Supporting Cell Involvement in Cochlear Damage and Repair: Novel Insights from a Quantitative Analysis of Cyclodextrin-Induced Ototoxicity in Mice

Abstract

The cochlea is vulnerable to various pathological conditions, with sensory cells typically being the primary targets of damage. However, supporting cells also experience significant impacts. Despite their critical role in maintaining the structural and functional integrity of the sensory epithelium, the supporting cell involvement in cochlear damage remains poorly understood. This study aimed to elucidate the susceptibility of supporting cells in cochlear damage and their role in structural repair, using a mouse model of ototoxicity induced by cyclodextrin—a cyclic oligomer of glucose that is known to preferentially damage outer hair cells at high doses. A morphological examination of the cochlea showed that cyclodextrin exposure caused significant sensory cell loss, particularly affecting outer hair cells across the cochlear spiral, except at the apex. Despite extensive hair cell damage, most supporting cells in the apical and middle cochlear regions survived. In the basal end, where substantial supporting cell loss occurred, certain Deiters’ cells survived even after losing their phalangeal processes. Additionally, our observations indicate that Hensen’s cells contribute to forming an epithelial layer over the basilar membrane when the organ of Corti collapses. Further quantitative analysis revealed location-dependent susceptibility among supporting cell types. Deiters’ cells demonstrated greater resilience than pillar cells. Notably, the three rows of Deiters’ cells displayed differential susceptibility: the third row showed a more significant loss in regions with sporadic Deiters’ cell loss, while the first row exhibited an increased loss in areas adjacent to regions of complete Deiters’ cell depletion. The reduction of Hensen’s cells started in the middle section of the cochlea, occurring at a greater level than the reduction observed in Deiters’ and pillar cells. However, in the extreme base, where both pillar and Deiters’ cells were largely or completely absent, some Hensen’s cells were still present. Together, these findings provide new insights into the varying vulnerability of supporting cells to cochlear damage and underscore their essential role in structural repair.

1. Introduction

The cochlea is the sensory organ responsible for hearing. Central to its sensory functions is the organ of Corti, a specialized structure that contains sensory cells and supporting cells. Outer hair cells (OHCs) amplify vibrations of the basilar membrane, while inner hair cells (IHCs) convert these mechanical signals into neural impulses. These sensory cells are vulnerable to a range of pathological conditions, including acoustic trauma, ototoxicity, age-related degeneration, and genetic abnormalities. In mammals, the inability of sensory cells to regenerate results in irreversible hearing loss upon their demise. Consequently, substantial research has focused on the mechanisms behind sensory cell death and survival.

Yet, understanding the role of supporting cells within the cochlea is equally critical. Under normal conditions, these cells play crucial roles in maintaining the normal structural and functional integrity of the organ of Corti. Under pathological conditions, supporting cells react to sensory cell pathogenesis in several ways. First, they play essential roles in tissue repair when sensory cell death occurs. Specifically, the phalangeal processes of Deiters’ cells expand when OHCs are dying (Ding and Salvi, 2005; Raphael, Y. and Altschuler, R. A., 1991; Raphael, Yehoash and Altschuler, Richard A, 1991; Raphael and Altschuler, 1992; Roberto and Zrro, 1988). These morphological changes in Deiters’ cells help to rapidly close the void in the reticular lamina caused by OHC loss, preserving its barrier function. Supporting cells also play a role in removing dead hair cells, a key component of cochlear recovery after damage (Abrashkin et al., 2006; Anttonen et al., 2014; Bucks et al., 2017; Hamernik, Roger P et al., 1984; Hu et al., 2018; Lee et al., 2021; Monzack et al., 2015). Moreover, supporting cells express a range of immune molecules and participate in inflammatory responses after damage (Hashimoto et al., 2005; Hertzano et al., 2010; Hu et al., 2018; Ladrech et al., 2013; Parker et al., 2011). Since the organ of Corti lacks immune cells under physiological conditions, supporting cells perform immune surveillance to maintain cochlear homeostasis (Hu et al., 2018). Supporting cell survival has also been shown to enhance long-term neuronal survival, suggesting that these cells regulate neuronal survival (Sugawara et al., 2005). Recent studies have investigated the transdifferentiation of supporting cells into sensory cells in the mammalian inner ear (Cox et al., 2014; McGovern et al., 2019; McLean et al., 2017; Mizutari et al., 2013; Rai et al., 2021). These studies have suggested that supporting cells, unlike sensory hair cells, have the potential to differentiate, offering a possible avenue for sensory cell regeneration and hearing restoration. Given the critical role of supporting cells, it is essential to have a detailed understanding of their susceptibility in cochlear damage.

Previous studies have revealed a general resilience of supporting cells to cochlear insults in various pathological conditions (Kamakura et al., 2018; Kaur et al., 2023; Sugawara et al., 2005). Many of these studies employed qualitative or semiquantitative analysis of cochlear sections, which limits the ability to define detailed differences among the cells. Confocal microscopy offers a more advanced tool for providing high-resolution, three-dimensional images, allowing for a more precise and comprehensive analysis of supporting cells. This method improves the detection of spatial changes in the organ of Corti’s architecture during cochlear damage and repair.

The present study aimed to characterize changes in supporting cells following cochlear damage. We employed a mouse model of ototoxicity, as supporting cells tend to survive under these conditions, making it an ideal approach for examining their response patterns. To induce cochlear damage, we used cyclodextrin, a family of cyclic oligosaccharides. Cyclodextrins have been used to reduce drug volatility, mask bitterness, and improve the solubility and stability of poorly water-soluble drugs. They are also employed in controlled drug release kinetics. While cyclodextrin is generally safe at low exposure levels, it exhibits ototoxic effects when administered in higher doses (Cronin et al., 2015; Crumling et al., 2017; Crumling et al., 2012; Liu et al., 2020). Previous studies have established that systemic administration of cyclodextrin leads to rapid induction of cochlear damage, with a single dose resulting in OHC death starting at the base of the cochlea and progressing toward the apex (Cronin et al., 2015; Crumling et al., 2017; Crumling et al., 2012). Cyclodextrin ototoxicity also affects IHCs (Liu et al., 2020), but to a lesser extent. This damage pattern resembles that caused by other ototoxic substances, such as platinum-based antineoplastic drugs and aminoglycosides (Kros and Steyger, 2019; Rybak and Ramkumar, 2007). We selected cyclodextrin for our study because, compared to other ototoxic agents like platinum-based drugs and aminoglycosides, the systemic impact of cyclodextrin treatment is notably less severe (Cronin et al., 2015; Crumling et al., 2017). Animals treated with cyclodextrin typically maintain overall well-being, with no overt systemic adverse effects, despite significant loss of hair cells. This selective damage to the inner ear makes cyclodextrin ototoxicity an advantageous model for investigating the long-term response of supporting cells following cochlear damage.

In this study, we conducted a quantitative analysis of supporting cells in mice treated with cyclodextrin. We observed that the loss of supporting cells aligns with the pattern of hair cell damage, with a more pronounced loss in the basal region of the cochlea. However, the extent of supporting cell loss was significantly less than that of hair cells. Notably, damage levels varied across different supporting cell populations. Deiters’ cells were more resilient than pillar cells, with susceptibility differences observed across individual rows. Reduction of Hensen’s cells began in the middle section of the cochlea, with losses exceeding those of Deiters’ and pillar cells. However, in the extreme base, Hensen’s cells remained present despite the complete loss of both pillar and Deiters’ cells. Additionally, we found that Hensen’s cells play a crucial role in repairing the region where complete loss of the organ of Corti structure occurs by forming a thin layer—previously referred to as flat epithelium—resolving prior uncertainty about the origin of this structure. We also identified novel markers for supporting cells, which will be invaluable for future research into their behavior and function. This study offers valuable insights to support future efforts in developing therapeutic strategies that target supporting cells for hearing restoration.

2. Material and Methods

2.1. Subjects

This study used 26 young and healthy CBA/CaJ mice (22–28 grams, 1.5 to 4 months of age with an equal number of males and females, The Jackson Laboratory, Bar Harbor, ME, USA). All mice were housed at the University at Buffalo Laboratory Animal Facility, employing a light cycle of 12 hours on and 12 hours off (6 am to 6 pm). Three to four mice of the same sex were housed per cage. The temperature and humidity of the room housing the animals were maintained at 22°C and 40–60% humidity, respectively.

Animals were randomly assigned to experimental groups based on the timing of tissue collection: 6 mice were included in the acute damage group (tissues collected at 1–4 days post-treatment), 12 in the chronic damage group (tissues collected at 6–9 weeks post-treatment), and 8 in the control group. The number of mice or cochleae used for each assessment parameter will be detailed in the Results section. The Institutional Animal Care and Use Committee of the State University of New York at Buffalo approved all procedures involving the use and care of animals.

2.2. Induction of cyclodextrin ototoxicity

2-Hydroxypropyl-β-cyclodextrin (referred to as cyclodextrin throughout this manuscript) was purchased from Cayman Chemical Company (Cat #: 16169, Ann Arbor, MI, USA). The solution was freshly prepared using normal saline and administered as a single subcutaneous injection at a dose of 8 mg per gram of body weight, with a total volume of 0.3 ml, injected at the midline of the back. The control mice received the same volume of saline solution only. The overall health of mice was monitored daily during the first week and then twice a week until the time of sacrifice.

2.3. Auditory brainstem responses (ABR)

ABRs were measured to assess the auditory function in control mice (without cyclodextrin treatment) and treated mice (6 weeks after the treatment, n=6 animals per condition). The ABR measurements were performed in a soundproof booth with the room temperature at 23°C. Mice were anesthetized with an intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg). Stainless steel needle electrodes were inserted subdermally over the vertex (active), posterior to the stimulated ear (reference), and contralateral to the non-stimulated ear (ground). Impedance between electrodes was maintained at less than 3 kΩ throughout testing which lasted for about 30–40 minutes per mouse.

The acoustic signals were generated, and the responses were recorded using Tucker-Davis Technologies (TDT, Alachua, FL, USA) hardware and software. The sound level was calibrated using a sound level meter (824, Larson Davis, ½ inch microphone). The electrodes used for ABR recordings were connected to a preamplifier (RA16LA, TDT) using a flexible, low-noise cable. The outputs of the preamplifier were sent to a digital signal processing module (RX5-2, Pentusa Base Station, TDT) and processed by TDT software (BioSigRP, TDT). The ABRs were elicited with tone bursts at 8, 16, 24 and 32 kHz (0.5 ms rise/fall Blackman ramp, 1 ms duration, alternating phase) at the rate of 21/s, which were generated digitally (SigGen, Alachua, FL, USA). The signals were sent to a closed-field loudspeaker (CF1; TDT) via a D/A converter (RP2.1; TDT; 100 kHz sampling rate), a programmable attenuator (PA5; TDT), and an amplifier (SA1; TDT). The responses were filtered (100 to 3000 Hz), amplified, and averaged using TDT hardware and software. ABRs were obtained over a range of intensities at each tested frequency. The stimulus was initially presented at high intensities to elicit a response and then gradually decreased in 10-dB steps and finally in 5-dB steps near the threshold. The threshold was defined as the lowest intensity at which the ABR wave I was reliably detected. When ABR responses were not detectable at the maximum stimulus level that our system could generate, we recorded the maximum level as the threshold.

2.4. Distortion-product otoacoustic emissions (DPOAE)

DPOAEs, an assessment of outer hair cell function, were measured in the same mice used for ABR measurements. The mice were anesthetized with a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg). DPOAEs were measured with an Extended-Bandwidth Acoustic Probe System (ER10X, Etymotic Research, Elk Grove Village, IL, USA) that includes a probe assembly with two loudspeakers and a low-noise microphone. Custom software designed using MATLAB (MATLAB, version 6.1, Natick, MA, USA) was used to generate the f1 and f2 stimuli and analyze the acoustic signals in the ear canal. The system was calibrated with a half-inch microphone (Model 2540, Larson Davis), a microphone preamplifier (Model 2221, Larson Davis), and custom sound level calibration software (MATLAB, version 6.1, Natick, MA, USA). The probe assembly was carefully inserted into the ear canal of the mouse. The intensities of f1 and f2 were calibrated in the ear canal using the probe microphone. The f2/f1 ratio was set to 1.2, and the intensity of L2 was 10 dB lower than that of L1. The intensity of L1 ranged from 80 to 25 dB SPL, and the intensity of L2 ranged from 70 to 15 dB SPL in 5-dB steps. F1 and f2 were digitally generated (192 kHz sampling rate, 24-bit D/A converter). Each stimulus was 90 ms in duration, presented at the rate of 5 Hz, and repeated 32 times. During the stimulus presentation, the acoustic signals in the ear canal were measured with the low-noise microphone, the outputs were digitized with a sound card (RME Babyface Pro, 192 kHz sampling rate, 24-bit A/D converter), and the amplitudes of f1, f2, and 2f1-f2 were computed using a fast Fourier transform (FFT). DPOAE input/output functions were constructed at f2 frequencies of 8, 16, 24, and 32 kHz by plotting the amplitude of 2f1-f2 as a function of the L2 intensity at each f2 frequency.

2.5. Cochlear collection

Mice were sacrificed at different time points after the cyclodextrin treatment. For the acute damage group, the mice were sacrificed at 1–4 days after the treatment (n=6 mice). For the chronic damage group, mice were sacrificed 6–9 weeks after treatment (n=12 mice), and the data from these mice were combined to form one group. Lastly, for the control group (n=8 mice), the mice were sacrificed at the same time as those in the experimental groups. Specifically, 3 mice were sacrificed alongside the acute group (1–4 days post-treatment), and 5 mice were sacrificed alongside the chronic group (6–9 weeks post-treatment). The mice were sacrificed using CO₂ asphyxiation, followed by decapitation. The cochleae were quickly removed from the skull and fixed overnight with 10% buffered formalin at 4°C. The cochleae were dissected in 10 mM phosphate-buffered saline (PBS) for the subsequent analyses.

2.6. Assessment of OHC damage

Before preparing whole-mount cochlear sensory epithelia for detailed examination using confocal microscopy, we performed an in situ OHC quantification for the mice in the chronic damage group (n=6 mice) and the control group (n=8 mice). This in situ observation reduced the possibility of dissection-induced tissue damage. Specifically, the apical bony shell of the cochlea was removed using a fine-tip diamond drill to expose the apical section of the cochlea. The cochleae were stained with Alexa Fluor 488 labeled phalloidin (1:100, Cat# A12379, Invitrogen, Waltham, MA, USA) in 10 mM PBS for 30 min at room temperature. The stained tissue was photographed in situ using an epifluorescence illumination microscope (Z6 APO apochromatic zoom system) equipped with a digital camera (DFC3000 G microscope camera) controlled by Leica Application Suite V4 PC-based software (Leica Microsystems, Buffalo Grove, IL, USA). After imaging the hair cells in the apical region, the apical turn of the cochlea was removed to expose the middle and basal sections. The cochlea was photographed again to collect images of hair cells in the middle and basal regions. Hair cell images were collected section by section from the apex to the base of the cochlea. Using Adobe Photoshop CS6 (RRID: SCR_014199), the sections were stitched together to form a panoramic view of the sensory epithelium. The number of missing OHCs was quantified along the cochlear spiral at intervals of 150-µm, and the results were presented as a cochleogram.

2.7. Whole-mount sensory epithelium preparations

After hair cell quantification, the remaining bony shell of the cochlea was removed using a fine-tip diamond drill. The cochlea was then decalcified with 10% ethylenediaminetetraacetic acid (EDTA) at 4°C for 1 day. After rinsing with 10 mM PBS, the sensory epithelia were removed and stored in 10 mM PBS before staining.

2.8. Cochlear staining for visualization of sensory cells and supporting cells

Sensory cells and supporting cells were visualized with a combination of protein and nuclear markers (Table 1). For immunohistological assays, the cochlear tissues were permeabilized with 0.5% Triton X-100 and 5% donkey serum (Cat #: 017-000-121, Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA) or 1% Bovine Serum Albumin in PBS (pH 7.4) for 1 hour at room temperature. The tissues were then incubated overnight at 4°C with one primary antibody or two primary antibodies (for details regarding the double-labeling, see the Results section). The primary antibody dilutions were conducted based on the manufacturer’s recommendation with minor adjustments.

Table 1.

Molecular probes used to identify target molecules

Probe	Targets	Catalog number	Company
IFIT3 Antibody	Deiters’ cells and pillar cells	PA522230	Invitrogen
Galectin-3 Antibody	Hensen’s cells	AF1197	Novus Biologicals
MYO7A Antibody	Hair cells	PA1-936	ThermoFisher Scientific
DAPI	Nuclei	62248	ThermoFisher Scientific
Alexa Fluor™ 488 Phalloidin	Actin	A12379	ThermoFisher Scientific
Alexa Fluor™ 594 Donkey anti-Goat IgG	Galectin-3 Antibody	A11058	ThermoFisher Scientific
Alexa Fluor™ 647 Donkey anti-Rabbit IgG (H+L)	IFIT3 Antibody	A31573	ThermoFisher Scientific

After incubation with the primary antibodies, the tissues were rinsed 3 times with 10 mM PBS and incubated in the dark with one secondary antibody or two secondary antibodies (see Table 1) for 2 hours at room temperature. These secondary antibodies were diluted with a blocking buffer (5% donkey serum (Cat #: 017-000-121, Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA)) or 1% bovine serum albumin in PBS. Then, the tissues were rinsed 3 times with 10 mM PBS. The tissues were further stained with Alexa Fluor 488 labeled phalloidin (1:100, Cat# A12379, ThermoFisher Scientific, Waltham, MA, USA) and 4’,6-diamidine-2’phenylindole dihydrochloride (DAPI, 10 µg/ml) in 10 mM PBS for 30 min at room temperature. After the staining, the tissues were rinsed 3 times in PBS, mounted on glass slides with glycerol mounting medium, and covered with a coverslip.

2.9. Confocal microscopy

All confocal images were obtained using an Andor Dragonfly spinning disk confocal microscope (Oxford Instruments, UK) mounted on an inverted Leica microscope base. Image visualization and acquisition were performed using a 40× water immersion lens with a numerical aperture (NA) of 1.3. All images were captured using the same parameter settings, including laser intensity and spectrum, zoom size, and step size. Each image area covered a tissue dimension of approximately 300 × 300 µm², encompassing the thickness of the entire sensory epithelium from the reticular lamina to the basilar membrane.

The sensory epithelium spiral was segmented into three sections for imaging: the apical section (approximately 0–45% of the distance from the apex), the middle section (approximately 45–85% of the distance from the apex), and the basal section (approximately 85–100% of the distance from the apex), with 2 to 4 images collected for each section.

2.10. Image analysis and supporting cell quantification

The collected images were examined and analyzed using Imaris software (version 10.2, Oxford Instruments, UK). The software’s 3D visualization feature was used to adjust the viewing angle for better visualization. The tissues were digitally sectioned to examine the target structures located deep within the organ of Corti. The section thickness was adjusted to optimize the visibility of structural details.

We quantified the number of Hensen’s cells, Deiters’ cells, inner pillar cells, and outer pillar cells in both control and cyclodextrin-treated cochleae using confocal images collected from the apical, middle, and basal sections of the cochlea. Each image covered an area of approximately 300 × 300 µm² of the organ of Corti. The cell counts were then normalized to a 300 µm length of the organ of Corti for consistent comparison.

2.11. Scanning electronic microscopy (SEM)

SEM was employed to examine the surface structures of the organ of Corti in two control cochleae and three cyclodextrin treated cochleae (six weeks post-treatment). The cochleae were fixed with 2% glutaraldehyde in 10 mM PBS at room temperature for 1 hour followed by refrigeration at 4°C for at least overnight. The cochleae were then decalcified with 10% ethylenediaminetetraacetic acid (EDTA) at 4°C for 5 days and dissected to expose the sensory epithelium in the middle and basal section of the cochlea (45–100% of the distance from the apex). The tissues were dehydrated in a graded series of ethanol (30%, 50%, 70%, 85%, 95%, and 100%) for 15 minutes each and then with 100% hexamethyldisilazane for 2 hours. The tissues were then mounted onto tissue stages and coated twice, first with evaporated carbon and then with gold, using a high vacuum evaporator (Denton 502 Evaporator, Denton Vacuum, LLC, Moorestown, NJ, USA). The tissues were examined and photographed using a field emission SEM (Hitachi SU-70, Tokyo, Japan) at 2.0 keV using an in-lens secondary electron detector at zero tilt or a lower detector at 70° tilt. A total of 4 cochleae (one from the acute damage group and three from the chronic damage group) were examined. To improve image contrast and highlight specific cells, we utilized Adobe Photoshop CS6 to color the cells of interest in the images presented in this paper.

2.12. Data analyses

Our initial analysis did not identify significant sex differences in supporting cell pathogenesis following the cyclodextrin treatment. Therefore, we combined data from both sexes for further analysis. For supporting cell analysis, we calculated the percentage of cells loss. Specifically, for Deiters’ cells and pillar cells, we counted both the number of missing cells and the number of remaining cells per 300 µm segment. The percentage loss was then determined using the following formula:

$$\%Loss=\frac{Thenumberofmissingcells}{Thenumberofmissingandremainingcells}x100$$

For Hensen’s cells analysis, we were unable to count the exact number of missing cells due to the lack of a well-organized structure in these cells. Thus, we obtained the average number of Hensen’s cells in normal control cochleae and compared it with the number of remaining cells counted in each confocal image. The percentage loss was estimated using the following formula:

$$\%Loss=\frac{Average\#ofcellsinnormalcontrolcochleae-\#ofRemainingcells}{Average\#ofcellsinnormalcontrolcochleae}x100$$

This percentage calculation was performed for each confocal image, and the group mean was then computed for each cochlear section (apical, middle, and basal).

All data are presented as mean ± 1 standard deviation. Statistical analyses were performed using SigmaStat (version 10.0.1.25, San Jose, CA, USA). Group means were statistically compared using an unpaired t-test or a two-way ANOVA. An α-level of 0.05 was chosen to denote significance for all statistical tests. Normative data and equal variance tests were performed for all statistical analyses. If these two criteria were not satisfied, non-parametric tests were performed. Multiple comparisons were performed using either the Bonferroni t-test or the Mann-Whitney U test. Sample sizes were determined using G*Power (version 3.1.92) if similar data from previous studies were available. Selected sample sizes gave 80% power to detect biologically significant changes. If no data were available for power analysis, we estimated the sample size based on pilot observations. If estimated sample sizes exceeded our experimental capacity, we used a sample size that allowed us to determine a trend of changes or differences.

3. Results

The organ of Corti contains sensory cells and multiple populations of supporting cells, ranging from medial to lateral: inner sulcus cells, inner border cells, inner phalangeal cells, inner pillar cells, outer pillar cells, Deiters’ cells, Hensen’s cells, Claudius’ cells, Boettcher’s cells, and outer sulcus cells. We used multiple markers to identify these cells (Fig. 1). Because cyclodextrin ototoxicity primarily targets OHCs, the current investigation focused on the supporting cells located adjacent to OHCs, including Hensen’s cells, Deiters’ cells, and inner and outer pillar cells.

Figure 1.

The organ of Corti structure and acute cyclodextrin-induced damage. (A) Schematic of sensory cells and supporting cell populations in the organ of Corti. The schematic focuses on supporting cells adjacent to OHCs, including Deiters’ cells, Hensen’s cells, inner pillar cells, and outer pillar cells. Cell specific markers are listed. (B) Cyclodextrin-induced ototoxicity in the mouse cochlea examined at 1 day post-treatment. Phalloidin staining of actin shows a significant loss of OHCs. The single arrows mark cuticular plates without phalloidin staining, while the double-arrow indicates a cuticular plate with phalloidin staining. The images show a middle region of the cochlea, about 60% distance from the apex. (C) Prestin immunolabeling reveals fragmented OHCs, indicating severe OHC damage (arrows). (D) Myosin 7α immunostaining demonstrates that IHCs remain largely intact (arrows). (E) DAPI staining supports the preferential damage to OHCs compared to IHCs. The asterisks indicate the areas of missing OHCs and the arrows point to IHC nuclei. Bars = 20 µm.

3.1. Cyclodextrin treatment results in significant hair cell damage and auditory dysfunction

Our previous studies have documented the ototoxic effect of cyclodextrin in rats, which primarily affects OHCs (Ding et al., 2021; Liu et al., 2020; Zhang et al., 2022). To determine whether this damage pattern is also present in mice, we examined the organ of Corti at 1 to 4 days after a single dose of cyclodextrin treatment (n=6 mice). We observed significant OHC damage starting from 1-day post-treatment. Phalloidin staining revealed loss of actin in the cuticular plates in the majority of OHCs in the middle and basal regions of the cochlea (Fig. 1B). Prestin immunolabeling showed a large amount of OHC fragments (Fig. 1C), suggesting significant OHC damage. In contrast to drastic OHC degradation, IHCs, illustrated by myosin 7α and DAPI staining, remained largely intact (Fig. 1D & E), except for the basal end of the cochlear spiral (approximately 80–100% distance from the apex) where IHC lesions were evident. The finding of acute sensory cell damage is consistent with the damage pattern previously reported in rat studies (Ding et al., 2021; Liu et al., 2020).

In addition to sensory cell loss, we observed loss of supporting cells in the extreme base of the cochlea during the acute phase of cochlear damage, indicating that cyclodextrin treatment may also be toxic to supporting cells. However, we did not conduct a detailed analysis of the acute phase, as our investigation primarily focused on the chronic phase of cyclodextrin ototoxicity (see subsequent sections for more details).

To assess the long-term impact of the cyclodextrin treatment on cochlear function, we examined ABRs and DPOAEs at 6 weeks post-treatment in a subset of mice (n=6 mice). This timeline was chosen to ensure observation beyond the acute phase of cochlear pathogenesis. DPOAEs were undetectable across all tested frequencies (8, 16, 24, and 32 kHz, data not shown). ABR measurements revealed significant threshold elevation at all tested frequencies (8, 16, 24, and 32 kHz), with high-frequency responses (24 and 32 kHz) becoming undetectable at the maximum intensity that our ABR equipment could deliver (Fig. 2A).

Figure 2.

Auditory dysfunction and cochlear sensory cell loss post-cyclodextrin treatment. (A) ABRs were measured in both cyclodextrin-treated (six weeks after cyclodextrin treatment) and control mice. In the treated mice, ABR thresholds are significantly elevated across all tested frequencies, with higher frequency responses (24 kHz and 32 kHz) becoming undetectable. n=the number of mice. (B) Cochleogram shows complete OHC loss in the basal region and partial loss in the middle and apical regions. (C) Comparison of the damage levels across the three rows of OHCs. The shaded areas marked the cochlear regions where three rows of OHC display different levels of damage, with the first row showing significantly greater loss. OHC1, OHC2, and OHC3 represent the first, second, and third rows of OHCs, respectively. n=the number of cochleae. (D) Comparison of the level of IHC loss among different cochlear regions from the apex to the base. IHC loss is most prominent in the basal region, with surviving IHCs in the middle and apical regions. Each dot represents one viewing area, and the data were collected from 6 cochleae.

To quantify the level of sensory cell loss, we counted the number of missing OHCs along the cochlear spiral at 6 weeks post-treatment in a portion of mice (n=6 mice with one cochlea from each mouse). As shown in the cochleogram (Fig. 2B), there was a complete loss of OHCs in the basal portion of the cochlea (60–100% distance from the apex). In the middle portion, both surviving and missing OHCs were noted. The apical portion of the cochlea showed only mild OHC loss. Furthermore, we observed a significant difference in the level of damage across the three rows of OHCs in the cochlear regions where both missing and surviving cells were present (the shaded area in Fig. 2C). The first row showed more OHC loss than the third row (n=6 mice, one-way ANOVA, F (2, 15) = 3.799, p = 0.046; Tukey’s multiple comparison test, first row vs. third row, p = 0.041).

We also examined IHCs. Since the level of IHC damage was significantly less than that of OHCs, we did not quantify the number of missing cells along the entire cochlear spiral. Instead, we selected representative areas along the cochlear spiral for analysis. As shown in Figure 2D, the pattern of IHC loss mirrored that of OHCs, being most pronounced at the basal end of the cochlea, where complete IHC loss was noted. In the middle portion of the cochlea, many IHCs survived even with complete loss of OHCs. In the apical section, virtually all IHCs were present. We also examined hair cells in normal control cochleae (n=6) and observed only a few instances of sporadic hair cell loss (data not shown). Collectively, the finding of significant loss of hair cells confirms that cyclodextrin-induced ototoxicity is an effective tool for studying supporting cell changes in cochleae with extensive hair cell damage.

3.2. Deiters’ cells in cyclodextrin-treated cochleae

Deiters’ cells provide structural support to OHCs at their apical and basal ends. The upward extension of the Deiters’ cell’s phalangeal process contacts the apex of OHCs, contributing to the formation of the reticular lamina. At the base, the cylindrical body of the Deiters’ cell creates a supportive structure, known as the Deiters’ cell cup, for the OHC.

We first examined the pattern of Deiters’ cell loss at 6 to 9 weeks after cyclodextrin treatment. Deiters’ cells were counted by observing their phalloidin-labeled foot plates because these structures were easily identifiable in confocal images. As expected, Deiters’ cell loss was scarce in the apical portion of the cochlea but became more frequent at the basal end of the cochlea. In this region, we found different vulnerabilities for the three rows of Deiters’ cells, and the difference was related to the damage level. In the region where Deiters’ cell loss was either sporadic or in small clusters involving only a few cells, the incidence of Deiters’ cell loss was more frequent in the third row (Fig. 3A & B, repeated one-way ANOVA, F (2, 24) = 29.43, p < 0.0001). Tukey’s multiple comparisons showed significant differences between row 1 and row 3 (Adjusted p < 0.001), and between row 2 and row 3 (Adjusted p < 0.0001), while there was no significant difference between row 1 and row 2 (Adjusted p > 0.05). This analysis was conducted using 13 confocal images from 10 cochleae. Conversely, in the region near the area of complete Deiters’ cell loss at the extreme base of the cochlea, the first row of Deiters’ cells exhibited a more significant loss (Fig. 3C & D, repeated one-way ANOVA, F (2, 10) = 12.42, p = 0.0019; Tukey’s multiple comparisons: row 1 vs. row 3, Adjusted p = 0.002; row 2 vs. row 3, Adjusted p = 0.010, and row 1 vs. row 2, Adjusted p > 0.05; n=6 confocal images from 6 cochleae). This observation revealed a shift in the pattern of Deiters’ cell loss, changing from predominantly affecting the third row to primarily impacting the first row as the lesion expanded toward the basal end of the cochlea.

Figure 3.

Patterns of Deiters’ cell loss following cyclodextrin treatment. (A) Image depicting a cochlear region where most Deiters’ cells are present. Note that the loss in this region is more frequent in the third row (highlighted with dotted circles). (B) Comparison of the percentage of cell loss for the three rows of Deiters’ cells. **** indicates significance with p < 0.0001. (C) Image illustrating a cochlear region adjacent to the area of complete Deiters’ cell loss (marked by the arrow). Notice that the loss is more prevalent in the first and second rows (highlighted with dotted circles). (D) Comparison of the percentage of cell loss for the three rows of Deiters’ cells. ** indicates significance with p < 0.01. Each dot represents the result from one confocal image, and the data were obtained from the analysis of 10 cochleae.

We then examined the structural integrity of surviving Deiters’ cells in the region where substantial OHC loss occurred. To visualize Deiters’ cells, we employed double staining with phalloidin and an antibody against IFIT3 (interferon-induced protein with tetratricopeptide repeats 3). IFIT3 is highly expressed in the central cytoskeletal bundles of Deiters’ cells, extending from the top of the phalangeal process to the conical foundation of the central stalk on the basement membrane. This marker allowed us to evaluate the structural integrity of Deiters’ cells (Fig. 4A–C). We first examined the Deiters’ cells in the middle portion of the cochlea, where significant loss of OHCs was evident, but the reticular lamina architecture remained largely unaltered (Fig. 4D). In this region, the central stalks of Deiters’ cells appeared normal (Fig. 4E & F) and maintained their upright position. This preservation is clearly demonstrated using a side-by-side comparison of Deiters’ cells beneath areas with missing OHCs and those still supporting surviving OHCs. As shown in Figures 4G–I, no observable difference was detected in Deiters’ cell stalks between two Deiters’ cells, one with and one without its supporting OHC. Moreover, both Deiters’ cells maintained their phalangeal processes, indicating continuous support for the reticular lamina.

Figure 4.

The structural integrity of Deiters’ cells in regions with severe OHC loss. (A) A 3D view of the organ of Corti stained for actin and IFIT3. (B and C) A confocal image created by digital sectioning of the organ of Corti to highlight Deiters’ cells. (D) Actin staining showing complete OHC loss in the middle region of a cyclodextrin-treated cochlea. (E) IFIT3 staining of the same region reveals relatively intact central stalks and phalangeal processes of Deiters’ cells. (F) Merged view of the images from (D) and (E). (G) Actin staining in a cochlear region with both missing and surviving OHCs. Thin lines highlight the contours of a missing (left) and a surviving OHC (right). (H and I) IFIT3 staining reveals that Deiters’ cells beneath both surviving and missing OHC regions show similar structures in their central stalks and phalangeal processes. The lines in panel H outline the contours of the central stalks of the two Deiters’ cells.

At the basal end of the cochlear spiral, we observed regions of the organ of Corti where the reticular lamina was no longer visible. These regions were covered by a layer of cells displaying asymmetrical and multi-angled shapes (marked by asterisks in Fig. 5A). Inspection of 3D confocal images revealed that Deiters’ cells beneath these areas could still survive but lost their characteristic upright structure (Fig. 5B). These cells exhibited IFIT3 immunoreactivity within their cell bodies, and Deiters’ cell cups remained identifiable. However, IFIT3 immunoreactivity in the phalangeal process was undetectable, suggesting loss of the phalangeal process. The numbers of these damaged cells varied from a few to several dozen. At the extreme base, Deiters’ cells were no longer visible beneath the layer of flat cells.

Figure 5.

Surface morphology in regions with complete loss of the reticular lamina. (A) A surface view of the organ of Corti shows flattened cells with multi-angled shapes covering areas where the reticular lamina is absent. Arrows point to the residual reticular lamina. Asterisks mark the region where the reticular lamina is no longer present and is now covered by a layer of cells with irregular polygonal shapes. (B) An enlarged view of panel A with a slight change in viewing angle. Surviving Deiters’ cells are visible with a complete loss of the reticular lamina. Asterisks indicate residual reticular lamina, and single arrows point to Deiters’ cells in the region where there is a complete loss of reticular lamina. IFIT3 immunoreactivity confirms surviving but structurally altered Deiters’ cells beneath this region. The double-arrows point to the region with complete loss of Deiters’ cells. (C) SEM image reveals a bumpy surface over the organ of Corti (lightly colored region) in a severely damaged basal region of the cochlea, indicating underlying Deiters’ cells despite the loss of the reticular lamina.

The surface morphology of the area with complete loss of the reticular lamina can also be observed using SEM. As shown in Figure 5C, the reticular lamina in the basal end of the cochlea was replaced by a layer of irregular polygon-shaped cells. Noticeably, this area exhibited a bumpy surface texture, which could suggest the possible presence of Deiters’ cells underneath.

Together, our observation revealed site-dependent changes in Deiters’ cells in cyclodextrin-treated cochleae. In the middle portion of the cochlea, despite significant OHC loss, surviving Deiters’ cells maintained their upright position, continuing to provide structural support to the reticular lamina. In the basal end of the cochlea, Deiters’ cell loss was common. While Deiters’ cells could survive, a small portion of surviving Deiters’ cells lose their phalangeal processes, which leads to the loss of the reticular lamina.

3.3. Hensen’s cells in cyclodextrin-treated cochleae

Hensen’s cells are located laterally to the OHCs in the organ of Corti. They are arranged in multiple rows. The first row, referred to as tectal cells, exhibits a distinct rectangular shape, whereas Hensen’s cells in the subsequent rows display irregular shapes. In the normal and cyclodextrin-treated cochleae with preserved reticular lamina structures, these cells can be identified through a surface view of confocal images (Fig. 6A & B). However, identifying Hensen’s cells became challenging when the reticular lamina structure was malformed due to the loss of, or morphological changes in, Deiters’ cells and pillar cells following cyclodextrin treatment. Thus, we included galectin-3 as a marker to identify Hensen’s cells. As shown in Figures 6C & D, strong galectin-3 immunoreactivity is observed in the cytoplasm and nuclei of Hensen’s cells, whereas neighboring supporting cells (Deiters’ cells, Claudius’ cells, and Boettcher’s cells) lack galectin-3 immunoreactivity.

Figure 6.

Identification and quantification of Hensen’s cells. (A) A surface view of the organ of Corti stained for actin in a normal cochlea. (B) A surface view of actin staining showing Hensen’s cells in the cochlear region with preserved phalangeal scars in the reticular lamina from a cyclodextrin-treated cochlea. Hensen’s cells are present. (C) A 3D view of the organ of Corti stained for galectin-3, actin, and DAPI in a normal ear. Hensen’s cells are clearly identified. (D) Galectin-3 immunoreactivity is present primarily in Hensen’s cells, whereas hair cells, Claudius’ cells, and Deiters’ cells lack galectin-3 immunoreactivity. Scale bars = 20 µm. (E) Comparison of the average number of Hensen’s cells per 300 µm across three regions along the cochlear spiral in normal control cochleae. Cell counts are comparable between the apical and middle regions, with slightly, but statistically significant, higher numbers in the basal region compared to the apical. (F) Comparison of the average Hensen’s cell count between control and cyclodextrin-treated ears across the same three cochlear regions. In cyclodextrin-treated cochleae, there is a significant reduction in the number of Hensen’s cells at the middle and basal regions of the cochlea (45–85% and 85–100% distance from the apex). A slight, but not statistically significant, reduction is observed in the apical regions (0–45% distance from the apex). Each dot represents the number of Hensen’s cells in a single confocal image. * indicates p < 0.05 and **** indicates p < 0.0001.

We quantified the number of Hensen’s cells in normal cochleae. As shown in Figure 6E, Hensen’s cells are relatively evenly distributed across the apical and middle regions, with a slightly higher count observed in the basal region compared to the apical region (n=10 cochleae, one-way ANOVA, F (2, 25) = 4.057, p = 0.0298; Tukey’s multiple comparison test, apical vs. basal, p = 0.029). In cyclodextrin-treated cochleae, we observed that Hensen’s cell numbers remained largely unchanged in the apical section of the cochlea. However, in the middle and basal sections, there was a significant reduction in the number of Hensen’s cells. A two-way ANOVA analysis confirmed these findings, showing significant interaction between location and treatment factors (two-way ANOVA, F (2,46) = 22.87, p < 0.0001; control vs. treated in the apical region, p = 0.336; control vs. treated in the middle region, p = 0.026; control vs. treated in the basal region, p < 0.0001; Fig. 6F, n=10 cochleae for each control and treated group). In contrast to the complete loss of Deiters’ cells observed in the extreme base of the cochlea, no regions with complete loss of Hensen’s cells were found along the cochlear spiral.

To explore the role of Hensen’s cells in tissue repair, we investigated their response to the loss of neighboring supporting cells. We began by observing how Hensen’s cells reacted when individual Deiters’ cells were lost. Figure 7 shows an area in a cyclodextrin-treated cochlea where a Deiters’ cell was lost, indicated by the absence of its actin-enriched footplate. Notably, an adjacent Hensen’s cell, marked by galectin-3 immunoreactivity, filled the space originally occupied by the Deiters’ cell. This observation suggests that Hensen’s cells are responsible for filling the gap caused by the loss of Deiters’ cells.

Figure 7.

Hensen’s cell response to Deiters’ cell loss. (A) Image showing Deiters’ cells in the basal region of a cyclodextrin-treated cochlea. The arrow points to a region of a missing Deiters’ cell. (B and C) Galectin-3 labeling shows a Hensen’s cell body, filling the gap left by a lost Deiters’ cell. This finding suggests that Hensen’s cells contribute to structural repair following Deiters’ cell loss.

The role of Hensen’s cells in repairing structural defects due to Deiters’ cell loss can also be observed by examining the reticular lamina. In the cochlear region where the three rows of Deiters’ cell-supported phalangeal scars were present, Hensen’s cells retained their normal surface morphology despite the complete loss of OHCs (Fig. 8A–C). Toward the basal end of the cochlea, the structure of the three rows of phalangeal scars gradually narrowed. In regions where only one or two rows of phalangeal scars remained, Hensen’s cells covered the areas of the missing reticular lamina (Fig. 8D–F). In regions where the reticular lamina was no longer visible, but Deiters’ cells were still present, Hensen’s cells modified their shape to flatter forms, effectively covering the entire organ of Corti (Fig. 8G–I). Progressing further toward the basal end, in regions where both Deiters’ cells and pillar cells had disappeared, Hensen’s cells became a layer of flat cells with polygonal shapes covering the areas of the organ of Corti (see the area marked in the asterisk in Fig. 8G & I). Notably, galectin-3 immunoreactivity enables the distinction between Hensen’s cells and the laterally positioned Claudius’ cells. The progression of the organ of Corti pathogenesis underscores the crucial role of Hensen’s cells in forming the epithelial layer that repairs the structure damage due to the loss of Deiters’ cells. However, while their role in structural repair is evident, such repair does not restore hearing function. This process is more comparable to wound healing, where the structural integrity is preserved, but functional recovery is not achieved. These limitations reduce the potential for Hensen’s cells to serve as a suitable target for drug development aimed at restoring hearing function.

Figure 8.

Morphological adaptations of Hensen’s cells during reticular lamina degeneration. (A, B, and C) Hensen’s cells, marked by the arrows, retain their typical shape in areas with complete OHC loss, while three rows of pharyngeal scars remain evident. (D, E, and F) With the loss of phalangeal scars, Hensen’s cells fill the void, causing a narrowing of the reticular lamina (as indicated by arrows). (G, H, and I) Hensen’s cells extend to cover the entire organ of Corti in areas where the reticular lamina structure is no longer visible. Arrows point to the region where galectin-3 positive Hensen’s cells cover the Deiters’ cells. Notably, Claudius’ cells lack galectin-3 immunoreactivity. The asterisk marks the region where all Deiters’ and pillar cells are absent, with the structure being covered by galectin-3 positive Hensen’s cells.

Changes in Hensen’s cell morphology can also be observed through SEM. Figure 9 provides surface views of the organ of Corti where the residual reticular lamina is present (Fig. 9A) and where it is absent (Fig. 9B). We reveal that as the reticular lamina deteriorates, Hensen’s cells expand medially to preserve the continuity of the surface structure. However, confidently identifying Hensen’s cells using SEM remains challenging in severely damaged regions (Fig. 9B).

Figure 9.

SEM analysis of Hensen’s cell adaptations. (A) Surface view of the organ of Corti in a region where the residual reticular lamina is present indicated by a lightly colored region pointed out by arrows. Hensen’s cells are located lateral to the residual reticular lamina. (B) Surface view of the organ of Corti in a region where the residual reticular lamina is absent. Hensen’s cells expand to maintain the continuity of the cochlear surface structure. It is important to note that identifying cell types using SEM in this region remains challenging.

3.4. Pillar cells in cyclodextrin-treated cochleae

Pillar cells are classified into inner and outer pillar cells, which connect at their tops to form an arched structure at the apex of the tunnel of Corti. We counted the number of pillar cells and found that the inner pillar cells slightly outnumber the outer pillar cells, with a ratio of approximately 1.38 to 1. Similar to Deiters’ cells, a majority of inner and outer pillar cells survived in the apical and middle regions of the cochlea despite significant loss of adjacent OHCs and IHCs after cyclodextrin treatment. However, we observed significant pillar cell loss in the basal region of the cochlea (Fig. 10A). Our findings revealed a strong positive correlation between the percentage of outer and inner pillar cell loss (Fig. 10B, Pearson correlation coefficient, r = 0.956, two-tailed p < 0.0001, n=19 confocal images from 10 cochleae). Notably, the extent of outer pillar cell loss was slightly, but significantly, higher than that of inner pillar cells (Fig. 10C, two-tailed paired Student’s t-test, t (18) = 4.178, p = 0.0006, n=19 images from 10 cochleae), suggesting that outer pillar cells are more vulnerable than inner pillar cells.

Figure 10.

Pillar cell loss in cyclodextrin-treated cochleae. (A) Actin staining showing pillar cells in the basal region of a cyclodextrin-treated cochlea. The asterisks mark the areas of missing outer pillar cells, and the cycles mark the areas of missing inner pillar cells. (B) Pearson correlation coefficient analysis reveals a strong and positive correlation in the percentage of loss between inner and outer pillar cells. (C) Comparison of the percentage of loss between inner and outer pillar cells. The level of outer pillar cells is slightly, but significantly, higher than that of inner pillar cells. *** indicates p < 0.001. IPC represents inner pillar cells, and OPC represents outer pillar cells.

We then assessed whether the surviving pillar cells continued to support the arch structure in the organ of Corti. As shown in Figures 11A & B, pillar cells retained their upright position and the arch over the tunnel of Corti remained in areas where OHCs were entirely absent in the middle section of the cochlea. However, in regions near the extreme base, where the reticular lamina was completely lost, the arch formed by the pillar cells appeared to be sagging (Fig. 11C & D). Overall, these observations demonstrated that despite significant loss of hair cells following cyclodextrin treatment, most inner and outer pillar cells in the apical and middle regions of the cochlea maintained their structural integrity. However, pillar cell loss and arch sagging were observed in the basal region.

Figure 11.

The structural integrity of pillar cells in regions with complete OHC loss. (A) Actin staining of the reticular lamina in a cyclodextrin-treated cochlea. All OHCs are absent, as indicated by the loss of actin staining in the cuticular plates of OHCs. (B) A side view of the region of the organ of Corti marked by the orange rectangle in panel A. The image was generated by digital sectioning of confocal images of the region. Note that the pillar cells maintain their vertical orientation (arrow), supporting the arch of the tunnel of Corti. (C) Actin staining of the reticular lamina in a cyclodextrin-treated cochlea. The image shows the boundary area between the region where pillar cells are visible and the region where they are no longer visible. The orange rectangle marks pillar cells next to the area of missing pillar cells. (D) The side view of the organ of Corti marked by the orange rectangle in panel C. Note that the surviving pillar cells near the areas of missing pillar cells sag (arrow), indicating compromised structural support.

3.5. Variability in damage levels among different supporting cell populations

Our final goal of the study was to assess the differences in damage levels among the supporting cell population examined in the current study: Deiters’ cells, pillar cells, and Hensen’s cells. To achieve this, we compared the percentage loss of these cells. For Deiters’ and pillar cells, because these cells are arranged in a well-defined pattern, we were able to count both the missing and remaining cells to calculate the percentage loss and then compare the damage levels between the two cell populations. In the apical and middle regions of the cochlea (0–85% distance from the apex), where supporting cell loss was sporadic, no significant differences were observed between these two cell types (Fig. 12A, two-tailed paired Student’s t-test, t (52) = 1.579, p = 0.120, n=53 confocal images from 10 cochleae). However, in the extreme base of the cochlea (85–100% distance from the apex), where significant loss of supporting cell loss occurred, the percentage of pillar cell loss was significantly higher compared to that of Deiters’ cells (Fig. 12B, two-tailed paired Student’s t-test, t (18) = 3.257, p = 0.0044, n=19 confocal images from 10 cochleae). The increased susceptibility of pillar cells was more apparent in areas adjacent to the complete loss of Deiters’ cells and pillar cells. As shown in Figure 12C, Deiters’ cells remained present in regions where both inner and outer pillar cells were already absent. Our findings suggest that pillar cells are more vulnerable to cyclodextrin treatment than Deiters’ cells.

Figure 12.

Comparative analysis of Deiters’ cell and pillar cell (inner and outer pillar cell) susceptibility to cyclodextrin treatment. (A) In the apical and middle regions, there are no significant differences in the percentage loss of Deiters’ cells (DC) and pillar cells (PC). Each dot represents data from one confocal image, and the data was obtained from the analysis of 10 cochleae. ** indicates p < 0.01. (B) In the basal region of the cochlea, pillar cells exhibit a significantly greater loss compared to Deiters’ cells. Each dot represents data from one confocal image, and the data was obtained from the analysis of 10 cochleae. (C) The image shows the juncture of the organ of Corti between the region with complete loss of pillar cells (marked by asterisks) and the region with surviving pillar cells marked by the double-arrow. Notice that Deiters’ cells are present (arrows) in the region where both inner and outer pillar cells are completely lost, highlighting the increased vulnerability of pillar cells to cyclodextrin-induced damage.

To assess Hensen’s cells, we quantified the cells present but could not determine the exact number of missing cells due to their disorganized distribution. Consequently, we estimated their cell loss by comparing the remaining cells in each viewing area with the average number of Hensen’s cells in normal control ears. Since our quantitative analysis of the three types of supporting cells was performed in identical viewing areas, we were able to directly compare the extent of their loss at these sites.

Figure 13 presents a comparison of the percentage loss of Deiters’ cells, pillar cells (both inner and outer pillar cells), and Hensen’s cells across the apical, middle, and basal regions of the cochlea. These regions correspond to distances of 0–45%, 45–85%, and 85–100% from the apex, respectively. The figure includes both average and individual percentage losses. The average loss was calculated for each cochlear region (0–45%, 45–85%, and 85–100% from the apex), while individual data points represent the percentage loss observed in each confocal image. The three dots along each vertical line signify the percentage loss of the three types of supporting cells within the same confocal image. On the x-axis, individual data points are arranged in order of increasing Deiters’ cell loss. The negative percentage values for Hensen’s cells indicate that the observed number is higher than in the control ears.

Figure 13.

Comparison of percentage loss across three supporting cell populations—Deiters’ cells, pillar cells (inner and outer pillar cells), and Hensen’s cells—in the apical (A), middle (B), and basal (C) regions of the cochlea, which correspond to distances of 0–45%, 45–85%, and 85–100% from the apex, respectively. In each plot, the left panel shows the comparison of the average percentage loss, and the right panel shows individual values from each confocal image. The three dots along the vertical lines represent the percentage loss of three types of supporting cells within the same confocal images. On the x-axis, individual data points are arranged in order of increasing Deiters’ cell loss. Notice that minimal loss is observed in Deiters’ and pillar cells within the apical and middle regions. In contrast, Hensen’s cells exhibit a greater degree of loss compared to Deiters’ and pillar cells in the middle region, despite considerable individual variation. This includes instances where negative values suggest a higher number of Hensen’s cells than in control ears. In the basal region, all three types of supporting cells showed an increased loss, with Hensen’s cell loss being greater than that of the pillar and Deiters’ cells. However, at the extreme base, while Deiters’ and pillar cell loss continued to increase, Hensen’s cell loss did not further increase. (D) Schematic representation of OHC and IHC loss across the three cochlear regions (apex, middle, and base). Unlike the localized loss of supporting cells, OHC and IHC loss extends throughout the cochlea, with the greatest loss occurring in the middle and basal regions, while the apical region exhibits greater survival, as indicated by the dark blue line for OHCs and the dark purple line for IHCs.

In the middle region (Fig. 13B), Hensen’s cells show a greater degree of loss compared to Deiters’ and pillar cells, despite considerable individual variation (Friedman test, χ² = 9.391, p = 0.009). This difference is not statistically significant in the apical region (Fig. 13A, Friedman test, χ² = 2.625, p = 0.2691). In the basal region, all three types of supporting cells exhibit increased loss (Fig. 13C), with Hensen’s cells showing the greatest loss (Friedman test, χ² = 16.67, p = 0.0002). However, at the extreme base, while Deiters’ and pillar cell loss continued to increase, Hensen’s cell loss did not further increase. Overall, supporting cell loss was confined to the basal region, while hair cell loss extended throughout the cochlea, with the most significant loss occurring in the middle and basal regions (Fig. 13D).

4. Discussion

4.1. Quantification of supporting cells reveals variations in their susceptibility to cochlear damage

Previous research has investigated the vulnerability of the organ of Corti, with significant emphasis on sensory cells. These studies have quantified sensory cell loss under diverse pathological conditions (Sugawara et al., 2005; Hamernik et al., 1984a). However, quantitative studies focusing on supporting cells remain scarce. The limited quantitative analysis of supporting cell populations is partly due to the lack of reliable methods, especially for quantifying Hensen’s cells, which lack a distinct distribution pattern. In this study, we employed confocal microscopy combined with 3D image analysis for tissue observation and analysis. Compared to the traditional tissue sectioning methods, 3D image analysis offers significant advantages by allowing the examination of tissue details without physical sectioning. It also surpasses conventional whole-mount preparation by providing a clearer view of hidden cells, such as Deiters’ cells, which are located deep within the organ of Corti. Without 3D analysis, quantifying these supporting cells would not be achievable.

In the current study, we quantified various supporting cell populations across different regions along the cochlear spiral. Our goal was to identify which supporting cells are more susceptible to damage and to determine how this susceptibility is distributed along the cochlear spiral. Our results revealed that pillar cells exhibit a heightened vulnerability, followed by Deiters’ cells. Notably, this finding differs from a previous study on human cochlear tissues that identified Deiters’ cells as the most vulnerable during aging degeneration (Kaur et al., 2023). While this discrepancy may be due to species differences, it could also stem from the limitations of the cochlear sectioning method used in the previous study, which may not have effectively identified Deiters’ cells. In contrast, our immunostaining method combined with 3D image analysis allows for accurate identification of these Deiters’ cells. Additionally, the previous study employed semiquantitative analysis, whereas our study utilized a quantitative approach, providing a more accurate assessment.

We also found that the two types of pillar cells displayed different levels of vulnerability, with outer pillar cells showing greater loss than inner pillar cells. This is consistent with findings from a previous study using semi-quantitative analysis of cochlear sections (Kaur et al., 2023). Interestingly, we found that the three rows of Deiters’ cells also exhibited different levels of damage depending on their location. In regions with sporadic or small clusters of Deiters’ cell loss, the third row experienced the most significant loss. However, in regions next to areas of complete Deiters’ cell loss, the first row was more severely affected. This finding is significant, as it suggests different vulnerabilities in the pathogenic progression of different supporting cell populations. Further investigation into the underlying mechanisms is essential for developing targeted interventions to mitigate the impact of ototoxicity on auditory health.

4.2. Deiters’ cell loss does not mirror the pattern of OHC loss

Deiters’ cells are supporting cells that have direct contact with OHCs. Given this proximity, it is reasonable to suspect that the survival of Deiters’ cell could be affected by the loss of the OHCs they support. However, our quantitative analysis indicates that the pattern of Deiters’ cell loss does not correspond to the pattern of OHC loss. Specifically, the vulnerability pattern of the three rows of OHCs differs from that of Deiters’ cells. For OHCs, the first row shows the most significant loss in the areas where surviving OHCs are present. However, Deiters’ cells in these areas display the opposite pattern, with the third row being the most vulnerable. Moreover, Deiters’ cells demonstrate considerable resilience, surviving even in regions where their associated OHCs have been lost. This resilience is particularly evident in the middle section of the cochlea, where significant OHC loss occurs, yet most Deiters’ cells remain intact and maintain their structural integrity. Even at the extreme base, where the organ of Corti has collapsed, some Deiters’ cells still manage to survive, though they become disoriented and lose their phalangeal processes. These observations suggest that the survival of Deiters’ cells is not directly dependent on the presence of OHCs. OHC loss and Deiters’ cell loss may involve different mechanisms of vulnerability and resilience.

4.3. Role of Hensen’s cells in the repair process in the organ of Corti

Previous research has explored the role of supporting cells in the structural repair of the organ of Corti following cochlear damage. Morphological observations have revealed a layer of epithelial cells covering the basilar membrane when the entire organ of Corti architecture collapses. Cross-sectioning views of the cochlea reveal that this epithelium consists of a thin layer of cells, while surface views observed using SEM indicate that these cells are large, flat, and irregularly shaped with multiple edges. Because these cells lack the typical morphology and intracellular cytoskeletal characteristics of supporting cells, their origin remains unclear. Some researchers speculate that this flat epithelium originates from Claudius’ cells (Hamernik, R. P. et al., 1984; Hamernik, Roger P et al., 1984; Roberto and Zrro, 1988), while others suggest that these cells exhibit the features of Hensen’ cells (Taylor et al., 2012; Taylor et al., 2008). Here, we provide compelling evidence to show that Hensen’s cells contribute to the formation of the epithelial layer.

We categorize the degradative process of the organ of Corti into four distinct phases using galectin-3 as a marker for Hensen’s cells. In the initial phase, the loss of three rows of OHCs is observed, though the reticular lamina remains unaltered. Hensen’s cells preserve their typical morphology at this stage (Fig. 10A–C). In the second phase, the reticular lamina becomes narrowed as the three rows of Deiters’ cell phalangeal scars become malformed. Hensen’s cells expand medially to maintain contact with the reticular lamina (Fig. 10D–F). The third phase is marked by the disappearance of the reticular lamina due to the loss of the phalangeal processes of Deiters’ cells and pillar cells. During this phase, Hensen’s cells undergo enlargement, forming an overarching structure over the surviving Deiters’ cells (Fig. 10G–I). The final phase is characterized by the complete absence of Deiters’ cells and pillar cells. Hensen’s cells are flattened and enlarged, taking on a flat and polygonal shape (Fig. 10G–I). Together, these findings suggest that Hensen’s cells change their morphology to form a protective layer over the basilar membrane as the organ of Corti degenerates.

4.4. Potential causal relationship between sensory cell damage and supporting cell damage

Our study demonstrates the loss of both sensory cells and supporting cells following cyclodextrin treatment, raising the question of whether these two events are causally related. One possibility is that high doses of cyclodextrin directly affect hair cells despite the presence of supporting cells, leading to damage and eventual loss of hair cells. The loss of sensory cells may then compromise the integrity of supporting cells. Alternatively, cyclodextrin could simultaneously target both hair cells and supporting cells, with varying levels of susceptibility among these cells. Another possibility is that cyclodextrin weakens supporting cells, diminishing their ability to sustain hair cells, subsequently leading to sensory cell damage. While our study cannot definitively answer these questions, the results suggest that supporting cell damage may largely be a consequence of sensory cell loss. This speculation is supported by the observation that sensory cell loss extends throughout the basal and middle sections of the cochlea, while supporting cell loss remains confined to the basal end. It is also possible that during the acute phase of cyclodextrin ototoxicity, both sensory and supporting cells are affected, as a small amount of supporting cell loss was observed within 1–4 days post-treatment. Further studies are needed to clarify whether traumatic events directly or indirectly impact both hair cells and supporting cells.

4.5. In summary

The present study aimed to elucidate the susceptibility of supporting cells to cyclodextrin-induced ototoxicity and their role in structural repair following cochlear damage. The findings indicate that supporting cell loss correlates with the location along the cochlear spiral: supporting cells at the basal end of the cochlea experienced the greatest loss, while only minimal loss was observed in the apical and middle regions despite substantial OHC loss. Notably, damage levels differed among supporting cell populations. Deiters’ cells demonstrated greater resilience than pillar cells, with varying susceptibility across specific rows. Reduction of Hensen’s cells began in the middle region, with losses surpassing those of both Deiters’ and pillar cells. However, at the extreme base, Hensen’s cells were present even when both pillar and Deiters’ cells were completely absent. Furthermore, our observations indicate that Deiters’ cells can survive even after losing their phalangeal processes. The study also shows the role of Hensen’s cells in tissue repair when the organ of Corti collapses, contributing to the formation of an epithelial layer over the basilar membrane. Collectively, these findings provide detailed insights into supporting cell involvement in cochlear damage and offer valuable information to guide future efforts in developing treatment strategies targeting these cells.

Highlights.

• Different supporting cell populations display varying levels of vulnerability to ototoxicity.

• Some Deiters’ cells can survive despite the loss of their phalangeal processes.

• Hensen’s cells, although susceptible to loss, avoid complete depletion in the extreme base.

• Hensen’s cells contribute to forming an epithelial layer over the basilar membrane when the organ of Corti collapses.

Abbreviation

OHC

Outer Hair Cells

IHC

Inner Hair Cells

PBS

Phosphate-Buffered Saline

ABR

Auditory Brainstem Response

DPOAE

Distortion-Product Otoacoustic Emissions

TDT

Tucker-Davis Technologies

SEM

Scanning Electron Microscopy

EDTA

Ethylenediaminetetraacetic Acid

DAPI

4’,6-Diamidino-2-Phenylindole

IFIT3

Interferon-Induced Protein with Tetratricopeptide Repeats 3

ANOVA

Analysis of Variance

RNA

Ribonucleic Acid