Here, we present a method for visualizing the uptake of 3 kDa Texas Red-labeled dextran in auditory hair cells with functional mechanotransduction channels. In addition, dextrans of 3–10 kDa can be used to study endocytosis in hair and supporting cells of the organ of Corti.
The hair cell mechanotransduction (MET) channel plays an important role in hearing. However, the molecular identity and structural information of MET remain unknown. Electrophysiological studies of hair cells revealed that the MET channel has a large conductance and is permeable to relatively large fluorescent cationic molecules, including some styryl dyes and Texas Red-labeled aminoglycoside antibiotics. In this protocol, we describe a method to visualize and evaluate the uptake of fluorescent dextrans in hair cells of the organ of Corti explants that can be used to assay for functional MET channels. We found that 3 kDa Texas Red-labeled dextran specifically labels functional auditory hair cells after 1–2 h incubation. In particular, 3 kDa dextran labels the two shorter stereocilia rows and accumulates in the cell body in a diffuse pattern when functional MET channels are present. An additional vesicle-like pattern of labeling was observed in the cell body of hair cells and surrounding supporting cells. Our data suggest that 3 kDa Texas-Red dextran can be used to visualize and study two pathways for cellular dye uptake; a hair cell-specific entry route through functional MET channels and endocytosis, a pattern also available to larger dextran.
The hair cells of the inner ear are the sensory cells that detect sound and covert the mechanically stimuli in electrical signals, which are ultimately interpreted by our brain. These cells have a staircase-shaped bundle of three rows of actin-based filaments, known as stereocilia, which protrude from their apical region1,2. The mechanical stimuli deflect the stereocilia filaments toward the longest row and trigger the opening of the mechanotransduction (MET) channels3. The opening of the MET channels leads to an influx of cations that depolarizes the cell and consequently signals the release of synapse vesicles at the basal region of the hair cell.
The biophysical properties of the MET channel essential for hearing have been extensively characterized. Among other properties, these channels are cationic selective and have a relatively large conductance (150–300 pS in low Ca2+)4,5,6,7,8,9,10. Remarkably, large fluorescent molecules such as FM1-43 and Texas Red-labeled aminoglycosides are permeant blockers of the MET channel, resulting in their accumulation in the hair cell body that can be visualized using fluorescence microscopy11,12,13,14. Conversely, the molecular identity and the structure of the MET channel and its permeation pathway have remained elusive. Increasing experimental evidence indicates that the transmembrane-like channel protein 1 (TMC1) is a component of the MET channel in mature hair cells15,16,17,18,19. Mutations in the transmembrane-like channel 1 (TMC1) alter the MET channel properties19,20,21,22 and cause deafness. In addition, TMC1 localizes to the site of the MET channel18,23 and interacts with the tip-link responsible for transmitting the mechanical force to the MET channel24,25. Furthermore, recent bioinformatics analysis has identified the TMC proteins as evolutionary related to the mechanosensitive channels TMEM63/OSCA proteins and the TMEM16 proteins, a family of calcium-activated chloride channels and lipid scramblases26,27,28. A structural model of TMC1 based on the relationship between these proteins revealed the presence of a large cavity at the protein-lipid interface27. This cavity harbors the two TMC1 mutations that cause autosomal dominant hearing loss (DFNA36)27,29,30,31,32, and selective modification of cysteine mutants for residues in the cavity alter MET channel properties28, indicating that it could function as the permeation pathway of the MET channel. The large size of this predicted cavity in TMC proteins could explain the ability of large molecules to permeate the MET channel. To test the prediction that the MET channel contains an unusually large permeation pathway and to push the limits of the size of the cavity observed in TMC1, we developed a protocol to perform uptake experiments in the organ of Corti explants with a larger molecule, 3 kDa dextran fluorescently labeled with Texas Red.
Dextran is a complex branched polysaccharide composed of many D-glucose molecules bound by alpha-1,6 glycosidic linkages. Its high solubility in water, low cell toxicity, and bioinertity make it a versatile tool to study several cellular processes. In addition, dextran is available in a wide range of sizes and fluorescently labeled with fluorophores of several colors. Fluorescently labeled dextrans are commonly used in cell and tissue permeability research33,34, to study endocytosis in multiple cellular systems35,36, and for neural tracing37,38. In the auditory field, dextran molecules have also been used to assess the disruption of the cell-cell junction and loss of the auditory sensory epithelium integrity after exposure to intense noise in the chinchilla organ of Corti39,40.
In this work, we exploited the properties of some of the smallest (3 and 10 kDa) fluorescent dextrans to perform uptake experiments in murine inner ear hair cells and explore the size of the permeation pathway of the inner ear hair cell MET channel. In addition, we used a laser-scanning confocal microscope (LSM) 880 equipped with an Airyscan detector to visualize and localize fluorescent dextran at the stereocilia and the cell body of auditory hair cells.
The animal care and experimental procedures were performed following the guidelines for the Care and Use of Laboratory Animals, which were approved by the Animal Care and Use Committee of the National Institute of Neurological Disorders and Stroke (Animal protocol #1336 to KJS).
1. Mice
2. Cochleae Dissection
3. Dextran Labeling
4. Sample Preparation for Imaging
5. Image Acquisition and Image Processing
NOTE: The confocal images were taken with a LSM 880 confocal microscope equipped with a 32 channel Airyscan detector in the super-resolution (SR) mode43 and a 63x objective.
We observed robust and specific labeling of hair cells after 2h incubation of organ of Corti explants from wild-type postnatal-day-6 (P6) mice with 3 kDa dextran fluorescently labeled with Texas Red (dextran-TR) (Figure 2A-B). Dextran labeling was observed in both inner and outer hair cells (IHC and OHC) at the basal, middle, and apical regions of the organ of Corti (Figure 2B).
Fluorescently labeled phalloidin was used to counterstain filamentous actin (F-actin) and visualize the actin-based hair cell stereocilia. We also performed similar experiments at shorter incubation times, and although we observed dextran-TR accumulation in the hair cell body, the signals were weaker and more variable than those at 2 h incubation (Figure 2C).
We next imaged both stereocilia and cell bodies of the hair cells and observed that only those cells that incorporated 3 kDa dextran-TR in their cell body showed fluorescent labeling of their stereocilia (Figure 3A). This relationship between stereocilia and cell body labeling was absent in hair cells from damaged tissue, which also presented unspecific labeling of dextran in several cell types of the sensory epithelium (yellow squares in Figure 2B, and Figure 3B). Importantly, we observed a uniform fluorescence signal along the stereocilia with enrichment at the tips of the shorter stereocilia rows where the MET channel is located18,23 (Figure 3C). Also, we noticed vesicle-like structures in the cell body of the hair cells and the neighboring supporting cells. The vesicle-like pattern of uptake in these cells suggests that dextran-TR can also be taken up by endocytosis (Figure 3C).
The protocol described here also allows for the examination of the uptake of larger dextrans and combinations of different dextrans. Larger dextran of 10 kDa labeled with Texas Red (dextran-TR) or fluorescein (dextran-FITC) also produced a vesicle-like pattern in the cell body of hair cells and supporting cells, in addition to accumulating around the hair cell membrane in a patchy pattern (Figure 4A,B). The 10 kDa dextran-TR also superficially labeled the three stereocilia rows of all the hair cells (Figure 4A, inset), probably due to the negatively charged surface of the hair cell plasma membrane44,45,46. We next examined the uptake of 3 kDa dextran-TR and 10 kDa dextran-FITC simultaneously in the organ of Corti explants. For the 3 kDa dextran-TR, we observed both a diffuse and a vesicle-like pattern. However, the 10 kDa dextran-FITC only displayed a vesicle-like pattern (Figure 4C). These data suggest that dextran uptake occurs by at least two distinct mechanisms that are dependent on the size of the dextran.
We next assessed whether functional MET channels were required for the uptake of 3 kDa dextran-TR. To do this, we tested dextran incorporation in hair cells from the organ of Corti explants in the presence of MET channel blockers (neomycin and amiloride)13,14,47 or the Ca2+ chelator BAPTA, which abolishes MET currents by breaking the tip-links and preventing the gating of the channel25,48,49. In these experiments, the tissue explants were previously incubated for 30 min with neomycin (500 mM), amiloride (150 mM), or with BAPTA (5mM) before the addition of 3 kDa dextran-TR in the presence of the corresponding MET blocker. The presence of BAPTA or the MET channel blockers prevented the stereocilia labeling (Figure 5A) and the uptake of 3 kDa dextran-TR in hair cells (Figure 5B). However, blockade of the MET channel preserved the vesicle-like pattern (Figure 5B), indicating that this pattern of uptake is independent of functional MET channels. Similar results have been observed in hair cells from TMC1/TMC2 double knock-out mice, which lack MET27. Intriguingly, these vesicle-like structures were not observable in the presence of amiloride, which is known to inhibit the Na+-H+ exchanger and thereby inhibit endocytosis50,51. These results indicate that 3kDa dextran-TR enters the hair cells through two different pathways, one common to larger dextrans involving vesicle-like structures and another that depends on functional MET channels.
Figure 1: Steps of the dissection of a murine organ of Corti. (A) Clean area and materials required for the dissection. (B) Exposed mouse cranium. Forceps are holding the skin, which has folded towards the nose. Black lines indicate the two incisions that are required for the removal of the brain. (C) Incised and opened cranium to allow for removal of the brain with a spatula (on the right). (D) Cranium and temporal bones after brain removal. (E) Excised skull, including the temporal bones (dashed black squares) in a P35 plate covered with media. (F) Excised cochleae. (G) Two dissected organs of Corti on the well of a glass depression plate. (H) Mounted sample on a glass coverslip ready for imaging. Please click here to view a larger version of this figure.
Figure 2: Hair cells uptake 3 kDa dextran-TR. (A) Schematic representation of Texas Red-labeled dextran (dextran-TR) containing six molecules of glucose corresponding to a molecular weight of 1.08 kDa. A succinimidyl ester reaction links a Texas Red molecule (magenta) to a glucose monomer. (B) Representative confocal image showing specific labeling of sensory hair cells (HC) with 3 kDa dextran-TR across the whole organ of Corti from a 6 day-old mouse. The basal (BA), middle (MD), and apical (AP) regions of the organ are indicated. Yellow squares indicate tissue damaged regions. (C) 3 kDa dextran-TR fluorescence after 30, 60, 90, or 120 min incubation with the organ of Corti explants is shown in magenta merged with F-actin in green (top images) and independently in grayscale (bottom images). The image display range was linearly adjusted in each one of the gray images independently for visualization of the 3 kDa dextran-TR fluorescence. Scale bar = 20 mm. This figure has been modified from27. Please click here to view a larger version of this figure.
Figure 3: Hair cell body and stereocilia labeling with 3 kDa dextran-TR. (A) Confocal images displaying 3 kDa dextran-TR fluorescence (magenta) at the cell body and stereocilia counterstained with phalloidin to label F-actin (green) to visualize hair cell boundaries and stereocilia. The three rows of outer hair cells (OHC) and one row of inner hair cells (IHC) are indicated. Scale bar = 20 mm. (B) Confocal images of a damaged tissue area displaying 3 kDa dextran-TR fluorescence (magenta) at the cell body and stereocilia. Yellow arrows indicate the labeling of non-sensory supporting cells. Scale bar = 20 mm. (C) Closer view of 3 kDa dextran-TR fluorescence at the cell body and stereocilia of outer hair cells (OHC, top) and inner hair cells (IHC, bottom). Scale bar = 5 mm. This figure has been modified from27. Please click here to view a larger version of this figure.
Figure 4: Labeling of larger 10 kDa dextran and a combination of 3 and 10 kDa dextran. (A) Confocal image of hair cells at the hair cell body (top) and stereocilia (bottom) levels incubated with 10 kDa dextran-TR. The dextran fluorescence signal is shown in magenta merged with F-actin in green and independently in grayscale. A couple of IHC are shown at higher magnification in the inset to appreciate the superficial labeling of stereocilia observed with the 10 kDa dextran-TR. Scale bar = 5 mm. (B) Confocal image of hair cells incubated with 10 kDa dextran-FITC represented as in A. (C) Confocal image of hair cells at the hair cell body (top) and stereocilia (bottom) levels incubated simultaneously with 3 kDa dextran-TR and 10 kDa dextran-FITC and imaged using independent channels and depicted separately in gray. In the right panel, F-actin (blue), 10 kDa dextran-FITC (green), and 3 kDa dextran-TR (magenta) are shown together along with phalloidin (blue). Scale bar = 20 mm, and the vesicle-like pattern of uptake is indicated with yellow arrows. This figure has been modified from27. Please click here to view a larger version of this figure.
Figure 5: Functional MET channels are required for 3 kDa dextran-TR uptake. Representative confocal images of hair cells at the stereocilia (A) or cell body (B) level after incubation with 3 kDa dextran-TR in the absence (control) or presence of BAPTA or the MET channel blockers amiloride and neomycin. 3 kDa dextran-TR fluorescence is shown in magenta. Tissue was counterstained with phalloidin to label F-actin for visualization of the hair cell stereocilia and boundaries (green). White arrows point to vesicle-like structures. The three rows of outer hair cells (OHC) and one row of inner hair cells (IHC) are indicated, Scale bar = 20 mm. This figure has been modified from27. Please click here to view a larger version of this figure.
This protocol describes how to perform uptake experiments in murine organ of Corti explants with 3 kDa dextran Texas Red. The goal of this method is to test whether molecules larger than others previously tested were also able to specifically label auditory hair cells and permeate through the MET channel. Similar experimental protocols have been previously used to evaluate the permeability of hair cells to other fluorescent dyes such as FM1-43 (0.56 kDa)12,19,20 and Texas Red-labeled aminoglycosides (1.29–1.43 kDa)13,52. The time needed for labeling hair cells using these fluorescent molecules varies with their size. The maximal labeling requires a few minutes for FM1-4313,53, 30-60 min for Texas Red-labeled aminoglycosides13, and 1.5 h for Texas Red-labeled 3 kDa dextran. Thus, long incubation times are crucial for this experiment as we saw only minimal labeling of the hair cell body at short incubation times (Figure 2C). We chose to study P6 mice for these experiments because, at this age, most apical and basal hair cells carry transduction currents and express both TMC1 and TMC2 proteins12,18,54,55. It should be possible to study younger animals as MET currents are present earlier than P6, in particular near the base of the cochlea12. Studying older animals may be more challenging because dissection of the bony cochlea becomes more difficult as this structure continues to ossify56.
There are several critical steps in this protocol. The most crucial one being the successful dissection and preparation of the organ of Corti tissue (Figure 1). The delicate structure of the organ of Corti and encasement in the boney cochlea presents a challenge for histological and biochemical analysis. Several detailed protocols for the successful dissection of this tissue have been previously published in this journal56,57,58 and others59. Disruption of the cell-cell junction and loss of the auditory sensory epithelium integrity may lead to the uptake of dextran of up to 40 kDa and the labeling of the affected cells39,40. In agreement with this, we observed unspecific labeling of hair cells and the surrounding supporting cells when the tissue was unintentionally damaged with the forceps during dissection before dextran incubation (Figure 2B and Figure 3B). There are no tricks or guidelines to reach proficiency at dissecting an intact organ of Corti other than practice and patience. However, the presence of 1–2 mM Ca2+ during tissue dissection and uptake experiments is crucial to maintain the integrity of the tip-link that transmits the mechanical force to the MET channel and thereby preserve MET channel function. Once the tissue is fixed with 4% PFA, the presence of calcium in the buffer or media becomes trivial.
One of the limitations of this study resides in the heterogeneous nature of the dextran molecules. The dextran elaborated by the Leuconostoc bacteria is a polymer of anhydroglucose composed of approximately 95% alpha-(1–6) D-linkages (Figure 2A). The remaining linkages account for the branching of dextran, which correlates with its length, so dextran of lower molecular weight are more rod-like and have a narrower size distribution while larger dextrans are polydisperse60. Therefore, the actual molecular weight of the molecules in each preparation of the 3 kDa dextran ranges from 1.5 to 3.0 kDa, including the fluorescent molecule61. Additionally, the fluorescently labeled dextran differ in the number of Texas Red molecules attached to the dextran (degree of labeling). It is advisable to use dextran with a degree of labeling of at least 0.4 since a lower degree of labeling will hinder the ability to detect the fluorescently labeled dextran by microscopy. The fluorescent dye coupled to the dextran and the lysine residues in the fixable versions, will determine the overall charge of the dextran. This is an important consideration since the cationic selectivity of the MET channel would preferably uptake cationic dextran27. Lastly, a lysine-fixable dextran is necessary to reach the resolution displayed in these experiments. A non-fixable dextran will continue to diffuse into the tissue hampering its accurate localization and visualization.
The LSM 880 confocal microscope equipped with an Airyscan detector used to image the organ of Corti samples provided us with twice the resolution, and better signal-to-noise ratio (SNR) compared to a conventional confocal microscope62. Therefore, this microscope allowed us not only to observe the robust dextran accumulation at the cell body but also to pinpoint the accumulation of the dextran at the stereocilia tips and the vesicle-like pattern at the cell body. A conventional confocal microscope would allow for the visualization and quantification of the accumulation of fluorescent dextran at the cell body, but it would difficult to distinguish the three rows of stereocilia and the precise localization of the dextran at the tips of the shorter stereocilia rows. A similar resolution to the one achieved with the LSM 880 Airyscan confocal could be reached using a conventional confocal microscope with oversampling and deconvolution, although the SNR would be lower than the one reached with the Airyscan detector.
In addition to the use of 3 kDa dextran-TR to assay for functional MET channels, this protocol could be further used to study endocytic processes in hair cells since all the dextrans tested reveal an intracellular vesicle-like pattern resembling endocytosis. This process was not the focus of our work, and additional studies would be required to fully characterize this phenomenon.
The authors have nothing to disclose.
We thank Vincent Schram from the NICHD microscopy and imaging core for assisting in the confocal image acquisition, and Tsg-Hui Chang for invaluable help with colony management and mice care. This research was supported by the Intramural Research Program of the NINDS, NIH, Bethesda, MD, to K.J.S. A.B. was supported by the Intramural Research Program of the NINDS, NIH, and by a Robert Wenthold Postdoctoral Fellowship from the intramural research program of the NIDCD.
#1.5 glass coverslips 18mm | Warner Instruments | 64-0714 | |
Alexa Fluor 488 Phalloidin | ThermoFisher | A12379 | |
Alexa Fluor 594 Phalloidin | ThermoFisher | A12381 | |
alpha Plan-Apochromat 63X/1.4 Oil Corr M27 objective | Carl Zeiss | 420780-9970-000 | |
Amiloride hydrochloride | EMD MILLIPORE | 129876 | |
Benchwaver 3-dimensional Rocker | Benchmarks scientific | B3D5000 | |
C57BL/6J wild-type mice | strain 000664 | The Jackson Laboratory | |
Cell impermeant BAPTA tetrapotassium salt | ThermoFisher | B1204 | |
Dextran, Fluorescein, 10,000 MW, Anionic, Lysine Fixable | ThermoFisher | D1820 | |
Dextran, Texas Red, 10,000 MW, Lysine Fixable | ThermoFisher | D1863 | |
Dextran, Texas Red, 3000 MW, Lysine Fixable | ThermoFisher | D3328 | |
Formaldehyde Aqueous Solution EM Grade | Electron microscopy science | 15710 | |
HBSS, calcium, magnesium, no phenol red | ThermoFisher | 14025 | |
HBSS, no calcium, no magnesium, no phenol red | ThermoFisher | 14170 | |
Image J or FIJI | NIH | http://fiji.sc/ | |
Immersol 518F oil immersion media | Carl Zeiss | 444970-9000-000 | |
Leibovitz's L-15 Medium, GlutaMAX Supplement | ThermoFisher | 31415029 | |
neomycin trisulfate salt hydrate | Sigma | N6386 | |
PBS (10X), pH 7.4 | ThermoFisher | 70011069 | |
Phalloidin-CF405M | Biotium | 00034 | |
ProLong Diamond antifade mounting media | ThermoFisher | P36970 | |
superfrost plus microscope slide | Fisherbrand | 22-037-246 | |
Triton X-100 | Sigma | T8787 | |
Zen Black 2.3 SP1 software | Carl Zeiss | https://www.zeiss.com/microscopy/us/products/microscope-software/zen.html |