Slight and hidden hearing loss differentially affect short- and long-term memory in young rats

Abstract

Mild forms of hearing loss (HL) have been linked to cognitive impairments in children, yet the neurobiological mechanisms underlying this connection remain unclear. Existing research using animal models mostly focuses on more severe levels of HL or investigates only limited aspects of cognition. To gain a broader understanding of how slight/hidden HL affects cognitive behaviors, we induced HL in 4-week-old Wistar rats through noise exposure. Auditory brainstem response measurements confirmed slight and hidden HL, but this auditory impairment did not alter the density of inner hair cells or their synapses with the spiral ganglion (primary auditory) neurons. Both short- and long- term memory formation were tested using the object location, novel object recognition, and social recognition task. Behaviorally, rats with slight/hidden HL performed better than normal hearing (NH) rats during short-term cognition tests. However, long-term memory was impaired in rats with slight/hidden HL when compared to NH controls. Slight/hidden HL also did not consistently affect (social) exploration. In conclusion, this study demonstrates that slight and hidden HL differentially affect short- and long-term cognitive processes in an animal model of early (noise-induced) HL, without affecting (social) exploration. These results suggest a nuanced relationship between slight and hidden HL and both short- and long-term memory formation, underscoring the importance of broader cognitive phenotyping and further investigation into the neurobiological structures linking hearing impairment with cognitive function.

Introduction

In recent years, a growing number of studies have examined the relationship between hearing loss (HL) and cognitive dysfunction (Bisogno et al., 2021). Despite the high prevalence of mild HL in children (Elbeltagy, 2020; Moore et al., 2020), research linking mild HL to cognitive impairments remains limited. Nevertheless, it is becoming increasingly evident that even mild HL can impact social and cognitive domains and be accompanied by developmental delays (Bess et al., 1998; Tharpe, 2008; Kamerer et al., 2019; Elbeltagy, 2020; le Clercq et al., 2020; Moore et al., 2020; Ji et al., 2023). These findings suggest possible effects on the underlying brain structures; however, the exact mechanisms remain unclear. Given the high prevalence of mild HL, there is a pressing need for further research to address this knowledge gap and understand its potential impact on cognitive development and the related neurobiological structures in children and young adults.

Numerous animal models have been developed to investigate the link between early-onset HL and cognitive performance, frequently using rodents, and especially rats and mice, due to their well-established behavioral paradigms and similarities in auditory systems to humans. Studies examining auditory deprivation in animals have demonstrated that HL impacts various cognitive functions, including spatial and object recognition memory (Liu et al., 2016, 2018; Park et al., 2016; Shukla et al., 2019; Manohar et al., 2020). Even speech-in-noise detection, a cognitively demanding task for humans, has been shown to be impaired in rats with severe HL (Riley et al., 2021).

While existing literature has provided valuable insight, the effects of early mild HL on cognition remain underexplored for several reasons. First, most research focuses on moderate to severe HL or investigates the effects of HL later in development or adulthood. Second, studies often examine specific cognitive domains, such as spatial memory or object recognition, in isolation, while other cognitive functions, like social recognition (SR) memory, receive far less attention. Additionally, many studies do not compare different intertrial intervals (ITIs) during cognitive tasks, leaving open the question of whether HL affects both short- and long-term memory formation. Thus, the effects of mild HL on the complexity of cognitive processes, which involve distinct neurobiological structures (Poldrack, 2010), remain underexplored.

To address these knowledge gaps and motivated by our previous findings that even mild HL in young rats can impair recognition memory in a novel object recognition (NOR) task (Jagersma et al., 2024), we assessed both short- and long-term cognitive performance in young rats with very mild—slight and hidden—HL. HL was induced through noise exposure and quantified functionally using auditory brainstem response (ABR) measurements and morphologically through immunohistochemical quantification of inner hair cells (IHCs) and their ribbon synapses. We employed a comprehensive test battery assessing object location (OL), NOR, and SR. Tests were conducted twice, with ITIs of 1 and 8 h, to evaluate both short- and long-term cognitive performance. In doing so, we aimed to understand the impact of very mild HL on short- and long-term spatial memory, recognition memory, and SR. Our results identify a complex relationship between very mild HL and both short- and long-term memory formation and emphasize the need for expanded cognitive profiling and further exploration of the neurobiological structures connecting hearing impairment with cognitive function.

Methods

Subjects

Twenty-three experimental male pups were born from six pregnant Wistar rats (Envigo Harlan, Horst, the Netherlands). Pregnant rats were delivered to our facility at 15 days of gestation. A separate pregnant Wistar rat was ordered to provide two stimulus animals, which were only used for the social preference and recognition test. After weaning, pups were housed in groups of three to four littermates per cage, mixing sham and noise-exposed rats to exclude maternal care effects. Housing and care were in accordance with Annex III of the Directive 2010/63/EU. Rats received ad libitum access to standard food, water, and cage enrichments (including nesting material and a cardboard tube). The animals were subjected to a 12 h:12 h light-dark cycle, with lights turning on at 07:00 a.m. and turning off at 07.00 p.m. Rats that did not exhibit appropriate hearing levels for either experimental group were excluded from subsequent behavioral testing. Similarly, for memory assessments, rats that explored objects or animals for less than 5 s or did not explore all stimuli were excluded from analyses. Animals that were excluded because of any of these criteria are mentioned in the corresponding section of the results. The experimental procedures were approved by and adhered to the guidelines and regulations of the University Medical Center Groningen and were conducted in agreement with the Law on Animal Experiments of The Netherlands. An overview of the procedures is shown in Fig. 1.

Fig. 1.

Overview of experimental procedures. The timeline illustrates the sequence of procedures experimental animals underwent. More detailed explanations of the individual tests are available in the corresponding sections of the methods.

Auditory brainstem response measurements

ABRs were measured using procedures and equipment described previously (Reijntjes et al., 2019). Animals were anesthetized using subcutaneous injections of 75 mg/kg ketamine and 0.25 mg/kg dexmedetomidine and then placed on a heating pad to maintain stable body temperature during recordings. Electrodes were placed subdermally on the mastoids below both ears and over the vertex on the forehead. ABRs were recorded from the right ear by presenting 2, 4, 8, 16, and 32 kHz pure tones and click stimuli at intensities ranging from 20 to 90 dB sound pressure level (SPL) in 5 dB increments. A total of 256 recordings were averaged to create the mean response. After ABR measurements were completed, atipamezole (1 mg/kg) was administered subcutaneously to expedite the recovery from anesthesia. Animals were returned to their home cage and had at least 5 days to recover before undergoing behavioral testing.

Absolute thresholds were determined manually as the first intensity at which the peak of wave I was detectable and then consistently present at all subsequent intensities. Threshold shifts were calculated by subtracting the absolute threshold measured at baseline from the absolute threshold measured either 2 or 7 weeks after noise exposure. Wave I amplitudes were measured using a custom-designed script (in R studio) that identified the maximum amplitude within a time window containing wave I.

Noise exposure

Four-week-old rats were anesthetized using subcutaneous injections of 75 mg/kg ketamine and 0.25 mg/kg dexmedetomidine and underwent (sham) noise exposure in an acoustic chamber. The 12 noise-exposed rats were subjected to continuous broadband white noise of 100 dB SPL for 2 h using a free field speaker (AVSL Piezo tweeter) connected to an amplifier (Dynavox ET-100). The 11 sham-exposed animals underwent the same handling procedures but were not exposed to noise inside the acoustic chamber. Following (sham) noise exposure, animals recovered in their home cage.

Behavioral apparatus and stimuli

A square arena (l/w/h: 100 × 100 × 50 cm) was used during the OL and the NOR tests. For the social preference and recognition test, a three-chamber arena was used (l/w/h: 120 × 80 × 40 cm) and consisted of three identical connected chambers (l/w/h: 40 × 80 × 40 cm). Two wired cups were placed in the outermost chambers to hold the stimulus animals. Visual cues on the walls of both arenas were used to facilitate spatial orientation. Common household and laboratory objects were used to test the cognitive performance of the rats (e.g. several types of bottles, soap dispensers, pepper grinders, spray bottles, and drinking cups). Objects were of similar volume but contained several distinguishable features such as material, texture, and shape. Although no inherent preference for specific objects was observed during the preliminary tests, the experiments still utilized a blocked design for presenting different sets of objects.

Object location

Spatial memory was tested in 5-week-old rats using the OL task. Animals were first habituated to the empty testing arena for 5 min. For the familiarization trial (FT), animals were reintroduced to the arena, which now contained two identical objects located in two corners of the arena, and were able to explore for 3 min. After the FT, animals were placed back into their home cage for an ITI of either 1 or 8 h. For the ITI of 1 h, home cages remained in a holding room close to the testing room. For the ITI of 8 h, home cages were returned to the housing room and then returned to the testing room. For the test trial (TT), animals were returned to the arena, in which one of the original objects was moved to a new location, and they were able to explore for 3 min.

Novel object recognition

Object recognition memory was tested in 8-week-old rats using the NOR test. Animals were first habituated to the testing arena for 5 min. For the FT, animals were reintroduced to the arena, which now contained two identical objects, and were able to explore for 3 min. After the FT, animals were placed back into their home cage for an ITI of either 1 or 8 h. For the ITI of 1 h home cages remained in a holding room close to the testing room. For the ITI of 8 h, home cages were returned to the housing room and then returned to the testing room. For the TT, animals were returned to the arena, in which one of the objects from the familiarization was replaced by a novel object, and observed for 3 min.

Social preference and recognition

Social preference and SR memory were tested in 10-week-old rats. Animals were first habituated to the testing arena, which contained two empty wire cups, for 10 min. For the FT, animals were reintroduced to the arena, which now contained an unfamiliar stimulus rat of the same age in one of the wire cups, and were able to explore for 10 min. This FT also served as the testing trial for the social preference task. After the FT, animals were placed back into their home cage for an ITI of either 1 or 8 h. For the ITI of 1 h home cages remained in a holding room close to the testing room. For the ITI of 8 h, home cages were returned to the housing room and then returned to the testing room. For the TT, animals were returned to the arena, in which one of the wire cups contained the now familiar rat from the FT and the other wire cup contained an unfamiliar rat of the same age. Animals were able to explore for 10 min.

Behavioral scoring

All recorded videos were scored by an observer blinded to the hearing status of the animals using BORIS (Friard and Gamba, 2016). Exploration time of objects or stimulus animals in the OL, NOR, and SR were scored in both the training and the TTs. A discrimination index (d2) [(time of exploration of the new object/animal during the TT – time of exploration of the familiar object/animal during the TT)/the total exploration time during the TT] was used to assess memory formation according to Akkerman et al. (2012).

Tissue isolation, immunofluorescence, and imaging

Eleven-week-old rats were anesthetized using isoflurane and then sacrificed by decapitation. Inner ears were dissected as described previously (Reijntjes et al., 2021) and postfixed in 4% paraformaldehyde for 2 h at 4 °C. Ears were stored in 0.01 M PBS at 4 °C until further processing. Organs of Corti were then isolated from the inner ears, immunofluorescently stained using an anti-CTBP2 primary antibody (IgG1, BD Biosciences, 612044) and anti-GluR2 primary antibody (IgG2a, Millipore, MAB 397) followed by goat anti-Mouse IgG1 Alexa Fluor 488 and goat anti-Mouse IgG2a Alexa Fluor 647 secondary antibodies (ThermoFisher, A-11001 and A-21241), and mounted as described previously (Reijntjes et al., 2021). Low-magnification micrographs were obtained using a Leica DM4000B fluorescent microscope. Cochlear frequency maps were generated for each organ of Corti using a freely available ImageJ plug-in from the Eaton-Peabody Laboratories Histology Core and the previously published place-frequency map (Müller, 1991; Reijntjes et al., 2021). High-magnification confocal micrographs were collected using a Leica SP8 confocal microscope. The region containing the entire row of IHCs and synapses was scanned at 600 Hz and a zoom of 1. 3D reconstructions were generated and used to count the number of IHCs and synapses using Imaris 6.4 software (Bitplane) as described previously (Reijntjes et al., 2021).

Statistical analyses

Analyses were performed in Prism 8 (GraphPad, La Jolla, California, USA). ABR measurements and synapse counts were analyzed with either a two-way ANOVA or a mixed-effects model approach followed by an appropriate post hoc test. Exploratory behaviors were compared between the sham- and noise-exposed groups using either independent t-tests or Mann–Whitney U tests. Discrimination indices were analyzed using one-sample t-tests against zero for sham- and noise- exposed groups. Specific statistical tests are also provided in the results section. All data are presented as mean ± standard error unless stated otherwise. For all tests, a P value lower than 0.05 was considered significant. Asterisks in figures corresponded to the following P values: *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, and ****P ≤ 0.0001.

Results

Noise-induced hearing loss

ABRs were measured 1 week prior and 2 and 7 weeks after (sham) noise exposure (when rats were respectively 3, 6, or 11 weeks old) to confirm comparable baseline hearing (at 3 weeks of age) and assess changes in hearing following (sham) noise exposure (at 6 and 11 weeks of age). Before sham/noise exposure (baseline), both experimental groups had comparable absolute thresholds for all tested stimuli; two-way ANOVA, [F(1,20) = 0.0494, P = 0.8264, Supplementary Figure 1a, Supplemental Digital Content 1, https://links.lww.com/BPHARM/A122]. Absolute thresholds of sham-exposed animals did not significantly differ when comparing thresholds measured before sham exposure to thresholds measured 2 and 7 weeks after sham exposure [two-way ANOVA, F(1.585, 15.85) = 0.9301, P = 0.3938, Supplementary Figure 2a, Supplemental Digital Content 1, https://links.lww.com/BPHARM/A122]. Therefore, threshold shifts of sham-exposed animals (Fig. 1, black line) are plotted as the average of the threshold shifts observed at 2 weeks and 7 weeks after sham exposure. As expected, absolute thresholds of noise-exposed animals differed significantly when comparing thresholds measured before noise exposure to thresholds measured 2 and 7 weeks after the noise exposure [two-way ANOVA, F(1.846, 20.30) = 47.14, P < 0.0001, Supplementary Figure 3a, Supplemental Digital Content 1, https://links.lww.com/BPHARM/A122]. When comparing threshold shifts between sham- and noise-exposed animals (Fig. 2a), two-way ANOVA indicates that there was a significant effect of noise exposure [F(2,30) = 27.68, P < 0.0001]. Compared to sham-exposed rats, noise-exposed rats showed threshold shifts of approximately 20 dB HL in response to pure tone stimuli of 16 kHz 2 weeks after noising (P = <0.0001). Seven weeks after noise exposure, absolute thresholds measured in response to 16 and 32 kHz pure tones were both elevated compared to sham-exposed rats (16 kHz: P = <0.0001, 32 kHz: P = 0.0002). One noise-exposed rat showed insufficient threshold shifts and was, therefore, excluded from the current analysis and subsequent experiments.

Fig. 2.

Noise exposure induces slight and hidden hearing loss. (a) Noise-exposed rats exhibited significantly elevated thresholds compared to sham-exposed rats at higher frequencies both 2 (open gray circles) and 7 weeks (filled gray circles) after exposure. Since there was no significant difference in thresholds between 2 and 7 weeks after sham exposure, the thresholds for sham-exposed rats are presented as an average (filled black circles). (b) Noise-exposed rats showed significantly reduced wave I amplitudes (measured at 80 dB sound pressure level) compared to sham-exposed rats across stimuli at both 2 and 7 weeks after noise. Again, since there was no significant difference in wave I amplitudes between 2 and 7 weeks after sham exposure, the amplitudes for sham-exposed rats are presented as an average. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, and ****P ≤ 0.0001.

We also examined ABR wave I amplitudes at 80 dB SPL to examine changes in suprathreshold responses. Before sham/noise exposure (baseline), both experimental groups had comparable wave I amplitudes at all tested stimuli [mixed-effects model, F(1,20) = 0.7127, P = 0.4085, Supplementary Figure 1b, Supplemental Digital Content 1, https://links.lww.com/BPHARM/A122]. To determine the effect of the development of the auditory system on the ABR results, we compared wave I amplitudes measured at baseline and both 2 and 7 weeks after sham/noise exposure (Supplementary Figures 2 and 3, Supplemental Digital Content 1, https://links.lww.com/BPHARM/A122). For both the sham and noise-exposed animals wave I amplitudes changed throughout development [sham: two-way ANOVA, F(1.384, 13.84) = 21.03, P = 0.0002; noised: mixed-effects model, F(1.384, 13.84) = 36.84, P < 0.0001, Supplementary Figures 2b and 3b, Supplemental Digital Content 1, https://links.lww.com/BPHARM/A122]. For the sham-exposed animals, wave I amplitudes measured at both 2 and 7 weeks after sham exposure were reduced compared to baseline measurements at 2, 4, 8, and 16 kHz (2 weeks post sham: 2 kHz: P = 0.0192, 4 kHz: P = 0.0202, 8 kHz: P = 0.0469, 16 kHz: P = 0.0018, 7 weeks post sham: 2 kHz: P = 0.0167, 4 kHz: P = 0.0096, 8 kHz: P = 0.0092, 16 kHz: P = 0.0006) but remained stable between 2 and 7 weeks after sham exposure. Therefore, the average wave I amplitudes at 2 and 7 weeks after sham exposure are shown in Fig. 2b.

Two weeks after noise exposure, the noise-exposed rats showed a significant decrease in wave I amplitudes at 8, 16, and 32 kHz compared to baseline measurements (8 kHz: P = 0.0130, 16 kHz: P = 0.0013, 32 kHz: P = 0.0005). Seven weeks after noise exposure, wave I amplitudes were significantly reduced across all tested stimuli compared to baseline measurements (click: P = 0.0012, 2 kHz: P = 0.0042, 4 kHz: P = 0.0016, 8 kHz: P = 0.0008, 16 kHz: P = 0.0003, 32 kHz: P = 0.0028). Furthermore, wave I amplitudes measured 7 weeks after noise exposure were significantly reduced compared to amplitudes measured 2 weeks after noising, indicating that hearing impairment continued to worsen over time (click: P = 0.0011, 2 kHz: P < 0.0001, 4 kHz: P = 0.0005, 8 kHz: P = 0.0002, 16 kHz: P = 0.0021).

When comparing wave I amplitudes at 80 dB between the sham- and noise-exposed groups we confirmed that noise exposure had a significant effect on wave I amplitudes [mixed-effects model, F(2.30) = 18.50, P < 0.0001, Fig. 2b]. Two weeks after noise exposure, the noise-exposed group showed significantly reduced wave I amplitudes compared to the sham-exposed group at 16 and 32 kHz (P = 0.0353 and P = 0.0001, respectively) and (unexpectedly) greater amplitudes at 2 kHz (P = 0.0479). However, by 7 weeks after noise exposure, wave I amplitudes at nearly all tested stimuli were significantly decreased in noise-exposed compared to sham-exposed groups. (clicks: P < 0.0001, 4 kHz: P = 0.0036, 8 kHz: P = 0.004, 16 kHz: P < 0.0001, 32 kHz: P < 0.0001). Furthermore, wave I amplitudes continued to decline over time following noise exposure, with most stimuli showing significantly reduced amplitudes at seven compared to 2 weeks after noise exposure (clicks: P < 0.0001, 2 kHz: P < 0.0001, 4 kHz: P < 0.0001, 8 kHz: P < 0.0001, 16 kHz: P = 0.0003).

Together, these results show that the noise exposure used in this study results in slight HL (thresholds shifts of ≤25 dB) and hidden HL (reduced wave I amplitudes without threshold shifts; Kohrman et al., 2020). This paradigm allowed us to investigate the effects of slight/hidden HL on cognitive function by comparing rats with normal hearing (NH, sham-exposed rats) to those with slight/hidden HL (noise-exposed rats).

Because hidden HL has been associated with loss of the ribbon synapses connecting the IHCs and spiral ganglion (primary auditory) neurons (Kujawa and Liberman, 2009), we compared the number of ribbon synapses per IHC in a subset of sham- versus noise-exposed rats (2–5 and 3–5 animals, respectively). Analysis was performed at various tonotopic locations (2, 8, 16, and 32 kHz) in isolated organs of Corti immunofluorescently stained with anti-CTBP2 as well as anti-GluR2 to visualize the IHC nuclei and presynaptic ribbons as well as glutamate receptor containing postsynapses (Fig. 3a). However, glutamate receptor immunostaining was not reliably detected. Additionally, due to dissection artifacts, not all tonotopic locations from each rat could be included in the statistical analysis. There were no significant differences in the number of ribbon synapses per IHC between sham- and noise-exposed rats across examined tonotopic locations [mixed-effects model: synapses/IHC: F(1,25) = 0.6252, P = 0.4365; Fig. 3b].

Fig. 3.

Noise exposure resulting in slight and hidden hearing loss is not associated with loss of IHCs or ribbon synapses. (a) Fluorescent micrographs of CTBP2-immunostained IHCs and ribbon synapses (green) in organs of Corti isolated (16 kHz region) from sham-exposed rats with normal hearing (NH) and noise-exposed rats with slight/hidden hearing loss (HL). (b) Ribbon synapse density (synapses/IHC) was not significantly different between sham- and noise-exposed animals at any of the examined tonotopic regions. IHCs, inner hair cells.

Object location memory

Short- and long-term location memory was tested in 5-week-old rats using the OL task (Fig. 4). Short-term memory, in which the TT was separated from the FT by 1 h, was examined in 9 rats with NH and 11 rats with slight/hidden HL. Two NH rats showed insufficient exploration of the objects during the TTs and were, therefore, excluded from subsequent analyses. Both groups of animals explored the objects for equal amounts of time during each trial [FT: t(20) = 0.4199, P = 0.6790, Fig. 4a; TT: U = 48, P = 0.9408, Fig. 4b]. When examining short-term OL memory, both groups of animals showed discrimination indices (Fig. 4c) that did not differ from zero [NH: t(8) = 0.0528, P = 0.9592; HL: t(10) = 0.7444, P = 0.4738].

Fig. 4.

Rats with slight and hidden hearing loss show no short- and long-term object location (OL) in contrast to mice with normal hearing (NH). (a and b) In the short-term OL task, total exploration time did not differ significantly between NH rats and rats with slight/hidden hearing loss (HL) during the familiarization trial (FT, a) or the test trial (TT, b). (c) There was no significant deviation from zero in the discrimination indices in either the NH or HL group. (d and e) In the long-term OL task, total exploration time also did not differ significantly between NH rats and rats with slight/hidden HL during either the FT (d) or the TT (e). (f) The discrimination index showed a significant positive deviation from zero in rats with NH. There was no significant deviation from zero in the discrimination index in rats with HL. *P ≤ 0.05.

Long-term OL memory, in which the TT was separated from the FT by 8 h, was examined in 10 animals per group. In these experiments, one animal per group explored one of the objects for less than 5 s during the TT and was, therefore, excluded from subsequent analyses. Both groups of animals again explored the objects for equal amounts of time during both trials [FT: t(20) = 0.1393, P = 0.8906; Fig. 4d; TT: t(18) = 0.3890, P = 0.7019; Fig. 4e]. However, for these experiments examining long-term OL memory (Fig. 4f), NH animals had a positive discrimination index [t(9) = 2.303, P = 0.0468], whereas the discrimination index of animals with slight/hidden HL did not significantly differ from zero [t(9) = 1.459, P = 0.1785].

Object recognition memory

Short- and long-term object recognition memory was tested in 5-week-old rats using the NOR task (Fig. 5). Both short- and long-term memory, in which the TT was separated from the FT by 1 or 8 h, respectively, were examined in 10 rats with NH and 11 rats with slight/hidden HL. Two rats were excluded from analyses: one rat with NH failed to explore both objects during the FT of the short-term NOR test and a second rat with NH showed insufficient exploration during the FT of the long-term NOR test.

Fig. 5.

Rats with slight and hidden hearing loss show short-term but not long-term novel object recognition (NOR). (a and b) In the short-term NOR task, total exploration time did not differ between normal hearing rats (NH) and rats with slight and hidden hearing loss (HL) during either the familiarization trial (FT, a) or the test trial (TT, b). (c) The discrimination index showed a significant positive deviation from zero in rats with HL. There was no significant deviation from zero in the discrimination index in rats with NH. (d) In the long-term NOR task, rats with HL showed significantly less total exploration time compared to rats with NH during the FT. (e) However, there was no significant difference in total exploration time between groups during the TT. (f) There was no significant deviation from zero in the discrimination indices in either the NH or HL group. *P ≤ 0.05 and **P ≤ 0.01.

During the short-term NOR test, both groups of animals explored the objects for equal amounts of time during each trial [FT: t(19) = 2.009, P = 0.0590, Fig. 5a; TT: t(19) = 1.509, P = 0.1477, Fig. 5b]. In these experiments examining short-term NOR memory (Fig. 5c), the discrimination index of animals with NH did not differ significantly from zero [t(9) = 1.248, P = 0.2435], whereas rats with slight/hidden HL had a significantly positive discrimination index [t(10) = 4.425, P = 0.0013].

During the long-term NOR test, rats with slight/hidden HL showed less exploration than NH rats during the FT [t(20) = 2.174, P = 0.0419, Fig. 5d], but this effect disappeared during the TT in which both groups of animals explored the objects for equal amounts of time [t(19) = 0.6096, P = 0.5493, Fig. 5e]. In these experiments examining long-term NOR memory (Fig. 5f), neither group of animals showed discrimination indices that deviated significantly from zero [NH: t(9) = 1.792, P = 0.1067; slight/hidden HL: t(10) = 1.535, P = 0.1559].

Social preference

Social preference (Fig. 6) was tested in 10-week-old rats using the FTs of the SR test. The data collected during the FTs of both the short- and long-term SR tests were averaged and used for analysis. A total of 11 animals per group were included. Both groups of animals explored for equal amounts of time [t(20) = 0.2784, P = 0.7836, Fig. 6a]. When examining social preference indices, both groups of animals showed positive discrimination indices [NH: t(10) = 11.64, P = <0.0001; HL: t(10) = 13.65, P = <0.0001, Fig. 6b].

Fig. 6.

Rats with slight and hidden hearing loss show social preferences. (a) Total exploration time did not differ between normal hearing rats (NH) and rats with slight/hidden hearing loss (HL). (b) The social preference index showed a significant positive deviation from zero in both rats with NH and rats with HL. ****P ≤ 0.0001.

Social recognition memory

Short- and long-term SR memory was tested in 10-week-old rats using the SR task (Fig. 7). After 1 or 8 h after the FT, short- and long-term memory was tested in 11 rats per group. During the short-term SR test, both groups of animals explored for equal amounts of time during the TT [t(20) = 0.07982, P = 0.9372, Fig. 7a]. In these experiments examining short-term SR memory (Fig. 7b), the discrimination index of animals with NH did not differ significantly from zero [t(10) = 0.8598, P = 0.4100], whereas rats with slight/hidden HL had a significantly positive discrimination index [t(10) = 3.702, P = 0.0041].

Fig. 7.

Rats with slight and hidden hearing loss show short- but no long-term social recognition (SR). (a) In the short-term SR task, total exploration time did not differ significantly between normal hearing rats (NH) and rats with slight/hidden hearing loss (HL). (b) The discrimination index showed a significant positive deviation from zero in rats with HL. There was no significant deviation from zero in the DI in rats with NH (F). (c) In the long-term SR task, total exploration time did not differ significantly between rats with NH and rats with HL. (d) The discrimination index showed a significant positive deviation from zero in rats with NH. There was no significant deviation from zero in the DI in rats with HL (F).*P ≤ 0.05.

During the long-term SR test, both groups of animals again explored for equal amounts of time during the TT [t(20) = 0.3210, P = 0.7515, Fig. 7c]. In these experiments examining long-term SR memory (Fig. 7d), the discrimination index of animals with NH had a significantly positive discrimination index [t(10) = 3.569, P = 0.0051], whereas rats with slight/hidden HL did not differ significantly from zero [t(10) = 0.0324, P = 0.9748].

Discussion

Overview

Mild HL is associated with cognitive impairments in children, yet the underlying neurobiological mechanisms remain unclear. Research has typically focused on animal models with severe HL, leading to a gap in understanding the neurobiological links between mild HL and cognitive impairment. Therefore, in this work, we aimed to establish a rat model to investigate cognitive deficits associated with mild HL. We used noise exposure to induce HL since noise exposure is a common cause of HL in the general population, including children, and since the intensity, frequency range, and duration of noise exposure can be controlled to manipulate the extent of HL. Measurement of ABRs was used to examine cochlear function across the experimental time course and confirmed slight/hidden HL in noise-exposed rats. Subsequent immunofluorescence to examine cochlear morphology revealed that slight/hidden HL was not associated with loss of either the IHCs or presynaptic ribbons. To assess cognitive function, we assessed both short- and long-term OL memory, object recognition memory, and SR memory. We found striking differences in the effects of HL on short- versus long-term memory. In general, noise-exposed rats with slight/hidden HL performed better than sham-exposed rats with NH on short-term memory tasks. However, the same rats with slight/hidden HL performed worse than rats with NH on long-term memory tasks. These results, including confounding factors, and their implications are discussed in more detail below.

Noise-induced slight/hidden hearing loss in young rats

To induce mild HL, we utilized a single round of noise exposure that resulted in slight HL at higher frequencies (≈15–20 dB at 16 and 32 kHz) and hidden HL at lower frequencies (4 and 8 kHz) and in response to click stimuli. In this work, hidden HL was identified as comparable absolute ABR wave I thresholds but reduced ABR wave I amplitudes (Bakay et al., 2018). Thus, our noise exposure paradigm results in rats with slight/hidden HL at young ages valuable for further behavioral examination. Our noise exposure paradigm differs from previous paradigms, which mostly rely on greater noise intensities (≥110 dB, Zheng et al., 2011; Lauer et al., 2018; Xu et al., 2022) or repeated noise exposure (Urán et al., 2014). Our approach, by using a less intense and singular noise exposure, provides a model of mild HL that is valuable for investigating cognitive deficits associated with mild HL.

In characterizing HL in this model, we observed two unexpected findings. First, we observed a steady decrease in ABR wave I amplitudes across stimuli in sham-exposed rats with NH across the experimental time course (at 3, 6, and 11 weeks of age; Supplementary Figures 1–3, Supplemental Digital Content 1, https://links.lww.com/BPHARM/A122). Developmental changes in ABR wave I amplitudes have not been previously measured in rats, and our work shows that these responses undergo maturation into the first 2 to 3 months of age, which is slightly longer than the auditory maturation timeline of rats that is reported by Pujol and Lavigne-Rebillard (2002). Second, previous work has linked noise-induced hidden HL to the loss of the synapses connecting the sensory IHCs to the spiral ganglion neurons (Kujawa and Liberman, 2009). We found no loss of CTBP2-immunostained presynaptic ribbons in the noise-exposed rats. CTBP2 is a well-established marker of presynaptic ribbons and is commonly used to assess ribbon loss following noise exposure. In these experiments, we also attempted to label postsynaptic glutamate receptors (see Methods), but the immunolabeling was unsuccessful for unknown reasons. Future studies should aim to quantify changes in both pre- and post-synaptic markers and assess their colocalization to better characterize synaptic alterations. Moreover, hidden HL has also been associated with disrupted myelination of the spiral ganglion neuron heminode (Wan and Corfas, 2017) as well as altered ion channel expression (Reijntjes et al., 2019), consistent with hidden HL likely arising from various etiologies (Kohrman et al., 2020). Future work should additionally characterize changes in myelination and ion channel distribution, which is known in mouse to undergo topographical changes in the spiral ganglion neuron heminode after the onset of hearing (Kim and Rutherford, 2016), to examine noise-induced as well as developmental changes in ABR wave I amplitudes in rats. Although more animals would be required, cochlear morphology could also be examined at various experimental timepoints (rather than at a single time point at the end of the experimental pipeline) to rule out possible structural recovery of lost synapses after noise exposure (Hickman et al., 2021).

Rats with slight/hidden hearing loss show exploratory behaviors that are comparable to rats with normal hearing

Unlike previous studies that have investigated the effect of more severe HL on exploratory behaviors (Johne et al., 2022), our findings suggest that rats with slight/hidden HL show normal objects or social exploration. Nevertheless, we did observe reduced object exploration during the FT of the long-term NOR task in rats with slight/hidden HL compared to rats with NH. However, this effect was not present in any of the other 11 assessments of exploratory behaviors across the OL, NOR, and SR tasks. Moreover, we do not believe the reduced exploration during the FT of the long-term NOR caused the lack of memory formation in rats with slight/hidden HL, since all animals explored above the cutoff threshold, ensuring sufficient exploration for memory formation.

Rats with slight/hidden hearing loss perform better on tasks testing short-term memory compared to rats with normal hearing

We found that rats with slight/hidden HL had relatively good short-term memory (i.e. positive discrimination indices) on the NOR and SR tasks. In contrast, rats with NH showed no discrimination on these tests. These results contradict previous observation of our group in which we found that similarly-aged rats exposed to higher intensity noise (120 dB SPL for 2 h) but with similar HL (also ≤25 dB HL) showed impaired short-term NOR compared to rats with NH (Jagersma et al., 2024). Difficulties replicating the effects of HL on cognitive function have been previously documented, with differences in noise exposure and duration of HL likely contributing to these discrepancies. For example, Uran et al. (2014) found that a single moderate noise exposure at postnatal day 15 did not impair short-term memory, whereas repeated exposures during the same developmental period did result in short-term deficits in NOR. In another pair of studies by the same research group using the same mouse model of HL, NOR impairments were initially observed 6 months after noise exposure (Park et al., 2016) but then later observed only 12 months after noise exposure (Park et al., 2018).

We suspect that the unexpectedly poor performance of the animals with NH results from the housing condition during the ITI (of 1 h) in a holding room that was conveniently near the experimental room but which may have inadvertently exposed the rats to activity from other researchers and animal caretakers. In rats with NH, this activity may have disrupted memory consolidation. Slight/hidden HL may have minimized the distraction from extraneous external noise and, thereby, unexpectedly, facilitated memory consolidation. Although these experiments need to be replicated with animals held in a quiet holding room during the ITI, they nevertheless suggest that slight/hidden HL may not impair some forms of short-term memory (OL and SR).

Lastly, neither group showed positive discrimination indices on the short-term OL task, which was the first memory test performed when animals were approximately 35 days old. When animals were a few days older, animals with NH showed positive discrimination indices on the long-term OL task (discussed further below). Since previous experiments have documented that spatial (object-location) memory in rats continues to mature between postnatal days 31 and 38 (Contreras et al., 2019), it may be that, in our experiments, the short-term OL test might have been performed too early to detect the impact of slight/hidden HL.

Rats with slight/hidden hearing loss perform worse on tasks testing long-term memory compared to rats with normal hearing

We found that rats with slight/hidden HL showed impaired long-term memory (i.e. discrimination indices that did not deviate significantly from zero) for both OL and SR tasks. In contrast, rats with NH showed normal (i.e. positive discrimination indices) on these tests. These results align with previous observations that also find that animal models with early onset, moderate to severe HL show impairments in cognition (Liu et al., 2016; Liu et al, 2018; Manohar et al., 2020; Shukla et al., 2019; Urán et al., 2014). In our experiments examining long-term memory tasks, animals were returned to their housing locations during the ITI (8 h), which were generally quieter than the holding locations and likely allowed for better sleep—a factor known to facilitate memory consolidation (Stickgold, 2005; Rasch and Born, 2013). If so, then rats with slight/hidden HL may indeed have disrupted memory consolidation that becomes apparent in long-term memory tasks.

Lastly, neither group demonstrated positive discrimination indices in the long-term NOR tasks. As the fourth and final memory-focused test, animals may have simply been less motivated to perform this particular task due to task fatigue and, in line with this possibility, rats with slight/hidden HL also showed reduced exploration compared to sham-exposed rats during the FT trial of the long-term NOR test. Interestingly, a different mouse model with early-onset moderate-to-severe HL showed normal NOR, both 3- and 6-months post-HL induction (Paciello et al., 2021). Collectively these findings suggest that long-term object recognition memory may be affected differently than long-term spatial or social memory and may be dependent on other factors. Further research is needed, especially since our control group did not exhibit long-term object recognition memory.

Additional considerations

There are other issues with the experimental design that should be considered. First, the behavioral tests examined in this study were conducted during the light phase. Roedel et al., 2006 showed that testing cognition during the light phase resulted in a cognitive disruption in DBA mice, a mouse model often used in HL research. This indicates that testing conditions may have influenced performance in our study. Second, the experimental design, which aimed to minimize the total number of animals needed, the sequence of tests or interaction across tests may have influenced results. Denninger et al. (2018) showed that first performing a test that relies on spatial cues can affect the outcome of tests that do not rely on such cues. Thus, the test sequence in our study—OL before NOR and SR—may have inadvertently primed the animals to focus on location cues during the OL task, causing them to pay less attention to object characteristics during the NOR task. Finally, ongoing maturation of auditory and/or cognitive function during the ages we examined (between 3- and 11-week-old) may influence results. Finally, performing individual tests at the same age in separate cohorts of animals would control for ongoing maturation of auditory and/or cognitive function during the ages we examined (between 3- and 11 weeks old).

Conclusion

In this study, we examined the effect of slight and hidden HL on cognitive function in young rats. Despite some acknowledged limitations, our results show that rats with slight/hidden hearing impacts perform differently than rats with NH on several tasks testing short- and long-term memory. Our findings motivate further research investigating the impact of HL on not only short- and long-term memory but also different stages of the learning and memory process, including encoding, storage, and retrieval (Melton, 1963), which rely on distinct neurobiological mechanisms (Bisaz et al., 2014). In addition to considering the magnitude of HL and investigating the impact of milder forms of HL as we did in this study, the impact of the onset and duration of HL on the extent of cognitive impairment should also be carefully examined as part of future research. A better understanding of the effects of HL on cognitive function and its underlying neurobiology is crucial for developing targeted interventions and strategies to address the complex consequences of HL.

Acknowledgements

We acknowledge the valuable assistance provided by Fanny Bunn and Christina Mavridou during their research internships.

The project was supported by the Young Academy Groningen (University of Groningen, the Netherlands) granted to J.D.A.O. and S.J.P., the Meisner Fonds granted to S.J.P., and the Heinsius Houbolt Foundation granted to S.J.P.

Conflicts of interest

There are no conflicts of interest.

Supplementary Material

bpharm-36-300-s001.docx ^{(167KB, docx)}