HCN channels are not required for mechanotransduction in sensory hair cells of the mouse inner ear

Abstract

The molecular composition of the hair cell transduction channel has not been identified. Here we explore the novel hypothesis that hair cell transduction channels include HCN subunits. The HCN family of ion channels includes four members, HCN1-4. They were originally identified as the molecular correlates of the hyperpolarization-activated, cyclic nucleotide gated ion channels that carry currents known as If, IQ or Ih. However, based on recent evidence it has been suggested that HCN subunits may also be components of the elusive hair cell transduction channel. To investigate this hypothesis we examined expression of mRNA that encodes HCN1-4 in sensory epithelia of the mouse inner ear, immunolocalization of HCN subunits 1, 2 and 4, uptake of the transduction channel permeable dye, FM1-43 and electrophysiological measurement of mechanotransduction current. Dye uptake and transduction current were assayed in cochlear and vestibular hair cells of wildtype mice exposed to HCN channel blockers or a dominant-negative form of HCN2 that contained a pore mutation and in mutant mice that lacked HCN1, HCN2 or both. We found robust expression of HCNs 1, 2 and 4 but little evidence that localized HCN subunits in hair bundles, the site of mechanotransduction. Although high concentrations of the HCN antagonist, ZD7288, blocked 50-70% of the transduction current, we found no reduction of transduction current in either cochlear or vestibular hair cells of HCN1- or HCN2- deficient mice relative to wild-type mice. Furthermore, mice that lacked both HCN1 and HCN2 also had normal transduction currents. Lastly, we found that mice exposed to the dominant-negative mutant form of HCN2 had normal transduction currents as well. Taken together, the evidence suggests that HCN subunits are not required for mechanotransduction in hair cells of the mouse inner ear.

Figure 1. Expression of HCN mRNA.

(A) Conventional RT-PCR was used to examine the expression of HCN1-4 in vestibular epithelia of wild-type mice. Arrows indicate the expected size of the PCR product: HCN1: 491 bp; HCN2: 337 bp; HCN3: 339 bp and HCN4: 419 bp. Lane 1 in each gel contained markers. Each subsequent lane contained the PCR product obtained using template cDNA harvested from the following sources. Lane 2: wild-type mouse brain; Lane 3: HCN1-deficient mouse brain; Lane 4: wild-type utricle; Lane 5: HCN1-deficient mouse utricle. (B) Quantitative RT-PCR was used to estimate total number of copies of messenger RNA for each of the four HCN subunits using wild-type mouse cochlea as template. Total copies in thousands of HCN mRNA transcripts per cochlea. Results were from a pool of 4 cochlea obtained from mice at P4.

Figure 2. Immunolocalization of HCN subunits in the inner ear.

All panels show confocal images of mouse inner ear epithelia with phalloidin staining shown on the left, HCN1 in the center and the merged image on the right. Phalloidin is shown in red and HCN1 in green. All scale bars indicate 5 µm. (A) Stereociliary bundles of wild-type mouse utricle at P8 stained with an antibody directed against the N-terminus of HCN1. (B) Basolateral hair cell membranes of wild-type mouse utricle stained with the N-terminal HCN1 antibody. (C) Stereociliary bundles of wild-type mouse cochlea harvested from the apex at P8 stained with same N-terminal HCN1 antibody. (D) Confocal image of the stereociliary bundles from a P8 utricle of a HCN1^−/− mouse stained the same HCN1 antibody shown in panels A–C. (E) Wild-type utricle focused at the hair bundle level stained with a different antibody that recognizes an epitope in the C-terminus. (F) Wild-type cochlear hair bundles stained with the antibody that recognizes the epitope in the C-terminus.

Figure 3. Mechanotransduction in wild-type and HCN-deficient hair cells.

Representative mechanotransduction currents evoked by hair bundle deflections. The left column of data were recorded from mouse utricle type II hair cells at P3 - P6. The data in the right column were recorded from mouse cochlear outer hair cells at P6 - P7. The scale bars at the top apply to all data in the column. The deflection protocol is shown at the bottom of trace panels I and J. Data were recorded under the following conditions: (A & B) wild-type control; (C & D) in the presence of 500 µM ZD7288; (E & F) HCN1^−/−; (G & H) HCN2^−/−; (I& J) HCN1^−/− and HCN2^−/−. Panels K-P show fluorescent images of hair cells following application of the transduction channel permeable dye, FM1-43, in utricle (P5 –P7) and organ of Corti (P6 – P9) sensory epithelia under the following conditions: (K & L) HCN1^−/−; (M & N) HCN2^−/−; (O& P) HCN1^−/− and HCN2^−/−. Uptake of FM1-43 appeared normal in all tissues examined. The scale bar in panel (O) equals 10 µm and also applies to panels K and M. The scale bar in panel (P) equals 5 µm and also applies to panels L and N. (Q) Summary of transduction currents recorded from vestibular and auditory hair cells. Maximum current amplitudes under each condition were averaged and were normalized relative to wild-type controls. Error bars show standard deviation; the number of samples is indicated above each bar.

Figure 4. Whole-cell currents recorded from control and transfected cells.

(A & B) Representative mechanotransduction currents recorded from mouse utricle type II hair cells at P3 - P6. Bundle deflections were evoked using the protocol shown at the bottom of figure 3I. Panel A shows data from a non-transfected control cell and panel B shows data from a GFP+ cell transfected with the HCN2-AYA construct. The scale bars in B apply to panels A and B. (C & D) Representative currents recorded in response to families of voltage steps that ranged between −124 mV and −64 mV in 10 mV increments. Capacitive transients and leak currents were subtracted for clarity. The scale bars in panel D apply to both panels C and D. Panel C shows I_h recorded from a non-transfected utricle type II hair cell from the same epithelium as that shown in panel A. Panel D shows a family of currents recorded from the same GFP+ shown in panel B. A family of voltage steps was used that was identical to those used to evoke the data shown in panel C. In this case, expression of the HCN2-AYA construct inhibited I_h. (E) A fluorescence image that revealed GFP expression was superimposed on a DIC image of the same field of cells. The recording pipette is visible to the right of the cell. Scale bar equals 5 µm.