P2X₂ Receptor Mediates Stimulation of Parasensory Cation Absorption by Cochlear Outer Sulcus Cells and Vestibular Transitional Cells

Abstract

Cochlear outer sulcus cells (OSC) and vestibular transitional cells (VTC) are part of the parasensory epithelium in the inner ear and are located in homologous positions between the sensory hair cells and the cation secretory epithelial cells in the cochlea and the vestibular labyrinth. OSC are known to sustain a reabsorptive transepithelial current and to contain an immunoreactivity for P2X₂purinergic receptors. This study addresses whether OSC and VTC share functional similarities and extends this hypothesis to the question of whether both cell types contain functional P2X₂ receptors. The current density (I_sc) was recorded with the vibrating probe technique and was found to be similar in VTC and OSC. Both gadolinium and flufenamic acid reducedI_sc in VTC, as reported previously for OSC.I_sc was stimulated by extracellular ATP but not by selective agonists of P2Y receptors. Purinergic receptor agonists increased I_sc with a potency order of ATP > 2′- and 3′-O-(4-benzoyl-benzoyl)adenosine 5′-triphosphate ≫ α,β-methyleneadenosine 5′-triphosphate in both OSC and VTC. In the presence of suramin (100 μm) or gadolinium (100 μm), the responses of ATP were inhibited significantly in both OSC and VTC. This pharmacological profile is consistent with that of the P2X₂receptor. These results demonstrate that VTC participate in vestibular parasensory cation absorption and that both OSC and VTC regulate their parasensory cation flux via P2X₂ receptors, which would regulate the endolymphatic concentration of the current-carrying ion species in auditory and vestibular transduction.

Auditory and vestibular transduction depend on the balance of secretion and absorption of cations by the epithelial cells bounding the endolymphatic spaces (Marcus, 2001). Potassium is secreted by strial marginal cells in the cochlea and by dark cells in the vestibular labyrinth. There is a quiescent and a stimulus-induced efflux of K⁺ from the endolymphatic space through cochlear and vestibular hair cells. Variations in the intensity and duration of acoustic and vestibular stimuli would cause fluctuations in endolymph cation composition if there were no regulation of the rates of secretion and/or absorption. Secretion is known to be under the control of several extracellular hormones and factors, including purinergic agonists (Marcus et al., 1997; Marcus and Scofield, 2001). It has been shown recently that the outer sulcus epithelial cells (OSC) in the cochlea provide a parasensory pathway in the cochlea that sustains an apical-to-basal transepithelial cation current (Marcus and Chiba, 1999; Chiba and Marcus, 2000, 2001). The OSC would therefore contribute to the ionic homeostasis of endolymph if they possess signaling pathways that regulate this cation current.

Vestibular transitional cells (VTC) occupy a position in the vestibular labyrinth analogous to that of OSC in the cochlea, lying between the K⁺ secretory cells and the sensory hair cells. More importantly, functional similarities have been reported at the cellular level. The basolateral membrane of both cell types is dominated by a large K⁺ conductance that has a similar and unusual pharmacologic profile (Wangemann and Marcus, 1989; Chiba and Marcus, 2001). These similarities suggest that VTC may provide a parasensory pathway in the vestibular labyrinth for cation absorption from endolymph.

P2X₂ purinergic receptors are ligand-gated nonselective cation channels that were found by immunohistochemistry to be expressed in OSC (Jarlebark et al., 2000). The expression of these and other purinergic receptors in several cells bordering the endolymphatic space in conjunction with a putative source of agonist and with ectoenzymes for agonist degradation have led to the proposition that the cochlea and vestibular labyrinth use paracrine and/or autocrine purinergic systems to maintain the homeostasis of endolymph (Housley et al., 1999; Marcus and Scofield, 2001).

The present study used the vibrating probe to determine whether VTC are homologous to OSC (i.e., sustain a transepithelial current directed from the apical to the basolateral side) and whether VTC and OSC regulate this current via P2X₂ purinergic ligand-gated ion channels. Our results demonstrate that VTC participate in vestibular parasensory cation absorption and that both OSC and VTC regulate their parasensory cation flux via P2X₂receptors. This flux would regulate the endolymphatic concentration of the current-carrying ion species in auditory and vestibular transduction.

MATERIALS AND METHODS

Tissue preparation. Gerbils (4–5 weeks of age) were anesthetized with sodium pentobarbital (50–100 mg/kg, i.p.) and killed under a protocol approved by the Institutional Animal Care and Use Committee of Kansas State University to remove the temporal bones. The methods for dissecting OSC and VTC epithelia have been described previously (Wangemann and Marcus, 1989; Chiba and Marcus, 2000). Briefly, the lateral wall from the upper cochlear turn was isolated and the stria vascularis was removed from spiral ligament to exclude any contribution of the marginal cells to the current density (I_sc). The lateral wall was folded with OSC facing outward (Fig.1A,B). Ampullas of the semicircular canals were isolated and a cut was made along the border between VTC and the vestibular hair cells. The tissue was folded with VTC facing outward (Fig. 1C,D). A potent inhibitor of dark cell I_sc, bumetanide (10 μm), was added to all bath solutions used for VTC experiments to exclude contaminating contributions from vestibular dark cells (VDC) (Marcus and Shipley, 1994). We confirmed this by observing that I_sc reversibly changed its sign from positive to negative when bumetanide was perfused (see Fig. 4D), consistent with its known inhibitory action on the basolateral Na⁺–2Cl⁻–K⁺cotransporter of the vestibular dark cells. Each tissue was mounted in a perfusion chamber on the stage of an inverted microscope (TE-300; Nikon, Tokyo, Japan) and continuously perfused at 37°C at an exchange rate of 1.1 times/sec.

Fig. 1.

Tissue preparation for OSC and VTC.A, C, Schematic illustration showing the location of OSC and VTC in the cochlea and semicircular canal ampulla, respectively.B, D, Prepared tissue for the measurement ofI_sc in OSC and VTC, respectively.HC, Vestibular hair cell (damaged); SL, spiral ligament; SMC, strial marginal cell;SV, stria vascularis; VP, vibrating probe. Scale bar, 50 μm. A andC adapted from graphics by P. Wangemann (Wangemann, 1997).

Fig. 4.

Dose–response relationships of ATP analogs in OSC and VTC. A, B, C, OSC; D, E, F, VTC. There is a discontinuity in B that indicates recordings from two different tissues. Numbersin C and F indicate numbers of observations at each concentration. In the dose–response curves (C, F), the data were normalized based on the response to 100 μm ATP. In the experiments on VTC, bumetanide (10 μm) was added to the bath solution to exclude influence from vestibular dark cells. αβ, αβmeATP; Bz, BzATP.

Voltage-sensitive vibrating probe. The vibrating-probe technique was chosen to measure transepithelial currents under short-circuit conditions because of the small extent of the epithelial domains of the OSC and VTC. In the upper turn of the cochlea, the apical membranes of OSC are exposed to endolymph, the luminal fluid, in a band that is two to four cells wide (Spicer and Schulte, 1996); VTC are similarly located in a narrow band between vestibular hair cells and dark cells in the ampulla (Oudar et al., 1988) (Fig. 1). The diameter of the vibrating-probe tip is ∼20 μm and allows detection of voltages in the low nanovolt range; vibration between two positions within the line of current flow yields voltages that correspond to current flow through the resistive physiological saline (Marcus, 1996).

The vibrating probe technique was identical to that described previously (Marcus and Shipley, 1994; Marcus, 1996). Briefly,I_sc was monitored by vibrating a platinum–iridium wire microelectrode that was insulated with parlene-C (Micro Electrodes, Gaithersburg, MD) and coated with Pt black on the exposed tip. The vibration was ∼20 μm along both a horizontal (x) and vertical (z) axis. Thex-axis was perpendicular to the face of the epithelium. The probe was positioned 20–30 μm from the apical surface of the epithelium with computer-controlled, stepper-motor manipulators (Applicable Electronics, Forestdale, MA) and specialized probe software (Automated Scanning Electrode Technique version 1.05; Science Wares, East Falmouth, MA). The bath references were 26-gauge Pt-black electrodes. Calibration was performed in physiologic saline (see below) using a glass microelectrode (tip, <1 μm outer diameter) filled with 3 m KCl as a point source of current. The frequencies of vibration were in the range of 200–400 Hz and were well-separated for the two orthogonal directions. The signals from the oscillators driving the probe were also fed to a dual-channel phase-sensitive detector. The asymmetry of the probe design yielded different resonant frequencies for the two directions of vibration. The signals of the X and Z detectors were connected to a 16 bit analog-to-digital converter (CIO-DAS1602/16; ComputerBoards, Mansfield, MA) in a Pentium III, 700 MHz computer. The sampling interval was 0.5 sec, which is the minimum for this software. The electrode was positioned where I_sc showed a maximumx value and minimum z value; data are expressed as the vector length of current density and were plotted with Origin software, version 6.1 (OriginLab Software, Northampton, MA).

The output from the vibrating probe depends not only on the specific short-circuit current of the epithelium but also on the position of the probe from the surface of the tissue and the exact geometry of each tissue sample. The current density reported here refers to the flux at the position of the probe and represents only a fraction of the current crossing the epithelium. No changes in the relative position of the probe attributable to swelling or shrinking of the tissue during experimental treatments were observed.

Solutions and chemicals. In all experiments, both sides of the epithelium were perfused with a perilymph-like physiologic saline containing (in mm): 150 NaCl, 3.6 KCl, 1 MgCl₂, 0.7 CaCl₂, 5 glucose, and 10 HEPES, pH 7.4. ATP (A-9187; Sigma, St. Louis, MO), UTP (U-4630; Sigma), 2′- and 3′-O-(4-benzoyl-benzoyl)adenosine 5′-triphosphate (BzATP) (B-6396; Sigma), α,β-methyleneadenosine 5′-triphosphate (αβmeATP) (M-6517; Sigma), suramin (S-2671; Sigma), and gadolinium chloride (G-7532; Sigma) were directly dissolved in physiologic saline just before use. UDP (U-4125; Sigma) and ADP (A-2754; Sigma) were preincubated for 1.5 hr at room temperature with hexokinase (1 U/ml; H-4502; Sigma) and glucose (5 mm) because the commercial preparations of UDP and ADP may be supplied with a minor component of UTP and ATP (Nicholas et al., 1996). Bumetanide (B-3023; Sigma) and flufenamic acid (F-9005; Sigma) were dissolved in DMSO and then diluted to 0.1% DMSO in the control solution before application. DMSO at this concentration had no effect on the short-circuit current. All purines and pyrimidines used here were applied to the bath only briefly (∼20 sec) to avoid desensitization of the receptors.

Data presentation and statistics. As an internal control, the response to 100 μm ATP was included in each experiment to compare the magnitude of effects among all of the purines and pyrimidines tested in this study. For the analysis, the peakI_sc was chosen, but when a peak was not unambiguously defined we used the averaged data for the 5 sec afterI_sc reached steady-state and compared these data with the averaged data for the 5 sec before the solution change. Data were expressed as the mean ± SEM (n= number of tissues) of the I_sc. Increases or decreases in I_sc were considered significant at a level of p < 0.05. A paired t test was used.

RESULTS

Cation absorption by VTC

The I_sc from VTC in physiologic saline was −4.2 ± 0.4 μA/cm²(n = 38). Perfusion of Gd³⁺ (1 mm) or flufenamic acid (100 μm) for 2 min each significantly decreased the I_sc by 54 ± 17% (from −3.1 ± 0.6 to −1.1 ± 0.4 μA/cm², n = 6) and by 52 ± 11% (from −2.8 ± 0.6 to −1.3 ± 0.3 μA/cm², n = 9), respectively (Fig. 2).

Fig. 2.

Effects of Gd³⁺ and flufenamic acid on I_sc in VTC. A, Gd³⁺ (1 mm, n = 6).B, Flufenamic acid (FFA) (0.1 mm, n = 9). Bumetanide (10 μm) was added to the bath solution to exclude influence from vestibular dark cells. The SEM bars are plotted only at intervals for clarity.

Modulation of absorptive cation flux by purinergic agonists: OSC and VTC

Perfusion of ATP (100 μm) increased theI_sc from −9.8 ± 1.0 to −26.4 ± 2.0 μA/cm² in OSC (n = 28) and from −5.1 ± 0.6 to −16.9 ± 1.2 μA/cm² in VTC (n = 23) (Figs. 3,4, 5). We first tested for mediation of this response by the P2Y family of purinergic receptors by perfusion of agonists for rodent P2Y₁ (ADP), P2Y₂ (UTP), P2Y₄ (UTP), and P2Y₆ (UDP). None of these agonists at a concentration of 100 μm changed theI_sc of either OSC (Fig. 3A,n = 5 each) or VTC (Fig. 3B,n = 5 each).

Fig. 3.

Comparison of effects of ATP, UTP, UDP, and ADP in OSC and VTC. A, OSC; B, VTC. Each drug was perfused at 100 μm. For the experiments in VTC, bumetanide (10 μm) was added to the bath solution to exclude influence from vestibular dark cells. UDP and ADP were preincubated with hexokinase (1 U/ml) in the presence of glucose for at least 1.5 hr (see Materials and Methods).

Fig. 5.

Effect of ATP in the presence of suramin or gadolinium on OSC and VTC. A, B, C, OSC; D, E, F, VTC. In the experiments on VTC, bumetanide (10 μm) was added to the bath solution to exclude influence from vestibular dark cells. Significance was tested based on the 100 μm ATP response (C, F).SUR, Suramin; Gd, gadolinium. *p < 0.05.

The subtypes of the P2X family of purinergic receptors are best identified by a comparison of the agonist potency of αβmeATP and BzATP with that of ATP (North and Surprenant, 2000). The results showed an increase in I_sc by agonists with a potency order (EC₅₀) of ATP > BzATP ≫ αβmeATP in both OSC and VTC (Fig. 4; Table1). The EC₅₀ values from the dose–response relationship for ATP, BzATP, and αβmeATP were 209, 511, and 7951 μm in OSC, and 180, 897, and 16,434 μm in VTC, respectively.

Table 1.

Effects of ATP, BzATP, and αβmeATP onI_sc (μA/cm²)

	ATP (μm)					BzATP (μm)			αβmeATP (μm)
	10	30	300	1000	3000	100	300	1000	100	1000
OSC	5.4  ± 1.3	10.4  ± 2.6	30.1  ± 6.7	37.2  ± 4.2	44.7  ± 6.8	9.5  ± 2.0	11.5  ± 1.5	24.5  ± 1.7	2.7  ± 0.5	5.8  ± 1.5
	(21.3  ± 4.5)	(21.3  ± 4.5)	(21.3  ± 4.5)	(17.0  ± 3.4)	(17.8  ± 3.7)	(19.3  ± 3.4)	(13.8  ± 2.7)	(13.8  ± 2.7)	(18.5  ± 4.0)	(9.8  ± 1.3)
	n = 5	n = 5	n = 5	n = 8	n = 7	n = 9	n = 4	n = 4	n = 8	n = 4
VTC		3.8  ± 0.8	14.2  ± 2.5	18.9  ± 2.4	23.6  ± 3.5	6.0  ± 1.0	9.7  ± 1.8	16.7  ± 3.6	2.1  ± 0.7	6.1  ± 1.2
		(7.9  ± 1.2)	(7.9  ± 1.2)	(7.9  ± 1.2)	(8.4  ± 1.3)	(10.9  ± 1.5)	(10.9  ± 1.5)	(10.9  ± 1.5)	(10.9  ± 1.5)	(10.9  ± 1.5)
		n = 6	n = 6	n = 6	n = 5	n = 5	n = 5	n = 5	n = 5	n = 5

Values were expressed as changes in I_sc by application of drugs. Values in parentheses indicate changes inI_sc by 100 μm ATP in the same tissues. These data were normalized and are plotted in Figure4C,F.

Suramin, a P2 receptor antagonist, was used to further assist in the identification of the P2X receptor mediating the response to ATP. Application of 100 μm suramin for 1 min resulted in no significant change in I_sc for either OSC or VTC. The I_sc before and after suramin was −10.6 ± 1.6 and −11.3 ± 1.7 μA/cm² (n = 5) in OSC and −6.0 ± 1.6 and −6.4 ± 1.8 μA/cm² (n = 6) in VTC (Fig. 5A,D). In the presence of 100 μm suramin, the stimulation of theI_sc by 100 μmATP was inhibited by 71 ± 5% (n = 5) in OSC and 81 ± 8% (n = 6) in VTC compared with the stimulation by ATP in the absence of suramin (Fig. 5A,C,D,F; Table 2).

Table 2.

Effects of ATP in the presence of SUR or Gd onI_sc (μA/cm²)

	ATP + SUR	ATP after wash	ATP + Gd	ATP after wash
OSC	6.2  ± 0.8*	21.9  ± 5.1 (NS)	1.4  ± 0.6*	9.1  ± 0.5 (NS)
	(23.7  ± 4.5, n = 5)	(23.7  ± 4.5, n = 5)	(11.9  ± 1.7)	(11.9  ± 1.7)
VTC	1.8  ± 0.6*	9.7  ± 1.0 (NS)	1.5  ± 0.5*	8.4  ± 1.1 (NS)
	(10.5  ± 1.1, n = 6)	(10.5  ± 1.1, n = 6)	(12.8  ± 1.5, n = 5)	(12.8  ± 1.5, n = 5)

Values were expressed as changes in I_sc by application of drugs. Values in parentheses indicate changes inI_sc by the first application of 100 μm ATP in the same tissues. SUR, Suramin; Gd, gadolinium.

^*p<0.05. Significance was tested based on the changes resulting from the first application of 100 μmATP. NS, Not significant.

Gd³⁺ is known to inhibit nonselective cation channels of several types, including P2X ligand-gated channels and the nonselective cation channels in the apical membrane of OSC (Marcus and Chiba, 1999; Chiba and Marcus, 2000). Application of Gd³⁺ (100 μm) for 3–4 min decreased the I_sc significantly, by 55 ± 9% (from −7.8 ± 1.7 to −3.2 ± 0.6 μA/cm², n = 5) in OSC and by 34 ± 3% (from −8.2 ± 0.9 to −5.4 ± 0.7 μA/cm², n = 5) in VTC (Fig. 5B,E). In the presence of 100 μm Gd³⁺, the stimulation of the I_sc by 100 μm ATP was inhibited by 89 ± 3% (n = 5) in OSC and 85 ± 5% (n = 5) in VTC compared with the stimulation by ATP in the absence of Gd³⁺ (Fig. 5B,C,E,F; Table2).

DISCUSSION

VTC are homologous to OSC

Previous examinations of a transitional cell epithelium posited a simple barrier function in maintaining ion gradients between endolymph and perilymph (Oudar et al., 1988). In contrast, our finding of a constitutive transepithelial current demonstrates an active role in endolymph homeostasis. The direction ofI_sc measured in the control solution was negative in both OSC and VTC (−9.8 and −5.1 μA/cm², respectively). The former value is similar to previous observations in OSC (Marcus and Chiba, 1999) and is accounted for by the absorption of cations (primarily Na⁺ in this perilymph-like solution, but primarily K⁺ under in vivoconditions) through the nonselective cation channels in the apical membrane of OSC (Chiba and Marcus, 2000).

The similar effects of Gd³⁺ and flufenamic acid on I_sc in OSC (Marcus and Chiba, 1999; Chiba and Marcus, 2000) and VTC (Fig. 2) support the notion of a strong homology in function of the two cell types. A minor difference was that there was no transient overshoot ofI_sc in VTC from flufenamic acid, which was caused in OSC by the additional activation of BK (large conductance, calcium-dependent) K⁺channels in the apical membrane (Chiba and Marcus, 2000). This difference in response suggests a lower density of BK channels in VTC than in OSC. Additional evidence for homology between VTC and OSC is the stimulation of I_sc by purinergic agonists in both cell types. The fact that suramin did not affect the constitutive I_sc but Gd³⁺ inhibited the current implies that there are two populations of nonselective cation channels, one sensitive and the other insensitive to extracellular ATP in both OSC and VTC.

Functional expression of P2X₂ receptor in OSC and VTC

Receptors for purines and pyrimidines have been investigated extensively in various systems and have been well summarized in recent reviews (Ralevic and Burnstock, 1998; North and Surprenant, 2000; Khakh et al., 2001). To date it is evident that there is no single agonist or antagonist that discriminates adequately between families of P2X and P2Y receptors. However, we used a series of strong agonists that together determined the absence of involvement of the P2Y receptor family in the stimulation of the I_scof OSC and VTC.

The P2Y family of purinergic receptors signals cellular events via G-protein-coupled signal pathways. In this study, the possibility of the coupling of P2Y receptors to stimulation ofI_sc could be excluded in both OSC and VTC because of the absence of response of I_sc to UTP, UDP, and ADP and because of the weaker response to BzATP than to ATP. Five mammalian P2Y receptors, P2Y₁, P2Y₂, P2Y₄, P2Y₆, and P2Y₁₁, have been cloned and are known to be valid members of the P2Y receptor family (Ralevic and Burnstock, 1998). ADP is well known as a potent agonist to P2Y₁, UTP is well known as a potent agonist to rodent P2Y₂ and P2Y₄, UDP is well known as a potent agonist to P2Y₆, and BzATP (BzATP > ATP) is well known as a potent agonist to P2Y₁₁ (Ralevic and Burnstock, 1998; Communi et al., 1999). One criterion that is sometimes used to distinguish the involvement of P2X from P2Y receptors is the faster onset of response (within 10 msec) of P2X, which is in contrast to an onset of ∼100 msec in P2Y receptors (Ralevic and Burnstock, 1998). It was not possible in our experiments to perfuse the tissue at rates that were sufficiently high enough to use this criterion.

Our results clearly showed that P2X receptors are functionally expressed in both OSC and VTC and that their activation stimulatesI_sc. By exclusion of the P2Y receptors and by previous immunolocalization of P2X₂receptors on OSC (Jarlebark et al., 2000), we predicted that the stimulation of I_sc by purinergic agonists occurred via a P2X₂ receptor. Sensitivity to αβmeATP and inhibition by suramin have been used as important tools to discriminate among the seven recombinant homomeric P2X receptors (Khakh et al., 2001). P2X₁ and P2X₃ receptors are not likely expressed in OSC and VTC because of the insensitivity ofI_sc to αβmeATP shown here, and P2X₄ and P2X₇ are not likely expressed in OSC and VTC because of the response ofI_sc to suramin shown here.

The remaining possibilities are P2X₂, P2X₅, and P2X₆ receptors. The agonist and antagonist profiles for P2X₂ are not known to be distinguished clearly from ones for P2X₅ (North and Surprenant, 2000). However, the characteristics of the P2X receptor in OSC and VTC are more consistent with those of the P2X₂ receptor subtype. First, the reported EC₅₀ of BzATP in the P2X₅ receptor was at least 50 times higher than that of ATP, which is at least a decade greater than found here (2.5 times in OSC and 5 times in VTC). In contrast, the EC₅₀ of BzATP in the heterologously expressed P2X₂ receptor was three times higher than that of ATP, in accordance with our results. Second, mRNA transcripts for P2X₁, P2X₅, and P2X₆ were not detected in the inner ear, including the cochlea and vestibular end organ, whereas transcripts for P2X₂, P2X₃, P2X₄, and P2X₇ were detected by reverse transcriptase-PCR (Brandle et al., 1999). Although our data are most consistent with functional P2X₂ receptors, we cannot completely exclude the possibility of the presence of heteromeric P2X receptors in OSC and VTC. P2X₂ subunits are known to coassemble with other subunits such as P2X₃ (Radford et al., 1997).

Interestingly, there was at least a one decade difference in the EC₅₀ values for ATP between those reported here in gerbil native tissues and those reported elsewhere in the rat recombinant homomeric P2X₂ receptor. The EC₅₀ for stimulation ofI_sc by ATP was near 200 μm in OSC and VTC, but it was reported to be in the range 1–30 μm in the recombinant homomeric P2X₂ receptors of rats (Khakh et al., 2001). Therefore, the EC₅₀ appears to be higher in native tissues than in expression systems, but the reason for this difference is not clear. Possible factors could be (1) the contribution of ectonucleotidase activity in the native tissue (Dunwiddie et al., 1997), (2) possible differences among species or among the cell types, or (3) differences in glycosylation states (Torres et al., 1998).

Ectonucleotidase activity has been found in the cochlea (Vlajkovic et al., 1998a,b) and, if present in our in vitro preparations, might be expected a priori to degrade the agonist to lower concentrations than originally supplied. However, this consideration does not likely apply in this study because our results showed much higher sensitivities to ATP than to the poorly metabolized αβmeATP. Any ectonucleotidase activity present was minimized by the relatively high exchange rate of the perfusion system; problems with enzymatic activity have primarily been noted in static chambers (Khakh et al., 2001).

Our findings of stimulation ofI_sc by purinergic agonists are most consistent with an apical membrane location of P2X₂. The ATP-insensitive nonselective cation channels found in OSC were located in the apical membrane, as shown by excised patch-clamp recordings, and those channels provided the primary pathway for the transepithelial current from the apical to the basolateral side that resulted in a negativeI_sc. P2X receptors are ligand-gated ion channels that are nonselective for cations (North and Surprenant, 2000). Because activation of these channels leads to an increase in the magnitude of the negative I_sc, it is highly likely that these receptor channels are also located in the apical membrane.

Physiologic significance

There is accumulating evidence that purinergic agonists such as ATP are used by the cochlea and vestibular labyrinth to regulate transduction processes. Elements of a complete signaling system have been identified; sources of agonists, receptors, and terminating enzymes have all been demonstrated in the cochlea, and functional receptors have been demonstrated in the vestibular labyrinth. A constitutive level of ATP in the perilymph and endolymph of the guinea pig cochlea was reported (Munoz et al., 1995) that increased significantly in the endolymph during noise exposure (Munoz et al., 2001). An increase in agonist during an increase in acoustic stimulation would lead to an increased parasensory flux. This signaling cascade would then serve as a protective mechanism to reduce the flux through the sensory pathway during intense stimulation. Our findings of purinergic stimulation of I_sc from VTC suggest that a similar regulatory system is operant in the vestibular labyrinth.

In conclusion, (1) VTC actively absorb cations by cellular mechanisms homologous to OSC rather than merely providing a simple barrier to sustain the high concentration differences of K⁺ and Na⁺between endolymph and perilymph; (2) both OSC and VTC serve as parasensory pathways to regulate K⁺ efflux through sensory hair cells during changes in the level of acoustic and vestibular stimulation; and (3) among the possible roles of extracellular nucleotides, one mechanism of regulation involves purinergic signaling via P2X₂ receptors in OSC and VTC, most likely in their apical membranes.

Note added in proof. P2X₂receptors have been localized recently in vestibular transitional cells by immunostaining (S. N. Syeda and A. Lysakowski, personal communication).