Developmental alterations in the biophysical properties of Ca_v1.3 Ca²⁺ channels in mouse inner hair cells

Abstract

Prior to hearing onset, spontaneous action potentials activate voltage-gated Ca_v1.3 Ca²⁺ channels in mouse inner hair cells (IHCs), which triggers exocytosis of glutamate and excitation of afferent neurons. In mature IHCs, Ca_v1.3 channels open in response to evoked receptor potentials, causing graded changes in exocytosis required for accurate sound transmission. Developmental alterations in Ca_v1.3 properties may support distinct roles of Ca_v1.3 in IHCs in immature and mature IHCs, and have been reported in various species. It is not known whether such changes in Ca_v1.3 properties occur in mouse IHCs, but this knowledge is necessary for understanding the roles of Ca_v1.3 in developing and mature IHCs. Here, we describe age-dependent differences in the biophysical properties of Ca_v1.3 channels in mouse IHCs. In mature IHCs, Ca_v1.3 channels activate more rapidly and exhibit greater Ca²⁺-dependent inactivation (CDI) than in immature IHCs. Consistent with the properties of Ca_v1.3 channels in heterologous expression systems, CDI in mature IHCs is not affected by increasing intracellular Ca²⁺ buffering strength. However, CDI in immature IHCs is significantly reduced by strong intracellular Ca²⁺ buffering, which both slows the onset of, and accelerates recovery from, inactivation. These results signify a developmental decline in the sensitivity of CDI to global elevations in Ca²⁺, which restricts negative feedback regulation of Ca_v1.3 channels to incoming Ca²⁺ ions in mature IHCs. Together with faster Ca_v1.3 activation kinetics, increased reliance of Ca_v1.3 CDI on local Ca²⁺ may sharpen presynaptic Ca²⁺ signals and improve temporal aspects of sound coding in mature IHCs.

Introduction

Voltage-gated Ca_v Ca²⁺ channels conduct inward Ca²⁺ currents that can depolarize the membrane potential and trigger Ca²⁺-dependent signal transduction in excitable cells. Multiple classes of Ca_v channels (Ca_v1.x-Ca_v3.x) have been characterized,¹ which play distinct roles often within the same cell. In most neurons, somatodendritic Ca_v1 channels mediate L-type Ca²⁺ currents that couple neuronal activity to changes in gene transcription,^2,3 while presynaptic Ca_v2 channels generating P/Q- and N-type Ca²⁺ currents regulate neurotransmitter release from nerve terminals.^4-6

In contrast to the diverse complement of Ca_v channnels in most neuronal cell-types, auditory hair cells express predominantly Ca_v1.3 channels, which mediate exocytosis of glutamate at ribbon synapses formed with primary afferent neurons.^7,8 Genetic inactivation of CACNA1D (Ca_v1.3 KO), which encodes the pore-forming α₁ subunit of Ca_v1.3, causes ~90% reduction in the whole-cell Ca²⁺ current in mouse inner hair cells (IHCs).^9-11 As a consequence, stimulus-secretion coupling is significantly impaired in Ca_v1.3 KO IHCs,¹¹ which contributes to deafness in these mice.^9,10 Spontaneous Ca²⁺-dependent action potentials, which support presensory afferent synaptic transmission,^12-14 are absent in Ca_v1.3 KO IHCs. The developmental upregulation of BK K⁺ channels and pruning of cholinergic efferents from the basal IHC membrane, which normally occurs around the onset of hearing,^13,15 are also absent in Ca_v1.3 KO IHCs.^11,16 Thus, Ca_v1.3 channels play distinct roles in regulating electrical activity and normal developmental programs in immature IHCs, and in transducing stimulus-evoked exocytosis in mature IHCs.

Comparisons of Ca_v1 currents in mature and immature auditory hair cells suggest differences in the properties of Ca_v1.3 channels that may tailor Ca²⁺ signals according to developmental stage. For example, in gerbil and rat IHCs, the extent to which Ca_v1 channels undergo Ca²⁺-dependent inactivation (CDI) declines after hearing onset.^17,18 CDI is a negative feedback regulation of Ca_v1 and Ca_v2 channels by incoming Ca²⁺ ions, which depends on calmodulin binding to the Ca_v α₁ subunit.^19,20 CDI is relatively weak for Ca_v1 channels in IHCs, which may be due to the antagonistic actions of calmodulin-like Ca²⁺ binding proteins expressed in IHCs.^17,21,22 The rates of Ca_v1 channel activation and inactivation can significantly influence IHC Ca²⁺ signals that regulate presensory action potentials or stimulus-secretion coupling. Whether Ca_v1.3 channels in mouse IHCs undergo developmental changes in their biophysical properties is incompletely characterized. Given the widespread use of mice as experimental models for IHC transmission, filling this void is essential for understanding Ca²⁺-dependent signaling in IHCs in this species.

In this study, we compared the properties of Ca²⁺ and Ba²⁺ currents (I_Ca and I_Ba, respectively) in IHCs from mice before and after hearing onset (~P12). We discovered that Ca_v1.3 channels exhibit significant differences in immature (P6–8) and mature (P16–18) IHCs in terms of activation and inactivation. In particular, CDI differs in magnitude and sensitivity to local rather than global Ca²⁺ elevations in Ca²⁺. Our results highlight age-dependent distinctions in the properties of Ca_v1.3 channels that may fine-tune Ca²⁺ signals required for the development and mature function of IHCs.

Results

Ca_v1.3 currents activate with faster kinetics in mature IHCs than in immature IHCs

Because a goal of this study was to compare CDI before and after hearing onset in mouse IHCs, we first analyzed activation of both I_Ca and I_Ba in P6–8 and P16–18 mouse IHCs. Whole-cell patch clamp recordings were performed at room temperature in solutions designed to isolate Ca_v1.3 currents from other ionic currents, particularly voltage-gated K⁺ currents (see Materials and Methods). The intracellular solution contained 2 mM EGTA, which is permissive for measuring CDI in IHCs.^17,18 Relatively high concentrations of extracellular Ca²⁺ or Ba²⁺ were used (5 mM) to amplify current amplitudes, thus increasing resolution of activation and inactivation kinetics. Consistent with previous studies of Ca_v1 channels in other cell-types,²³ I_Ca underwent a progressive increase in amplitude (run-up) upon attaining whole-cell configuration in immature IHCs (Fig. 1). For this reason, we first monitored the amplitude of I_Ca evoked by steps to +10 mV every 20 sec until there was no further run-up. Interestingly, run-up was not observed in mature IHCs, which was confirmed by the same protocol. Thus, data were collected ~5 min after whole cell patch rupture in immature IHCs, when steady-state I_Ca amplitudes were achieved. In mature IHCs, data were collected ~1 min after patch rupture to ensure equilibration with intracellular solution components.

Figure 1. Run-up of I_Ca in immature mouse IHCs. (A) Representative current traces for I_Ca evoked by 20-ms test pulses to +10 mV, 10 sec and ~4 min after patch rupture. (B) I_Ca was normalized to that at 4 min after patch rupture and plotted against time.

As noted previously,^11,14 the most dramatic difference in Ca_v1.3 currents was the smaller amplitude of I_Ca and I_Ba in mature IHCs compared with immature IHCs. In current-voltage (I-V) relationships, the maximum current amplitude was obtained with a −10-mV pulse for I_Ca and was significantly smaller in mature IHCs than in immature IHCs (19.4 ± 15.4%; Figure 2A–D). A similar difference was observed for I_Ba (32.9 ± 10.2%) except that the peak I_Ba was obtained between −10 and −0 mV, due to changes in surface charge screening when Ba²⁺ rather than Ca²⁺ is used as the permeant ion.²⁴ As has been reported for Ca_v1.3 in heterologous expression systems,²⁵ there was some rectification in the I-V relationship at positive voltages, probably due to limited permeability of Ca_v1.3 channels to Cs⁺ ions. Thus for greater accuracy, we measured parameters for voltage-dependent activation in normalized tail current-voltage relationships fit with the Boltzmann equation. By this analysis, the voltage for half-maximal activation (V_h) was more positive and there was a significant increase in the slope factor (k) with age for I_Ca and I_Ba (Table1, Fig. 2E and F). In addition, the activation kinetics of I_Ca were significantly faster in mature IHCs than in immature IHCs (~5–29% from −40 mV to +40 mV, p < 0.001 by two-way ANOVA; Fig. 3A). The acceleration in Ca_v1.3 activation depended on Ca²⁺ as the charge carrier since it was not observed for I_Ba (p = 0.34 by two-way ANOVA; Fig. 3B). These results indicate that, in addition to a reduction in Ca_v1.3 current density, the intrinsic biophysical properties of Ca_v1.3 channels also change with development.

Figure 2. Voltage-dependent properties of I_Ca and I_Ba in mature IHCs and in immature IHCs. (A and B) Representative I_Ca (upper panels) and I_Ba (lower panels). Currents were evoked by 15– ms depolarizing pulses from a holding voltage of −84 mV. (C and D) I-V relationships for currents in (A). (E and F) G-V relationships for currents in (A and B). Tail current amplitudes were normalized to that obtained with a pulse to +50 mV (Norm. tail) and plotted against test voltage. Smooth lines in (C–F) represent curve fits from the Boltzmann equation. The parameters for the fits shown were: in (C), V_h = −26.6, k = −5.9, V_rev = 56.4 mV, G_max = 4.3 nS for P6–8 (n = 22), V_h = −27.7, k = −5.4, V_rev = 56.1 mV, G_max = 3.2 nS for P16–18 (n = 15); in (D), V_h = -32.8, k = −4.9, V_rev = 55.5 mV, G_max = 4.3 for P6–8 (n = 22), V_h = −30.1 mV, k = −4.7, V_rev = 48.0 mV, G_max = 3.3 for P16–18 (n = 15); in (E), V_h = −24.1 mV, k = 9.5 for P6–8 (n = 22), V_h = −19.0 mV, k = 13.1 for P16–18 (n = 9); in (F), V_h = −29.2 mV, k = 8.4 for P6–8 (n = 11), V_h = −23.2 mV, k = 9.6 for P16–18 (n = 15).

Table 1. Parameters for voltage-dependent activation of Ca_v1.3 currents in mouse IHCs.

		P6–8	n	P16–18	n	p-value1
ICa (2 mM EGTA)	Peak current density2 (pA/pF)	−35.6 ±1.8 (at −10 mV)	22	−23.0 ±1.2 (at −10 mV)	15	< 0.001
	V1/2 (mV)3	−24.1 ± 0.9		−19.0 ± 1.8	9	0.009
	k3	9.5 ± 0.3		13.1 ± 0.7		< 0.001
IBa (2 mM EGTA)	Peak current density2 (pA/pF)	−36.2 ± 1.8 (at −10mV)	11	−21.7 ± 1.2 (at −10mV)	15	< 0.001
	V1/2 (mV)3	−29.2 ± 1.1		−23.2 ± 0.9		< 0.001
	k3	8.4 ± 0.3		9. ± 0.3		0.011
ICa (10 mM BAPTA)	Peak current density2 (pA/pF)	−34.5 ± 1.3 (at −10 mV)	25	−20.0 ± 0.8 (at −10 mV)	8	< 0.001
	V1/2 (mV)3	−24.4 ± 0.6		−18.6 ± 0.9		< 0.001
	k3	8.3 ± 0.3		12.7 ± 0.8		< 0.001

¹ Comparison between P6–8 and P16–18 by t-test;²obtained from I-V relationships;³obtained from G-V relationships.

Figure 3. I_Ca activation kinetics are faster in mature IHCs than immature IHCs. (A and B) Upper panels, representative traces for I_Ca and I_Ba for P6–8 (gray) or P16–18 (black) IHCs. Currents were evoked by a 15-ms step to -10 mV from a holding potential of −84 mV. Dashed lines represent exponential fits of the current trace. The time constants (τ, ms) for the displayed current traces were: for I_Ca (A) 0.62 (P6–8) and 0.28 (P16–18); for I_Ba (B) 0.50 (P6–8) and 0.38 (P16–18). Lower panels, time constants for I_Ca (A) and I_Ba (B) were plotted against test voltage. For P6–8, n = 22 for I_Ca, 15 for I_Ba. For P16–18, n = 11 for I_Ca, 15 for I_Ba. *p < 0.001 by two-way ANOVA.

Ca²⁺-dependent inactivation is greater in mature IHCs than immature IHCs

To determine if Ca_v1.3 inactivation was also subject to developmental regulation in mouse IHCs, we compared the extent to which I_Ca and I_Ba inactivated in response to sustained depolarizations. Like other Ca_v1 channels, Ca_v1.3 undergoes inactivation due to Ca²⁺- or voltage-dependent mechanisms (CDI and VDI, respectively). CDI manifests as stronger inactivation of I_Ca compared with I_Ba, due to Ca²⁺ binding to the calmodulin that is associated with the channel protein.²⁰ Unlike I_Ca, which undergoes both CDI and VDI, I_Ba exhibits purely VDI since Ba²⁺ substitutes poorly for Ca²⁺ in binding to calmodulin.²⁶ We measured CDI and VDI with a triple-pulse voltage-protocol in which I_Ca or I_Ba was evoked by test pulses before (p1) and after (p2) a 200-ms prepulse to varying voltages. Inactivation was measured as the ratio of the p2/p1 current amplitudes; this ratio was less than 1 for prepulses inducing inactivation (Fig. 4A and B). As expected for CDI, the prepulse-voltage dependence of I_Ca inactivation was U-shaped and was greatest for prepulses evoking the maximal inward I_Ca. In contrast, inactivation of I_Ba with this protocol was not evident across the full prepulse voltage range. In mature IHCs, I_Ba underwent facilitation (p2/p1 > 1) at positive prepulse voltages, likely due to voltage-dependent facilitation (VDF).²⁷ Since I_Ca is thought to undergo VDI as well as VDF, we isolated the effects of CDI on I_Ca as the difference in p2/p1 for I_Ca and I_Ba (F_CDI, Fig. 4C). By this metric, CDI was significantly greater in mature IHCs than in immature IHCs (p < 0.001 by two-way ANOVA).

Figure 4. CDI is greater in mature IHCs than immature IHCs. (A) Top panel, Voltage protocol in which test currents were evoked by pulses (p1, p2) from −84 mV to −20 mV (for I_Ca) or −30 mV (for I_Ba) before and after a 200-ms prepulse to various voltages. Bottom panels, representative traces for I_Ca or I_Ba recorded in P6–8 or P16–18 mouse IHCs. Currents were evoked by p1 (black) and p2 (red) overlaid for comparison. Scale bars, 0.2 nA (vertical), 2.5 ms (horizontal). (B) Inactivation was measured as the ratio of P2/P1 current amplitude and plotted against prepulse voltage. Results are from mouse IHCs from P6–8 (n = 20 for I_Ca, n = 6 for I_Ba) or P16–18 (n = 18 for I_Ca, n = 8 for I_Ba). (C) CDI was measured as the difference in p2/p1 for I_Ca and I_Ba (F_CDI) and is plotted against prepulse voltage for P6–8 and P16–18 IHCs. *p < 0.05 by two-way ANOVA and post-hoc Bonferroni t-test.

To address the underlying mechanism(s) for the age-dependent increase in CDI, we analyzed the onset of I_Ca inactivation by obtaining p2/p1 ratios in voltage protocols with prepulses of varying duration (Fig. 5A). The prepulse was set to −20 mV for I_Ca and −30 mV for I_Ba, to partially compensate for the negative shift in activation voltages for I_Ba compared with I_Ca (Table 1). The duration of the p1 and p2 test pulse was increased to 5 ms in these experiments to ensure that test currents reached steady-state levels, as long prepulses tended to slow the activation kinetics of the p2 current. The onset of inactivation was obtained by plotting p2/p1 ratios for I_Ca against prepulse duration. With this protocol, I_Ca inactivated with a double exponential time course in both immature and mature IHCs. Both fast and slow inactivation were likely influenced by CDI: the fast component was not observed for I_Ba, and the slow component of I_Ca inactivation was significantly faster than that for I_Ba (τ_slow 2.0 ± 0.4 sec for I_Ca vs. 3.1 ± 0.6 sec for I_Ba, P6–8, p = 0.15; τ_slow = 0.8 ± 0.13 sec for I_Ca vs. 2.3 ± 0.4 sec for I_Ba, P16–18, p < 0.001; both by t-test). While there was no age-dependent difference in the time constants for I_Ba inactivation (p = 0.3), both the slow and fast time constants were significantly faster for I_Ca inactivation in mature IHCs than in immature IHCs (Fig. 5C, p < 0.05). However, the amplitudes of the slow and fast components of I_Ca inactivation did not vary with age (A_fast = −0.25 ± 0.01 for P6–8 vs. −0.23 ± 0.02 for P16–18, p = 0.20; A_slow = −0.40 ± 0.03 for P6–8 vs. −0.34 ± 0.01 for P16–18, p = 0.31; both by t-test). These results indicate that CDI is governed by fast and slow processes, both of which occur more rapidly in mature IHCs compared with in immature IHCs.

Figure 5. Onset of CDI is faster in mature IHCs than in immature IHCs. (A) Top panel, Voltage protocol in which test currents were evoked by pulses (p1, p2) from −84 mV to −20 mV (for I_Ca) or −30 mV (for I_Ba) before and after a prepulse to −10 mV of varying durations. Bottom panels, representative traces for I_Ca or I_Ba recorded in P6–8 or P16–18 mouse IHCs with protocols using the indicated prepulse durations. Currents evoked by p1 (black) and p2 (red) are overlaid for comparison. Scale bars, 0.8 nA (vertical), 2.5 ms (horizontal). (B) Onset of inactivation was measured by plotting p2/p1 values against prepulse duration. Smooth line represents fit with a double (I_Ca) or single (I_Ba) exponential function. For P6–8, curve fit parameters for I_Ca were y₀ = 0.34, A_fast = 0.25, τ_fast = 32.9 ms, A_slow = 0.40, τ_slow = 1990 ms (n = 22); for I_Ba, y₀ = 0.91, A = 56.5, τ = 3108 ms (n = 8). For P16–18, curve fit parameters for I_Ca were y₀ = 0.37, A_fast = 0.28, τ_fast = 20.4 ms, A_slow = 0.34, τ_slow = 795 ms (n = 12); for I_Ba, y₀ = 0.79, A = 30.1, τ = 2313 ms (n = 11). (C) Average values from exponential fits of I_Ca data in (B). *p < 0.02 by Mann–Whitney rank sum test.

To analyze recovery from inactivation, p2 test currents were measured at varying intervals after the inactivating prepulse and the ratio of p2/p1 plotted against the recovery interval (Fig. 6A). The duration of the prepulse was limited to 200 ms since longer prepulses induced VDI, particularly in mature IHCs (Fig. 5A); this would complicate analyses of recovery from CDI. In single exponential fits of these data, there was no difference in the amplitude (A = 0.32 ± 0.02 for P6–8 vs. 0.37 ± 0.03 for P16–18, p = 0.40, by t-test) or time constant (τ = 425.9 ± 28.3 ms for P6–8 vs. 486.4 ± 87.7 ms for P16–18, p = 0.08, by Mann-Whitney rank sum test; Fig. 6B). These results indicate that greater CDI in mature IHCs results primarily from faster onset, rather than slower recovery, of I_Ca from inactivation.

Figure 6. Recovery from CDI does not differ in mature and immature IHCs. (A) Top panel, Voltage protocol in which I_Ca was evoked by pulses (p1, p2) from −84 mV to −20 mV before and at variable recovery intervals after a 200-ms prepulse to −10 mV. Bottom panel, representative traces for I_Ca recorded in P6–8 or P16–18 mouse IHCs. Currents evoked by p1 (black) and p2 (red) at the indicated recovery intervals are overlaid for comparison. Scale bars, 0.8 nA (vertical), 2.5 ms (horizontal). (B) Recovery of I_Ca from inactivation was measured by plotting p2/p1 for data obtained as in (A) against recovery interval. Data were fit with an exponential function (smooth line). Curve fit parameters were: for P6–8 (n = 19), y₀ = 0.99, A; −0.321, τ = 425 ms; for P16–18 (n = 8), y₀ = 1.0, A = −37.8, τ = 486 ms.

Ca²⁺-dependent inactivation depends on local Ca²⁺ in mature IHCs but not immature IHCs

In heterologous expression systems, CDI of Ca_v1 channels is not affected by relatively high concentrations of intracellular Ca²⁺ buffers (i.e., 10 mM BAPTA), due to a reliance on rapid, local Ca²⁺ influx through individual channels.²⁸ However, CDI of Ca_v1 currents in immature rat IHCs but not in mature gerbil IHCs is suppressed by high intracellular concentrations (10 mM) of EGTA or BAPTA.^17,18 These results suggest that the Ca²⁺ sensitivity of Ca_v1 CDI may be subject to developmental regulation. To test this, we analyzed the impact of increasing Ca²⁺ buffering strength on CDI in mature and immature mouse IHCs using triple-pulse protocols with varying prepulse voltages. To rapidly buffer incoming Ca²⁺ ions, we used relatively high concentrations (10 mM) of BAPTA in the intracellular recording solution. Since BAPTA (10 mM) does not influence inactivation of I_Ba mediated by Ca_v1.3,²⁹ we restricted analysis to I_Ca in these experiments. Compared with our standard solution with EGTA (2 mM), the BAPTA-containing solution significantly blunted CDI in immature IHCs across a range of prepulse voltages (up to ~30% between −30 and +20 mV, p < 0.05 by two-way ANOVA; Fig. 7A and B). Remarkably, this effect of BAPTA was not observed in mature IHCs (p = 0.16 by two-way ANOVA; Fig. 7A–C). These effects of BAPTA were not due to altered voltage-dependence of activation, since G-V parameters were not significantly different with BAPTA or low intracellular EGTA (Table 1, Fig. S1).

Figure 7. CDI is inhibited by BAPTA in immature IHCs but not mature IHCs. (A) Top, Voltage protocol in which test currents were evoked by pulses (p1, p2) from −84 mV to −20 mV before and after a 200-ms prepulse to various voltages. Bottom, Representative traces for I_Ca recorded in P6–8 or P16–18 mouse IHCs with protocols using the indicated prepulse voltages. Intracellular recording solution contains 2 mM EGTA or 10 mM BAPTA. Currents evoked by p1 (black) and p2 (red) are overlaid for comparison. Scale bars, 0.2 nA (vertical), 2.5 ms (horizontal). (B) p2/p1 was plotted against prepulse voltage for P6–8 (n = 20 for 2 EGTA, n = 25 for 10 BAPTA) and P16–18 (n = 18 for 2 EGTA, n = 8 for BAPTA). *p < 0.05 for EGTA vs. BAPTA, by two-way ANOVA and Bonferroni t-test. (C) Representative normalized current traces obtained with 2 EGTA or 10 BAPTA. I_Ca was evoked by a 200-ms test pulse to −10 mV from −84 mV. (D) For P6–8 IHCs, p2/p1 ratios were plotted against I_Ca amplitude for 2 EGTA or 10 BAPTA. Lines represent fits by linear regression.

The BAPTA-sensitivity of CDI in immature but not mature IHCs suggests that Ca²⁺ elevations that support CDI are not restricted to local Ca²⁺ influx through single channels but rather global Ca²⁺ signals fueled by multiple neighboring channels. In this respect, Ca_v1.3 CDI in immature IHCs may be similar mechanistically to CDI of Ca_v2.1 channels, which is strongly suppressed by high concentrations of BAPTA or EGTA.^30-34 If so, CDI should increase with channel density in immature IHCs. Consistent with this prediction, p2/p1 ratios decreased (i.e., inactivation increased) with increasing I_Ca amplitude (Fig. 7D). BAPTA caused a significant increase in the slope of the p2/p1 vs. I_Ca relationship (4 × 10⁻⁴ for EGTA vs. 9.5 × 10⁻⁴ for BAPTA, p < 0.03 by ANCOVA; Fig. 7D). These results show that BAPTA increases the current-amplitude dependence of CDI, such that small-amplitude I_Ca showed weaker CDI (and BAPTA-sensitivity) than large-amplitude I_Ca.

To determine if the sensitivity of CDI to BAPTA in immature IHCs was due to effects of BAPTA on the onset and/or recovery kinetics of I_Ca inactivation, we compared these parameters using either 2 mM EGTA or 10 mM BAPTA in the intracellular recording solution. With BAPTA, the onset of inactivation exhibited a double exponential time course as was found for EGTA (Figs. 5 and 8A–C). In these experiments, BAPTA did not affect either the slow or fast time constants (τ_fast = 48.1 ± 9.2 ms for BAPTA vs. 32.9 ± 3.9 ms for EGTA, p = 0.06; τ_slow = 1.2 ± 0.3 ms for BAPTA vs. 2.0 ± 0.5 ms for EGTA, p = 0.07, both by Mann–Whitney rank sum test; Fig. 8C). However, BAPTA significantly decreased the corresponding amplitudes (A_fast = 0.15 ± 0.02 for BAPTA vs. 0.25 ± 0.01 for EGTA, p < 0.001; A_slow = 0.20 ± 0.01 for BAPTA vs. 0.40 ± 0.03 for EGTA, p < 0.001, both by Mann-Whitney rank sum test) and increased the available current (Y₀ = 0.66 ± 0.02 for BAPTA vs. 0.34 ± 0.04 for EGTA, p < 0.001, by Mann–Whitney rank sum test; Fig. 8C). In addition, BAPTA accelerated recovery of I_Ca from inactivation (τ = 271.9 ± 28.3 ms for BAPTA vs. 425.9 ± 28.3 ms for EGTA, p < 0.001 by t-test; Fig. 8D–F). Thus, BAPTA increases Ca_v1.3 channel availability in immature IHCs by inhibiting the amount of I_Ca inactivation as well as by speeding recovery from CDI. Our results highlight age-dependent differences in the intrinsic properties and modulation of Ca_v1.3 channels in mouse IHCs that may be important for the normal function and/or maturation of these cells as sound-transducers in the inner ear.

Figure 8. BAPTA slows the onset and accelerates recovery of CDI in immature IHCs. (A) Voltage protocol and representative p1 and p2 current traces for I_Ca with the indicated prepulse durations. Scale bars, 0.2 nA (vertical), 2.5 ms (horizontal). (B) Graph shows data plotted as in Figure 4B for I_Ca recorded with 2 EGTA (n = 22) or 10 BAPTA (n = 14). Smooth line represents double exponential curve fit (for 10 BAPTA, y₀ = 0.66, A_fast = 0.14, τ_fast = 48.1 ms, A_slow = 0.20, τ_slow = 1251 ms; see Figure 4B for parameters for 2 EGTA). (C) Average parameters for exponential fits of data in (B).*p < 0.001 by Mann-Whitney rank sum test. (D) Voltage protocol and representative p1 and p2 current traces for I_Ca with the indicated recovery intervals. Scale bars, 0.2 nA (vertical), 2.5 ms (horizontal). (E) Graph shows data plotted as in Figure 5B for I_Ca recorded with 2 EGTA (n = 8) or 10 BAPTA (n = 14). Smooth line represents exponential curve fit (y₀ = 1.0, A = −0.25, τ = 271 ms see Fig. 5B for parameters for 2 EGTA). (F) Average parameters for exponential fits of data in (E).*p < 0.001 by t- test.

Discussion

The aim of the present study was to characterize the biophysical properties of Ca_v1.3 channels in mouse IHCs before and after the onset of hearing. While comprehensive analyses of Ca_v1 currents have been performed in auditory hair cells from other species,^17,18,35 details regarding the function and modulation of Ca_v1.3 channels in immature and mature mouse IHCs is necessary to understand the sequelae associated with dysregulated Ca²⁺ signaling in mouse models of deafness. Our results indicate that Ca_v1.3 channels undergo significant changes in activation and inactivation that may support distinct roles of these channels at different stages of development.

Technical considerations and comparisons with other species

Due to the concentration of Ca²⁺ or Ba²⁺ (5 mM) in our extracellular recording solution and the fact that our recordings were performed at room temperature (~25°C), parameters for I_Ca and I_Ba activation and inactivation in the present study differ from those expected under physiological conditions (~1.3 mM extracellular Ca²⁺, ~37°C). In addition to increasing the amplitudes of I_Ca and I_Ba, the higher than physiological concentration of extracellular permeant ions would cause a positive shift in the half-maximal activation voltage (V_h) due to increased neutralization of cell surface charges.³⁶ Subsequently, the voltage-dependence of time constants for activation and inactivation should be positively shifted with 5 mM compared with 1.3 mM extracellular Ca²⁺ or Ba²⁺. However, charge-screening effects of higher divalent concentrations should not influence the voltage-sensitivity of Ca_v1.3 activation, which is reflected by the steepness of the G-V curves (k from Boltzmann fits).^37,38 Assuming that Ca_v1.3 properties in mature and immature IHCs would be equally affected by the increased Ca²⁺ or Ba²⁺ concentration, we would expect our results to accurately report the age-dependent difference, if not the absolute values, pertaining to Ca_v1.3 properties.

As shown previously,^17,18 recordings at room temperature (~25°C) would cause smaller Ca_v1 current amplitudes and slower activation and inactivation kinetics than at physiological temperature (~37°C). At the same time, signaling processes that post-translationally alter the intrinsic biophysical properties of Ca_v1.3 channels would proceed more rapidly at 37°C than at 25°C. This could be a concern given that Ca_v1 amplitudes in immature IHCs show greater temperature sensitivity (larger Q₁₀) than in mature IHCs in the gerbil.¹⁸ Notably, we observed greater run-up of I_Ca upon initiation of whole-cell recordings in immature compared with mature IHCs (Fig. 1). Ca_v1 channel run-up has been reported in turtle and frog hair cells^35,39 and in cardiac myocytes (see for example ref. 40). While the underlying mechanisms are not entirely clear, run-up of cardiac Ca_v1 currents is associated with patch rupture in whole-cell recordings and exhibits a rapid (< 5 min) and late (> 10 min) phase, the latter of which is not observed at lower than physiological temperatures.⁴⁰ Therefore, our recordings at room temperature may minimize alterations in channel properties resulting from intracellular dialysis.

Our results in mouse IHCs may not be entirely consistent with those reported for gerbil and rat IHCs due to methodological and/or species-related differences. For example, the age-dependent acceleration in I_Ca activation kinetics in our study (Fig. 3) agrees with results from gerbil IHCs¹⁸ but not rat IHCs, in which I_Ca activation was slower after hearing onset than before.¹⁷ In addition, while we found CDI to be greater in mature IHCs than in immature IHCs (Fig. 4), there was no such age-related difference in gerbil IHCs,¹⁸ and CDI was actually greater in immature IHCs than in mature IHCs in the rat.¹⁷ Our results regarding the Ca²⁺ buffer-sensitivity of CDI in immature but not in mature mouse IHCs (Figs. 7 and 8) are consistent with these previous studies: high concentrations of BAPTA (5 mM) did not affect CDI in mature gerbil IHCs¹⁸ but did significantly suppress CDI in immature rat IHCs.¹⁷ Therefore, the shift in CDI dependence on global Ca²⁺ signals in immature IHCs, to local Ca²⁺ signals in mature IHCs, may be a fundamental feature of IHC development.

Possible mechanisms for developmental alterations in Ca_v1.3 function

The properties of Ca_v1 channels are subject to multiple forms of regulation, which could account for the difference in Ca_v1.3 properties we noted during mouse IHC development. We showed previously that the PDZ-domain containing protein harmonin interacts with Ca_v1.3 channels in mouse IHCs after hearing onset and inhibits I_Ba.²⁷ In mature mouse IHCs, harmonin enhances proteosomal degradation of Ca_v1.3, which could contribute to the age-dependent decrease in I_Ca or I_Ba current density (Fig. 2C and D). A second possibility is related to alternative splicing of the pore-forming Ca_v1.3 α₁ subunit (α₁1.3), which produces variants that differ in their biophysical properties and modulation.^25,41-47 After hearing onset, there may be an upregulation of α₁1.3 variants lacking portions of a C-terminal modulatory domain (CTM), which exhibit decreased current density relative to variants with the intact CTM.^41,48 However, it is also possible that factors regulating the transcription/translation of Ca_v1.3 channels could also vary in mouse IHCs with age.

Some CTM-lacking variants exhibit enhanced voltage-dependence of activation and stronger CDI than variants containing the entire CTM.⁴⁴ Such variants could account for increased CDI in mature IHCs (Fig. 4), but their hyperpolarized activation properties relative to the CTM-containing variants are inconsistent with our findings of weaker voltage-dependent activation of Ca_v1.3 currents in mature than in immature mouse IHCs (Table 1). Alternative splicing and RNA editing of the calmodulin-binding IQ domain α₁1.3 also produce Ca_v1.3 variants with limited CDI,^47,49 although the extent to which these variants are present in immature mouse IHCs is unknown. Ca_v1.3 channels interact with a variety of regulatory proteins, which may also influence CDI.⁵⁰ Rab 3-interacting molecule (RIM2α) inhibits CDI when coexpressed with Ca_v1.3 channels in HEK293 cells.⁵¹ RIM2α transcripts were detected in immature but not mature IHCs⁵¹ and so could be responsible for the more moderate CDI in mouse IHCs prior to hearing onset (Fig. 4). Other possibilities include Ca²⁺ binding proteins related to calmodulin (CaBPs),^17,21,22 and various synaptic proteins,⁵² which also suppress inactivation of Ca_v1.3 in heterologous expression systems. Whether such Ca_v1.3 modulatory proteins undergo developmental changes in expression remains to be elucidated.

The BAPTA-sensitivity of CDI in immature IHCs (Figs. 7 and 8) contrasts with the notion that Ca_v1 channel CDI is largely mediated by local Ca²⁺ influx through single channels.⁵³ In heterologous systems, Ca_v1.3 CDI is spared by high concentrations (10 mM) intracellular BAPTA due to a molecular determinant in the cytoplasmic N-terminal domain of α₁1.3 (N-terminal spatial Ca²⁺ transforming element, NSCaTE).²⁹ Binding of the N-terminal lobe of calmodulin to NSCaTE is thought to alter the Ca²⁺ binding properties of calmodulin such that CDI is determined by rapid local increases in Ca²⁺ that cannot be buffered by BAPTA.²⁹ Deletion of NSCaTE from α₁1.3 transforms the spatial selectivity of Ca_v1.3 CDI such that it can be blunted by BAPTA.²⁹ While α₁1.3 splice variants lacking NSCaTE remain to be identified, their presence in immature but not mature IHCs could explain the BAPTA sensitivity of CDI in the former and not the latter. A second possibility relates to the increase in synaptic clustering of Ca_v1.3 channels in mouse IHCs after hearing onset.^38,54 In frog auditory hair cells, synchronous multivesicular release occurs even in the presence of 10 mM BAPTA, which is attributed to a role of the ribbon as a barrier for Ca²⁺ diffusion.⁵⁵ Ca_v1.3 channels anchored near the ribbon may therefore be influenced by rapidly accumulating Ca²⁺ nanodomains that promote single-channel CDI in mature IHCs. In contrast, a large fraction of Ca_v1.3 channels are extrasynaptically localized prior to hearing onset,³⁸ where the absence of the ribbon may allow for faster Ca²⁺ diffusion away from open channels and slower elevations in Ca²⁺ that can be buffered by BAPTA.

Physiological significance of age-dependent changes in Ca_v1.3 properties and modulation

The distinct properties of Ca_v1.3 channels we describe in immature and mature mouse IHCs may support the varying roles of these channels during development. Prior to hearing onset, Ca_v1.3 channels mediate spontaneous Ca²⁺-dependent action potentials,¹¹ which support exocytosis^12,56 and synaptic activity in the auditory pathway.⁵⁷ Raising the concentration of Ca²⁺ in the extracellular recording solution from 1.3 to 5 mM increases the amplitude and accelerates the upstroke and repolarization, and enhances the frequency of the presensory action potentials in mouse IHCs.⁵⁶ Therefore, the increased I_Ca density and activation and stronger voltage-dependent activation in the immature compared with in mature IHCs (Fig. 2C and D; Table 1) may be required to maintain the shape and timing of the action potential waveforms. Ca_v1.3 Ca²⁺ influx enhances repolarization of the action potential through coupling to the activation of SK Ca²⁺ activation K⁺ channels.⁵⁸ Based on the ability of BAPTA to suppress Ca_v1.3-dependent activation of SK channels, Ca_v1.3 channels may be localized ~40 nm from SK channels.⁵⁸ This is in contrast to the tighter coupling of nicotinic acetylcholine receptors (~13 nm distance),⁵⁸ which also conduct Ca²⁺ that activates SK channels.^13,59 In this context, the reduced CDI prior to hearing onset (Fig. 4) would allow for more prolonged Ca²⁺ signals that support SK channel activation and efficient Ca²⁺ spiking. In addition, sustained I_Ca in immature IHCs may support activity-dependent gene transcription, a hallmark function of Ca_v1 channels in the central nervous system.⁶⁰ Ca_v1.3 channels are required for normal levels of BK K⁺ channel transcription in mouse IHCs,¹⁶ and BK channel expression in muscle is regulated by the Ca²⁺-dependent gene transcription factor, nuclear factor of activated T-cells (NFAT3).⁶¹ Future studies analyzing gene expression differences in mature and immature IHCs would help identify Ca²⁺-regulated transcripts that may be important for IHC development.

In mature IHCs, the developmental enhancement of Ca_v1.3 activation kinetics (Fig. 3) and CDI (Fig. 4) may improve the temporal aspects of sound coding. In paired, pre- and postsynaptic recording at the IHC synapse, faster Ca_v1.3 activation correlates with decreased latency of postsynaptic responses,⁶² which supports the ability of auditory nerve fibers to fire at particular times during low-frequency stimuli (phase-locking).⁶³ While Ca_v1.3 CDI would be expected to limit exocytosis by restricting Ca²⁺ influx, CDI is expected to contribute modestly to depression at the IHC synapse.^62,64,65 Rather, CDI may help shape Ca²⁺ signals required for appropriate rates of vesicle replenishment at the ribbon^64,65 and/or help limit Ca²⁺ loads that could provoke metabolic stress.

In summary, we have described key functional distinctions in the properties of Ca_v1.3 channels in immature and mature mouse IHCs. An understanding of how these developmental differences in Ca_v1.3 Ca²⁺ signals contribute to the maturation and function of IHCs in the auditory pathway remains an important challenge for future studies.

Materials and Methods

Ethical approval

All procedures were approved by the Institutional Animal Care and Use Committee at the University of Iowa in accordance with National Institutes of Health guidelines.

Preparation of mouse cochlear tissue

C57bl/6 mice (Harlan Laboratories; P6–8 or P16–18 males or females) were euthanized with ketamine (100 mg/kg) and xylazine (9 mg/kg), the skull was opened, and the temporal bones were dissected out and immersed in dissection solution (MEM/Glutamax-1 (Invitrogen) supplemented with 10 mM HEPES) which was preincubated at 37°C in a humidified incubator with 5% CO₂ . The apical turn was dissected in fresh dissection solution and the spiral ligament removed. The tissue was then secured under a glass pin on a 15 mm round coverslip in dissection solution and incubated for 2–5 h at 37°C in a humidified incubator with 5% CO₂ prior to recording.

Electrophysiological recordings

IHCs in the apical cochlear turn were visualized on an upright microscope (BX51WI, Olympus) with a 40X water-immersion objective with DIC optics. The basolateral membrane of IHCs was subject to whole-cell patch clamp recording with electrodes pulled from thick-walled borosilicate glass capillaries (1B150F, Warner Instruments). Before recording, the viability of the IHCs was confirmed by the following morphological features: uniform cell shape with a narrow neck, basal location of the nucleus, membrane birefringence, and intact stereocilia. The internal solution contained (in mM): 100 Cs-gluconate, 30 TEA-Cl, 1 MgCl₂, 4 MgATP, 0.3GTP, 5 HEPES and 2 EGTA; pH was adjusted to 7.35 with CsOH; osmolarity~305 mOsm. In some experiments, 10 mM BAPTA was substituted for 2 EGTA and Cs-gluconate was reduced to 92 mM. External solution contained (in mM): 105 NaCl, 2.8 KCl, 3 CsCl, 35 TEA-Cl, 5 CaCl₂ or BaCl₂, 1 MgCl₂, 10 glucose, and 10 HEPES supplemented with MEM Vitamins and Amino Acids at 1X; pH was adjusted to 7.4 with TEA-OH; osmolarity~310 mOsm. Under these conditions, the free intracellular Ca²⁺ was estimated to be less than 0.5 nM (with 2 mM intracellular EGTA) or less than 0.1 nM (with 10 mM intracellular BAPTA). On the day of recording, 4-aminopyridine (10 mM), apamin (0.3 mM) and TTX (0.5 mM, for P6–8 IHCs) were added to the external solution. Electrode resistances were 3–5 MΩ in the external solution. Data were acquired at room temperature with a HEKA EPC-10 amplifier controlled by Patchmaster software (HEKA Elektronik). Leak subtraction was done online with a P/8 protocol. Series resistance was compensated with the patch clamp circuitry (50–70%); average uncompensated series resistance was 9.64 ± 0.38 (n = 121). Currents were low-pass filtered at 5 kHz and sampled at 20 kHz. Voltages were not corrected for the liquid junction potential of −7 mV in the external recording solution.

Electrophysiological data were analyzed with custom routines in IgorPro software (Wavemetrics). Current-voltage (I-V) relationships were fit with the equation: I = [G_max (V − V_rev)]/ [1 + exp [(V_h − V)/k], where I is the current, G_max is maximum chord conductance, V is voltage, V_rev is reversal potential, V_h is half-maximal activation voltage, and k is the slope factor. Conductance-voltage relationships were fit with the equation; G/G_max = 1/ [1 + exp [(V_h − V)/k]]. Time constants for activation and parameters for recovery from inactivation were obtained by fitting with the equation: I(t) = y_o + A[exp (−t/τ)], where I(t) is current at time = t, y₀ is the offset (asymptote), A is the amplitude, and τ is the time constant. The onset of I_Ca inactivation was fit with a double exponential function: I(t) = y₀ + A_fast [exp (−t/τ _fast)] + A_slow [exp (−t/τ _slow)]where I(t) is current at time = t, y₀ is the offset (asymptote), A_fast and A_slow are the amplitudes, and τ _fast and τ _slow are the time constants. Average data are expressed as mean ± SEM. Statistical comparisons were done as indicated using SigmaPlot software (Systat).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.