Age‐related changes in P2Y receptor signalling in mouse cochlear supporting cells

Ethics statement

All procedures were approved by the Home Office, in line with the Animals (Scientific Procedures) Act 1986, and were approved by the Ethical Review Committee at the University of Sheffield (180 626_Mar). Mice had an unlimited access to food and water. Schedule 1 cervical dislocation, and subsequent decapitation, was used to humanely kill mice used for ex vivo experiments. For in vivo recordings of auditory brainstem responses (ABRs), mice were anaesthetised using an intraperitoneal (i.p.) injection of ketamine (100 mg kg⁻¹ body weight; Fort Dodge Animal Health, Fort Dodge, IA, USA) and xylazine (10 mg kg⁻¹ body weight, Rompun 2%; Bayer HealthCare LLC, Tarrytown, NY, USA). At the end of the ABR experiments, mice were either killed by cervical dislocation or recovered from anaesthesia (i.p. injection of atipamezole: 1 mg kg⁻¹ body weight). During the recovery from anaesthesia, mice were returned to their cage placed on a thermal mat and monitored over the following 2–5 h. Littermate mice of either sex were used for experiments.

Mouse strains

We used three strains of mice known to display differing severities of progressive hearing loss. C57BL/6N (6N) exhibit early onset hearing loss due to a hypomorphic allele in Cadherin 23 (Cdh23^ahl ) (Jeng et al., 2021; Jeng, Ceriani et al., 2020; Jeng, Johnson et al., 2020; Johnson et al., 1997; Kane et al., 2012; Noben‐Trauth et al., 2003). These mice show early onset hearing loss that is already evident at 3 months of age in the high‐frequency region and continues to worsen over time (Jeng, Ceriani et al., 2020). Cdh23 encodes for cadherin‐23 that, together with protocadherin‐15, forms the tip links that are required to gate the mechanoelectrical transducer channels (Kazmierczak et al., 2007). The C57BL/6N ^Cdh23+ (6N‐Repaired) strain is co‐isogenic to 6N, except for having their Cdh23 mutation corrected via CRISPR/Cas9‐mediated homology directed repair (Mianné et al., 2016). 6N‐Repaired mice retain normal high‐frequency hearing thresholds until late in life, but they still exhibit low‐frequency hearing loss comparable to 6N mice (Jeng, Ceriani et al., 2020). The third strain is C3H/HeJ (C3H), which is known to retain good hearing function until later in life (Jeng et al., 2021; Jeng, Ceriani et al., 2020; Jeng, Johnson et al., 2020; Ohlemiller et al., 2016).

Auditory brainstem responses

Following the onset of anaesthesia (see Ethics statement above) and the loss of the retraction reflex with a toe pinch, mice were placed onto a heat mat (37°C) in a soundproof chamber (MAC‐3 acoustic chamber, IAC Acoustic, UK). Subdermal electrodes were placed under the skin behind the pinna of each ear (reference and ground electrode) and on the vertex of the mouse (active electrode) as previously described (Ingham et al., 2011). Auditory brainstem responses (ABRs) were recorded from male and female mice from the strains listed above. Sound stimuli were delivered to the ear by calibrated loudspeakers (MF1‐S, Multi Field Speaker, Tucker‐Davis Technologies, USA) placed 10 cm from the animal's pinna. Sound pressure was calibrated with a low‐noise microphone probe system (ER10B+, Etymotic, USA). Experiments were performed using a customised software (Ingham et al., 2011) driving an RZ6 auditory processor (Tucker‐Davis Technologies). Auditory thresholds were estimated from the resulting ABR waveform and defined as the lowest sound level (measured in decibel, dB) where any recognisable feature of the waveform was visible. Responses were measured for clicks and stimulus pure tones of frequencies at 3, 6, 12, 18, 24, 30, 36 and 42 kHz. Stimulus sound pressure levels were typically 0–95 dB SPL, presented in steps of 5 dB SPL. The brainstem response signal was averaged over 256 repetitions. Tone bursts were 5 ms in duration with a 1 ms on/off ramp time, which was presented at a rate of 42.6/s.

Immunofluorescence microscopy

Inner ears were dissected out and perfused with 4% paraformaldehyde solution for 20 min at room temperature. The fixed samples were then micro‐dissected to retrieve the apical turn of the organ of Corti. These were then blocked in 5% horse serum for an hour at room temperature before being incubated at 35°C overnight in primary antibody solution. The primary antibodies used included anti‐P2Y₁ (Alomone Labs, cat. no. APR‐009), anti‐P2Y₂ (Alomone Labs, cat. no. APR‐010) and anti‐P2Y₄ (Alomone Labs, cat. no. APR‐006) all at a concentration of 1:800. The following day, samples were washed in phosphate buffer saline (PBS) and incubated for an hour at 35°C in secondary antibody solution. This solution contained goat anti‐rabbit IgG 490 (Alexa fluor 488, A11034, Thermo Fisher) and Texas Red‐X Phalloidin for staining of F‐actin (1:1000, T7471, Thermo Fisher). The samples were then washed in PBS and mounted in VECTASHIELD (H‐1000). Images of the cochleae were taken with a Zeiss LSM 880 Airyscan microscope. The z‐stack images were taken with 0.5 μm incremental steps. Fiji ImageJ software was used to process the images and generate maximum intensity projections (https://imagej.net/Fiji).

Tissue preparation and dye loading

Experiments were performed from the 9–12 kHz region of the cochlear apical coil (Müller et al., 2005). The apical coil was dissected out in extracellular solution composed of (in mm): 135 NaCl, 5.8 KCl, 1.3 CaCl₂, 0.9 MgCl₂, 0.7 NaH₂PO₄, 5.6 d‐glucose, 10 HEPES‐NaOH. Sodium pyruvate (2 mm), amino acids and vitamins were added from concentrates (Thermo Fisher Scientific, UK). The pH was adjusted to 7.48 (∼308 mmol kg⁻¹). Once dissected, the organ of Corti was transferred into a calcium dye loading solution and incubated at 37°C for 30 min. The loading solution contained DMEM/F12, 0.1% pluronic F‐127 (Thermo Fisher Scientific, UK), 250 μm sulfinpyrazone to avoid dye sequestration (Sigma‐Aldrich) and the cell‐permeant fluorescent calcium indicator Fura‐2‐AM at a concentration of 10 μM (Thermo Fisher Scientific, UK).

Calcium imaging

After loading the cochlea with the Ca²⁺ dye, the dissected apical coil was transferred to a microscope chamber, immobilised via a nylon mesh attached to a stainless‐steel ring, and then viewed with an upright microscope (Olympus BX51WI, Japan) equipped with Nomarski Differential Interference Contrast (DIC) optics (Olympus LUMPlanFl 60× water immersion objective, 1.0 NA) and a 15× eyepiece. The microscope chamber was continuously perfused with extracellular solution by a peristaltic pump (Cole‐Palmer, UK). All experiments were performed at room temperature. Preparations were perfused with extracellular solution for 15 min before imaging to allow for dye de‐esterification.

For pharmacological experiments, ATP, ADP, UTP, MRS2500 (Tocris Bioscience, UK), Thapsigargin (Tocris Bioscience, UK), AR‐C 118925XX (Tocris Bioscience, UK), and MRS4062 (Tocris Bioscience, UK) were diluted from concentrates in extracellular solution and applied to the preparation using either a Pico‐injector (PLI‐100, Harvard Apparatus, USA) or a gravity‐fed perfusion system through a multi‐barrel microcapillary. Pressure was kept at a minimum to avoid motion during imaging and the triggering of mechanically induced Ca²⁺ responses.

Fura‐2 fluorescence was excited at two alternating wavelengths by using two light‐emitting diodes (LEDs) with peak wavelengths of 365 nm and 385 nm, respectively (Thorlabs, USA) filtered through narrow band filters (FF360/23 and FF392/23, respectively Semrock, USA). Fura‐2 emission was separated from excitation light through a long‐pass dichromatic mirror (T495lpxr, Chroma, USA) and filtered through a bandpass interference filter (ET525/50M, Chroma). Fluorescence images were formed on a scientific grade CMOS camera (Hamamatsu ORCA Flash 4.0 V2, Japan) operating in ‘external trigger’ mode and captured on a computer using Micro‐Manager (Edelstein et al., 2014). The timing of camera exposure and LED illumination were controlled by a microcontroller (Arduino Mega, Italy) and set by custom‐built Python software (Python 3.10, Python Software Foundation). LED light was switched on only during frame exposure by the camera to minimise phototoxicity. The exposure time was set at 65 ms for the 365 nm LED and 25 ms for the 385 nm LED. The exposures at the two wavelengths were separated by 100 ms and repeated every 1 s (final framerate for the ratio signal: 1 Hz). The framerate was chosen to minimise UV light exposure to the cells, while maintaining enough temporal resolution to sample the relatively slow purinergic‐induced Ca²⁺ changes in supporting cells. A brightfield image of the preparation was generally acquired at the end of each recording to aide with regions of interest (ROIs) placement (see below).

Images were analysed off‐line using software and scripts written in Python and ImageJ (NIH). Square ROIs (20 × 20 pixels, or 3.5 μm × 3.5 μm) were manually placed using a custom software on identified supporting cells in the proximity of hair cells (Fig. 5) using an average of fluorescence images captured using 385 nm excitation or a brightfield image as reference. The background fluorescence value was calculated as the average pixel value from a timelapse acquisition in a cochlear explant not loaded with the Fura‐2 Ca²⁺ dye and subtracted from each fluorescence image.

Ca²⁺ changes were quantified as fractional dye emission ratio changes:

where R(t) is the ratio between the fluorescence emission values for excitation at 365 nm and 385 nm at time t. R ₀ denotes the baseline R value calculated as an average during the 3 s before stimulation.

To quantify Ca²⁺ responses, we measured the maximum and average of the ΔR/R ₀ signal from individual ROIs during the application of the different compounds. To compare recordings in which compounds were applied for different durations, we limited the analysis to the first 30 s after the onset of drug application. Calcium oscillations were identified by detecting local maxima in the Ca²⁺ trace using a peak‐finding algorithm (function find_peaks of the scipy.signal Python module). The frequency was calculated as the number of Ca²⁺ oscillations divided by the duration of the stimulus. Individual traces were subsequently inspected manually to remove any spurious peaks and re‐analysed.

Statistical analysis

Statistical comparisons between two paired groups were made by Wilcoxon signed‐rank test. For statistical comparisons of unpaired data, analysis of variance (one‐way or two‐way ANOVA followed by an appropriate post hoc test), the non‐parametric Kruskal‐Wallis test (followed by pairwise Wilcoxon rank‐sum test with Bonferroni correction for multiple comparisons) or the non‐parametric aligned ranks transformation (ART) two‐way ANOVA (followed by Tukey's post hoc test or the pairwise Wilcoxon rank‐sum test with Bonferroni correction) were used. P < 0.05 was selected as the criterion for statistical significance. Statistical tests were performed using R software (R Core Team 2022). Only mean values with a similar variance between groups were compared. Values are quoted in text and figures as means ± SD. Animals of either sex were randomly assigned to the different experimental groups. No statistical methods were used to define sample size, which was determined based on previous published similar work on cochlear ageing and supporting cell physiology (e.g. Ceriani et al., 2016; Ceriani et al., 2019; Jeng et al., 2021; Jeng, Ceriani et al., 2020; Jeng, Johnson et al., 2020). Animals were taken from several cages and breeding pairs over a period of several months.

Progression of ABR thresholds between C57BL/6N and C3H/HeJ mice

Auditory brainstem responses (ABRs) were used to test hearing sensitivity of 1‐month and 20‐month‐old C57Bl/6N (6N) and C3H/HeJ (C3H) mice (Fig. 1 A and B). Thresholds for clicks recorded from 6N mice largely increased between young adult and aged mice (P < 0.0001, post hoc test from one‐way ANOVA), but not in C3H mice (P = 0.7607) (Fig. 1 C). Pure‐tone evoked ABRs thresholds (Fig. 1 D and E) were also found to increase with age in 6N mice (P < 0.0001: two‐way ANOVA) but less so in C3H mice (P = 0.0155). These data corroborate previous finding indicating that the hypomorphic allele in Cdh23^ahl present in 6N mice, leads to the almost complete loss of hearing function in aged mice (Jeng et al., 2021; Jeng, Ceriani et al., 2020; Jeng, Johnson et al., 2020; Johnson et al., 1997; Kane et al., 2012; Noben‐Trauth et al., 2003). However, a recent study has shown that the correction of the Cdh23^ahl allele using CRISPR/Cas9 (Mianné et al., 2016) was able to rescue high‐frequency, but not low‐frequency hair cell loss in the aged co‐isogenic 6N‐Repaired mice (<18 kHz: Jeng, Ceriani et al., 2020). This indicates that other changes, rather than the Cdh23^ahl allele, are likely to be responsible for the progressive low‐frequency hearing loss in the 6N mouse strain. Since supporting cells are essential for the normal development, survival and function of hair cells and their neurons (e.g. Jagger & Forge, 2015; Wan et al., 2013), we investigated whether they undergo strain‐specific changes in the ageing cochlea.

Changes in purinergic receptor expression in the ageing cochlea

In the mature cochlea, supporting cells are divided into two compartments delimited by the tunnel of Corti: the medial compartment, where the IHCs are located and the outer compartment, which contains the OHCs (Jagger & Forge, 2006). The supporting cells located in the medial compartment are organised in a single layer and include the inner phalangeal cells (IPhC) and the inner border cells (IBC), which surround the IHCs, and the cells forming the so‐called inner sulcus (IS, Fig. 2 A) (Fritzsch & Elliott, 2018). This area develops from the greater epithelial ridge (GER), which contains the Kölliker's organ that is transiently present during pre‐hearing developmental stages (Fig. 2 A). Conversely, supporting cells in the lateral region of the organ of Corti are part of the lesser epithelial ridge (LER) or outer sulcus.

Among the different types of P2Y receptors expressed in the mammalian cochlea, P2Y₁, P2Y₂, and P2Y₄ have been shown to be the primary isoforms associated with the generation of both spontaneous and induced Ca²⁺ signals (Babola et al., 2020; Huang et al., 2010; Piazza et al., 2007; Prades et al., 2021). Therefore, we investigated whether the level of expression and localization of these three P2Y receptors change in the apical‐coil of the cochlear sensory epithelium (6−12 kHz, Müller et al., 2005) from ageing mice using immunostaining. P2Y₁ receptors were highly expressed in the supporting cells of the GER during pre‐hearing stages of development (P7) in both 6N (Fig. 2 B, D and F, top row) and C3H mice (Fig. 2 C, E and G, top row). However, compared to P7, the number of P2Y₁ puncta were largely decreased in the supporting cells of young adult (1 month) and aged mice from both 6N (Fig. 2 D and F) and C3H mice (Fig. 2 E and G). P2Y₂ and P2Y₄ puncta were also largely reduced between P7 and 1 month of age in both 6N (P2Y₂: Fig. 3 A and C; P2Y₄: Fig. 4 A and C) and C3H mouse strains (P2Y₂: Fig. 3 B and D; P2Y₄: Fig. 4 B and D). However, P2Y₂ and P2Y₄ puncta reappeared in the supporting cells of 6N aged mice (P2Y₂: Fig. 3 C; P2Y₄: Fig. 4 C). These immunolabelling experiments provide a qualitative indication that the expression of the three P2Y receptors is likely to change during the maturation and ageing of the cochlea, and also between the 6N and C3H strains for P2Y₂ and P2Y₄.

Age‐related changes in purinergic‐mediated Ca²⁺ signalling in cochlear supporting cells

To obtain a functional readout of purinergic receptors in cochlear supporting cells, we performed ratiometric Ca²⁺ imaging with the dye Fura‐2 in acute explants of the apical‐coil cochlear region (6–12 kHz) of 6N, 6N‐Repaired and C3H mice. In the pre‐hearing developing cochlea, supporting cells of the GER release ATP that, by binding to G‐protein coupled P2Y auto‐receptors, leads to the generation and propagation of intercellular Ca²⁺ waves (Piazza et al., 2007; Tritsch et al., 2007). Although some previous studies have used micromolar concentrations of ATP to investigate purinergic receptor activity in the GER (1–100 μm: Horváth et al., 2016; Lahne & Gale, 2008; Rabbitt & Holman, 2021; Tritsch & Bergles, 2010), we found that 100 nm ATP was sufficient to trigger Ca²⁺ elevations in the supporting cells of the GER in the close proximity of the IHCs (Fig. 5).

Calcium dynamics in the GER and IS were measured from the supporting cells next to the IHCs (Fig. 6 A). In 6N mice, we observed age‐dependent changes in Ca²⁺ dynamics following application of 100 nm ATP (Fig. 6 B–E, N). Both sustained Ca²⁺ oscillations and repetitive Ca²⁺ ‘spikes’ (i.e. oscillations where Ca²⁺ levels returned to baseline after every cyclic increase) following application of ATP were recorded from the supporting cells of the GER of P7 pre‐hearing mice (Fig. 6 B). However, both average and maximum Ca²⁺ responses were greatly reduced in young adult mice (1 month‐old, P < 0.0001 for both comparisons, pairwise Wilcoxon rank sum test, ART two‐way ANOVA, Fig. 6 O and P). This change in Ca²⁺ response with cochlear maturation is consistent with previous studies showing a reduction of purinergic‐induced responses (Tritsch & Bergles, 2010) and down‐regulation of P2Y receptor expression in the inner sulcus after the onset of hearing (Huang et al., 2010). Although ATP‐induced Ca²⁺ responses did not change in size between 1 and 6 months of age (average: P = 0.1246, maximum: P > 0.9999), by 12 months of age they started to increase (1 vs. 12 months, P < 0.0001 for both average and maximum) and by 18–24 months resembled those recorded during pre‐hearing stages (Fig. 6 O and P).

To determine whether the increase in ATP sensitivity was due to the hearing loss phenotype of 6N mice, or a more general ageing phenomenon in the supporting cells, we performed Ca²⁺ imaging experiments in C3H mice, which at 20 months of age still exhibit good hearing thresholds (Fig. 1). We found that cochlear supporting cells from mature C3H mice showed a very small age‐related increase in the size of ATP‐induced Ca²⁺ responses (Fig. 6 F–I, N), which did not significantly increase after 12 months of age (P > 0.9999 for both average and maximum, pairwise Wilcoxon rank sum test, ART two‐way ANOVA, Fig. 6 O and P). Strain comparison showed that both the average and maximum Ca²⁺ responses recorded from supporting cells were significantly higher in 6N compared to C3H mice at 18–24 months of age (P < 0.0001 for both comparisons), but not at P7–P8 (P = 0.1647 and P > 0.9999, respectively) (Fig. 6 N and O). These data suggest that the increase in ATP sensitivity is linked to the more severe hearing loss phenotype of the 6N strain compared to the C3H, indicating a possible role in the progression of age‐related hearing loss.

To further corroborate this conclusion, and to find whether the observed phenotype was linked to Cdh23^ahl allele, we performed Ca²⁺ imaging experiments in 6N‐Repaired mice, which, despite their better overall hearing sensitivity, share a similar low‐frequency hearing loss with the co‐isogenic 6N strain (Jeng, Ceriani et al., 2020). In 6N‐Repaired mice, ATP‐induced Ca²⁺ responses in the supporting cells also increased with age, with a similar progression as the one observed in 6N mice (Fig. 6 J–N) since no significant difference was found when comparing both their average and maximum Ca²⁺ responses (P = 0.7176 and P > 0.9999, respectively: Tukey's post hoc test, ART two‐way ANOVA, Fig. 6 O and P). Beside the increase in size, a prominent feature of Ca²⁺ responses to prolonged ATP application was the presence of both sustained Ca²⁺ oscillations and repetitive Ca²⁺ spikes in several supporting cells. The frequency of these Ca²⁺ oscillations in the supporting cells increased with age in both 6N and 6N‐Repaired adult mice (Fig. 6 Q). Conversely, this increase in oscillatory behaviour was much less prominent in the C3H strain (Fig. 6 Q).

To measure the sensitivity of Ca²⁺ responses to extracellular ATP, we applied different concentrations of the purinergic agonist onto the supporting cells of 18‐ to 24‐month‐old 6N mice (Fig. 7). Calcium oscillations were visible in supporting cells in the presence of 30–300 nm ATP, while Ca²⁺ responses tended to plateau after an initial peak at a higher concentration (1 μm, Fig. 7 A and B). The concentration of ATP able to trigger 50% of the average and maximum Ca²⁺ responses was around 60–70 nm (Fig. 7 C and D), which is reasonably close to that previously reported for the supporting cells of the LER of pre‐hearing mice (23 nm, Gale et al., 2004) and that present in extracellular cochlear fluids at rest (10–20 nm, Muñoz et al., 1995; Muñoz et al., 2001).

ATP‐induced Ca²⁺ responses depend on Ca²⁺ released from intracellular stores

In the immature cochlea, ATP‐dependent activation of G‐coupled P2Y receptors promotes the release of Ca²⁺ from intracellular stores (Ceriani et al., 2016; Gale et al., 2004; Piazza et al., 2007). We therefore tested whether ATP‐induced Ca²⁺ signals in supporting cells from the aged cochlea depend on a similar mechanism. Thapsigargin is a known inhibitor of sarcoplasmic/endoplasmic reticulum Ca²⁺‐ATPase (SERCA) (Thastrup, 1990), thus preventing the refilling of Ca²⁺ into the stores. We found that incubation of the cochlear preparation with 2 μm thapsigargin completely abolished the ATP‐dependent Ca²⁺ responses in the supporting cells of aged mice (P < 0.0001, Wilcoxon signed‐rank test, Fig. 8 A and B), as also observed in the pre‐hearing cochlea (P < 0.0001, Fig. 8 B, see also Babola et al., 2020; Piazza et al., 2007). This suggests that the Ca²⁺ responses observed in the aged cochlea using sub‐micromolar concentrations of ATP are mainly due to the activation of metabotropic purinergic (P2Y) receptors functionally linked to intracellular Ca²⁺ stores.

Calcium responses mediated by P2Y₁ receptors do not change with age

Having established that P2Y are the primary purinergic receptors underlying the increase in Ca²⁺ responses in the supporting cells of the aged cochlea, we determined whether all three P2Y receptor subtypes identified in the pre‐hearing cochlea (P2Y₂ and P2Y₄: Huang et al., 2010; Piazza et al., 2007; P2Y₁: Babola et al., 2020; see also Figs 2, 3, 4) were involved.

To investigate the possible contribution of P2Y₁ receptors, we used adenosine 5′‐diphosphate (ADP), which is a known agonist for the P2Y₁, P2Y₁₂ and P2Y₁₃ receptors (Waldo et al., 2004, von Kügelgen et al., 2016). We found that 1 μm ADP elicited large Ca²⁺ increases in the supporting cells of 6N mice at all ages tested, including in the young adult cochlea (Fig. 9 A–C), an age when they could not be triggered by the application of physiological concentrations of ATP (Fig. 6). Both the average and maximum of the Ca²⁺ responses measured from supporting cells in the presence of ADP were found to be not significantly different between young adult (1–2 months) and aged mice (average Ca²⁺ response: P = 0.6912; maximum Ca²⁺ response: P = 0.9999, Wilcoxon rank sum test, Kruskal Wallis test, Fig. 9 D–F). The extracellular application of the specific P2Y₁ receptor antagonist MRS2500 (Hechler et al., 2006) fully blocked ADP‐induced Ca²⁺ responses, further confirming that P2Y₁ receptors underlie these (maximum Ca²⁺ response: P < 0.0001, Wilcoxon signed‐rank test, Fig. 9 G and H). Overall, as the responses to ADP showed no age‐related relationship after the onset of hearing, these results indicate that P2Y₁ receptors do not contribute to the re‐appearance of ATP‐induced Ca²⁺ responses seen in the aged cochlea.

P2Y₂‐ and P2Y₄‐mediated Ca²⁺ responses increase in aged mice

The application of 300 nm UTP, a selective agonist of P2Y₂ and P2Y₄ receptors that mobilises Ca²⁺ from intracellular stores (Piazza et al., 2007), triggered large Ca²⁺ responses in both pre‐hearing and aged supporting cells from the 6N mouse strain, with some cells displaying prominent Ca²⁺ oscillations (Fig. 10 A–C). These UTP‐induced responses followed a similar age‐dependent time course as those induced by ATP (Fig. 6 O and P), with a reduction in amplitude at 1–2 months of age and a subsequent increase in 18‐ to 24‐month‐old mice. A similar effect was also observed in ageing C3H mice (Fig. 10 D–F), although UTP‐induced Ca²⁺ responses in 18‐ to 24‐month‐old mice were much smaller compared to age‐matched 6N (average and maximum Ca²⁺ response: P < 0.0001, pairwise Wilcoxon rank‐sum test, ART two‐way ANOVA, Fig. 10 G–I). These results confirm that purinergic‐induced Ca²⁺ signals also reappear in the ageing cochlea from C3H mice, albeit at later ages compared to the 6N strain, as also suggested when ATP was used (Fig. 6). To investigate whether both P2Y₂ and P2Y₄ were involved in the UTP‐induced Ca²⁺ signals in the aged cochlea, we performed some experiments using the selective P2Y₂ antagonist AR‐C 118925XX (Rafehi et al., 2017) and P2Y₄ agonist MRS4062 (Maruoka et al., 2011). We found that AR‐C 118925XX abolished the UTP‐induced Ca²⁺ responses (Fig. 11 A and B), while MRS4062 produced similar Ca²⁺ signals to those elicited by UTP (Fig. 11 C and D). These findings confirm the involvement of both P2Y₂ and P2Y₄ receptors in the generation of Ca²⁺ responses in the supporting cells of the aged cochlea.

Metabotropic purinergic receptors in the supporting cells of the developing, adult and aged mouse cochlear inner sulcus

Metabotropic P2Y receptors show widespread expression in the cochlea, both in sensory and supporting cells (Huang et al., 2010; Köles et al., 2019). During pre‐hearing stages of cochlear development, the P2Y receptors expressed in the supporting cells of the GER (Babola et al., 2021; Huang et al., 2010) play a crucial role in the propagation of slow spontaneous intercellular Ca²⁺ waves across the immature epithelium (Babola et al., 2020; Johnson et al., 2017; Tritsch & Bergles, 2010; Tritsch et al., 2007). This ATP‐dependent Ca²⁺ signalling is mediated by P2Y receptors and has been shown to regulate not only the development and fine tuning of sensory hair cells and their innervation (Ceriani et al., 2019; Johnson et al., 2017), but also the refinement of tonotopic maps in the brain (Babola et al., 2018; Babola et al., 2021; Kersbergen et al., 2022). After the onset of hearing, the cochlea reaches functional maturity, and the inner sulcus (IS) replaces the GER. Supporting cells in the IS of the mouse cochlea no longer elicit the spontaneous Ca²⁺ waves and the slow ATP‐induced currents characteristic of P2Y receptor activation (Sirko et al., 2019). Indeed, previous work has shown that in rats the expression of P2Y receptors in the supporting cells of the inner sulcus is reduced compared to that in the immature GER (Huang et al., 2010).

Previous investigations have shown that P2Y‐mediated Ca²⁺ responses in cochlear supporting cells are elicited by micromolar concentrations of ATP (e.g. Horvath et al., 2016; Lahne & Gale, 2008; Rabbitt & Holman, 2021; Sirko et al., 2019; Tritsch & Bergles 2010). We showed that Ca²⁺ responses can also be elicited by nanomolar ATP, which approaches the concentration present in extracellular cochlear fluids in vivo (Muñoz et al., 1995; Muñoz et al., 2001). Although the contribution of P2Y receptors to ATP‐mediated Ca²⁺ responses is well established, the nature of the receptor subtype is still debated. Classical studies have provided evidence showing the involvement of P2Y₂ and P2Y₄ receptors (Huang et al., 2010; Piazza et al., 2007). However, more recent findings have shown that P2Y₁ receptors are the main subtype driving purinergic‐mediated spontaneous Ca²⁺ responses in the supporting cells of the GER in the prehearing cochlea (Babola et al., 2021).

Our data show that Ca²⁺ responses in the supporting cells nearby the IHCs of the pre‐hearing cochlea can be elicited not only by the P2Y₁ agonist ADP, but also by the P2Y₂ and P2Y₄ receptor agonist UTP. This finding indicates that all three purinergic subtypes may contribute to the Ca²⁺ responses in the supporting cells of the pre‐hearing cochlea (see also Babola et al., 2020; Huang et al., 2010). Since P2Y₁ affinity for ADP is 100 times that of ATP (von Kügelgen & Hoffmann, 2016), it is possible that ATP in the prehearing cochlea is readily converted into ADP by ectonucleotidases (O'Keefe et al., 2010). Alternatively, several members of the P2Y receptor family could combine to form hetero‐oligomers, altering the receptors properties, including their pharmacology and trafficking (Ecke et al., 2008; Nakata et al., 2010). It is thus possible that interactions between the different P2Y receptor subtypes could influence their response to agonists/antagonists. Although our Ca²⁺ imaging approach combined with purinergic pharmacology directly implicates P2Y₂ and P2Y₄ in the reappearance of ATP‐induced Ca²⁺ signalling in the aged cochlea, it is very likely that other mechanisms such as cytoplasmic Ca²⁺ buffering or Ca²⁺ clearance may change with age. Although very little is known about these additional mechanisms in the aged mouse cochlea, a dysregulation of neuronal Ca²⁺ dynamics has previously been associated with ageing of the nervous system (Mattson et al., 2018). Moreover, it remains to be determined whether other mechanisms known to be involved in the propagation of intercellular Ca²⁺ responses in the immature cochlea, such as diffusion of second messengers (e.g. IP₃) between supporting cells through gap junction channels (Anselmi et al., 2008; Ceriani et al., 2016), are also changed in the ageing cochlea.

Role of purinergic receptors in the cochlea

Several lines of evidence indicate that one of the most common causes of age‐related hearing loss in mice is the change in the morphology and function of the sensory hair cells and their innervation within the cochlea (Lauer et al., 2012; Liberman, 2017; Jeng et al., 2021; Jeng, Ceriani et al., 2020; Jeng, Johnson et al., 2020). In this study we showed that by 12 months of age, the supporting cells located in the low‐frequency cochlear region (9–12 kHz) of 6N and the co‐isogenic strain 6N‐Repaired mice upregulate P2Y₂ and P2Y₄ receptors. Both 6N and 6N‐Repaired mice exhibit a comparable progressive low‐frequency hearing loss for frequencies below 18 kHz (Jeng, Ceriani et al., 2020). Different from the 6N and 6N‐Repaired mice, the C3H strain retains very good hearing thresholds across the frequency range even at older ages, and their hair cells and associated innervation have normal structure and function until at least 15–18 months of age (Jeng et al., 2021; Jeng, Ceriani et al., 2020; Jeng, Johnson et al., 2020). The supporting cells present in the cochlea of C3H mice showed minimal age‐related changes in the expression of P2Y receptors and mediated Ca²⁺ signalling. These findings indicate that, like hair cells (Jeng et al., 2021; Jeng, Ceriani et al., 2020; Jeng, Johnson et al., 2020), the supporting cells undergo specific functional changes during ageing that signify the level of progressive loss of auditory function.

Supporting cells can act as phagocytes by engulfing and removing dying hair cells (Anttonen et al., 2014; Monzack et al., 2015). Moreover, the release of ATP from the hair cells has previously been proposed as the ‘eat me’ signal that activates their removal by nearby supporting cells (Bird et al., 2010). Therefore, the increased activity of purinergic receptors in the aged cochlea could participate in control of hair cell death, for example during immune responses, which may be associated with age‐related hearing loss (Köles et al., 2019; Verschuur et al., 2014). This mechanism, if present, could be driven by P2Y₂ receptors since they have been linked to immune cell migration to the site of inflammation, with ATP released from damaged cells acting as an attracting chemotactic signal (Chen et al., 2006). Interestingly, the ATP‐induced Ca²⁺ responses in the supporting cells of aged mice resemble those present in the developing cochlea (Piazza et al., 2007), in which ATP can be released following damage to hair cells (Gale et al., 2004; Lahne & Gale, 2008; Lahne & Gale, 2010; Piazza et al., 2007). However, there is very little or no loss of hair cells in the cochlear region investigated in this study (9–12 kHz), at least up to about 12 months of age (Jeng et al., 2021; Jeng, Johnson et al., 2020), a time when the purinergic signalling is already upregulated. Therefore, the involvement of the observed age‐related changes in P2Y receptors in cochlear immune responses seems unlikely.

Changes in P2Y‐mediated signalling in the aged cochlea could potentially act as a protective mechanism to limit or to avoid further damage caused by the several morphological and functional changes occurring in the sensory hair cells (e.g. Jeng et al., 2021; Jeng, Ceriani et al., 2020; Jeng, Johnson et al., 2020; Lauer et al., 2012), which may be perceived as cochlear insults. Indeed, it has been shown that the concentration of ATP in both endolymph and perilymph, and the induced Ca²⁺ signals in the supporting cells, can increase following pathological insults such as noise, hypoxia, or ischaemia (Chan & Rouse, 2016; Muñoz et al., 1995; Muñoz et al., 2001). This increase as been interpreted as ATP having a protecting role on the cochlea during insults by mediating adaptation of cochlear responses (Housley et al., 2013) and regulating cochlear mechanics (Skellet et al., 1997).

As mentioned above, there are remarkable similarities between the ATP‐induced Ca²⁺ responses in the supporting cells of aged mice and those present in the developing cochlea (Babola et al., 2020; Johnson et al., 2017; Tritsch & Bergles 2010), which have been suggested to be involved, in addition to controling cell death, in modulating gene expression (Ortolano et al., 2008). A similar process in which the cochlear tissue seemingly ‘recapitulates’ the developmental configuration has been seen in the hair cells, with the efferent system re‐forming direct axo‐somatic contacts with the IHCs in aged mice (Jeng et al., 2021; Lauer et al., 2012). The upregulation of P2Y‐mediated Ca²⁺ signalling in the supporting cells occurred after some of the age‐related changes in the hair cells, such as a reduction in their surface area and the size of K⁺ currents, which were already evident at 6 months of age (Jeng et al., 2021; Jeng, Johnson et al., 2020). In the future, it will be interesting to understand whether the age‐related changes in the efferent system and purinergic signalling are simply a consequence to the progressive deterioration of the mammalian cochlea with age, or instead an attempt to ‘recapitulate’ early developmental ages. The latter mechanisms, which are normally used to control the functional and structural remodelling of the maturing epithelium, could potentially be used to repair or limit the damage caused by the malfunctioning hair cells during ARHL.

Competing interests

The authors declare no conflict of interest.

Author contributions

All authors helped with the interpretation of the data and commenting on the manuscript. S.A.H. and F.C. performed the experiments. S.A.H., J.‐Y. J., D.J., W.M., and F.C. contributed to the writing of the paper. W.M. and F.C. conceived and coordinated the study. All authors approved the final version of the manuscript. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

This work was supported by the BBSRC (BB/V006681/1 to F.C and W.M). S.A.H. is funded by a PhD studentship from the Royal National Institute for Deaf People (RNID; S58 to D.J and W.M.). J‐Y.J. was supported by the RNID & Dunhill Medical Trust Fellowship (PA28).