Mitochondrial alterations in the cochlea of Cdk5rap1‐knockout mice with age‐related hearing loss

Previous studies have revealed that age‐related hearing loss (AHL) in Cdk5 regulatory subunit‐associated protein 1 (Cdk5rap1)‐knockout mice is associated with pathology in the cochlea. Here, we aimed to identify mitochondrial alterations in the cochlea of Cdk5rap1‐knockout mice with AHL. Mitochondria in the spiral ganglion neurons (SGNs) and hair cells (HCs) were normal despite senescence; however, the mitochondria of types I, II, and IV spiral ligament fibrocytes were ballooned, damaged, and ballooned, respectively, in the stria vascularis. Our results suggest that the accumulation of dysfunctional mitochondria in the lateral wall, rather than the loss of HCs and SGNs, leads to the onset of AHL. Our results provide valuable information regarding the underlying mechanisms of AHL and the relationship between aberrant tRNA modification‐induced hearing loss and mitochondrial dysfunction.

Previous studies have revealed that age-related hearing loss (AHL) in Cdk5 regulatory subunit-associated protein 1 (Cdk5rap1)-knockout mice is associated with pathology in the cochlea. Here, we aimed to identify mitochondrial alterations in the cochlea of Cdk5rap1-knockout mice with AHL. Mitochondria in the spiral ganglion neurons (SGNs) and hair cells (HCs) were normal despite senescence; however, the mitochondria of types I, II, and IV spiral ligament fibrocytes were ballooned, damaged, and ballooned, respectively, in the stria vascularis. Our results suggest that the accumulation of dysfunctional mitochondria in the lateral wall, rather than the loss of HCs and SGNs, leads to the onset of AHL. Our results provide valuable information regarding the underlying mechanisms of AHL and the relationship between aberrant tRNA modification-induced hearing loss and mitochondrial dysfunction.
Age-related hearing loss (AHL) is a multifactorial condition. Some studies have indicated that altered mitochondrial DNA, imbalanced mitochondrial redox, reactive oxygen species production, and modified antioxidant capacity occur due to age-related oxidative stress and are related to senescence of the cochlea and AHL in mice and humans [1][2][3][4][5]. Mitochondria are crucial for energy supply, maintaining cellular redox homeostasis, programmed cell death, and cell signaling regulation [6]. Translational precision by mitochondrial ribosome affects nuclear gene expression and cytoplasmic proteostasis, both of which are determinants of aging [7][8][9][10]. The post-transcription alteration of tRNA nucleotides occurs through numerous enzymatic reactions that are crucial for accurate and efficient decoding [11], mainly because they contribute to the binding of the tRNA codon [12,13]. Several diseases, including mitochondrial diseases and type II diabetes, have been associated with alterations of mitochondrial tRNAs (mt-tRNAs) [9][10][11][14][15][16][17][18][19][20].
Mitochondrial tRNAs in mammals are modified by nuclear tRNA-associated enzymes, including Cdk5 regulatory subunit-associated protein 1 (CDK5RAP1), which induces 2-methylthio (ms 2 ) modifications in mammalian mt-tRNAs [16,21]. Incomplete deficiency in ms 2 modification markedly weakens mitochondrial protein synthesis in stressful situations, leading to AHL and respiratory defects, as seen in Cdk5rap1-knockout (KO) mice, which are predisposed to stress-induced mitochondrial remodeling [22,23]. In the inner ear of Cdk5rap1-KO mice, deficiency in ms 2 modification induces an initial decline in cochlear function owing to type II and IV spiral ligament (SLi) fibrocyte degeneration and stria vascularis (SV) senescence preceding spiral ganglion neuron (SGN) and hair cell (HC) loss (Fig. 1). The mitochondria of type II and IV SLi fibrocytes of Cdk5rap1-KO mice were disrupted (i.e., mitochondria ballooning and damaged cristae) in the initial stages of AHL. These findings suggest that a reduction in the modification of mt-tRNAs leads to SLi mitochondrial dysfunction and oxidative stress throughout the cochlea, thereby inducing cytotoxicity in fibrocytes and an early decrease in endolymphatic potential (EP) during the development of AHL.
The extracellular fluid present in the cochlear duct and ionic homeostasis of the endolymph is crucial for hearing and plays an important role in sound transduction [24]. The SV transports K + ions and SLi fibrocytes travel from the organ of Corti via HCs back to the SV and SLi by transduction and then back into the scala media during this process [24,25]. Chronic EP changes produced by ion content variations enforced by standing potentials may change the ionic composition of the endolymph [24]. Such putative ionic composition abnormalities of the local endolymph may lead to HC and secondary SGN corrosion, thereby causing detectable modifications in the dysfunctional lateral wall [24,26]. Hearing is corrected by the ion-transport system and mitonuclear protein homeostasis remodeling, which could restrain oxidative stress in young animals [6,27].
Gradual dysfunction of the respiratory complexes in SLi fibrocytes occurs with age and affects the quality control of mitochondria, leading to SLi and SV cellular senescence and chronic EP reduction. The subsequent deformities in the ionic composition of the local endolymph may lead to HC deterioration and secondary SGN damage. Some studies have revealed that dysregulated tRNA modifications in humans precede aging [9,10]. Thus, cochlear energy loss may cause AHL, possibly due to the reduction in post-transcriptional mt-tRNA modifications in humans.
However, previous studies did not further investigate outer HCs (OHCs), SGNs, marginal SV cells, or type I SLi fibrocytes even though damage or senescence was observed, which is crucial for histopathological assessment in AHL [24,[28][29][30] (Fig. 1). Therefore, the present study evaluated the cochlear mitochondria in the inner ears of Cdk5rap1-KO mice that have reduced ms 2 mt-tRNA modifications.

Transmission electron microscopy
Cdk5rap1-KO and CNT mice (male, five per group) were anesthetized via intraperitoneal injection of 4 mgÁkg À1 of xylazine (Bayer, Shawnee, KS, USA) and 120 mgÁkg À1 of ketamine-HCl (Daiichi Sankyo, Tokyo, Japan) in NaCl (0.9%) at 4, 12, 20, or 48 weeks after birth. The mice were perfused with 2.5% glutaraldehyde (GLUT) and 2 mM CaCl 2 formed in 0.1 M sodium cacodylate (C 2 H 6 AsNaO 2 ) (pH 7.4) via the ascending aorta. The inner ears of mice were examined, and samples were maintained at 25°C in a mixture of 2% paraformaldehyde and 2% GLUT in phosphate-buffered saline (pH 7.4). Using 2% osmium tetroxide, postfixation was conducted on ice for 30 min. Consequently, the specimens were stained with 1.5% uranyl acetate for 1 h at 4°C, desiccated in a 100% propylene oxide and ethanol gradient (50, 70, 90, and 100% in ascending order), inserted in Epon-Araldite, and polymerized for 48 h at 60°C. Samples were cut into 60-65-nm sections with an ultramicrotome (UC7i; Leica, Munich, Germany) and stained with fresh Reynolds' lead citrate and 1.5% uranyl acetate at 25°C. Transmission electron microscopy (TEM) (HT7700; Hitachi, Tokyo, Japan; n = 5 for each genotype) was used to analyze the samples. Twenty images of each sample were acquired at each magnification in the same plane for an accurate representation. The amount of damaged and normal mitochondria in each image was determined using the mitochondrion cristae score, which indicates structural abnormalities, in IMAGEJ software (National Institutes of Health, Bethesda, MD, USA) [31], and the proportion of wounded to normal mitochondria was calculated. In addition, we measured the dimensions of oblong mitochondria in each image (n = 5 for each genotype) by tracing the outer mitochondrial membrane of each mitochondrion using the freehand tool in IMAGEJ software (National Institutes of Health) [31].

Statistical analyses
Eighty female Cdk5rap1-KO mice and CNT littermates were selected for genotyping. The results were assessed using a two-way ANOVA and a post hoc Tukey test. Statistical analyses were evaluated with GRAPHPAD PRISM 8.0.0 for Windows (GraphPad Software, San Diego, CA, USA). Data were corrected via Bonferroni analysis using the family-wise error rate. The information validated the statistical test assumptions. The sample size and statistical power were estimated before and after the collection of data with 'PS: Power and Sample Size Calculation', Ver. 3.1.6 (Biostatistics Dept, Vanderbilt University, Nashville, TN, USA) [32]. Data are denoted as the mean AE standard error. P < 0.05 indicated statistical significance.

Results
Degenerating mitochondria were absent in HCs and SGNs of Cdk5rap1-KO mice Transmission electron microscopy revealed no differences in subcellular localization and its morphology, mitochondrial damage, and the sizes of HCs and SGNs according to age in Cdk5rap1-KO compared with CNT mice (

Degenerated mitochondria in SLi fibrocytes of Cdk5rap1-KO mice
Transmission electron microscopy showed that type I fibrocyte mitochondria in the SLi of Cdk5rap1-KO mice had a loss of the cristae in the inner mitochondrial membrane (IM) starting at 12 weeks of age, which was earlier than that in CNT mice ( Fig. 5A-E). At 12, 20, and 48 weeks of age, the mitochondria of Cdk5rap1-KO mice were more noticeably impaired than those of CNT mice ( Fig. 5D; 12 weeks, P = 0.004; 20 weeks, P < 0.001; 48 weeks, P < 0.001; F[3, 32] = 9.299, P < 0.001). However, the mitochondria of Cdk5rap1-KO mice were not much bigger than those of CNT mice at any age (Fig. 5E). Regarding type II and IV SLi fibrocytes in relation to the findings of another study, the Cdk5rap1-KO mice in our research began to lose the cristae in the IM at the age of 20 weeks (Fig. 5A,B,F). Furthermore, at 20 and 48 weeks of age, the mitochondria of Cdk5rap1-KO mice were considerably more damaged than those of CNT mice

Discussion
We used TEM in Cdk5rap1-KO mice to analyze the impact of insufficient ms 2 modification of mt-tRNAs, which results in poor mitochondrial protein translation in tissues in the cochlea during AHL. In 4-week-old Cdk5rap1-KO mice, TEM revealed no differences in the mitochondrial structure and size compared with CNT mice. However, in older Cdk5rap1-KO mice, TEM revealed changes in mitochondrial damage and size in the SV and SLi but not in the HCs and SGNs.
In a previous study, SGN and SV marginal cell senescence, SLi deterioration of fibrocytes (type II and IV), reduced production of Na + /K + ATPase a1 (fibrocytes, type II and IV) (Fig. 1), and a slight reduction in EPs was observed in younger Cdk5rap1-KO mice. Conversely, reduced EPs and hearing loss were observed in older Cdk5rap1-KO mice. The Cdk5rap1-KO mice had SGN and HC loss as well as downregulation of Na + /K + -ATPase a1 (fibrocytes, type II and IV) and Cx26 (type I fibrocytes) in the SLi at 20 weeks of age. Remarkably, compared with CNT mice, these alterations occurred sooner in Cdk5rap1-KO mice (Fig. 1). In the aforementioned study, Cdk5rap1-KO mice with SLi fibrocyte degeneration, low Na + /K + -ATPase 1 expression, and SV senescence prior to the loss of HCs and SGNs showed an early decline in cochlear function. Therefore, we hypothesized that fibrocytes from the SLi counterbalance fewer EPs by moving ions through fibrocytes in the SLi and SV in younger Cdk5rap1-KO mice.
Interestingly, in the present study, no discernible differences were found in mitochondrial damage and the sizes of HCs and SGNs at different ages between Cdk5rap1-KO and CNT mice. AHL may be induced by HC and SGN damage according to the histopathology of the cochlea in age-graded humans and mice [33]. HC damage is exacerbated by age, whereas SGN damage is constant with the disintegration that occurs after the loss of HCs. Although mitochondrial damage is considered the cause of age-related degeneration of HCs and SGNs, our study suggests that mitochondrial damage in HCs and SGNs does not lead to the onset of AHL. We speculated that the reason for these results was tissue damage due to chronic EP drop, as described below.
In addition to the loss of HCs and their innervation via SGNs, cochlear histopathology has also reported widespread degeneration of fibrocytes from SLi [2,24] and SV degeneration. Our study showed that although mitochondrial damage in type I SLi was observed in Cdk5rap1-KO mice at 12 weeks of age, the ballooning was not significant compared with CNT. Moreover, at the same stage, mitochondria in type II and IV SLi were disrupted (i.e., mitochondria ballooning and cristae loss). As a result, HC and SGN damage was preceded by the cell damage identified in the SLis of Cdk5rap1-KO mice as well as EP reduction at a younger age, indicating that the fibrocyte pathology resulting in the degeneration of sensory cells may be the cause of early AHL since such tendencies were seen at older ages in CNT mice. Previous studies have reported that the administration of a permanent complex II blocker or a mitochondrial toxin, such as 3-nitropropionic acid, caused inflammatory responses in the lateral wall following acute mitochondrial dysfunction [34,35]. Moreover, these previous studies have also described lateral wall histopathology and mitochondrial dysmorphology.
Regarding SV, in our previous study, senescence markers were expressed in the marginal SV cells; however, despite its significance in the pathophysiology of AHL, we did not observe any modifications in the thickness of the SV with aging [24,36]. TEM analysis showed ballooned, although undamaged, mitochondria in the marginal SV cells in 20-week-old Cdk5rap1-KO mice. We speculate that this was caused by SLi compensation, as the SV functions as a barrier between the lateral wall and the endolymphatic area.
In Cdk5rap1-KO mice, dysfunction of the mitochondria is counterbalanced by mitonuclear protein homeostasis [22] through the initiation of the unfolded protein response of mitochondria that applies a defensive effect [6,22]. The amount of mitochondrial damage probably exceeds the ability of mitophagy to maintain the mitochondrial network, which results in mitochondrial malfunction [22,37]. Additionally, elevated complex I deficiencies were linked to a slight increase in oxidative stress, which results in the unfolded protein response and cytotoxicity [37]. Our research corroborates the idea that, depending on its severity, mitochondrial malfunction under sedentary conditions may not instantly result in a pathogenic phenotype. These findings imply that the lack of mt-tRNA alterations generates oxidative stress throughout the cochlea and failure of mitochondria in SV and SLi, which triggers fibrocyte cytotoxicity and reduces EPs early on in the advancement of AHL.  Based on our research, we hypothesized that AHL in Cdk5rap1-KO mice is caused by the depletion of the ion-transport system in and around the SV (i.e., SLi fibrocytes) affected by reduced mitochondrial activity as a result of the lack of ms 2 modifications in mt-tRNAs. By altering the ion-transport system and homeostasis of mitonuclear proteins, hearing is compensated, which may prevent oxidative stress in young animals [6,27]. As people age, the increasing disruption of the respiratory complexes affects the quality control of the mitochondria in SLi fibrocytes, leading to senescence of the SLi and SV cells and a persistent drop in EPs. The loss of secondary SGNs and subsequent HC degeneration may be facilitated by the ensuing anomalies in the local endolymphatic ion composition. Previous research has demonstrated that dysregulated tRNA changes occur before human aging [9,10]. As a result, reduced post-transcriptional alterations in mt-tRNAs in humans may be the source of AHL.
The C57BL/6 mouse model might not accurately depict AHL, even though the pathology and etiology of hearing loss are identical to those of AHL. This is a major limitation of our in vivo investigation. Due to a Cdh23 mutation, C57BL/6 mice are more susceptible to AHL [38]. Ideally, we would have used strains such as BALB/c or CBA that were backcrossed with a strain without any type of hearing issues. Moreover, TEM artifacts were observed in this study and vacuolization from inadequate fixation can be seen in microstructural TEM images as artifacts frequently. Therefore, adequate settings and conditions with sufficient fixation are necessary for future studies.
The findings of previous and current studies suggest that ms 2 modifications of mt-tRNAs may cause oxidative stress and damage to the mitochondria in the SLi and accelerate aging, thereby causing AHL. In our study, the mitochondrial findings in HCs and SGNs were normal, despite senescence. Mitochondrial findings in the SV revealed only ballooning, and those in the SLi type I, II, and IV fibrocytes indicated damage and ballooning. Generally, the presence of mitochondria has been reported in inner hair cells (IHCs), OHCs, SGNs, SV, and SLi in the cochlea. Mitochondria are located at the top and around the nucleus in IHCs, whereas mitochondria are located at the top and along the outer wall of the OHCs. Mitochondria are located in the fine membranous processes of the marginal cells in the SV [39,40].
Mitochondria play a crucial role in providing energy for the specialized sensory HCs that detect sound and enable hearing. In addition to their energy-producing role, mitochondria in the inner ear also play a role in regulating the calcium levels in HCs, which is important for their function. Mitochondria are also involved in the process of apoptosis, which can be triggered by certain types of damage or stress to the HCs. Overall, the physiological function of mitochondria in the inner ear is critical for the proper function of the sensory HCs and, thereby, for hearing [39]. Our study suggested that mitochondria in the SLi were also crucial for hearing.
We demonstrated that a lack of CDK5RAP1, which catalyzes modifications of mt-tRNAs (ms 2 ), results in mitochondrial malfunction. Cell senescence and AHL are caused by mitochondrial malfunction. Our findings imply that defective mitochondria may contribute to the advancement of AHL. Our analysis provides valuable insights regarding the underlying mechanisms of AHL and the relationship between aberrant tRNA modification-induced hearing loss and mitochondrial dysfunction. Future treatment approaches aimed at reducing the consequences of AHL could be guided by elucidating the processes causing AHL.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article.