, despite the inability of steroid to prevent Endo G nuclear translocation. Alternatively, in the presence of steroid, Endo G nuclear translocation may be insufficient to lead to apoptosis due other steroid-induced protein changes.

One possible limitation of the presented in vitro model presented here is that the effect of blood circulation on the wound healing process is not present; this may alter the levels of migrating monocytes/macrophages and of certain growth factors and cytokines. Another possible limitation is the denervation of central projections, but this should not have a large effect on short-term explants.

Currently, CIs are mostly offered to totally, bilaterally deaf patients, but many people with severe hearing losses do poorly with conventional hearing aids and are not currently CI candidates. Therefore, it is important to define the inflammatory process that can result from placement of a CI electrode array and to test drug treatment therapies to control inflammation-initiated HC death and fibrosis within the scala tympani because these events can limit the efficacy of CI as their use expands to patients with residual hearing (Bas et al., 2012). This model can provide the quantitative data necessary to assist in accomplishing this goal.

In summary, in a later stage, the inflammatory process is reduced and is overtaken by a proliferative response (Li et al., 2007). Our results suggest that mechanical trauma-initiated loss of auditory HCs in our explants occurs mainly via programmed cell death mediated primarily by activated caspase-3.

CONCLUSIONS

In this innovative in vitro model of EIT-induced cell death, we have described two different but overlapping phases that form part of the wound healing process. First, an inflammatory phase appears to last between 24 and 48 hr and is characterized by induction of pro-inflammatory enzymes and the release of cytokines, quantifiable in our model. This inflammatory response is followed by a proliferative response of wound repair. During the proliferative phase, there is growth of new tissue in an attempt to repair and remodel the wound space. This tissue formation response is known to involve angiogenesis, proliferation of fibroblasts, deposition of collagen, and other extracellular matrices. The proliferative phase process, in the context of cochlear implantation, can lead a fibrosis and scar tissue formation around the electrode array with the undesirable consequence of requiring an increase in the level of the stimulating current and therefore current spread. The in vitro EIT model described here can be useful for drug screening in order to achieve residual hearing preservation, for the prevention of excessive levels of fibrosis within the scala tympani. It is a feasible and reproducible in vitro model that can be used for drug identification and testing prior to the start of in vivo studies.