Methionine sulfoxide reductase B3 (MsrB3) is a protein repair enzyme that specifically reduces methionine-R-sulfoxide to methionine. A recent genetic study showed that the MSRB3 gene is associated with autosomal recessive hearing loss in human deafness DFNB74. However, the precise role of MSRB3 in the auditory system and the pathogenesis of hearing loss have not yet been determined. This work is the first to generate MsrB3 knockout mice to elucidate the possible pathological mechanisms of hearing loss observed in DFNB74 patients. We found that homozygous MsrB3−/− mice were profoundly deaf and had largely unaffected vestibular function, whereas heterozygous MsrB3+/− mice exhibited normal hearing similar to that of wild-type mice. The MsrB3 protein is expressed in the sensory epithelia of the cochlear and vestibular tissues, beginning at E15.5 and E13.5, respectively. Interestingly, MsrB3 is densely localized at the base of stereocilia on the apical surface of auditory hair cells. MsrB3 deficiency led to progressive degeneration of stereociliary bundles starting at P8, followed by a loss of hair cells, resulting in profound deafness in MsrB3−/− mice. The hair cell loss appeared to be mediated by apoptotic cell death, which was measured using TUNEL and caspase 3 immunocytochemistry. Taken together, our data suggest that MsrB3 plays an essential role in maintaining the integrity of hair cells, possibly explaining the pathogenesis of DFNB74 deafness in humans caused by MSRB3 deficiency.
The cochlea (the auditory portion of the inner ear) plays a vital role in the conversion of physical vibrations generated by sound waves into chemical signals in the process of hearing. The mammalian cochlea contains ∼16 000 sensory hair cells that are arranged as three rows of outer hair cells (OHCs) and one row of inner hair cells (IHCs) (1,2). Because these hair cells cannot be regenerated during the lifetime of an animal, damage to or loss of hair cells due to various factors, such as genetic defects, noise and drugs, is irreversible and leads to permanent hearing loss (3).
Methionine is one of the most sensitive amino acids to oxidation by reactive oxygen species (ROS). The oxidation of methionine generates a diastereomeric mixture of methionine-R-sulfoxide and methionine-S-sulfoxide (4), which may cause significant changes to protein structure and function (5,6). However, oxidized methionine can be repaired using the enzyme methionine sulfoxide reductase (Msr). Two families of Msrs, MsrA and MsrB, have evolved to catalyze the stereospecific reduction of methionine sulfoxide to methionine in proteins (7–9). MsrA is specific for the S-form of methionine sulfoxide, whereas MsrB only reduces the R-form (10,11). The functions of Msrs are to repair oxidatively damaged proteins, to protect against oxidative stress and to regulate protein function.
In mammals, three genes encode MsrB proteins (MsrB1, 2 and 3, which have different subcellular locations), whereas a single gene encodes MsrA (12–14). MsrB1 is a cytosolic and nuclear selenoenzyme, MsrB2 is targeted to the mitochondria and methionine sulfoxide reductase B3 (MsrB3) exists in two forms in humans, MSRB3A and MSRB3B, which contain different N-terminal regions generated via alternative first exon splicing. MSRB3A is targeted to the endoplasmic reticulum (ER), whereas MSRB3B resides in the mitochondria (13). However, no evidence for the existence of alternatively spliced MsrB3 forms in mice has been reported. Instead, mouse MsrB3 contains consecutive ER and mitochondrial signal peptides in the N-terminal region and is targeted to the ER (15).
Recently, Ahmed et al. (16) reported that the MSRB3 gene is associated with human DFNB74, a locus for autosomal recessive hearing loss. Two mutations in MSRB3, one missense and one nonsense mutation, were identified in families with the DFNB74 locus. The missense mutation (p.Cys89Gly) is a loss-of-function allele that results in a complete lack of MSRB3 enzymatic activity. The nonsense mutation (p.Arg19X) occurs in the mitochondrial targeting sequence of MSRB3B. These data suggest that functional MSRB3 is essential for hearing in humans (16). However, the lack of animal models for MsrB3 deficiency has hindered our understanding of the function of MSRB3 and the pathogenesis of DFNB74 hearing loss.
Here, we present the first report of the generation and evaluation of MsrB3 knockout mice as an animal model for DFNB74 and propose a possible pathological mechanism leading to hearing loss caused by MSRB3 deficiency in DFNB74 patients.
Generation of MsrB3 knockout mice
In mice, the MsrB3 gene is located on chromosome 10 and comprises ∼120 kb of genomic DNA. MsrB3 knockout mice were generated using homologous recombination. The knockout construct was designed to replace exon 7 encoding the catalytic Cys residue with a Neo cassette (Fig. 1A); correct gene targeting was verified with the Southern blot analysis (Supplementary material, Fig. S1).
In homozygous MsrB3−/− mice, the MsrB3 protein was not detected in the heart, skeletal muscle or testis, tissues in which MsrB3 proteins are highly expressed in wild-type mice (Fig. 1B). In addition, heterozygous MsrB3+/− mice exhibited lower MsrB3 levels than wild-type MsrB3+/+ mice. We also examined MsrA, MsrB1 and MsrB2 proteins expression levels in MsrB3 knockout mice. MsrB1 levels were higher in the heart and skeletal muscle in MsrB3−/− mice than in MsrB3+/+ mice (Fig. 1B). MsrB2 was expressed at similar levels in the heart but at slightly decreased levels in the skeletal muscle and testis. The levels of MsrA did not differ in any of the tissues examined (Fig. 1B).
MsrB3 knockout mice are deaf
Because the MSRB3 gene is associated with DFNB74 deafness in humans (16), we sought to ascertain whether the MsrB3 knockout mice also exhibited a hearing impairment phenotype. We measured the hearing thresholds using an auditory brainstem response (ABR) test in MsrB3+/− and MsrB3−/− littermates at postnatal day (P) 20, an age when the hearing system is considered to have largely matured. Heterozygous MsrB3+/− mice exhibited an ABR threshold and waveform with an average hearing threshold of 29.5 decibels (dB) using click stimuli (Fig. 2, left panel) and average hearing thresholds of 27.5, 26.3 and 30.3 dB at 8, 16 and 32 kHz (n = 15), respectively, using tone bursts (Supplementary Material, Fig. S2, left panel). Homozygous MsrB3−/− mice did not respond to click stimuli or tone bursts (n = 15), indicating profound hearing impairment (Fig. 2, right panel, Supplementary Material, Fig. S2, right panel). However, we did not observe any abnormal vestibular behaviors such as circling, head tossing or walking disability in MsrB3−/− mice. These data clearly demonstrate that the deletion of MsrB3 leads to deafness in mice.
MsrB3 is expressed in the sensory epithelia of cochlear and vestibular tissues
MsrB3 expression was analyzed immunohistochemically in mouse inner ears from embryonic day (E) 12.5 to P18. The inner ear sections contained both the cochlea and vestibular end organs including the cochlear duct, utricle, saccule and crista ampullaris. No detectable MsrB3 signal was observed at E12.5 in any inner ear sections (Fig. 3A and I). MsrB3 expression began to be detected at E13.5 in the sensory epithelia of vestibular organs, such as the maculae of the utricle, saccule and all three cristae (Fig. 3J and data not shown). MsrB3 expression in the cochlea was not detected until E15.5, at which point MsrB3 signals were found primarily in the basal and middle turns of the cochlea, and later in all cochlear turns (Fig. 3D). These temporal and spatial expression patterns of MsrB3 follow the sequence of inner ear hair cell differentiation. The MsrB3 signals in the cochlea and vestibule were continuously detected until at least P18, when the mouse hearing ability matures (Fig. 3D–G and J–O). Magnified views of P18 specimens showed that the MsrB3 signals were present in the cochlear epithelium from the inner to the outer sulcus cells, encompassing the auditory hair cells (Fig. 3G and H) and the sensory epithelia of the vestibular organs (Supplementary Material, Fig. S3).
Cellular localization of MsrB3 at the base of the stereocilia and in the hair cell body
To assess the intracellular localization of MsrB3 in the hair cells, we performed whole-mount immunostaining of the organ of Corti from wild-type mice. In a top view of the organ of Corti, MsrB3 immunoreactivity was viewed as dot-like structures at the base of the stereocilia of both IHCs and OHCs (Fig. 4A and D). However, side views of the IHCs revealed that MsrB3 proteins were expressed in more rod-like patterns at the base of each stereocilium (Fig. 4B and C). Similar localization of MsrB3 proteins at the base of stereocilia was also observed in vestibular hair cells (data not shown). MsrB3 immunoreactivity was completely absent in the MsrB3−/− organ of Corti, whereas heterozygote MsrB3+/− cochlea exhibited expression patterns similar to those of wild-type mice (Supplementary Material, Fig. S4). In addition, faint and scattered MsrB3 staining was also detected in the cell bodies of hair cells.
Msr level and activity in the inner ear
We analyzed the expression levels and activities of MsrBs and MsrA in the inner ears of MsrB3+/+, MsrB3+/− and MsrB3−/− mice. The expression of MsrB3, which was examined using western blot experiments, was reduced in the inner ears of MsrB3+/− mice and completely abolished in those of MsrB3−/− mice (Fig. 5A). The expression levels of MsrA and MsrB1 in the inner ears of MsrB3+/− and MsrB3−/− mice appeared slightly higher than those in the inner ears of MsrB3+/+ mice.
Surprisingly, inner ear MsrB activities were not reduced in heterozygote MsrB3+/− or homozygote MsrB3−/−mice compared with wild-type mice (Fig. 5B). Similarly, no significant reductions in MsrB activities were observed in the heart, skeletal muscle or testis in MsrB3−/− mice (data not shown). These results suggest that MsrB3 enzymatic function is either compensated for by increased MsrB1 isozyme activity in MsrB3+/− and MsrB3−/− mice or that this enzymatic function does not play a major role in MsrB activity.
The inner ears of MsrB3 knockout mice exhibit no significant changes in protein oxidation
To determine whether the observed MsrB3 deficiency increases cellular protein oxidation in the inner ear, we measured the levels of methionine sulfoxide and carbonyl proteins in MsrB3+/+, MsrB3+/− and MsrB3−/− littermates at P10, P20 and P30 using western blot analysis. H2O2-treated P10 wild-type cochlear tissues exhibited some increased band intensities and size shifts compared with non-treated cochlear tissues, validating our methodology for measuring methionine sulfoxide levels (Supplementary Material, Fig. S5). However, when we compared the inner ear proteins of MsrB3+/− and MsrB3−/− mice, no significant differences were observed in either the band intensity or size shift at all three stages tested (Supplementary Material, Fig. S5).
Furthermore, no increase in the carbonyl level of proteins was detected in the inner ears of MsrB3−/− mice compared with MsrB3+/+ and MsrB3+/− mice at all stages tested (Supplementary Material, Fig. S6), and no significant difference was observed in the carbonyl content of the heart tissues of MsrB3+/+ (or MsrB3+/−) and MsrB3−/− mice. Together, these data suggest that MsrB3 deletion has no effect on protein oxidation in inner ear tissues.
The stereocilia of inner ear hair cells progressively degenerate in MsrB3−/− mice
Our observations that MsrB3 proteins are strongly expressed in the basal part of stereociliary bundles and that MsrB enzymatic activity is largely normal in the inner ear of MsrB3−/− mice led us to examine the possible structural role of MsrB3 in stereociliary bundles. Using scanning electron microscopy, we examined the integrity of the stereociliary bundles of MsrB3−/− mice and compared them with those of MsrB3+/+ and MsrB3+/− mice (Fig. 6). Stereociliary bundles normally display three rows of OHCs with ‘V’- or ‘W’-like arrays and one IHCs row with a linear or crescent-like arrangement (Fig. 6A, left panel). We observed no obvious changes or defects in the stereociliary bundles of heterozygote MsrB3+/− mice at any tested stage (P5, P8, P13, P20 and P50) (Fig. 6A, middle panel). In contrast, alterations in the stereociliary arrangement exhibiting an Ω (ohm)-like shape were observed in the OHCs starting at P8 in MsrB3−/− mice (Fig. 6A, right panel). Subsequent degeneration of stereociliary bundles in both IHCs and OHCs was first observed at P13 in MsrB3−/− mice but not MsrB3+/+ or MsrB3+/− mice (Fig. 6B). Stereociliary degeneration became significantly more severe at P20, and no stereociliary bundles were observed in the OHCs of MsrB3−/− mice at P50. Similar degeneration (although slightly less severe) was also observed in the stereociliary bundles of IHCs (Fig. 6). Interestingly, the degeneration appeared to be more severe in the apical turns than in the middle or basal turns (Supplementary Material, Fig. S7). Although the degeneration of hair cells was clearly identified in the inner ear at P20, the histological analysis showed no obvious morphological defects in the inner ear at this stage (Supplementary Material, Fig. S8). A magnified view of the middle turn of the midmodiolar sections showed an intact organ of Corti, stria vascularis and spiral ligament in the MsrB3−/− mice (Supplementary Material, Fig. S8B). Taken together, our results show that the lack of MsrB3 resulted in the stereociliary degeneration of both OHCs and IHCs, which may explain the profound hearing loss observed in MsrB3−/− mice and DFNB74 patients.
Deletion of MsrB3 leads to apoptotic cell death in the organ of Corti
We next examined whether stereociliary degeneration is associated with hair cell loss in MsrB3−/− mice using the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay and by immunolabeling active caspase 3 on the cochlear sections of MsrB3+/− and MsrB3−/− mice at P20. Although no TUNEL-positive cells or active caspase 3 signals were observed in the MsrB3+/− cochlea (Fig. 7A and C), the MsrB3−/− cochlea exhibited TUNEL-positive cells in the organ of Corti (including the hair and supporting cells) (Fig. 7B and B′). Active caspase 3 was also detected in the organ of Corti of MsrB3−/− mice (Fig. 7D and D′). We quantified the increase in apoptotic cells by counting TUNEL(+) cells in the organ of Corti of MsrB3+/− and MsrB3−/− mice. The number of TUNEL(+) cells was significantly increased in the cochlea of MsrB3−/− mice, with an average of 61.5 cells, compared with that in the cochlea of MsrB3+/−mice, with an average of 5 cells (n = 4, P < 0.008, Supplementary Material, Fig. S9). These data suggest that MsrB3 plays a protective role against apoptotic cell death in the cochlea.
Reduction and oxidation (redox) homeostasis is essential in almost all living organisms to delicately control the balance between the beneficial and harmful effects of free radicals such as ROS and reactive nitrogen species. The deleterious actions of these free radicals have been associated with the pathogenesis of a variety of diseases including cancer, diabetes mellitus, inflammatory diseases, neurodegenerative disorders and aging (17). A recent report demonstrating genetic mutations in the MSRB3 gene as a cause of prelingual hearing loss in DFNB74 patients also strongly suggested a crucial role of redox homeostasis in the development or functioning of the inner ear (16).
To explore the pathophysiological mechanisms of deafness caused by the null alleles of MSRB3, we generated MsrB3−/− mice. Similar to the phenotypes observed in the DFNB74 patients (16), the homozygote MsrB3−/− mice displayed profound hearing loss with normal vestibular function (Supplementary Material, Fig. S10), suggesting that our MsrB3−/− mice were excellent animal models for human DFNB74 hearing loss. Thus far, knockout models for some other Msr genes have been reported. MsrA−/− mice exhibited various neurological abnormalities, such as tiptoe walking, abnormal behavior and lower locomotor activity, as well as a dramatically shortened life span; however, the hearing ability of MsrA−/− mice has not been described (18–20). MsrB1−/− mice appeared generally normal, although the liver and kidney were slightly affected (21). These results suggest that MsrB3 is an important Msr protein involved in hearing function in mammals.
Consistent with the hearing loss phenotype of MsrB3−/− mice, we observed that MsrB3 proteins are densely localized at the base of the stereocilia on the apical surface of the hair cells in the organ of Corti. Recently, TRIOBP, an actin-bundling protein, has been reported to be localized at the base of stereocilia and is essential for hearing due to its role in the formation of stereocilia rootlets (22). Because Triobp−/− mice showed stereociliary degeneration phenotypes similar to those observed in MsrB3−/− mice, we considered whether MsrB3 could also be associated with the formation of rootlets. However, rootlet structures were clearly observed in the hair cells of MsrB3−/− mice at P8 using transmission electron microscopy (Supplementary Material, Fig. S11). In addition, quantitative real-time PCR analysis showed no significant differences in the mRNA levels of TRIOBP between the inner ears of MsrB3+/− and MsrB3−/− mice at P5, P10 or P20 (data not shown). Therefore, it is unlikely that MsrB3 is functionally associated with TRIOBP in regulating the formation of stereociliary bundle rootlets. Instead, MsrB3 appears to control the maturation and/or maintenance of stereociliary bundles because the stereociliary bundles of MsrB3−/− mice exhibited morphological alterations starting at P8 and progressively degenerated during time points spanning the completion of hearing development in mice. Mechanotransduction activity was evaluated with FM1-43 dye in the hair cells of MsrB3+/− and MsrB3−/− mice at P3. We observed that FM1-43 uptake was normal in both MsrB3+/− and MsrB3−/− mice but completely abolished by adding 5 mm EGTA, which has been shown to disrupt tip links at the stereocilia (data not shown). These results indicate that the hair cells of the MsrB3−/− mice develop normally and have functional mechanotransduction channels until at least P3 and further suggest that degeneration rather than maldevelopment of the hair cells is the primary cause of hearing impairments in the MsrB3−/− mice.
However, the molecular mechanisms by which MsrB3 deficiency leads to degeneration of the stereocilia remain to be resolved. Recently, Hung et al. (23) showed that Mical, a monooxygenase, can selectively oxidize two methionines (46th and 49th Met) of the 16 Met residues in the actin polypeptide, thereby inducing F-actin to disassemble. Moreover, Lee et al. (24) reported that both MsrB1 and MsrB2 have been shown to reduce the oxidized Met46 and Met49 in Mical-treated F-actin. Based on these reports, we hypothesize that MsrB3 may play a crucial role in the redox regulation of F-actin in hair cells and that the deficiency of MsrB3 may result in increased oxidation of F-actin by Mical, which accelerates the disassembly of F-actin in stereocilia, causing their degeneration.
MsrB3 plays a protective role against oxidative stress in MsrB3 transgenic flies and yeast cells (25,26). However, no significant accumulation of carbonyl derivatives was observed in the inner ears of MsrB3−/− mice compared with MsrB3+/− and MsrB3+/+mice. These data suggest that oxidative stress may not be a major cause of hair cell degeneration induced by MsrB3 deficiency. Notably, no significant differences in MsrB enzyme activity were observed between the inner ears of MsrB3−/− and MsrB3+/+ (or MsrB3+/−) mice. We were not able to examine the enzymatic activities of MsrB1, MsrB2 and MsrB3 individually because all three MsrB enzymes act on the same substrates, and their catalytic activities are similar (27). Interestingly, the expression levels of MsrB1 proteins in the inner ear were slightly higher in MsrB3+/− and MsrB3−/− mice than in the wild-type. Thus, it is possible that the increased activity of the MsrB1 isozyme compensated for the loss of MsrB3 activity in MsrB3+/− and MsrB3−/− mice. Taken together, these results suggest that the loss of MsrB3 activity controlling redox homeostasis is not a major pathological cause leading to hearing loss in MsrB3−/− mice.
Our TUNEL and anti-active caspase immunoassays demonstrated that MsrB3 deficiency induces apoptotic cell death in inner ear tissues. We consistently observed that knocking down MsrB3 expression in mammalian cell lines resulted in apoptotic cell death, confirming the anti-apoptotic role of MsrB3 (data not shown). The apoptosis of sensory hair cells is an important factor contributing to many types of hearing impairments including age-related, noise-induced, ototoxic and monogenic forms of hearing loss (28–30). These results suggest that the role of MsrB3 in protecting cochlear hair cells might not be directly mediated by its well-known protein reduction activities.
In summary, we report the generation and characterization of an MsrB3−/− mouse model, which is useful for studying the pathological mechanisms of DFNB74 hearing loss in humans and understanding the function of MsrB3 in auditory hair cells. We found that MsrB3 deficiency led to the progressive degeneration of stereociliary bundles, followed by the apoptotic loss of hair cells, which ultimately resulted in profound hearing loss. Our results provide novel insights into the pathogenesis of human deafness caused by MSRB3 mutations.
MATERIALS AND METHODS
Generation of MsrB3 knockout mice
The MsrB3 knockout-targeting construct was designed to abolish exon 7 within the gene by replacing this exon with a neomycin resistance gene. Exon 7 harbors the active site of the enzyme, including the catalytic Cys residue. To clone the 5′ left arm, a first 1509-bp (LA1) fragment, a second 2778-bp (LA2) fragment downstream of LA1 and a third 3003-bp (LA3) fragment downstream of LA2 were amplified using PCR from 129/SvJ mouse genomic DNA and inserted into the NotI/NheI, NheI/XhoI and XhoI/SalI sites, respectively, in the pDrive cloning vector. A 2681-bp fragment (RA) coding for the 3′ right arm was amplified using PCR and inserted into the SalI/NheI sites in the pDrive cloning vector. All cloned fragments were verified using DNA sequencing. The LA1 fragment containing the NotI/NheI sites was then ligated to the same sites of pDrive containing LA2. The MsrB3 knockout construct was prepared as follows: the RA fragment containing the SalI/NheI sites was cloned into the XhoI/NheI sites of the pOSDupDel-Neo vector. The LA3 fragment containing the XbaI/SalI sites was then cloned into the same sites of pOSDupDel containing the RA fragment. The LA1/2 fragment containing the NotI/XhoI sites was then cloned into the same sites of pOSDupDel containing the RA and LA3 fragments. The targeting construct contained a 7.69-kb 5′ left arm, a 1.3-kb neomycin cassette and a 2.68-kb 3′ right arm. We electroporated 129/SvJ embryonic stem (ES) cells with the targeting construct after it had been linearized using NotI. Positive recombinant ES cells were identified using PCR and then microinjected into C57BL/6 blastocysts. The resulting chimeric mice were then mated with C57BL/6 mice. Germ line transmission was confirmed using Southern blot analysis (Supplementary Material, Fig. S1). Genomic DNA was digested with SphI and then probed with a 651-bp 3′ probe. The knockout allele yielded a 5.1-kb band, whereas the wild-type allele generated a 6.9-kb band.
Animal preparation and experiments
All mice used in the animal experiments were obtained by breeding heterozygous MsrB3 knockout mice of a 129/SvJ and C57BL/6 mixed background. The mouse genotypes were confirmed using PCR analysis. The animal experiments were conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee of Kyungpook National University and Yeungnam University.
Genomic DNA was isolated from mouse tails. PCR genotyping was performed using the following primers: one forward primer (5′-CAAAGTCTGTCTGGAGCCATC-3′) and two reverse primers for the wild-type (5′-AGAACTGCCATCTAGACGAGAGAG-3′) and knockout (5′-GATGATCTCGTCGTGACCCATG-3′) alleles. The PCR products were 538 bp for the wild-type allele and 393 bp for the knockout allele.
Auditory brainstem response
The animals were intramuscularly anesthetized using a mixture of tiletamine–zolazepam (1.8 mg/100 g) and xylazine hydrochloride (0.7 mg/100 g). Body temperature was monitored using a rectal thermometer, and the animals were maintained on a heating pad at 37°C. All experiments were performed in a soundproof room. Sound-evoked ABRs were measured using a Tucker Davis Technology (TDT) system as previously described (31). Briefly, needle electrodes from the head stage (RA4LI, TDT) were connected to a pre-amplifier (RA4PA, TDT) to record ABRs and inserted into the vertex (+ charge), mastoid (− charge), and hind leg (ground). Acoustic stimuli were calibrated at 90 dB SPL (sound pressure level) with 8, 16 and 32 kHz tone bursts using calibration software (SigCalRP) in TDT System 3 and the probe microphone system. Tone burst stimuli with a 1-ms rise/fall and a 5-ms plateau or transient click stimuli at frequencies of 8, 16 and 32 kHz were generated using signal design software (SigGenRP) and applied. The signals needed for the stimuli were generated using SigGenRP and an RP2.1 real-time processor and then transmitted in sequence through a programmable attenuator (PA5, TDT), speaker driver (ED1, TDT) and electrostatic speaker (EC1, TDT). Stimuli were generated every 36.1 ms for 500 repetitions from 90 dB SPL to the acoustic threshold in 5-dB decrements at every frequency. The phase of the stimuli was reversed after each stimulus to reduce noise caused by repetitive stimuli.
Four different polyclonal anti-MSRB3 antibodies were tested and their specificities were compared: two new antisera prepared by immunizing rabbits with a recombinant human MSRB3A, a rabbit polyclonal antiserum against human MSRB3 [as previously described (15)] and a rabbit polyclonal anti-human MSRB3 obtained from Sigma-Aldrich. To validate specificity against mouse MSRB3, we performed western blotting and immunohistochemistry analyses using MsrB3+/− and MsrB3−/− mouse tissues and cells transfected with a mouse MsrB3–GFP fusion construct (15). Among the evaluated antibodies, the previously described anti-human MSRB3 antibody (15) exhibited the highest specificity for mouse MsrB3 (Supplementary Material, Figs S4 and S12). Thus, this antibody was used for further study.
Inner ears were isolated from E12.5, E13.5, E14.5, E15.5 and E18.5, as well as P0 and P18 mice and fixed with 4% paraformaldehyde in phosphate buffered saline (PBS) for 1 day at 4°C. The inner ears from P18 mice were then decalcified in 10% ethylenediaminetetraacetic acid in PBS for 2 days at 4°C. All tissues were dehydrated with 30% sucrose in PBS and a 1:1 30% sucrose:OCT compound (Optimal Cutting Temperature compound, Surgipath FSC22, Leica Microsystems) for 1 day each at 4°C. The inner ears were then embedded in OCT compound. Serial sections were prepared by cutting the tissues into 10 μm sections using a −27°C cryostat and dried for 30 min at room temperature (25°C). After permeabilization with 0.5% Triton X-100 in PBS for 5 min, the tissue sections were blocked with a blocking solution (5% normal goat serum and 2% bovine serum albumin in PBS) for 1 h and incubated overnight at 4°C with the rabbit anti-MsrB3 (1:50) and mouse anti-Myo7a antibody (1:100, DSHB 138-1) diluted in the blocking solution. Goat anti-rabbit IgG antibody conjugated with Alexa 555 fluorescein (Molecular Probes/Invitrogen) and goat anti-mouse IgG antibody conjugated with Alexa 488 fluorescein (Molecular Probes/Invitrogen) were used as secondary antibodies and diluted in the blocking solution (1:1000); the diluted antibody was incubated with the tissue sections for 1 h at room temperature. The tissue sections were counter-stained with 4′-6-diamidino-2-phenylindole (DAPI) for 5 min at room temperature and visualized using a Zeiss DE/AX10 Imager A1 fluorescence microscope system (Carl Zeiss).
Inner ears were isolated from P6 and P15 MsrB3+/+ mice and P5 and P13 MsrB3+/− and MsrB3−/− mice; the ears were rapidly fixed by infusion with 4% paraformaldehyde in PBS via the round window and immersed in the same fixative solution at 4°C for 1 h. After fixation, the organ of Corti and vestibule were dissected using a dissecting microscope. The dissected samples were immunostained with an anti-Msr3B antibody following the immunostaining procedure described above. Next, F-actin was labeled by incubation with Alexa Fluor 488 phalloidin (Molecular Probes/Invitrogen). After mounting in Fluoromount (Sigma-Aldrich), the samples were visualized using a confocal laser scanning microscope (LSM700, Carl Zeiss) and the ZEN 2009 program.
Scanning electron microscopy
Cochleas were immediately isolated from the euthanized mice and perfused carefully through the round window using 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 m sodium cacodylate buffer (pH 7.4) after a hole was made at the top of the cochlea. The perfused cochleas were immersed in the same fixation mixture for 1 h at room temperature. The lateral wall, Reissner's membrane and the tectorial membrane were removed via dissection under a dissecting microscope and fixed overnight at 4°C in 0.1 m sodium cacodylate buffer (pH 7.4) containing 2.5% glutaraldehyde, 2 mm calcium chloride and 3.5% sucrose. After fixation, the dissected specimens were rinsed three times for 20 min at 4°C with 0.1 m sodium cacodylate buffer containing 2 mm calcium chloride. The specimens were treated post-fixation using the osmium tetroxide (OsO4)-thiocarbohydrazide method developed by Hunter-Duvar (32). Briefly, specimens were immersed in 1% OsO4 for 1 h at 4°C and then placed in saturated thiocarbohydrazide for 20 min at room temperature. This specimen treatment was repeated three times at room temperature. The specimens were then dehydrated using a graded series of ethanol solutions, dried using a critical point drier (HCP-2, Hitachi), affixed on a stub and coated with platinum using a sputter coater (E1030, Hitachi). The coated specimens were mounted on a stub holder and viewed using a cold-field emission scanning electron microscope (S-4300, Hitachi) operated at 15 kV.
Transmission electron microscopy
For ultra-thin sectioning, the temporal bones were rapidly removed and fixed in 2.5% glutaraldehyde and 2% paraformaldehyde in 1× PBS (pH 7.4) for 24 h. After fixation, the samples were decalcified in 0.2 m EDTA (pH 7.4) with 2.5% glutaraldehyde and 2% paraformaldehyde at 4°C for 2 weeks and then post-fixed using osmium tetroxide (OsO4) for 30 min. The specimens were rinsed three times in 0.1 m cacodylate buffer at 4°C and then dehydrated using a graded series of ethanol solutions. To remove residual ethanol, the dehydrated specimens were treated twice with propylene oxide at room temperature. Then, the specimens were placed in capsule caps filled with fresh Epon resin and incubated overnight at 65°C. Ultra-thin sections (60 nm) were obtained with a Leica Ultracut UCT ultramicrotome (Leica Microsystems) using a diamond knife (Diatome) and collected on 15 nickel grids (400 mesh). Grids were examined on a transmission electron microscope (H-7500, Hitachi).
Inner ears were isolated from P20 mice and fixed with 4% paraformaldehyde in PBS for 1 day at 4°C. After fixation, the inner ears were decalcified in 10% ethylenediaminetetraacetic acid in PBS for 2 days at 4°C. The samples were then dehydrated using a series of ethanol solutions, cleared in xylene and embedded in paraffin. Five-micrometer sections were cut on a RM2125 RT microtome (Leica Microsystems) and stained with hematoxylin and eosin (H&E).
TUNEL assay and active caspase 3 detection
Cochlear sections obtained from P20 MsrB3+/− and MsrB3−/− mice were prepared as described above. The TUNEL assay was performed using an In Situ Cell Death Detection Kit with Fluorescein (Roche Diagnostics) according to the manufacturer's instructions. Activated caspase 3 was detected using an anti-cleaved caspase 3 antibody (Cell Signaling Technology) using the immunostaining procedure described above. The number of apoptotic cells was quantified by counting TUNEL(+) cells in the organ of Corti of MsrB3+/− and MsrB3−/− mice, and the difference was calculated by a t-test (SPSS for Windows, version 18, SPSS Inc.).
Msr enzyme assay
The Msr enzyme assay used protein from the whole otic capsule including both whole membranous and bony labyrinth of MsrB3+/+, MsrB3+/− and MsrB3−/− mice. The reaction mixture (100 μl) contained 50 mm sodium phosphate (pH 7.5), 50 mm NaCl, 20 mm dithiothreitol, 200 μm dabsylated methionine-R-sulfoxide (for MsrB) or methionine-S-sulfoxide (for MsrA) and 200 μg of crude protein. The reaction was conducted at 37°C for 30 min, and the reaction product dabsyl-Met was quantified using high performance liquid chromatography.
Analysis of methionine sulfoxide levels
The tissues used in the Msr enzyme assay were homogenized in a lysis buffer containing 50 mm Tris–HCl (pH 7.5), 150 mm NaCl, 1% Triton X-100, 10% glycerol and 1% 2-mercaptoethanol including Protease Inhibitor Cocktail Set 1 (Calbiochem). Supernatants were obtained via centrifugation. Protein concentrations were determined using a BCA protein assay kit (Thermo Fisher Scientific). The hydrogen peroxide (H2O2)-treated samples were also incubated in 100 mm H2O2 for 2 h at 37°C in the presence of 2% SDS to inhibit peroxidase activity. The protein samples were resolved using 10% SDS–polyacrylamide gel electrophoresis. The levels of methionine sulfoxide in the proteins were determined using a methionine sulfoxide immunoblotting kit (Cayman Chemical).
Measurement of carbonylated proteins
Carbonylated protein levels in the tissues used in the Msr enzyme assay were measured using the Oxyblot kit (Chemicon) according to the manufacturer's instructions. The blots were quantitatively analyzed using ImageJ software.
Conflict of Interest statement. None declared.
This work was supported by the National Research Foundation of Korea Grant (2011-0006178 to H.Y.K.) and the Korea Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (A111774 to K.Y.L.).