Complexes of vertebrate TMC1/2 and CIB2/3 proteins form hair-cell mechanotransduction cation channels

Calcium and integrin-binding protein 2 (CIB2) and CIB3 bind to transmembrane channel-like 1 (TMC1) and TMC2, the pore-forming subunits of the inner-ear mechano-electrical transduction (MET) apparatus. Whether these interactions are functionally relevant across mechanosensory organs and vertebrate species is unclear. Here we show that both CIB2 and CIB3 can form heteromeric complexes with TMC1 and TMC2 and are integral for MET function in mouse cochlea and vestibular end organs as well as in zebrafish inner ear and lateral line. Our AlphaFold 2 models suggest that vertebrate CIB proteins can simultaneously interact with at least two cytoplasmic domains of TMC1 and TMC2 as validated using nuclear magnetic resonance spectroscopy of TMC1 fragments interacting with CIB2 and CIB3. Molecular dynamics simulations of TMC1/2 complexes with CIB2/3 predict that TMCs are structurally stabilized by CIB proteins to form cation channels. Overall, our work demonstrates that intact CIB2/3 and TMC1/2 complexes are integral to hair-cell MET function in vertebrate mechanosensory epithelia.

Methods
Labeling zebrafish hair cells with FM 1-43, FM 4-64, and FM 1-43FX

To label lateral-line hair cells with FM 1-43, larvae were immersed in 1 µM FM 1-43 (ThermoFisher, T3163) in E3 for 30 s, and then washed 3 times in E3. After washing, larvae were mounted in 1% low melt agarose containing an anesthetic, 0.02% tricaine methanesulfonate (Syndel, ANADA 200-226). FM 1-43 labeled hair cells were then imaged on an A1 Nikon upright laser-scanning confocal microscope with a 60× 1.0 NA water objective. FIJI was used to process image Z-stacks. To estimate the number of hair cells per neuromast, laser scanning DIC images were used to count the number of hair bundles per neuromast. To quantify FM 1-43 label, a 5 µm circular region of interest (ROI) was placed on the soma of each hair cell with a hair bundle in the DIC image. These ROIs were used to quantify the mean FM 1-43 intensity per hair cell. These intensity values were averaged to calculate the average FM 1-43 intensity per neuromast.

FM 4-64 (ThermoFisher, T13320) was used to label hair cells in the crista within the zebrafish inner ear. For this labeling, Tg(myo6b:memGCaMP6s)idc1larvae were mounted on their side on Sylgard chamber in E3 embryo media containing 0.02% MESAB. Approximately 2-3 nL of 10 µM FM 4-64 in 0.1 M KCl were injected into the otic capsule as previously described (Smith et al., 2020). FM 4-64 labeled hair cells of the medial crista were then imaged on an A1 Nikon upright laser-scanning confocal microscope with a 25× 1.10 NA or 60× 1.0 NA water objective. FIJI was used to process image Z-stacks. To assess tall versus short hair cells in each crista memGCaMP6s label was used.

To correlate hair bundle orientation in lateral line hair cells with intact mechanotransduction, larvae were immersed in 3 µM FM 1-43FX (ThermoFisher, F25255) in E3 for 30 s, and then washed 3 times in E3. After washing, larvae were fixed with 4% paraformaldehyde in PBS at 4°C overnight. After fixation, larvae were washed 4 × 5 min in 0.1% Tween in PBS (PBST). Larvae were then placed in blocking solution (2% goat serum, 1% bovine serum albumin, 2% fish skin gelatin in PBS + 0.3% triton) for 2 h at room temperature. After block Alexa Fluor 633 Phalloidin (ThermoFisher, A22284) was added at 1:100 in PBST. Larvae were incubated for 4 h at room temperature and then washed 4 × 5 min in PBST. Larvae were mounted on glass slides with Prolong Gold (ThermoFisher Scientific) using No. 1.5 coverslips. To image hair bundle orientation and FM 1-43FX label in lateral-line hair cells, a Zeiss LSM780 confocal microscope with Airyscan was used. Hair cells were imaged using a 63× 1.3 NA oil objective and processed with an Airyscan processing factor of 7.0. For analysis, Z-stacks were max-projected in FIJI. Images were background subtracted using a rolling ball correction. Hair cells with a FM 1-43FX intensity more than twice the background intensity were scored as FM 1-43FX positive.

Zebrafish Ca2+ imaging and behavior

For imaging of mechanosensitive Ca2+ signals in neuromast hair bundles, Tg(myo6b:memGCaMP6s) animals were mounted on their side on Sylgard chamber in E3 embryo media containing 0.2% MESAB (Lukasz and Kindt, 2018). Larvae were then pinned and a solution containing α-bungarotoxin (125 µM, Tocris) was injected into the heart cavity for paralysis during imaging. After paralysis, larvae were immersed in extracellular imaging solution (in mM: 140 NaCl, 2 KCl, 2 CaCl2, 1 MgCl2 and 10 HEPES, pH 7.3, OSM 310 +/-10). To image Ca2+ signals in the hair bundles we used a Bruker Swept-Field confocal system equipped with a Rolera EM-C2 CCD camera (QImaging, Surrey, Canada) and a Nikon CFI Fluor 60×1.0 NA water immersion objective as previously described (Q. Zhang et al., 2018). The system includes a band-pass 488/561 nm filter set (59904-ET, Chroma) and is controlled using Prairie View software (Bruker Corporation). A piezoelectric motor (PICMA P-882.11-888.11 series, PI Instruments) attached to the objective and used to acquire 5-plane Z-stacks every 0.5 µm using a 35 µm slit at a 50-Hz frame rate for a 10-Hz volume rate. A fluid-jet composed of a glass capillary attached to a pressure clamp system (HSPC-2-SB, ALA) was used to deliver a 500-ms anterior and posterior step stimulation to deflect hair bundles along their axis of sensitivity.

To quantify GCaMP6s fluorescent intensity changes during stimulation, we registered the Z-stacks and then average projected the Z-stacks into a single plane. We then loaded the projected images into FIJI and used the Time Series Analyzer V3 plugin to create circular ROIs with a ∼1.7 μm diameter. ROIs were placed on the center of each individual bundle. We then measured and plotted change in the mean intensity (ΔF/F0) within the region during the recording period. The mean intensity within each ROI was computed for each hair bundle. A hair bundle with a signal magnitude (peak value of intensity change upon stimulation) above 10% ΔF/F0 was considered mechanosensitive. To create the AVG hair bundle GCaMP6s (ΔF/F0) per neuromast plot in Figure 5-figure supplement 1C, only traces from mechanosensitive cells were averaged.

A Zantiks MWP behavioral system was used to examine startle responses. The Zantiks system tracked and monitored behavioral responses via a built-in infrared camera at 30 frames per second. A 12-well plate was used to house larvae during behavioral analysis. Each well was filled with E3 and 1 larva. All fish were acclimated in the plate within the Zantiks chamber in dark for 15 min before each test. To induce startle, an integrated stepper motor was used to drive a vibration-induced startle response. A vibrational stimulus that triggered a maximal % of animals startling in controls without any tracking artifacts (due to the vibration) was used for our analyses. Each larva was presented with the vibrational stimulus 5 times with 100 s between trials. During the tracking and stimulation, a Cisco router connected to the Zantiks system was used to relay x, y coordinates of each larva every frame. To qualify as a startle response, a distance above 4 pixels or ∼1.9 mm was required within 2 frames after stimulus onset. Animals were excluded from our analysis if no tracking data was recorded for the animal.