Cochlear function in mice with only one copy of the prestin gene
1 Department of Communication Sciences and Disorders
2 Department of Neurobiology and Physiology, Northwestern University, Evanston, IL 60208, USA
3 Department of Developmental Neurobiology, St Jude Children's Research Hospital, Memphis, TN 38105, USA
4 Section on Structural Cell Biology, NIDCD, NIH, Bethesda, MD 20892, USA
Abstract
, 百拇医药
Targeted deletion of the prestin gene reduces cochlear sensitivity and eliminates both frequency selectivity and outer hair cell (OHC) somatic electromotility. In addition, it has been reported by Liberman and colleagues that F2 generation heterozygotes exhibit a 6 dB reduction in sensitivity, as well as a decrease in protein and electromotility. Considering that the active process is non-linear, a halving of somatic electromotility would be expected to produce a much larger change in sensitivity. We therefore re-evaluated comparisons between heterozygotes and wildtype mice using both in vivo and in vitro electrophysiology, as well as molecular biology. Data reported here for F3–F5 generation mice indicate that compound action potential thresholds and tuning curves, as well as the cochlear microphonic, are similar in heterozygotes and wildtype controls. Measurements of non-linear capacitance in isolated OHCs demonstrate that charge density, as well as the voltage dependence and sensitivity of motor function, is indistinguishable in the two genotypes, as is somatic electromotility. In addition, both immunocytochemistry and western blot analysis in young adult mice suggest that prestin protein in heterozygotes is near normal. In contrast, prestin mRNA is always less than in wildtype mice at all ages tested. Results from F3–F5 generation mice suggest that one copy of the prestin gene is capable of compensating for the deleted copy and that heterozygous mice do not suffer peripheral hearing impairment.
, 百拇医药
Introduction
The discovery of prestin (Zheng et al. 2000), the motor protein in cochlear outer hair cells (OHC), resulted in several efforts to determine its functional significance. For example, targeted deletion of the prestin gene reduces sensitivity (Liberman et al. 2002) and eliminates frequency selectivity (Cheatham et al. 2004). It is also known that OHCs harvested from knockout mice do not exhibit somatic electromotility (Liberman et al. 2002). In other words, the length changes associated with voltage-dependent conformational changes in the motor protein (Brownell et al. 1985; Kachar et al. 1986; Ashmore, 1987; Santos-Sacchi & Dilger, 1988) are not observed. Because of the significance of gene dosage for inherited disease, it is also of interest to compare wildtype controls, not only with homozygous mice, but also with heterozygous mice having only one copy of the prestin gene. These evaluations are required in order to learn if prestin protein is reduced in heterozygotes and, if so, if this reduction is associated with changes in cochlear function. Liberman et al. (2002) suggested that somatic electromotility in heterozygotes was about half that in wildtype controls, cochlear sensitivity was decreased by 6 dB and prestin protein levels were reduced. These results are puzzling because a halving of motility would be expected to result in a much greater change in sensitivity based upon what is known about cochlear amplification (Mountain et al. 1983; Patuzzi et al. 1989; Fettiplace & Ricci, 2003; Santos-Sacchi, 2003).
, http://www.100md.com
In this project, we re-evaluate comparisons between heterozygotes and wildtype controls to determine if they display different phenotypes. If auditory function were compromised, as suggested by the Liberman et al. (2002) report, then heterozygotes would exhibit haploinsufficiency. In this condition, a single copy of the prestin gene would be incapable of providing sufficient protein to assure normal auditory function. In order to test this hypothesis, compound action potential (CAP) threshold curves and CAP tuning curves were acquired from a recording electrode placed on the round window to determine if cochlear sensitivity and frequency selectivity are different in heterozygotes and wildtype mice. In addition to the neural responses, the cochlear microphonic (CM) was also recorded. Although inner hair cells (IHC) can contribute to this response, the CM is thought to primarily reflect OHC receptor currents (Dallos & Cheatham, 1976b). Following euthanasia, animals were evaluated in three additional experiments. First, whole-cell, voltage-clamp recordings were used to determine the amplitude of somatic electromotility and the voltage-dependent non-linear capacitance (NLC), its electrical correlate (Ashmore, 1990; Santos-Sacchi, 1991). Second, prestin protein expression was assessed using immunocytochemistry and laser confocal microscopy, as well as western blot analysis. Third, mRNA was quantified using real-time RT-PCR. Results from these experiments at the molecular, cellular and systems levels were compared to determine the influence of one versus two copies of the prestin gene on cochlear function.
, http://www.100md.com
Methods
In the in vivo physiological experiments, F4 generation wildtype and heterozygous mice were used as subjects while both F3 and F4 generations were combined for the in vitro studies. Wildtype mice were generated by mating two wildtype parents, while breeding a wildtype with a prestin knockout mouse produced the heterozygotes. For the molecular biology experiments, including immunocytochemistry, western blot analysis and real-time RT-PCR, wildtype mice were bred with heterozygotes to obtain F4 and F5 litters with about equal portions of the two genotypes in a single litter. All animals were on a mixed 129SvEv/C57BL6 background. The details of how the knockout mice were generated are available in a previous report (Liberman et al. 2002). All procedures were approved by the National Institutes of Health and by Northwestern University's Institutional Review Board.
, 百拇医药
In vivo physiology
Cochlear in vivo physiology was studied in wildtype (WT, +/+) and heterozygous (Het, +/–) mice anaesthetized with sodium pentobarbital (80 mg kg–1, I.P.). Additional doses were given throughout the experiment to maintain a surgical level of anaesthesia. During both surgery and data collection, the headholder was heated to prevent cooling of the cochlea (Shore & Nuttall, 1985; Ohlemiller & Siegel, 1992). Electrical potentials were recorded using a round window electrode. At the end of each experiment, the animals were humanely terminated with an overdose of anaesthetic.
, http://www.100md.com
Compound action potential thresholds were acquired using a tracking program, which determined the sound pressure level necessary to generate a 10 μV N1/P1 voltage (Taylor & Creelman, 1967; Gummer et al. 1987). The CAP tuning curves were collected using a simultaneous, tone-on-tone masking paradigm (Dallos & Cheatham, 1976a). In this procedure, the level of a 12 kHz probe tone, presented alone, was adjusted to generate a response of 45 μV measured between the first negative and subsequent positive peak of the CAP. Masker frequency and level were then varied to produce a 3 dB decrease in the probe response, i.e. to 32 μV. The masker was presented in alternating phase to minimize the CM. For all neural measurements a custom-made, programmable filter was used to bandpass filter electrical signals between 0.2 and 3.0 kHz. In contrast, CM input–output functions were obtained using lowpass filtering at 49.9 kHz to prevent aliasing. In all cases, however, a differential amplifier, with gain of 1000 and a passband between less than 1 Hz and 50 000 Hz, was utilized. Peak-to-peak CM voltages were determined offline by fast Fourier transformation (FFT) of averaged response waveforms. Further details about these procedures are provided elsewhere (Pearce et al. 2000; Cheatham et al. 2004).
, http://www.100md.com
Non-linear capacitance
In vitro physiological experiments were performed on OHCs harvested from 11 heterozygous and 12 wildtype mice following euthanasia with sodium pentobarbital (200 mg kg–1, I.P.). Sense organs from both ears were subjected to enzymatic treatment with type IV Collagenase (Sigma-Aldrich, St Louis MO, USA) in Leibovitz's L-15 solution (Gibco, Invitrogen, Carlsbad CA, USA), at a concentration of 0.5 mg ml–1 for 10 min. Gentle pipetting allowed individual OHCs to be separated from the cellular matrix. Gigohm seals were obtained at the basal pole of dissociated OHCs. The cells were then whole-cell voltage clamped at room temperature using an Axopatch 200A amplifier (Axon Instruments, Union City, CA, USA). Capacitive currents were studied in isolation by using an ionic blocking solution composed of (mM): 99.2 NaCl, 20 TEA, 20 CsCl, 2 CoCl2, 1.47 MgCl2, 2 CaCl2, 10 Hepes and 5 dextrose. The solution was adjusted to pH 7.2 with CsOH and to 300 mosmol l–1 with dextrose. The solution inside the patch pipette contained (mM): 140 CsCl, 10 EGTA, 2 MgCl2 and 10 Hepes and was also adjusted to pH 7.2 with CsOH and to 300 mosmol l–1 with dextrose. Capacitance measurements were obtained with a high-resolution (2.56 ms) protocol (Santos-Sacchi & Navarrete, 2002) using a two-sine voltage stimulus (10 mV peak at 390.6 and 781.2 Hz) in which sinusoids were superimposed on one cycle of a 2.5 Hz sinusoidal voltage ranging between ±175 mV and starting from a holding potential of 0 mV. Subsequent FFT analysis utilized parameter variations provided by Pusch & Neher (1998). All data collection and some analyses were performed with jClamp (Scisoft, CT, USA) using a National Instruments PCI-6052E board. Fits to the capacitance data were made in IgorPro (Wavemetrics, Lake Oswego, OR, USA). The maximum charge transferred through the membrane's electric field (Qmax), the slope factor of the voltage dependence (), the voltage at peak capacitance (V) and the linear membrane capacitance (Clin) were determined. Because Clin reflects the total surface area of the membrane, we also obtained an adjusted Clin value, ClinAdj, to reflect only that part of the membrane containing motors (Huang & Santos-Sacchi, 1993). This adjusted datum was used to compute the charge density (Qmax/ClinAdj), thereby normalizing for cell size (Oliver & Fakler, 1999). In other words, charge density was used as a normalized metric for the voltage-mediated charge transfer through only that part of the membrane containing motor protein.
, http://www.100md.com
Somatic electromotility
A Leica DM-IRB inverted microscope with differential interference contrast (DIC) optics was used for observing voltage-clamped OHCs. Images were acquired at 30 fps/360 x 240 pixel resolution with a Panasonic WVCD22 video camera and digitized using a Hauppage WinTV GO PCI card. Script protocols were written in jClamp to acquire audio video interleaved (avi) images in synchrony with the voltage stimulus used for extracting the capacitance data. Following transfer from avi to QuickTimevideo format and enlargement to 640 x 480 pixel resolution, the files were analysed using DIAS software (Soll Technologies, Iowa city, IA, USA) running on a Macintosh G4 computer. Cell edges were either determined manually or automatically by the software from individual movie frames. To compensate for the variability in cell length, normalization to each cell's maximum axial length was performed to obtain the amplitude of movement as a percentage. The analytical limit of this video imaging technique was also calculated to evaluate the statistical validity of the procedure. For example, measurement of each OHC's length, at a constant holding potential of 0 mV, always gave a standard deviation of less than 0.5 pixels using at least 10 frames for each cell. Thus, the motile length changes that we observed, ranging from 3 to 5 pixels, are not attributed to analytical or artifactual fluctuations.
, 百拇医药
Immunocytochemistry
Mice were cardiac perfused first with heparinized saline and then with 4% paraformaldehyde. After 4 h post-fixation at room temperature, cochleae from +/+ and +/– mice were dissected, decalcified overnight in 10% EDTA and treated with 0.3% Triton X-100/PBS. Ovalbumin was used to block non-specific binding. The tissue was then exposed to an antimouse prestin antibody (1 : 2000) raised against an epitope in prestin's C-terminal (Matsuda et al. 2004) followed by the secondary antibody (1 : 200), Alexa488-conjugated antirabbit IgG (Molecular Probes, Eugene, OR, USA). Samples from the apical half of the cochlea were mounted on glass slides with cilia pointing up using Fluoromount-G and viewed on a Leica confocal microscope equipped with an argon laser with excitation at 488 nm; emission at 510 nm. Photometric gain was standardized so that the brightest signal in the labelled region was below the maximum (grayscale intensity 4095), ensuring that there were no pixels at saturation. Confocal fluorescent images of OHCs from wildtype and heterozygous mice were taken at the same time and with the same fixed photometric gain. A series of z-sections from the apex to the base of the cell, at intervals of 1.0 μm, were acquired at a resolution of 1024 x 1024 with 12 bits per pixel through a 40x objective (NA 1.25). A single pixel corresponded to a 183 x 183 nm square in the specimen plane. In order to obtain a semiquantitative measure of prestin protein expression in the OHC lateral membrane, captured fluorescent images were analysed with the Integrated Intensity feature of the MetaMorph imaging system (Universal Imaging, Molecular Devices Corp., Downingtown, PA, USA). The threshold was set to 1000 to avoid non-specific signals. Areas of immunostaining were then rendered in number of pixels so that the fluorescent intensity/area between +/+ and +/– OHCs could be calculated. Protein expression was determined at approximately the same distance from the helicotrema in wildtype and heterozygous mice at 3 and at approximately 8 weeks of age. Only sections immunostained in the same experiment were compared to avoid interexperimental variability.
, http://www.100md.com
Western blot analysis
In order to ensure that comparisons were made between intact cochlear samples, cochleae were carefully dissected and trimmed by the same experimenter. The samples were then collected in CelLytic MT Cell Lysis Reagent (C3228 Sigma-Aldrich, St. Louis MO, USA) supplemented with phenylmethylsulphonyl fluoride (PMSF 100 μg ml–1), a protease inhibitor cocktail (1 : 50, P8340, Sigma-Aldrich, St. Louis, MO, USA), thimerosal (1 : 10 000) and DNase (10 μg ml–1). After homogenization with a Kontes pellet pestle (Fisher Scientific International, Hampton NJ, USA), lysates were centrifuged at 3000 g for 15 min at 4°C. The cochlear sample was then diluted in an equal volume of 2x loading buffer (lithium dodecyl sulphate (LDS) Laemmli buffer with 100 mM dithiothreitol (DTT)), heated to 100°C for 5 min and spun 10 min before loading onto a 7.5% LDS-acrylamide gel. After electrophoresis, the gel proteins were electrotransferred onto a nitrocellulose membrane, which was blocked with blocking solution (2% non-fat dry milk/2% bovine serum albumin in PBS) and then incubated with the antimouse prestin antibody, also in blocking solution. After washing with Tris-buffered saline containing 0.2% Tween 20 (TBST), the membrane was incubated with an HRP-conjugated secondary antibody (1 : 10,000, donkey antirabbit IgG, Jackson ImmunoResearch Laboratories, West Grove, PA, USA) in blocking solution containing 1% normal donkey serum. Finally, after thoroughly washing with TBST, prestin expression was detected using an enhanced chemiluminescence western blotting detection system (Amersham Biosciences, Piscataway, NJ, USA). The apparent molecular masses were calculated by non-linear curve fitting of the molecular mass standards and indicated on each gel. After development, the film was photographed and analysed using Kodak ID Image Analysis Software. This program allows one to select regions of interest around each band and to compare these to similar regions representing the background. Means and standard deviations were obtained for the integrated intensity minus the background in arbitrary units. Protein expression was quantified at postnatal days 9 (P9) and 21 (P21), using two cochleae per sample, and at postnatal day 63 (P63) using a single cochlea.
, http://www.100md.com
Because quantification of prestin protein using enhanced chemiluminescence can be compromised by enzyme/substrate kinetics, we determined the range where antiprestin is able to detect quantitative differences in its antigen. Different amounts of prestin protein synthesized in yeast hosts were applied to LDS-PAGE and blotted with antiprestin. The intensities of prestin blots were then acquired at three different exposure times. By comparing the band intensity of our cochlear samples with the yeast-prestin data, it was determined that our results fell within the quasi-linear range. In addition, at P63 when only one cochlea was used, similar results were obtained on the opposite ear, indicating a high degree of reproducibility.
, http://www.100md.com
RNA isolation
Both cochleae from each mouse were carefully dissected and trimmed in RNAlater solution (Ambion, Austin, TX, USA) by the same experimenter. The two cochleae were homogenized in 300 μl lysis buffer using the Absolutely RNA RT-PCR Miniprep Kit (Stratagene, La Jolla, CA, USA). Total RNA was immediately isolated and stored at –80°C. Even though isolated RNA was treated with DNase during the RNA isolation procedure, the PCR was performed to check for genomic DNA contamination. The forward primer was 5'-CATGCTGAAGAAAATGAAATCCCTGCAGAGACC; the reverse primer, 5'-GCTGGAGTACCCCAGTGCTTAT-GCCCGAGACC. Cycling conditions were as follows: denature at 95°C for 1.5 min followed by 45 cycles at 95°C for 1 min, 60°C for 1 min and then at 72°C for 2 min. Only samples without contamination were used in real-time PCR reactions.
, 百拇医药
Real-time RT-PCR
Quantitative polymerase chain reaction with reverse transcription (RT-PCR) was used to assay prestin mRNA from litters at P9, P21 and P49. The reverse transcription reaction was performed on a 15 μl sample at 42°C for 90 min with StrataScript reverse transcriptase (Stratagene), in a 1x buffer containing 6 ng μl–1 of random primer (9mers), 1 mM dNTP and 300 ng cochlear RNA from each mouse. PCR reactions were performed on an iCycler machine (Bio-Rad, Hercules, CA, USA) using Brilliant QPCR Master mix (Stratagene, La Jolla, CA, USA). Details of the cycling conditions were described by Liberman et al. (2002). Specific prestin primer pairs for real-time PCR were 5'-TCGGGCATAAGCACTGGG-3' (forward), 5'-ACGGCTGCCAGCATGG-3' (reverse) and, for the TaqMan probe, 5'/56FAM-TACTCCAGCTT-CCCCAAGGCTTAGCCT-/3BHQ-1–3'. Universal 18S internal standards were purchased from Ambion. The sequence of the 18S probe was 5'-/56-FAM/TGGC-TGAACGCCACTTGTCCCTCTAA/3BHQ-1-3'. Each real-time PCR was performed at the same time on mice from the same litter except for the data at P21 where samples from two litters were combined. Samples from each mouse were repeated in triplicate. Prestin cDNA from one of the samples was diluted and used to generate standard curves. The relative quantities of prestin mRNA were calculated from the standard curves and normalized to 18S (Stankovic & Corfas, 2003).
, http://www.100md.com
Results
In vivo physiology
Cochlear sensitivity was measured in wildtype and heterozygous mice by collecting CAP thresholds at frequencies between 1.8 and 50.0 kHz. The CAP thresholds in Fig. 1 provide a physiological representation of the animal's audiogram. In Fig. 1A, mean CAP thresholds are shown for both genotypes. Fig. 1B provides the threshold difference between animals having one versus two copies of the prestin gene, with positive values indicating that heterozygotes display better sensitivity than controls. Not only are the differences relatively small, they are less than the standard deviations. The latter are shown for both genotypes with error bars representing standard deviations in heterozygotes and broken lines representing standard deviations in wildtype mice. The larger deviations at high frequencies are probably related to the age-related hearing loss associated primarily with the 129SvEv strain (Henry & Chole, 1980; Shnerson & Pujol, 1981; Li & Borg, 1991; Zheng et al. 1999; He et al. 2004), which supplied the embryonic stem cells used for gene targeting.
, http://www.100md.com
A, mean CAP thresholds for wildtype (, broken lines) and heterozygous (, continuous lines) mice. The average age for wildtype mice was 51 days; that for heterozygotes, 47 days. B, the difference in CAP thresholds (continuous line). Standard deviations for heterozygotes are shown with error bars in the positive direction only; for wildtype mice, with the broken lines.
Figure 2A displays CAP tuning curves (Dallos & Cheatham, 1976a) for a probe frequency at 12 kHz for wildtype and heterozygous mice. CAP tuning curves represent average iso-response functions, obtained in the frequency–intensity plane, for a small group of auditory nerve fibres with characteristic frequencies corresponding to the 12 kHz place. As such, they represent the frequency selectivity of the peripheral auditory system. The average probe level, which generated the 45 μV N1/P1 voltage for the 12 kHz probe alone, was 53 dB for both genotypes. Tuning was evaluated by measuring Q10 values, which were 5.3 in both wildtype and heterozygous mice. These values were obtained by dividing the bandwidth at 10 dB above the tip of the tone-on-tone masking function by probe frequency. Figure 2B shows the difference in decibels between the two curves. Again, the differences are less than the standard deviations, indicating that frequency selectivity is similar in the two genotypes. For both the CAP thresholds and CAP tuning curves, the differences between wildtype and heterozygous mice are probably related to normal variability and not to how many copies of the prestin gene are available.
, 百拇医药
A, average CAP tuning curves for wildtype (broken lines) and heterozygous (continuous lines) mice. Isolated symbols indicate the probe level, which was 53 dB for both genotypes. The average age for wildtype mice was 51 days; for heterozygotes, 48 days. As in Fig. 1, the difference between the tuning curves is given in B along with the standard deviations.
In addition to measurements of neural responses, the CM was also recorded. This gross-potential response is thought to represent OHC receptor currents (Dallos & Cheatham, 1976b). Figure 3 provides CM input–output functions recorded at 6, 12, 16 and 32 kHz for the F4 generation mice used in the in vivo experiments. In order to facilitate comparisons with Figs 1 and 2, the magnitudes are plotted as decibels re: 1 μV. The lower panel of Fig. 3 shows differences in CM responses between wildtype and heterozygous mice. The differences for F4 generation mice are smaller than the standard deviations and always less than 6 dB.
, http://www.100md.com
The top row provides CM data at 6, 12, 16 and 32 kHz. The average age for wildtype (heterozygous) mice was 49 (47) days. Plots at the bottom display the difference between CM input–output functions obtained in the two genotypes (continuous line). Standard deviations are represented as bars (broken lines) for heterozygous (wildtype) mice.
Taken together, the sensory and neural responses recorded from the round window in wildtype and heterozygous mice are very similar. In fact, we tested each individual data point statistically (Student's t test), and there was no difference between genotypes (P < 0.05). This statement applies to all three measurements, i.e. the CAP thresholds, the CAP tuning curves and the CM input–output functions. We also tested differences between CAP thresholds in +/+ and +/– animals using a mixed factor design analysis of variance (Collyer & Enns, 1987). The F score between the two groups was 0.004, which indicates that responses in the two genotypes are indistinguishable at the 95% level. A similar computation was performed for CM input–output functions obtained at 16 kHz. Comparisons yielded a vanishingly small F score and P= 0.99, again attesting to no difference between physiological responses in wildtype and heterozygous mice.
, 百拇医药
Non-linear capacitance
It is well known that somatic motility is associated with the translocation of electric charges across the OHC membrane, and that this charge movement imparts a bell-shaped voltage dependence to the membrane capacitance (Ashmore, 1990; Santos-Sacchi, 1991; Dallos & Fakler, 2002). A representative example from a wildtype and a heterozygous mouse is provided in Fig. 4A. The capacitance functions peak at the membrane voltage that is most effective in generating a motile response, i.e. around –70 mV.
, http://www.100md.com
A, capacitance functions are provided for a representative wildtype control (broken lines) and a heterozygote (solid lines). Parameter fits to the data appear as thin continuous lines. The Qmax was 872.7 (702.1), 29.4 (29.9), V–66.3 (–69.6) and Clin 6.27 (5.7) for the wildtype mouse (heterozygote). B, normalized functions are provided for wildtype (broken line) and heterozygous (bold line) mice. The functions represent the derivatives of a first-order Boltzmann function, normalized to ClinAdj in the form:
, 百拇医药
The two plots were obtained from the average values of the parameters (Qmax, , V and ClinAdj) shown in Table 1. C, somatic electromotility. Video images of individual OHCs were used to obtain measures of somatic electromotility. The peak voltage-dependent length change was expressed as a percentage of each OHC's maximum length. The data were then normalized by giving the wildtype mean percentage change in length a value of 1.0 and then plotting the mean for heterozygotes relative to this value.
, 百拇医药
The capacitance data for each OHC were fitted with the derivative of a first-order Boltzmann function (Santos-Sacchi, 1991; Oliver et al. 2001). The means and standard deviations of the capacitance function parameters used to fit the data are provided in Table 1, along with P values obtained using the Student's t test and Het/WT ratios. The slope factor, alpha, and the voltage at peak capacitance, V, are virtually identical in the two populations. In contrast, the maximum charge displacement (Qmax) and the linear capacitance (Clin) are significantly different (P\#8810; 0.01, Student's t test). Because the linear component of the membrane capacitance is proportional to the cell surface area (length) (Huang & Santos-Sacchi, 1993; Santos-Sacchi & Navarrete, 2002), the smaller average Clin in heterozygotes suggests that cells harvested from wildtype mice were, on average, somewhat longer than OHCs harvested from mice having only one copy of the prestin gene. This difference in cell length does not necessarily imply that OHCs at a given location are shorter in heterozygotes, although a very small difference was reported by Liberman et al. (2002) in F2 generation mice. This caution is advised because OHCs are harvested from the entire cochlea and data are obtained from only a few cells per mouse. It is therefore possible that wildtype OHCs originated from a more apical location.
, 百拇医药
The difference in Clin between genotypes suggests that these data should be adjusted to reflect only that part of the lateral membrane containing prestin. Although this information is not available for mice, the distribution of OHC somatic motility was determined using electrical amputation by Huang & Santos-Sacchi (1993). For cells harvested from the apex of the guinea pig cochlea (length 70 μm; radius 5 μm), 27% of the OHC's surface area was found to lack motor function. This result is consistent with the immunocytochemistry, which shows that prestin is not expressed in the hair bundle or in the nuclear region (Weber et al. 2002). In order to compute the OHC surface area lacking prestin, Huang & Santos-Sacchi (1993) modelled the basal region of the cell as a half sphere with a 2 μm extension. The apical area was composed of a 5 μm section surrounding the cuticular plate plus the surface area of the stereocilia. Each OHC was assumed to have 50 stereocilia, each with a radius of 0.15 μm and a length of 6 μm. In order to obtain an estimate of the surface area for short mouse OHCs (length 25 μm), we made the following adjustments. First, we examined radial sections of the organ of Corti from the apical half of the mouse cochlea and determined that the cells had a radius of 2.7 μm and a cuticular plate thickness of 2 μm (unpublished observations). Because stereocilia increase in number at the base of the mammalian cochlea (Lim, 1986), 100 were assumed to reside in a given mouse OHC. This number is consistent with our unpublished observations for basal-turn guinea pig OHCs. We also used a shorter stereocilia length of 0.8 μm (Lim, 1986). By substituting these mouse parameters and following the same computations as Huang & Santos-Sacchi (1993), we determined that 34.3% of the surface area in mouse OHCs is without prestin protein. Based on this information, Clin values were adjusted to remove that part of the linear capacitance not associated with motor function. This adjustment was made by multiplying the average Clin value in wildtype mice (6.18 pF) by 0.343 to obtain the capacitance value of the inactive region, i.e. 2.12 pF. For each cell, this latter value was subtracted from Clin to obtain the adjusted value, ClinAdj. The charge density, computed by dividing each cell's Qmax by ClinAdj, is virtually identical in heterozygotes and wildtype controls. This similarity is reflected in the capacitance functions shown in Fig. 4B. These functions, obtained using the values in Table 1, represent the average of the fits for wildtype (broken line) and heterozygous (continuous lines) mice. In this graph, the voltage-dependent portion of the capacitance is shown, normalized to ClinAdj. In other words, the capacitance values were adjusted, by removing the linear capacitance associated with the OHC's inactive region, prior to normalization. After adjustment, the curves are virtually identical for the two genotypes. Hence, these results demonstrate that there is no difference in the electrophysiology of OHCs harvested from heterozygous mice, i.e. there is no difference in motor function.
, http://www.100md.com
Somatic electromotility
Some OHCs were recorded on video, allowing us to obtain a measure of somatic electromotility. Five video files from each genotype were appropriate for analysis, as shown in Fig. 4C. The average parameters in these cells (WT: Qmax 837.4 ± 130.7 fC, 33.2 ± 3.6 mV, V–59.6 ± 19.9 mV, Clin 5.95 ± 0.49 pF; Het: Qmax 697.1 ± 56 fC, 34 ± 2.8 mV, V–55.9 ± 21.8 mV, Clin 5.42 ± 0.39 pF) showed that they formed a representative sample. The peak amplitude of movement was obtained as a percentage of maximum cell length in order to adjust for variations in the size of individual OHCs. These data were normalized by giving the +/+ mean percentage length change (5.47 ± 0.65%) a value of 1.0 and plotting the +/– mean percentage length change (5.48 ± 1.0%) relative to this. The bar graph demonstrates that the amplitude of movement in OHCs harvested from wildtype versus heterozygous mice was not significantly different, consistent with the data obtained from the capacitance measurements shown in Fig. 4B.
, 百拇医药
Immunocytochemistry
Wildtype and heterozygous mice show antiprestin antibody staining localized to the lateral membrane, where the OHC motor protein is expressed (Belyantseva et al. 2000; Zheng et al. 2001). In mice at P21, immunofluorescence is similar for both genotypes, as shown in Fig. 5A. These confocal images were taken 1.5 mm from the helicotrema in both ears. Although not shown here, similar images are observed at approximately 8 weeks of age. A semiquantitative measure of prestin protein expression was determined by obtaining the intensity of fluorescence per area. In order to avoid background staining, pixel intensity was integrated over a thresholded area restricted to the lateral membrane of the OHCs. In other words, the immunofluorescence was restricted to prestin protein targeted to the lateral membrane. Figure 5B provides normalized protein levels obtained using immunocytochemistry at 3 and approximately 8 weeks of age. The average integrated intensity per thresholded area in wildtype mice is given a value of 1.0. Results from heterozygous mice at the same age are then plotted relative to this normalized wildtype value. Standard deviations appear as error bars. For all results, the values are 1.0 indicating that wildtype and heterozygous mice show similar levels of prestin protein expression. Therefore, prestin protein is not reduced by half in heterozygotes that are at least 21 days of age.
, 百拇医药
A, both images were obtained 1.5 mm from the helicotrema in littermates at 21 days of age. The image at the top (bottom) is from a heterozygous (wildtype) mouse. B, results are shown for wildtype (filled bars) and heterozygous (shaded bars) mice killed at three weeks of age and for three groups of young adult mice. Data at 8 weeks were obtained from two litters: a litter of wildtype mice aged 7 weeks 5 days and a litter of heterozygotes aged 8 weeks 5 days. In all other cases, the data are collected from a single litter. The data are normalized by giving the average intensity/area for the wildtype mice in any given litter a value of 1.0. The average heterozygote intensity/area is then plotted relative to this value. Standard deviations appear as error bars.
, 百拇医药
Western blot analysis
In order to obtain a more quantitative estimate of prestin protein expression, we also performed western blot analysis in heterozygous mice and their controls. These animals were produced by mating wildtype with heterozygous mice, allowing comparisons to be made between animals in a single litter. To ensure that each sample was composed of a complete cochlea, the dissections were performed by the same experimenter. In addition, a whole cochlea (P63), or a pair of cochleae (P9 and P21), was loaded into one well of LDS-PAGE. This was done to ensure that comparisons were made between total prestin protein, even though the total amount of protein in each well might differ due to variations in the sizes of cochlear samples in different animals. After blotting cochlear samples with antiprestin, immunobands are displayed at 100 kDa, the prestin monomer in its natural glycosylated form (Matsuda et al. 2004). An example is shown in Fig. 6A for mice at P21. Prestin bands are similar in heterozygous and wildtype mice but absent in the prestin knockout mouse shown on the left. Results comparing prestin protein levels in animals with one versus two copies of the prestin gene are provided in Fig. 6B for litters at P9, P21 (including the mice shown in Fig. 6A) and P63. As in Fig. 5B, the average amount of prestin protein in wildtype mice at any given age is given a value of 1.0, thereby normalizing the results. Data from heterozygous mice in the same litter are then plotted relative to the wildtype datum. At P9, prestin protein expression in heterozygotes is 63% of that in controls; at P21 and P63, it is 92%. The asterisk indicates that the difference in prestin protein is statistically significant but only at P9 (Student's t test P < 0.01). Because the P21 litter contained only one wildtype mouse, error bars are not provided. Theoretically, one would expect 50% wildtype pups when crossing wildtype and heterozygous mice. However, in any given litter, the proportions can vary as happened here at p21. Unfortunately, no additional litters were available to allow us to obtain more data. In contrast to the immunofluorescence (Fig. 5A), western blots document total protein levels, i.e. protein in the cytoplasm as well as that targeted to the lateral membrane. In spite of this difference, both methods indicate that prestin protein levels are not reduced in heterozygous mice older than 3 weeks. However, because the average standard deviation for western blots was 16% of the mean, a small difference between genotypes would be difficult to determine.
, 百拇医药
A, an example of the western blots used to quantify prestin protein expression in animals having one versus two copies of the prestin gene. The wildtype and heterozygous mice were from the same litter and were 21 days of age. The molecular mass of monomeric prestin was estimated based on a linear regression of the molecular mass ladder. No band appears for the knockout mouse shown on the left. B, the amount of prestin protein expressed at P9, P21 and P63. Regions of interest were drawn around each of the bands at 100 kDa. A similar region of interest was used to define the intensity of the background, which was then subtracted from each prestin-associated band. The measurements reflect the integrated intensity of the 100 kDa band measured relative to that of the background. In contrast to results at P9 and P21, only one cochlea was required to generate results at P63. As in Fig. 5B, the data are normalized to the mean wildtype intensity and standard deviations appear as error bars. The latter are absent for the P21 control because the litter contained only one wildtype mouse. *Protein levels are statistically different between heterozygous and wildtype mice but only at P9 (Student's t test, P < 0.01).
, 百拇医药
Real-time RT-PCR
prestin mRNA was quantified using real-time RT-PCR beginning at P9 when prestin mRNA approaches its mature value in wildtype mice (Reisinger et al. 2005). Data in Fig. 7 confirm that mRNA is lower in heterozygotes at 9 days of age (Liberman et al. 2002). In other words, one copy of prestin generates about half the mRNA and half the protein (see Fig. 6B) of WT controls at this early stage of development. Data at P21 and P49 also indicate that prestin mRNA remains relatively low in young adult heterozygotes. Differences between genotypes are statistically significant at P9 and P21 (Student's t test, P < 0.01). In addition, the values at P49 approach significance with a P value of 0.052. Although the mRNA in heterozygotes is less than in controls at all ages tested, protein levels are near normal in young adult mice (see Figs 5B and 6B). Therefore, cochlear physiology and prestin protein levels are wildtype-like, in spite of the fact that mRNA levels are reduced in mice with only one copy of the prestin gene.
, 百拇医药
This figure provides prestin mRNA values for wildtype (filled bars) and heterozygous (shaded bars) mice obtained using real-time RT-PCR. Data are provided for litters at P9, P21 and P49. The values are normalized by plotting data from heterozygotes relative to the mean wildtype RNA level, which is given a value of 1.0. Standard deviations appear as error bars. The P values obtained using the Student's t test are also appended.
Discussion
, http://www.100md.com
The first report on prestin knockout mice (Liberman et al. 2002) stated that prestin mRNA in heterozygotes was half that in controls, antibody labelling for prestin protein was reduced and somatic electromotility was decreased by nearly one half. In addition, cochlear sensitivity, assessed using distortion product otoacoustic emissions (DPOAEs) and auditory brainstem responses (ABRs), was decreased by 6 dB for +/– mice when compared with wildtype controls. These results are unexpected because a halving of OHC motor function would be expected to reduce cochlear sensitivity by much more than 6 dB, based upon what is known about cochlear amplification (Mountain et al. 1983; Patuzzi et al. 1989; Fettiplace & Ricci, 2003; Santos-Sacchi, 2003). For example, the approximate halving of electric driving force, due to a large reduction in the endocochlear potential (EP) by furosemide, is known to change cochlear mechanics by reducing sensitivity and frequency selectivity (Ruggero & Rich, 1991). Sewell (1984) reported similar results in single unit responses recorded from the auditory nerve. More recently, Gow et al. (2004) demonstrated that Claudin-11 null mice, which lack basal cell tight junctions in the stria vascularis, have EPs below 30 mV and hearing thresholds elevated by 50 dB. These experimental results are consistent with a simple model of active OHCs (Patuzzi et al. 1989). This theoretical analysis indicates that abolition of the positive EP would reduce the driving force across the apex of cochlear OHCs by about one half, thereby reducing their receptor potentials, which provide the driving force for somatic electromotility. According to Patuzzi et al. (1989), these changes are predicted to produce a 46 dB change in sensitivity at the single unit level. A more elaborate hydrodynamic model (Neely & Kim, 1986) predicts even larger threshold shifts. It is therefore surprising that the changes in sensitivity reported by Liberman et al. (2002) were only 3–6 dB for DPOAEs and 1–8 dB for ABRs, considering that somatic motility and prestin protein expression were stated to be reduced by nearly one half.
, http://www.100md.com
In contrast to Liberman et al. (2002), data in this report demonstrate that CAP thresholds, CAP tuning curves and CM input–output functions are similar in heterozygotes and wildtype mice. Not only are the differences between genotypes small, they are less than 6 dB and less than the standard deviations. Even after testing each data point, the small difference in sensitivity, seen in mice with one versus two copies of the prestin gene, was not statistically significant at the 0.05 level. We also assessed OHC motor function by measuring NLC and somatic electromotility. Results indicate that differences between wildtype and heterozygous mice for peak electromotility, as well as , V and charge density (Qmax/ClinAdj) are not statistically different (Student's t test). In other words, there is no alteration in the sensitivity or voltage dependence of electromechanical transduction or in the maximum charge movement or the OHC length change in animals having one versus two copies of the prestin gene. We also evaluated data obtained by Jia et al. (2003) on F2 generation prestin heterozygotes and their controls, to obtain the linear capacitance associated with the active region only. Using their mean values, the adjusted charge density in mice with only one copy of the prestin gene was 95.4% of wildtype. These results from two different laboratories indicate near-normal OHC motor function in F2–F4 generation heterozygotes and their controls. All of these results are consistent with immunocytochemical and western blot analyses showing that prestin protein expression is wildtype-like in adolescent heterozygous mice. Therefore, we conclude that protein expression in heterozygotes is not one-half of that expressed in wildtype controls. Although prestin mRNA in heterozygotes is less than in wildtype controls at all ages tested (Fig. 7), near normal protein levels are observed (Figs 5B and 6B) at a time when the in vivo and in vitro physiology is similar in the two genotypes.
, 百拇医药
One might argue that differences between the results of Liberman et al. (2002) and those reported here could relate to the larger variability associated with measurements at the round window, i.e. to the more invasive procedures required to record cochlear potentials as suggested by Liberman et al. (2004). If true, then the differences between wildtype and heterozygous mice are very small and clinically insignificant. They are so small in fact, that they cannot be resolved because normal variability exceeds the difference between wildtype and heterozygous mice. Another possibility is that differences between the two studies relate to genetic drift (Wolfer et al. 2002), i.e. the use of F2 generation mice by Liberman et al. (2002)versus the F3–F5 generation mice used here. Here it is important to mention that DPOAE amplitudes in heterozygotes were similar to wildtype controls in F3–F5 generation mice (Liberman et al. 2004). In other words, DPOAEs were reduced in F2 but not F3–F5 heterozygotes. However, when considering this issue, it should also be stated that the NLC data from Jia et al. (2003) show normal OHC charge densities in heterozygous mice from the F2 generation when Clin is adjusted to reflect only that part of the membrane that contains motors. In other words, the NLC measured independently by two different groups is wildtype-like in F2–F4 heterozygotes. This observation argues against generational differences/genetic drift being the explanation for differences between the results reported here and those of Liberman et al. (2002). Perhaps subsequent prestin animal models, including those backcrossed to strains without age-related hearing loss, will help to resolve the discrepancy between the data presented here and those published by Liberman et al. (2002).
, 百拇医药
The physiological results in this report imply that OHCs compensate for the deleted prestin gene, thereby minimizing deleterious effects on peripheral auditory function. In other words, heterozygous mice do not exhibit haploinsufficiency. Because the safety margin that separates normal from pathological conditions is large, the hearing loss due to prestin removal in mice requires loss-of-function mutations in both alleles, i.e. it is recessive. These results must be considered in light of information on human subjects with a particular prestin mutation (Liu et al. 2003). Although 3.2% of the deaf probands examined showed heterozygosity, the audiological findings were difficult to interpret because hearing losses were highly variable. In fact, some individuals demonstrated moderate-to-profound hearing losses, suggesting haploinsufficiency, while one heterozygote reported no hearing loss at all. In addition, the data are few, the age of the participants was between 10 and 67 years, and the age of onset for hearing loss was from birth to 35 years. These complications, along with data showing near normal cochlear physiology and protein levels in adolescent heterozygous mice, implies that by itself one copy of the prestin gene does not necessarily result in reduced protein levels to a level that is insufficient to produce normal auditory function. If the deafness in human heterozygotes results from interactions with modifier genes, then these are probably more numerous than in mice. In fact, the mixing of backgrounds in human populations results in a large number of modifier genes that can increase variability by influencing the age of onset, progression and/or severity of hearing loss (Friedman et al. 2000). This contrasts with the limited mixing of C57BL6 and 129 mouse strains used in gene targeting. It is therefore not unexpected that mouse models differ from human disease conditions.
, 百拇医药
The data in this report also suggest that prestin protein expression is autoregulated. This possibility relates to the fact that prestin mRNA is well below wildtype at all ages tested, while the protein expression is wildtype-like in adolescent mice. We speculate that near-normal protein levels are attained prior to P33. This latter estimate is suggested because heterozygotes display wildtype-like physiology at P33, the youngest age tested. It is also difficult to know if prestin's autoregulation occurs at the transcriptional or translational level. Because this is not a developmental study, we do not have measurements at numerous points during development. It is therefore possible to miss a surge in prestin mRNA that might occur over a relatively short time frame. However, if there is no surge, i.e. the mRNA is 50% of wildtype at all ages, then it might be tempting to argue that regulation is at the translational level. Such a conclusion, however, would be premature because we are measuring steady-state RNA levels and know nothing about transcriptional rates or the stability of newly synthesized prestin mRNA (Bustin, 2002). We also know nothing about the stability of prestin protein, which could be influenced by glycosylation (Matsuda et al. 2004) and/or phosphorylation (Deak et al. 2005), or the details of membrane targeting and oligomerization. Additional experiments are therefore required in order to understand how prestin is regulated.
, 百拇医药
References
Ashmore JF (1987). A fast motile response in guinea-pig outer hair cells: The cellular basis of the cochlear amplifier. J Phyiol 388, 323–347.
Ashmore JF (1990). Forward and reverse transduction in the mammalian cochlea. Neurosci Res Suppl 12, S39–S50.
Belyantseva IA, Adler HJ, Curi R, Frolenkov GI & Kachar B (2000). Expression and localization of Prestin and the sugar transporter GLUT-5 during development of electromotility in cochlear outer hair cells. J Neurosci 20, RC116, 1–5.
, http://www.100md.com
Brownell WE, Bader CR, Bertrand D & Ribaupierre Y (1985). Evoked mechanical responses of isolated outer hair cells. Science 227, 194–196.
Bustin SA (2002). Quantification of mRNA using real-time reverse transcription PCR (RT-PCR): trends and problems. J Mol Endocrinol 29, 23–39.
Cheatham MA, Huynh KH, Gao J, Zuo J & Dallos P (2004). Cochlear function in Prestin knockout mice. J Physiol 560, 821–830.
, 百拇医药
Collyer CE & Enns JT (1987). Analysis of Variance, the Basic Designs. Nelson-Hall, Chicago IL.
Dallos P & Cheatham MA (1976a). Compound action potential (AP) tuning curves. J Acoust Soc Am 59, 591–597.
Dallos P & Cheatham MA (1976b). Production of cochlear potentials by inner and outer hair cells. J Acoust Soc Am 60, 510–512.
Dallos P & Fakler B (2002). Prestin, a new type of motor protein. Nature Rev 3, 104–111.
, 百拇医药
Deak L, Zheng J, Du GG, Orem A, Aguiaga S, Matsuda K et al. (2005). Effects of cyclic nucleotides on the function of prestin. J Physiol 563, 483–496.
Fettiplace R & Ricci AJ (2003). Adaptation in auditory hair cells. Curr Opin Neurol 13, 446–451.
Friedman T, Battey J, Kachar B, Riazuddin S, Noben-Trauth K, Griffith A et al. (2000). Modifier genes of hereditary hearing loss. Curr Opin Neurobiol 10, 487–493.
, 百拇医药
Gow A, Davies C, Southwood CM, Frolenkov G, Chrustowski M, Ng L et al. (2004). Deafness in Claudin-11 null mice reveals the critical contribution of basal cell tight junctions to stria vascularis function. J Neurosci 24, 7051–7062.
Gummer AW, Smolders JWT & Klinke R (1987). Basilar membrane motion in the pigeon measured with the Mssbauer technique. Hear Res 29, 63–92.
He DZZ, Cheatham MA, Pearce M & Vetter DE (2004). Mouse outer hair cells lacking the 9 Ach receptor are motile. Brain Res Dev Brain Res 148, 19–25.
, 百拇医药
Henry KR & Chole RA (1980). Genotypic differences in behavioral, physiological and anatomical expressions of age-related hearing loss in the laboratory mouse. Audiol 19, 369–383.
Huang G & Santos-Sacchi J (1993). Mapping the distribution of the outer hair cell motility voltage sensor by electrical amputation. Biophys J 65, 2228–2236.
Jia S, Kachar B, Gao J, Beisel K, Feng F, He DZZ et al. (2003). Properties of OHC motors in Prestin-mutant mice. Abstr 26th Midwinter Res Meet Assoc Res Otolaryngol 104.
, 百拇医药
Kachar B, Brownell WE, Altschuler R & Fex J (1986). Electrokinetic shape changes of cochlear outer hair cells. Nature 322, 365–368.
Li HS & Borg E (1991). Age-related loss of auditory sensitivity in two mouse genotypes. Acta Otolaryngol 111, 827–834.
Liberman MC, Gao J, He DZZ, Wu X, Jia S & Zuo J (2002). Prestin is required for electromotility of the outer hair cell and for the cochlear amplifier. Nature 419, 300–304.
, http://www.100md.com
Liberman MC, Zuo J & Guinan JJ (2004). Otoacoustic emissions without somatic motility: Can stereocilia mechanics drive the mammalian cochleaJ Acoust Soc Am 116, 1649–1655.
Lim D (1986). Cochlear anatomy related to cochlear micromechanics. A review. J Acoust Soc Am 67, 1686–1695.
Liu ZX, Ouyang XM, Xia XJ, Zheng J, Du Pandya ALL, Welch KO, Petit C, Smith RJ, Webb BT, Yan D, Arnos KS, Corey D, Dallos P, Nance WE & Chen ZY (2003). Prestin, a cochlear motor protein, is defective in non-syndromic hearing loss. Human Mol Genet 12, 1155–1162.
, http://www.100md.com
Matsuda K, Zheng J, Du GG, Klocker N, Madison LD & Dallos P (2004). N-linked glycosylation sites of the motor protein prestin: effects on membrane targeting and electrophysiological function. J Neurochem 89, 928–938.
Mountain DC, Hubbard AE & McMullen TA (1983). Electromechanical processes in the cochlea. In Mechanics of Hearing, ed. de Boer E, Viergever MA, pp. 119–126. Delft University Press, Delft, the Netherlands.
, 百拇医药 Neely ST & Kim DO (1986). A model for active elements in cochlear biomechanics. J Acoust Soc Am 79, 1472–1480.
Ohlemiller KK & Siegel JH (1992). The effects of moderate cooling on gross cochlear potentials in the gerbil: basal and apical differences. Hear Res 63, 79–89.
Oliver D & Fakler B (1999). Expression density and functional characteristics of the outer hair cell motor protein are regulated during postnatal development in rat. J Physiol 519, 791–800.
, http://www.100md.com
Oliver D, He DZZ, Klcker N, Ludwig J, Schulte U, Waldeggar S et al. (2001). Intracellular anions as the voltage sensor of prestin, the outer hair cell motor protein. Science 292, 2340–2343.
Patuzzi RB, Yates GK & Johnstone BM (1989). Outer hair cell receptor currents and sensorineural hearing loss. Hear Res 42, 47–72.
Pearce M, Richter C-P & Cheatham MA (2000). A reconsideration of sound calibration in the mouse. J Neurosci Meth 106, 57–67.
, 百拇医药
Pusch M & Neher E (1998). Rates of diffusional exchange between small cells and a measuring patch pipette. Pflugers Arch 411, 204–211.
Reisinger E, Zimmermann U, Knipper M, Ludwig J, Klcker N, Fakler B et al. (2005). Cod106, a novel synaptic protein expressed in sensory hair cells of the inner ear and in CNS neurons. Mol Cell Neurosci 28, 106–117.
Ruggero MA & Rich NC (1991). Furosemide alters organ of Corti mechanics: evidence for feedback of outer hair cells upon the basilar membrane. J Neurosci 11, 1057–1067.
, 百拇医药
Santos-Sacchi J (1991). Reversible inhibition of voltage-dependent outer hair cell motility and capacitance. J Neurosci 11, 3096–3110.
Santos-Sacchi J (2003). New tunes from Corti's organ: the outer hair cell boogie rules. Curr Opin Neurobiol 13, 1–10.
Santos-Sacchi J & Dilger JP (1988). Whole-cell currents and mechanical responses of isolated outer hair cells. Hear Res 35, 143–150.
Santos-Sacchi J & Navarrete E (2002). Voltage-dependent changes in specific membrane capacitance caused by prestin, the outer hair cell lateral membrane motor. Pflugers Arch 444, 99–106.
, 百拇医药
Sewell WF (1984). The effects of furosemide on the endocochlear potential and auditory-nerve fiber tuning curves in cats. Hear Res 14, 305–314.
Shnerson A & Pujol R (1981). Age-related changes in the C57BL/6J mouse cochlea. I. Physiological findings. Dev Brain Res 2, 65–75.
Shore SE & Nuttall AL (1985). The effects of cochlear hypothermia on compound action potential tuning. J Acoust Soc Am 77, 590–598.
, 百拇医药
Stankovic KM & Corfas G (2003). Real-time quantitative RT-PCR for low-abundance transcripts in the inner ear: analysis of neurotrophic factor expression. Hear Res 185, 97–108.
Taylor MM & Creelman CD (1967). PEST: Efficient estimates on probability functions. J Acoust Soc Am 41, 782–787.
Weber T, Zimmerman U, Winter H, Mack A, Kpschall Rohbock K et al. (2002). Thyroid hormone is a critical determinant for the regulation of the cochlear motor protein prestin. Proc Nat Acad Sci U S A 99, 2901–2906.
, 百拇医药
Wolfer DP, Crusio WE & Lipp H-P (2002). Knockout mice: simple solutions to the problems of genetic background and flanking genes. Trends Neurosci 25, 336–340.
Zheng QY, Johnson KR & Erway LC (1999). Assessment of hearing in 80 inbred strains of mice by ABR threshold analyses. Hear Res 130, 94–109.
Zheng J, Long KB, Shen W, Madison LD & Dallos P (2001). Prestin topology: localization of protein epitopes in relation to the plasma membrane. Neuroreport 12, 1929–1935.
Zheng J, Shen W, He DZZ, Long KB, Madison LD & Dallos P (2000). Prestin is the motor protein of cochlear outer hair cells. Nature 405, 149–155., 百拇医药(M. A. Cheatham, J. Zheng,)
2 Department of Neurobiology and Physiology, Northwestern University, Evanston, IL 60208, USA
3 Department of Developmental Neurobiology, St Jude Children's Research Hospital, Memphis, TN 38105, USA
4 Section on Structural Cell Biology, NIDCD, NIH, Bethesda, MD 20892, USA
Abstract
, 百拇医药
Targeted deletion of the prestin gene reduces cochlear sensitivity and eliminates both frequency selectivity and outer hair cell (OHC) somatic electromotility. In addition, it has been reported by Liberman and colleagues that F2 generation heterozygotes exhibit a 6 dB reduction in sensitivity, as well as a decrease in protein and electromotility. Considering that the active process is non-linear, a halving of somatic electromotility would be expected to produce a much larger change in sensitivity. We therefore re-evaluated comparisons between heterozygotes and wildtype mice using both in vivo and in vitro electrophysiology, as well as molecular biology. Data reported here for F3–F5 generation mice indicate that compound action potential thresholds and tuning curves, as well as the cochlear microphonic, are similar in heterozygotes and wildtype controls. Measurements of non-linear capacitance in isolated OHCs demonstrate that charge density, as well as the voltage dependence and sensitivity of motor function, is indistinguishable in the two genotypes, as is somatic electromotility. In addition, both immunocytochemistry and western blot analysis in young adult mice suggest that prestin protein in heterozygotes is near normal. In contrast, prestin mRNA is always less than in wildtype mice at all ages tested. Results from F3–F5 generation mice suggest that one copy of the prestin gene is capable of compensating for the deleted copy and that heterozygous mice do not suffer peripheral hearing impairment.
, 百拇医药
Introduction
The discovery of prestin (Zheng et al. 2000), the motor protein in cochlear outer hair cells (OHC), resulted in several efforts to determine its functional significance. For example, targeted deletion of the prestin gene reduces sensitivity (Liberman et al. 2002) and eliminates frequency selectivity (Cheatham et al. 2004). It is also known that OHCs harvested from knockout mice do not exhibit somatic electromotility (Liberman et al. 2002). In other words, the length changes associated with voltage-dependent conformational changes in the motor protein (Brownell et al. 1985; Kachar et al. 1986; Ashmore, 1987; Santos-Sacchi & Dilger, 1988) are not observed. Because of the significance of gene dosage for inherited disease, it is also of interest to compare wildtype controls, not only with homozygous mice, but also with heterozygous mice having only one copy of the prestin gene. These evaluations are required in order to learn if prestin protein is reduced in heterozygotes and, if so, if this reduction is associated with changes in cochlear function. Liberman et al. (2002) suggested that somatic electromotility in heterozygotes was about half that in wildtype controls, cochlear sensitivity was decreased by 6 dB and prestin protein levels were reduced. These results are puzzling because a halving of motility would be expected to result in a much greater change in sensitivity based upon what is known about cochlear amplification (Mountain et al. 1983; Patuzzi et al. 1989; Fettiplace & Ricci, 2003; Santos-Sacchi, 2003).
, http://www.100md.com
In this project, we re-evaluate comparisons between heterozygotes and wildtype controls to determine if they display different phenotypes. If auditory function were compromised, as suggested by the Liberman et al. (2002) report, then heterozygotes would exhibit haploinsufficiency. In this condition, a single copy of the prestin gene would be incapable of providing sufficient protein to assure normal auditory function. In order to test this hypothesis, compound action potential (CAP) threshold curves and CAP tuning curves were acquired from a recording electrode placed on the round window to determine if cochlear sensitivity and frequency selectivity are different in heterozygotes and wildtype mice. In addition to the neural responses, the cochlear microphonic (CM) was also recorded. Although inner hair cells (IHC) can contribute to this response, the CM is thought to primarily reflect OHC receptor currents (Dallos & Cheatham, 1976b). Following euthanasia, animals were evaluated in three additional experiments. First, whole-cell, voltage-clamp recordings were used to determine the amplitude of somatic electromotility and the voltage-dependent non-linear capacitance (NLC), its electrical correlate (Ashmore, 1990; Santos-Sacchi, 1991). Second, prestin protein expression was assessed using immunocytochemistry and laser confocal microscopy, as well as western blot analysis. Third, mRNA was quantified using real-time RT-PCR. Results from these experiments at the molecular, cellular and systems levels were compared to determine the influence of one versus two copies of the prestin gene on cochlear function.
, http://www.100md.com
Methods
In the in vivo physiological experiments, F4 generation wildtype and heterozygous mice were used as subjects while both F3 and F4 generations were combined for the in vitro studies. Wildtype mice were generated by mating two wildtype parents, while breeding a wildtype with a prestin knockout mouse produced the heterozygotes. For the molecular biology experiments, including immunocytochemistry, western blot analysis and real-time RT-PCR, wildtype mice were bred with heterozygotes to obtain F4 and F5 litters with about equal portions of the two genotypes in a single litter. All animals were on a mixed 129SvEv/C57BL6 background. The details of how the knockout mice were generated are available in a previous report (Liberman et al. 2002). All procedures were approved by the National Institutes of Health and by Northwestern University's Institutional Review Board.
, 百拇医药
In vivo physiology
Cochlear in vivo physiology was studied in wildtype (WT, +/+) and heterozygous (Het, +/–) mice anaesthetized with sodium pentobarbital (80 mg kg–1, I.P.). Additional doses were given throughout the experiment to maintain a surgical level of anaesthesia. During both surgery and data collection, the headholder was heated to prevent cooling of the cochlea (Shore & Nuttall, 1985; Ohlemiller & Siegel, 1992). Electrical potentials were recorded using a round window electrode. At the end of each experiment, the animals were humanely terminated with an overdose of anaesthetic.
, http://www.100md.com
Compound action potential thresholds were acquired using a tracking program, which determined the sound pressure level necessary to generate a 10 μV N1/P1 voltage (Taylor & Creelman, 1967; Gummer et al. 1987). The CAP tuning curves were collected using a simultaneous, tone-on-tone masking paradigm (Dallos & Cheatham, 1976a). In this procedure, the level of a 12 kHz probe tone, presented alone, was adjusted to generate a response of 45 μV measured between the first negative and subsequent positive peak of the CAP. Masker frequency and level were then varied to produce a 3 dB decrease in the probe response, i.e. to 32 μV. The masker was presented in alternating phase to minimize the CM. For all neural measurements a custom-made, programmable filter was used to bandpass filter electrical signals between 0.2 and 3.0 kHz. In contrast, CM input–output functions were obtained using lowpass filtering at 49.9 kHz to prevent aliasing. In all cases, however, a differential amplifier, with gain of 1000 and a passband between less than 1 Hz and 50 000 Hz, was utilized. Peak-to-peak CM voltages were determined offline by fast Fourier transformation (FFT) of averaged response waveforms. Further details about these procedures are provided elsewhere (Pearce et al. 2000; Cheatham et al. 2004).
, http://www.100md.com
Non-linear capacitance
In vitro physiological experiments were performed on OHCs harvested from 11 heterozygous and 12 wildtype mice following euthanasia with sodium pentobarbital (200 mg kg–1, I.P.). Sense organs from both ears were subjected to enzymatic treatment with type IV Collagenase (Sigma-Aldrich, St Louis MO, USA) in Leibovitz's L-15 solution (Gibco, Invitrogen, Carlsbad CA, USA), at a concentration of 0.5 mg ml–1 for 10 min. Gentle pipetting allowed individual OHCs to be separated from the cellular matrix. Gigohm seals were obtained at the basal pole of dissociated OHCs. The cells were then whole-cell voltage clamped at room temperature using an Axopatch 200A amplifier (Axon Instruments, Union City, CA, USA). Capacitive currents were studied in isolation by using an ionic blocking solution composed of (mM): 99.2 NaCl, 20 TEA, 20 CsCl, 2 CoCl2, 1.47 MgCl2, 2 CaCl2, 10 Hepes and 5 dextrose. The solution was adjusted to pH 7.2 with CsOH and to 300 mosmol l–1 with dextrose. The solution inside the patch pipette contained (mM): 140 CsCl, 10 EGTA, 2 MgCl2 and 10 Hepes and was also adjusted to pH 7.2 with CsOH and to 300 mosmol l–1 with dextrose. Capacitance measurements were obtained with a high-resolution (2.56 ms) protocol (Santos-Sacchi & Navarrete, 2002) using a two-sine voltage stimulus (10 mV peak at 390.6 and 781.2 Hz) in which sinusoids were superimposed on one cycle of a 2.5 Hz sinusoidal voltage ranging between ±175 mV and starting from a holding potential of 0 mV. Subsequent FFT analysis utilized parameter variations provided by Pusch & Neher (1998). All data collection and some analyses were performed with jClamp (Scisoft, CT, USA) using a National Instruments PCI-6052E board. Fits to the capacitance data were made in IgorPro (Wavemetrics, Lake Oswego, OR, USA). The maximum charge transferred through the membrane's electric field (Qmax), the slope factor of the voltage dependence (), the voltage at peak capacitance (V) and the linear membrane capacitance (Clin) were determined. Because Clin reflects the total surface area of the membrane, we also obtained an adjusted Clin value, ClinAdj, to reflect only that part of the membrane containing motors (Huang & Santos-Sacchi, 1993). This adjusted datum was used to compute the charge density (Qmax/ClinAdj), thereby normalizing for cell size (Oliver & Fakler, 1999). In other words, charge density was used as a normalized metric for the voltage-mediated charge transfer through only that part of the membrane containing motor protein.
, http://www.100md.com
Somatic electromotility
A Leica DM-IRB inverted microscope with differential interference contrast (DIC) optics was used for observing voltage-clamped OHCs. Images were acquired at 30 fps/360 x 240 pixel resolution with a Panasonic WVCD22 video camera and digitized using a Hauppage WinTV GO PCI card. Script protocols were written in jClamp to acquire audio video interleaved (avi) images in synchrony with the voltage stimulus used for extracting the capacitance data. Following transfer from avi to QuickTimevideo format and enlargement to 640 x 480 pixel resolution, the files were analysed using DIAS software (Soll Technologies, Iowa city, IA, USA) running on a Macintosh G4 computer. Cell edges were either determined manually or automatically by the software from individual movie frames. To compensate for the variability in cell length, normalization to each cell's maximum axial length was performed to obtain the amplitude of movement as a percentage. The analytical limit of this video imaging technique was also calculated to evaluate the statistical validity of the procedure. For example, measurement of each OHC's length, at a constant holding potential of 0 mV, always gave a standard deviation of less than 0.5 pixels using at least 10 frames for each cell. Thus, the motile length changes that we observed, ranging from 3 to 5 pixels, are not attributed to analytical or artifactual fluctuations.
, 百拇医药
Immunocytochemistry
Mice were cardiac perfused first with heparinized saline and then with 4% paraformaldehyde. After 4 h post-fixation at room temperature, cochleae from +/+ and +/– mice were dissected, decalcified overnight in 10% EDTA and treated with 0.3% Triton X-100/PBS. Ovalbumin was used to block non-specific binding. The tissue was then exposed to an antimouse prestin antibody (1 : 2000) raised against an epitope in prestin's C-terminal (Matsuda et al. 2004) followed by the secondary antibody (1 : 200), Alexa488-conjugated antirabbit IgG (Molecular Probes, Eugene, OR, USA). Samples from the apical half of the cochlea were mounted on glass slides with cilia pointing up using Fluoromount-G and viewed on a Leica confocal microscope equipped with an argon laser with excitation at 488 nm; emission at 510 nm. Photometric gain was standardized so that the brightest signal in the labelled region was below the maximum (grayscale intensity 4095), ensuring that there were no pixels at saturation. Confocal fluorescent images of OHCs from wildtype and heterozygous mice were taken at the same time and with the same fixed photometric gain. A series of z-sections from the apex to the base of the cell, at intervals of 1.0 μm, were acquired at a resolution of 1024 x 1024 with 12 bits per pixel through a 40x objective (NA 1.25). A single pixel corresponded to a 183 x 183 nm square in the specimen plane. In order to obtain a semiquantitative measure of prestin protein expression in the OHC lateral membrane, captured fluorescent images were analysed with the Integrated Intensity feature of the MetaMorph imaging system (Universal Imaging, Molecular Devices Corp., Downingtown, PA, USA). The threshold was set to 1000 to avoid non-specific signals. Areas of immunostaining were then rendered in number of pixels so that the fluorescent intensity/area between +/+ and +/– OHCs could be calculated. Protein expression was determined at approximately the same distance from the helicotrema in wildtype and heterozygous mice at 3 and at approximately 8 weeks of age. Only sections immunostained in the same experiment were compared to avoid interexperimental variability.
, http://www.100md.com
Western blot analysis
In order to ensure that comparisons were made between intact cochlear samples, cochleae were carefully dissected and trimmed by the same experimenter. The samples were then collected in CelLytic MT Cell Lysis Reagent (C3228 Sigma-Aldrich, St. Louis MO, USA) supplemented with phenylmethylsulphonyl fluoride (PMSF 100 μg ml–1), a protease inhibitor cocktail (1 : 50, P8340, Sigma-Aldrich, St. Louis, MO, USA), thimerosal (1 : 10 000) and DNase (10 μg ml–1). After homogenization with a Kontes pellet pestle (Fisher Scientific International, Hampton NJ, USA), lysates were centrifuged at 3000 g for 15 min at 4°C. The cochlear sample was then diluted in an equal volume of 2x loading buffer (lithium dodecyl sulphate (LDS) Laemmli buffer with 100 mM dithiothreitol (DTT)), heated to 100°C for 5 min and spun 10 min before loading onto a 7.5% LDS-acrylamide gel. After electrophoresis, the gel proteins were electrotransferred onto a nitrocellulose membrane, which was blocked with blocking solution (2% non-fat dry milk/2% bovine serum albumin in PBS) and then incubated with the antimouse prestin antibody, also in blocking solution. After washing with Tris-buffered saline containing 0.2% Tween 20 (TBST), the membrane was incubated with an HRP-conjugated secondary antibody (1 : 10,000, donkey antirabbit IgG, Jackson ImmunoResearch Laboratories, West Grove, PA, USA) in blocking solution containing 1% normal donkey serum. Finally, after thoroughly washing with TBST, prestin expression was detected using an enhanced chemiluminescence western blotting detection system (Amersham Biosciences, Piscataway, NJ, USA). The apparent molecular masses were calculated by non-linear curve fitting of the molecular mass standards and indicated on each gel. After development, the film was photographed and analysed using Kodak ID Image Analysis Software. This program allows one to select regions of interest around each band and to compare these to similar regions representing the background. Means and standard deviations were obtained for the integrated intensity minus the background in arbitrary units. Protein expression was quantified at postnatal days 9 (P9) and 21 (P21), using two cochleae per sample, and at postnatal day 63 (P63) using a single cochlea.
, http://www.100md.com
Because quantification of prestin protein using enhanced chemiluminescence can be compromised by enzyme/substrate kinetics, we determined the range where antiprestin is able to detect quantitative differences in its antigen. Different amounts of prestin protein synthesized in yeast hosts were applied to LDS-PAGE and blotted with antiprestin. The intensities of prestin blots were then acquired at three different exposure times. By comparing the band intensity of our cochlear samples with the yeast-prestin data, it was determined that our results fell within the quasi-linear range. In addition, at P63 when only one cochlea was used, similar results were obtained on the opposite ear, indicating a high degree of reproducibility.
, http://www.100md.com
RNA isolation
Both cochleae from each mouse were carefully dissected and trimmed in RNAlater solution (Ambion, Austin, TX, USA) by the same experimenter. The two cochleae were homogenized in 300 μl lysis buffer using the Absolutely RNA RT-PCR Miniprep Kit (Stratagene, La Jolla, CA, USA). Total RNA was immediately isolated and stored at –80°C. Even though isolated RNA was treated with DNase during the RNA isolation procedure, the PCR was performed to check for genomic DNA contamination. The forward primer was 5'-CATGCTGAAGAAAATGAAATCCCTGCAGAGACC; the reverse primer, 5'-GCTGGAGTACCCCAGTGCTTAT-GCCCGAGACC. Cycling conditions were as follows: denature at 95°C for 1.5 min followed by 45 cycles at 95°C for 1 min, 60°C for 1 min and then at 72°C for 2 min. Only samples without contamination were used in real-time PCR reactions.
, 百拇医药
Real-time RT-PCR
Quantitative polymerase chain reaction with reverse transcription (RT-PCR) was used to assay prestin mRNA from litters at P9, P21 and P49. The reverse transcription reaction was performed on a 15 μl sample at 42°C for 90 min with StrataScript reverse transcriptase (Stratagene), in a 1x buffer containing 6 ng μl–1 of random primer (9mers), 1 mM dNTP and 300 ng cochlear RNA from each mouse. PCR reactions were performed on an iCycler machine (Bio-Rad, Hercules, CA, USA) using Brilliant QPCR Master mix (Stratagene, La Jolla, CA, USA). Details of the cycling conditions were described by Liberman et al. (2002). Specific prestin primer pairs for real-time PCR were 5'-TCGGGCATAAGCACTGGG-3' (forward), 5'-ACGGCTGCCAGCATGG-3' (reverse) and, for the TaqMan probe, 5'/56FAM-TACTCCAGCTT-CCCCAAGGCTTAGCCT-/3BHQ-1–3'. Universal 18S internal standards were purchased from Ambion. The sequence of the 18S probe was 5'-/56-FAM/TGGC-TGAACGCCACTTGTCCCTCTAA/3BHQ-1-3'. Each real-time PCR was performed at the same time on mice from the same litter except for the data at P21 where samples from two litters were combined. Samples from each mouse were repeated in triplicate. Prestin cDNA from one of the samples was diluted and used to generate standard curves. The relative quantities of prestin mRNA were calculated from the standard curves and normalized to 18S (Stankovic & Corfas, 2003).
, http://www.100md.com
Results
In vivo physiology
Cochlear sensitivity was measured in wildtype and heterozygous mice by collecting CAP thresholds at frequencies between 1.8 and 50.0 kHz. The CAP thresholds in Fig. 1 provide a physiological representation of the animal's audiogram. In Fig. 1A, mean CAP thresholds are shown for both genotypes. Fig. 1B provides the threshold difference between animals having one versus two copies of the prestin gene, with positive values indicating that heterozygotes display better sensitivity than controls. Not only are the differences relatively small, they are less than the standard deviations. The latter are shown for both genotypes with error bars representing standard deviations in heterozygotes and broken lines representing standard deviations in wildtype mice. The larger deviations at high frequencies are probably related to the age-related hearing loss associated primarily with the 129SvEv strain (Henry & Chole, 1980; Shnerson & Pujol, 1981; Li & Borg, 1991; Zheng et al. 1999; He et al. 2004), which supplied the embryonic stem cells used for gene targeting.
, http://www.100md.com
A, mean CAP thresholds for wildtype (, broken lines) and heterozygous (, continuous lines) mice. The average age for wildtype mice was 51 days; that for heterozygotes, 47 days. B, the difference in CAP thresholds (continuous line). Standard deviations for heterozygotes are shown with error bars in the positive direction only; for wildtype mice, with the broken lines.
Figure 2A displays CAP tuning curves (Dallos & Cheatham, 1976a) for a probe frequency at 12 kHz for wildtype and heterozygous mice. CAP tuning curves represent average iso-response functions, obtained in the frequency–intensity plane, for a small group of auditory nerve fibres with characteristic frequencies corresponding to the 12 kHz place. As such, they represent the frequency selectivity of the peripheral auditory system. The average probe level, which generated the 45 μV N1/P1 voltage for the 12 kHz probe alone, was 53 dB for both genotypes. Tuning was evaluated by measuring Q10 values, which were 5.3 in both wildtype and heterozygous mice. These values were obtained by dividing the bandwidth at 10 dB above the tip of the tone-on-tone masking function by probe frequency. Figure 2B shows the difference in decibels between the two curves. Again, the differences are less than the standard deviations, indicating that frequency selectivity is similar in the two genotypes. For both the CAP thresholds and CAP tuning curves, the differences between wildtype and heterozygous mice are probably related to normal variability and not to how many copies of the prestin gene are available.
, 百拇医药
A, average CAP tuning curves for wildtype (broken lines) and heterozygous (continuous lines) mice. Isolated symbols indicate the probe level, which was 53 dB for both genotypes. The average age for wildtype mice was 51 days; for heterozygotes, 48 days. As in Fig. 1, the difference between the tuning curves is given in B along with the standard deviations.
In addition to measurements of neural responses, the CM was also recorded. This gross-potential response is thought to represent OHC receptor currents (Dallos & Cheatham, 1976b). Figure 3 provides CM input–output functions recorded at 6, 12, 16 and 32 kHz for the F4 generation mice used in the in vivo experiments. In order to facilitate comparisons with Figs 1 and 2, the magnitudes are plotted as decibels re: 1 μV. The lower panel of Fig. 3 shows differences in CM responses between wildtype and heterozygous mice. The differences for F4 generation mice are smaller than the standard deviations and always less than 6 dB.
, http://www.100md.com
The top row provides CM data at 6, 12, 16 and 32 kHz. The average age for wildtype (heterozygous) mice was 49 (47) days. Plots at the bottom display the difference between CM input–output functions obtained in the two genotypes (continuous line). Standard deviations are represented as bars (broken lines) for heterozygous (wildtype) mice.
Taken together, the sensory and neural responses recorded from the round window in wildtype and heterozygous mice are very similar. In fact, we tested each individual data point statistically (Student's t test), and there was no difference between genotypes (P < 0.05). This statement applies to all three measurements, i.e. the CAP thresholds, the CAP tuning curves and the CM input–output functions. We also tested differences between CAP thresholds in +/+ and +/– animals using a mixed factor design analysis of variance (Collyer & Enns, 1987). The F score between the two groups was 0.004, which indicates that responses in the two genotypes are indistinguishable at the 95% level. A similar computation was performed for CM input–output functions obtained at 16 kHz. Comparisons yielded a vanishingly small F score and P= 0.99, again attesting to no difference between physiological responses in wildtype and heterozygous mice.
, 百拇医药
Non-linear capacitance
It is well known that somatic motility is associated with the translocation of electric charges across the OHC membrane, and that this charge movement imparts a bell-shaped voltage dependence to the membrane capacitance (Ashmore, 1990; Santos-Sacchi, 1991; Dallos & Fakler, 2002). A representative example from a wildtype and a heterozygous mouse is provided in Fig. 4A. The capacitance functions peak at the membrane voltage that is most effective in generating a motile response, i.e. around –70 mV.
, http://www.100md.com
A, capacitance functions are provided for a representative wildtype control (broken lines) and a heterozygote (solid lines). Parameter fits to the data appear as thin continuous lines. The Qmax was 872.7 (702.1), 29.4 (29.9), V–66.3 (–69.6) and Clin 6.27 (5.7) for the wildtype mouse (heterozygote). B, normalized functions are provided for wildtype (broken line) and heterozygous (bold line) mice. The functions represent the derivatives of a first-order Boltzmann function, normalized to ClinAdj in the form:
, 百拇医药
The two plots were obtained from the average values of the parameters (Qmax, , V and ClinAdj) shown in Table 1. C, somatic electromotility. Video images of individual OHCs were used to obtain measures of somatic electromotility. The peak voltage-dependent length change was expressed as a percentage of each OHC's maximum length. The data were then normalized by giving the wildtype mean percentage change in length a value of 1.0 and then plotting the mean for heterozygotes relative to this value.
, 百拇医药
The capacitance data for each OHC were fitted with the derivative of a first-order Boltzmann function (Santos-Sacchi, 1991; Oliver et al. 2001). The means and standard deviations of the capacitance function parameters used to fit the data are provided in Table 1, along with P values obtained using the Student's t test and Het/WT ratios. The slope factor, alpha, and the voltage at peak capacitance, V, are virtually identical in the two populations. In contrast, the maximum charge displacement (Qmax) and the linear capacitance (Clin) are significantly different (P\#8810; 0.01, Student's t test). Because the linear component of the membrane capacitance is proportional to the cell surface area (length) (Huang & Santos-Sacchi, 1993; Santos-Sacchi & Navarrete, 2002), the smaller average Clin in heterozygotes suggests that cells harvested from wildtype mice were, on average, somewhat longer than OHCs harvested from mice having only one copy of the prestin gene. This difference in cell length does not necessarily imply that OHCs at a given location are shorter in heterozygotes, although a very small difference was reported by Liberman et al. (2002) in F2 generation mice. This caution is advised because OHCs are harvested from the entire cochlea and data are obtained from only a few cells per mouse. It is therefore possible that wildtype OHCs originated from a more apical location.
, 百拇医药
The difference in Clin between genotypes suggests that these data should be adjusted to reflect only that part of the lateral membrane containing prestin. Although this information is not available for mice, the distribution of OHC somatic motility was determined using electrical amputation by Huang & Santos-Sacchi (1993). For cells harvested from the apex of the guinea pig cochlea (length 70 μm; radius 5 μm), 27% of the OHC's surface area was found to lack motor function. This result is consistent with the immunocytochemistry, which shows that prestin is not expressed in the hair bundle or in the nuclear region (Weber et al. 2002). In order to compute the OHC surface area lacking prestin, Huang & Santos-Sacchi (1993) modelled the basal region of the cell as a half sphere with a 2 μm extension. The apical area was composed of a 5 μm section surrounding the cuticular plate plus the surface area of the stereocilia. Each OHC was assumed to have 50 stereocilia, each with a radius of 0.15 μm and a length of 6 μm. In order to obtain an estimate of the surface area for short mouse OHCs (length 25 μm), we made the following adjustments. First, we examined radial sections of the organ of Corti from the apical half of the mouse cochlea and determined that the cells had a radius of 2.7 μm and a cuticular plate thickness of 2 μm (unpublished observations). Because stereocilia increase in number at the base of the mammalian cochlea (Lim, 1986), 100 were assumed to reside in a given mouse OHC. This number is consistent with our unpublished observations for basal-turn guinea pig OHCs. We also used a shorter stereocilia length of 0.8 μm (Lim, 1986). By substituting these mouse parameters and following the same computations as Huang & Santos-Sacchi (1993), we determined that 34.3% of the surface area in mouse OHCs is without prestin protein. Based on this information, Clin values were adjusted to remove that part of the linear capacitance not associated with motor function. This adjustment was made by multiplying the average Clin value in wildtype mice (6.18 pF) by 0.343 to obtain the capacitance value of the inactive region, i.e. 2.12 pF. For each cell, this latter value was subtracted from Clin to obtain the adjusted value, ClinAdj. The charge density, computed by dividing each cell's Qmax by ClinAdj, is virtually identical in heterozygotes and wildtype controls. This similarity is reflected in the capacitance functions shown in Fig. 4B. These functions, obtained using the values in Table 1, represent the average of the fits for wildtype (broken line) and heterozygous (continuous lines) mice. In this graph, the voltage-dependent portion of the capacitance is shown, normalized to ClinAdj. In other words, the capacitance values were adjusted, by removing the linear capacitance associated with the OHC's inactive region, prior to normalization. After adjustment, the curves are virtually identical for the two genotypes. Hence, these results demonstrate that there is no difference in the electrophysiology of OHCs harvested from heterozygous mice, i.e. there is no difference in motor function.
, http://www.100md.com
Somatic electromotility
Some OHCs were recorded on video, allowing us to obtain a measure of somatic electromotility. Five video files from each genotype were appropriate for analysis, as shown in Fig. 4C. The average parameters in these cells (WT: Qmax 837.4 ± 130.7 fC, 33.2 ± 3.6 mV, V–59.6 ± 19.9 mV, Clin 5.95 ± 0.49 pF; Het: Qmax 697.1 ± 56 fC, 34 ± 2.8 mV, V–55.9 ± 21.8 mV, Clin 5.42 ± 0.39 pF) showed that they formed a representative sample. The peak amplitude of movement was obtained as a percentage of maximum cell length in order to adjust for variations in the size of individual OHCs. These data were normalized by giving the +/+ mean percentage length change (5.47 ± 0.65%) a value of 1.0 and plotting the +/– mean percentage length change (5.48 ± 1.0%) relative to this. The bar graph demonstrates that the amplitude of movement in OHCs harvested from wildtype versus heterozygous mice was not significantly different, consistent with the data obtained from the capacitance measurements shown in Fig. 4B.
, 百拇医药
Immunocytochemistry
Wildtype and heterozygous mice show antiprestin antibody staining localized to the lateral membrane, where the OHC motor protein is expressed (Belyantseva et al. 2000; Zheng et al. 2001). In mice at P21, immunofluorescence is similar for both genotypes, as shown in Fig. 5A. These confocal images were taken 1.5 mm from the helicotrema in both ears. Although not shown here, similar images are observed at approximately 8 weeks of age. A semiquantitative measure of prestin protein expression was determined by obtaining the intensity of fluorescence per area. In order to avoid background staining, pixel intensity was integrated over a thresholded area restricted to the lateral membrane of the OHCs. In other words, the immunofluorescence was restricted to prestin protein targeted to the lateral membrane. Figure 5B provides normalized protein levels obtained using immunocytochemistry at 3 and approximately 8 weeks of age. The average integrated intensity per thresholded area in wildtype mice is given a value of 1.0. Results from heterozygous mice at the same age are then plotted relative to this normalized wildtype value. Standard deviations appear as error bars. For all results, the values are 1.0 indicating that wildtype and heterozygous mice show similar levels of prestin protein expression. Therefore, prestin protein is not reduced by half in heterozygotes that are at least 21 days of age.
, 百拇医药
A, both images were obtained 1.5 mm from the helicotrema in littermates at 21 days of age. The image at the top (bottom) is from a heterozygous (wildtype) mouse. B, results are shown for wildtype (filled bars) and heterozygous (shaded bars) mice killed at three weeks of age and for three groups of young adult mice. Data at 8 weeks were obtained from two litters: a litter of wildtype mice aged 7 weeks 5 days and a litter of heterozygotes aged 8 weeks 5 days. In all other cases, the data are collected from a single litter. The data are normalized by giving the average intensity/area for the wildtype mice in any given litter a value of 1.0. The average heterozygote intensity/area is then plotted relative to this value. Standard deviations appear as error bars.
, 百拇医药
Western blot analysis
In order to obtain a more quantitative estimate of prestin protein expression, we also performed western blot analysis in heterozygous mice and their controls. These animals were produced by mating wildtype with heterozygous mice, allowing comparisons to be made between animals in a single litter. To ensure that each sample was composed of a complete cochlea, the dissections were performed by the same experimenter. In addition, a whole cochlea (P63), or a pair of cochleae (P9 and P21), was loaded into one well of LDS-PAGE. This was done to ensure that comparisons were made between total prestin protein, even though the total amount of protein in each well might differ due to variations in the sizes of cochlear samples in different animals. After blotting cochlear samples with antiprestin, immunobands are displayed at 100 kDa, the prestin monomer in its natural glycosylated form (Matsuda et al. 2004). An example is shown in Fig. 6A for mice at P21. Prestin bands are similar in heterozygous and wildtype mice but absent in the prestin knockout mouse shown on the left. Results comparing prestin protein levels in animals with one versus two copies of the prestin gene are provided in Fig. 6B for litters at P9, P21 (including the mice shown in Fig. 6A) and P63. As in Fig. 5B, the average amount of prestin protein in wildtype mice at any given age is given a value of 1.0, thereby normalizing the results. Data from heterozygous mice in the same litter are then plotted relative to the wildtype datum. At P9, prestin protein expression in heterozygotes is 63% of that in controls; at P21 and P63, it is 92%. The asterisk indicates that the difference in prestin protein is statistically significant but only at P9 (Student's t test P < 0.01). Because the P21 litter contained only one wildtype mouse, error bars are not provided. Theoretically, one would expect 50% wildtype pups when crossing wildtype and heterozygous mice. However, in any given litter, the proportions can vary as happened here at p21. Unfortunately, no additional litters were available to allow us to obtain more data. In contrast to the immunofluorescence (Fig. 5A), western blots document total protein levels, i.e. protein in the cytoplasm as well as that targeted to the lateral membrane. In spite of this difference, both methods indicate that prestin protein levels are not reduced in heterozygous mice older than 3 weeks. However, because the average standard deviation for western blots was 16% of the mean, a small difference between genotypes would be difficult to determine.
, 百拇医药
A, an example of the western blots used to quantify prestin protein expression in animals having one versus two copies of the prestin gene. The wildtype and heterozygous mice were from the same litter and were 21 days of age. The molecular mass of monomeric prestin was estimated based on a linear regression of the molecular mass ladder. No band appears for the knockout mouse shown on the left. B, the amount of prestin protein expressed at P9, P21 and P63. Regions of interest were drawn around each of the bands at 100 kDa. A similar region of interest was used to define the intensity of the background, which was then subtracted from each prestin-associated band. The measurements reflect the integrated intensity of the 100 kDa band measured relative to that of the background. In contrast to results at P9 and P21, only one cochlea was required to generate results at P63. As in Fig. 5B, the data are normalized to the mean wildtype intensity and standard deviations appear as error bars. The latter are absent for the P21 control because the litter contained only one wildtype mouse. *Protein levels are statistically different between heterozygous and wildtype mice but only at P9 (Student's t test, P < 0.01).
, 百拇医药
Real-time RT-PCR
prestin mRNA was quantified using real-time RT-PCR beginning at P9 when prestin mRNA approaches its mature value in wildtype mice (Reisinger et al. 2005). Data in Fig. 7 confirm that mRNA is lower in heterozygotes at 9 days of age (Liberman et al. 2002). In other words, one copy of prestin generates about half the mRNA and half the protein (see Fig. 6B) of WT controls at this early stage of development. Data at P21 and P49 also indicate that prestin mRNA remains relatively low in young adult heterozygotes. Differences between genotypes are statistically significant at P9 and P21 (Student's t test, P < 0.01). In addition, the values at P49 approach significance with a P value of 0.052. Although the mRNA in heterozygotes is less than in controls at all ages tested, protein levels are near normal in young adult mice (see Figs 5B and 6B). Therefore, cochlear physiology and prestin protein levels are wildtype-like, in spite of the fact that mRNA levels are reduced in mice with only one copy of the prestin gene.
, 百拇医药
This figure provides prestin mRNA values for wildtype (filled bars) and heterozygous (shaded bars) mice obtained using real-time RT-PCR. Data are provided for litters at P9, P21 and P49. The values are normalized by plotting data from heterozygotes relative to the mean wildtype RNA level, which is given a value of 1.0. Standard deviations appear as error bars. The P values obtained using the Student's t test are also appended.
Discussion
, http://www.100md.com
The first report on prestin knockout mice (Liberman et al. 2002) stated that prestin mRNA in heterozygotes was half that in controls, antibody labelling for prestin protein was reduced and somatic electromotility was decreased by nearly one half. In addition, cochlear sensitivity, assessed using distortion product otoacoustic emissions (DPOAEs) and auditory brainstem responses (ABRs), was decreased by 6 dB for +/– mice when compared with wildtype controls. These results are unexpected because a halving of OHC motor function would be expected to reduce cochlear sensitivity by much more than 6 dB, based upon what is known about cochlear amplification (Mountain et al. 1983; Patuzzi et al. 1989; Fettiplace & Ricci, 2003; Santos-Sacchi, 2003). For example, the approximate halving of electric driving force, due to a large reduction in the endocochlear potential (EP) by furosemide, is known to change cochlear mechanics by reducing sensitivity and frequency selectivity (Ruggero & Rich, 1991). Sewell (1984) reported similar results in single unit responses recorded from the auditory nerve. More recently, Gow et al. (2004) demonstrated that Claudin-11 null mice, which lack basal cell tight junctions in the stria vascularis, have EPs below 30 mV and hearing thresholds elevated by 50 dB. These experimental results are consistent with a simple model of active OHCs (Patuzzi et al. 1989). This theoretical analysis indicates that abolition of the positive EP would reduce the driving force across the apex of cochlear OHCs by about one half, thereby reducing their receptor potentials, which provide the driving force for somatic electromotility. According to Patuzzi et al. (1989), these changes are predicted to produce a 46 dB change in sensitivity at the single unit level. A more elaborate hydrodynamic model (Neely & Kim, 1986) predicts even larger threshold shifts. It is therefore surprising that the changes in sensitivity reported by Liberman et al. (2002) were only 3–6 dB for DPOAEs and 1–8 dB for ABRs, considering that somatic motility and prestin protein expression were stated to be reduced by nearly one half.
, http://www.100md.com
In contrast to Liberman et al. (2002), data in this report demonstrate that CAP thresholds, CAP tuning curves and CM input–output functions are similar in heterozygotes and wildtype mice. Not only are the differences between genotypes small, they are less than 6 dB and less than the standard deviations. Even after testing each data point, the small difference in sensitivity, seen in mice with one versus two copies of the prestin gene, was not statistically significant at the 0.05 level. We also assessed OHC motor function by measuring NLC and somatic electromotility. Results indicate that differences between wildtype and heterozygous mice for peak electromotility, as well as , V and charge density (Qmax/ClinAdj) are not statistically different (Student's t test). In other words, there is no alteration in the sensitivity or voltage dependence of electromechanical transduction or in the maximum charge movement or the OHC length change in animals having one versus two copies of the prestin gene. We also evaluated data obtained by Jia et al. (2003) on F2 generation prestin heterozygotes and their controls, to obtain the linear capacitance associated with the active region only. Using their mean values, the adjusted charge density in mice with only one copy of the prestin gene was 95.4% of wildtype. These results from two different laboratories indicate near-normal OHC motor function in F2–F4 generation heterozygotes and their controls. All of these results are consistent with immunocytochemical and western blot analyses showing that prestin protein expression is wildtype-like in adolescent heterozygous mice. Therefore, we conclude that protein expression in heterozygotes is not one-half of that expressed in wildtype controls. Although prestin mRNA in heterozygotes is less than in wildtype controls at all ages tested (Fig. 7), near normal protein levels are observed (Figs 5B and 6B) at a time when the in vivo and in vitro physiology is similar in the two genotypes.
, 百拇医药
One might argue that differences between the results of Liberman et al. (2002) and those reported here could relate to the larger variability associated with measurements at the round window, i.e. to the more invasive procedures required to record cochlear potentials as suggested by Liberman et al. (2004). If true, then the differences between wildtype and heterozygous mice are very small and clinically insignificant. They are so small in fact, that they cannot be resolved because normal variability exceeds the difference between wildtype and heterozygous mice. Another possibility is that differences between the two studies relate to genetic drift (Wolfer et al. 2002), i.e. the use of F2 generation mice by Liberman et al. (2002)versus the F3–F5 generation mice used here. Here it is important to mention that DPOAE amplitudes in heterozygotes were similar to wildtype controls in F3–F5 generation mice (Liberman et al. 2004). In other words, DPOAEs were reduced in F2 but not F3–F5 heterozygotes. However, when considering this issue, it should also be stated that the NLC data from Jia et al. (2003) show normal OHC charge densities in heterozygous mice from the F2 generation when Clin is adjusted to reflect only that part of the membrane that contains motors. In other words, the NLC measured independently by two different groups is wildtype-like in F2–F4 heterozygotes. This observation argues against generational differences/genetic drift being the explanation for differences between the results reported here and those of Liberman et al. (2002). Perhaps subsequent prestin animal models, including those backcrossed to strains without age-related hearing loss, will help to resolve the discrepancy between the data presented here and those published by Liberman et al. (2002).
, 百拇医药
The physiological results in this report imply that OHCs compensate for the deleted prestin gene, thereby minimizing deleterious effects on peripheral auditory function. In other words, heterozygous mice do not exhibit haploinsufficiency. Because the safety margin that separates normal from pathological conditions is large, the hearing loss due to prestin removal in mice requires loss-of-function mutations in both alleles, i.e. it is recessive. These results must be considered in light of information on human subjects with a particular prestin mutation (Liu et al. 2003). Although 3.2% of the deaf probands examined showed heterozygosity, the audiological findings were difficult to interpret because hearing losses were highly variable. In fact, some individuals demonstrated moderate-to-profound hearing losses, suggesting haploinsufficiency, while one heterozygote reported no hearing loss at all. In addition, the data are few, the age of the participants was between 10 and 67 years, and the age of onset for hearing loss was from birth to 35 years. These complications, along with data showing near normal cochlear physiology and protein levels in adolescent heterozygous mice, implies that by itself one copy of the prestin gene does not necessarily result in reduced protein levels to a level that is insufficient to produce normal auditory function. If the deafness in human heterozygotes results from interactions with modifier genes, then these are probably more numerous than in mice. In fact, the mixing of backgrounds in human populations results in a large number of modifier genes that can increase variability by influencing the age of onset, progression and/or severity of hearing loss (Friedman et al. 2000). This contrasts with the limited mixing of C57BL6 and 129 mouse strains used in gene targeting. It is therefore not unexpected that mouse models differ from human disease conditions.
, 百拇医药
The data in this report also suggest that prestin protein expression is autoregulated. This possibility relates to the fact that prestin mRNA is well below wildtype at all ages tested, while the protein expression is wildtype-like in adolescent mice. We speculate that near-normal protein levels are attained prior to P33. This latter estimate is suggested because heterozygotes display wildtype-like physiology at P33, the youngest age tested. It is also difficult to know if prestin's autoregulation occurs at the transcriptional or translational level. Because this is not a developmental study, we do not have measurements at numerous points during development. It is therefore possible to miss a surge in prestin mRNA that might occur over a relatively short time frame. However, if there is no surge, i.e. the mRNA is 50% of wildtype at all ages, then it might be tempting to argue that regulation is at the translational level. Such a conclusion, however, would be premature because we are measuring steady-state RNA levels and know nothing about transcriptional rates or the stability of newly synthesized prestin mRNA (Bustin, 2002). We also know nothing about the stability of prestin protein, which could be influenced by glycosylation (Matsuda et al. 2004) and/or phosphorylation (Deak et al. 2005), or the details of membrane targeting and oligomerization. Additional experiments are therefore required in order to understand how prestin is regulated.
, 百拇医药
References
Ashmore JF (1987). A fast motile response in guinea-pig outer hair cells: The cellular basis of the cochlear amplifier. J Phyiol 388, 323–347.
Ashmore JF (1990). Forward and reverse transduction in the mammalian cochlea. Neurosci Res Suppl 12, S39–S50.
Belyantseva IA, Adler HJ, Curi R, Frolenkov GI & Kachar B (2000). Expression and localization of Prestin and the sugar transporter GLUT-5 during development of electromotility in cochlear outer hair cells. J Neurosci 20, RC116, 1–5.
, http://www.100md.com
Brownell WE, Bader CR, Bertrand D & Ribaupierre Y (1985). Evoked mechanical responses of isolated outer hair cells. Science 227, 194–196.
Bustin SA (2002). Quantification of mRNA using real-time reverse transcription PCR (RT-PCR): trends and problems. J Mol Endocrinol 29, 23–39.
Cheatham MA, Huynh KH, Gao J, Zuo J & Dallos P (2004). Cochlear function in Prestin knockout mice. J Physiol 560, 821–830.
, 百拇医药
Collyer CE & Enns JT (1987). Analysis of Variance, the Basic Designs. Nelson-Hall, Chicago IL.
Dallos P & Cheatham MA (1976a). Compound action potential (AP) tuning curves. J Acoust Soc Am 59, 591–597.
Dallos P & Cheatham MA (1976b). Production of cochlear potentials by inner and outer hair cells. J Acoust Soc Am 60, 510–512.
Dallos P & Fakler B (2002). Prestin, a new type of motor protein. Nature Rev 3, 104–111.
, 百拇医药
Deak L, Zheng J, Du GG, Orem A, Aguiaga S, Matsuda K et al. (2005). Effects of cyclic nucleotides on the function of prestin. J Physiol 563, 483–496.
Fettiplace R & Ricci AJ (2003). Adaptation in auditory hair cells. Curr Opin Neurol 13, 446–451.
Friedman T, Battey J, Kachar B, Riazuddin S, Noben-Trauth K, Griffith A et al. (2000). Modifier genes of hereditary hearing loss. Curr Opin Neurobiol 10, 487–493.
, 百拇医药
Gow A, Davies C, Southwood CM, Frolenkov G, Chrustowski M, Ng L et al. (2004). Deafness in Claudin-11 null mice reveals the critical contribution of basal cell tight junctions to stria vascularis function. J Neurosci 24, 7051–7062.
Gummer AW, Smolders JWT & Klinke R (1987). Basilar membrane motion in the pigeon measured with the Mssbauer technique. Hear Res 29, 63–92.
He DZZ, Cheatham MA, Pearce M & Vetter DE (2004). Mouse outer hair cells lacking the 9 Ach receptor are motile. Brain Res Dev Brain Res 148, 19–25.
, 百拇医药
Henry KR & Chole RA (1980). Genotypic differences in behavioral, physiological and anatomical expressions of age-related hearing loss in the laboratory mouse. Audiol 19, 369–383.
Huang G & Santos-Sacchi J (1993). Mapping the distribution of the outer hair cell motility voltage sensor by electrical amputation. Biophys J 65, 2228–2236.
Jia S, Kachar B, Gao J, Beisel K, Feng F, He DZZ et al. (2003). Properties of OHC motors in Prestin-mutant mice. Abstr 26th Midwinter Res Meet Assoc Res Otolaryngol 104.
, 百拇医药
Kachar B, Brownell WE, Altschuler R & Fex J (1986). Electrokinetic shape changes of cochlear outer hair cells. Nature 322, 365–368.
Li HS & Borg E (1991). Age-related loss of auditory sensitivity in two mouse genotypes. Acta Otolaryngol 111, 827–834.
Liberman MC, Gao J, He DZZ, Wu X, Jia S & Zuo J (2002). Prestin is required for electromotility of the outer hair cell and for the cochlear amplifier. Nature 419, 300–304.
, http://www.100md.com
Liberman MC, Zuo J & Guinan JJ (2004). Otoacoustic emissions without somatic motility: Can stereocilia mechanics drive the mammalian cochleaJ Acoust Soc Am 116, 1649–1655.
Lim D (1986). Cochlear anatomy related to cochlear micromechanics. A review. J Acoust Soc Am 67, 1686–1695.
Liu ZX, Ouyang XM, Xia XJ, Zheng J, Du Pandya ALL, Welch KO, Petit C, Smith RJ, Webb BT, Yan D, Arnos KS, Corey D, Dallos P, Nance WE & Chen ZY (2003). Prestin, a cochlear motor protein, is defective in non-syndromic hearing loss. Human Mol Genet 12, 1155–1162.
, http://www.100md.com
Matsuda K, Zheng J, Du GG, Klocker N, Madison LD & Dallos P (2004). N-linked glycosylation sites of the motor protein prestin: effects on membrane targeting and electrophysiological function. J Neurochem 89, 928–938.
Mountain DC, Hubbard AE & McMullen TA (1983). Electromechanical processes in the cochlea. In Mechanics of Hearing, ed. de Boer E, Viergever MA, pp. 119–126. Delft University Press, Delft, the Netherlands.
, 百拇医药 Neely ST & Kim DO (1986). A model for active elements in cochlear biomechanics. J Acoust Soc Am 79, 1472–1480.
Ohlemiller KK & Siegel JH (1992). The effects of moderate cooling on gross cochlear potentials in the gerbil: basal and apical differences. Hear Res 63, 79–89.
Oliver D & Fakler B (1999). Expression density and functional characteristics of the outer hair cell motor protein are regulated during postnatal development in rat. J Physiol 519, 791–800.
, http://www.100md.com
Oliver D, He DZZ, Klcker N, Ludwig J, Schulte U, Waldeggar S et al. (2001). Intracellular anions as the voltage sensor of prestin, the outer hair cell motor protein. Science 292, 2340–2343.
Patuzzi RB, Yates GK & Johnstone BM (1989). Outer hair cell receptor currents and sensorineural hearing loss. Hear Res 42, 47–72.
Pearce M, Richter C-P & Cheatham MA (2000). A reconsideration of sound calibration in the mouse. J Neurosci Meth 106, 57–67.
, 百拇医药
Pusch M & Neher E (1998). Rates of diffusional exchange between small cells and a measuring patch pipette. Pflugers Arch 411, 204–211.
Reisinger E, Zimmermann U, Knipper M, Ludwig J, Klcker N, Fakler B et al. (2005). Cod106, a novel synaptic protein expressed in sensory hair cells of the inner ear and in CNS neurons. Mol Cell Neurosci 28, 106–117.
Ruggero MA & Rich NC (1991). Furosemide alters organ of Corti mechanics: evidence for feedback of outer hair cells upon the basilar membrane. J Neurosci 11, 1057–1067.
, 百拇医药
Santos-Sacchi J (1991). Reversible inhibition of voltage-dependent outer hair cell motility and capacitance. J Neurosci 11, 3096–3110.
Santos-Sacchi J (2003). New tunes from Corti's organ: the outer hair cell boogie rules. Curr Opin Neurobiol 13, 1–10.
Santos-Sacchi J & Dilger JP (1988). Whole-cell currents and mechanical responses of isolated outer hair cells. Hear Res 35, 143–150.
Santos-Sacchi J & Navarrete E (2002). Voltage-dependent changes in specific membrane capacitance caused by prestin, the outer hair cell lateral membrane motor. Pflugers Arch 444, 99–106.
, 百拇医药
Sewell WF (1984). The effects of furosemide on the endocochlear potential and auditory-nerve fiber tuning curves in cats. Hear Res 14, 305–314.
Shnerson A & Pujol R (1981). Age-related changes in the C57BL/6J mouse cochlea. I. Physiological findings. Dev Brain Res 2, 65–75.
Shore SE & Nuttall AL (1985). The effects of cochlear hypothermia on compound action potential tuning. J Acoust Soc Am 77, 590–598.
, 百拇医药
Stankovic KM & Corfas G (2003). Real-time quantitative RT-PCR for low-abundance transcripts in the inner ear: analysis of neurotrophic factor expression. Hear Res 185, 97–108.
Taylor MM & Creelman CD (1967). PEST: Efficient estimates on probability functions. J Acoust Soc Am 41, 782–787.
Weber T, Zimmerman U, Winter H, Mack A, Kpschall Rohbock K et al. (2002). Thyroid hormone is a critical determinant for the regulation of the cochlear motor protein prestin. Proc Nat Acad Sci U S A 99, 2901–2906.
, 百拇医药
Wolfer DP, Crusio WE & Lipp H-P (2002). Knockout mice: simple solutions to the problems of genetic background and flanking genes. Trends Neurosci 25, 336–340.
Zheng QY, Johnson KR & Erway LC (1999). Assessment of hearing in 80 inbred strains of mice by ABR threshold analyses. Hear Res 130, 94–109.
Zheng J, Long KB, Shen W, Madison LD & Dallos P (2001). Prestin topology: localization of protein epitopes in relation to the plasma membrane. Neuroreport 12, 1929–1935.
Zheng J, Shen W, He DZZ, Long KB, Madison LD & Dallos P (2000). Prestin is the motor protein of cochlear outer hair cells. Nature 405, 149–155., 百拇医药(M. A. Cheatham, J. Zheng,)