Genetic deletion of SK2 channels in mouse inner hair cells prevents the developmental linearization in the Ca²⁺ dependence of exocytosis

Tissue preparation

Apical-coil IHCs (n = 124) of SK2 Δ/Δ mutant mice (Bond et al. 2004) and their littermate controls (+/Δ and +/+) were studied in acutely dissected organs of Corti from postnatal day 2 (P2) to P50, where the day of birth is P0. Mice were killed by cervical dislocation in accordance with UK Home Office regulations. The cochleae were dissected in normal extracellular solution (mm): 135 NaCl, 5.8 KCl, 1.3 CaCl₂, 0.9 MgCl₂, 0.7 NaH₂PO₄, 5.6 d-glucose, 10 Hepes–NaOH. Sodium pyruvate (2 mm), amino acids and vitamins (Eagle's MEM) were also added from concentrates (Invitrogen, UK). The pH was adjusted to 7.5 (osmolality ∼308 mmol kg⁻¹). The dissected cochleae were transferred to a microscope chamber, immobilized under a nylon mesh attached to a stainless steel ring, and continuously perfused with a peristaltic pump using the above extracellular solution. The organs of Corti were viewed using an upright microscope (Leica DMLFS, Germany) with Nomarski optics (×63 objective). In situ recordings were made by exposing the basolateral surfaces of IHCs using a suction pipette (tip diameter about 4 μm), which was filled with extracellular solution. In some experiments 300 nm apamin (Calbiochem, UK) was locally applied to block the SK current (I_SK). Extracellular solutions containing drugs were applied through a multibarrelled pipette positioned near the preparation. Animals were genotyped as previously described (Bond et al. 2004).

Electrical recording

Patch clamp recordings were performed near body temperature (34–37°C) using an Optopatch amplifier (Cairn Research Ltd, UK). Patch pipettes were made from soda glass capillaries (Harvard Apparatus Ltd, UK) and had a typical resistance in extracellular solution of 2–3 MΩ. In order to reduce the electrode capacitance, patch electrodes were coated with surf wax (Mr Zoggs SexWax, USA). Patch electrodes were positioned on the IHC membrane using a PatchStar micromanipulator (Scientifica, UK). The whole-cell patch pipette filling solution was (mm): 131 KCl, 3 MgCl₂, 1 EGTA–KOH, 5 Na₂ATP, 5 Hepes–KOH, 10 Na₂-phosphocreatine (pH 7.3, 294 mmol kg⁻¹). The perforated-patch technique (Rae et al. 1991) was used on a few immature IHCs (n = 9) in order to support the whole-cell voltage response findings. For these experiments the pipette filling solution contained (mm): 21 KCl, 110 potassium aspartate, 3 MgCl₂, 5 Na₂ATP, 1 EGTA–KOH, 5 Hepes–KOH, 10 sodium phosphocreatine (pH 7.3, 295 mmol kg⁻¹). The antibiotic amphotericin B (Calbiochem, UK) was dissolved in dry dimethylsulfoxide (DMSO) prior to its dilution into the above intracellular solution to a final concentration of 120 μg ml⁻¹ or 240 μg ml⁻¹ (DMSO final dilution was 1: 500 or 1: 250, respectively). The patch pipette was tip-filled with the above normal potassium aspartate intracellular solution before back-filling with the amphotericin B containing solution. This was performed in order to prevent leakage of the antibiotic into the bath prior to sealing onto the IHC membrane. Data were acquired (sampling: 5 kHz, filtering: 2.5 kHz, 8-pole Bessel) using pCLAMP software and a Digidata 1320A (Molecular Devices, USA) and stored on computer for off-line analysis (Origin: OriginLab Corp., USA; Mini Analysis Program: Synaptosoft, USA). Statistical comparisons of means (quoted ± s.e.m.) were made by Student's two-tailed t test where P < 0.05 indicates statistical significance.

Current recordings were corrected off-line for the linear leak conductance (g_leak). In immature IHCs (P2–P6) g_leak was typically calculated between −84 mV and −74 mV (+/+ and +/Δ: 1.3 ± 0.2 nS, n = 15; Δ/Δ: 1.5 ± 0.2 nS, n = 9), as the outward K⁺ currents activate positive to around −60 mV. For adult IHCs (P20) g_leak (+/+ and +/Δ: 1.0 ± 0.2 nS, n = 4; Δ/Δ: 0.9 ± 0.2 nS, n = 5) was usually calculated between −84 mV and −94 mV, as the outward K⁺ currents I_K,s and I_K,f activate positive to about −80 mV. g_leak was not significantly different between control and mutant IHCs. In voltage-clamp recordings, membrane potentials were corrected for residual series resistance (R_s) after compensation (immature IHCs: +/+ and +/Δ: 5.2 ± 0.4 MΩ, n = 31; Δ/Δ: 4.4 ± 0.5 MΩ, n = 19; adult: +/+ and +/Δ: 1.2 ± 0.1 MΩ, n = 4; Δ/Δ: 1.1 ± 0.1 MΩ, n = 5) and for a liquid junction potential (LJP), measured between electrode and bath solutions. The LJP was −4 mV for the KCl-based and −11 mV for the potassium aspartate-based intracellular solution. Voltage recordings in current clamp were also corrected for the LJP.

Membrane capacitance measurement

Real-time changes in membrane capacitance (ΔC_m) were measured using the Optopatch as previously described (Johnson et al. 2002, 2005). Briefly, a 4 kHz sine wave of 13 mV RMS was applied to IHCs from −81 mV and was interrupted for the duration of the voltage step. The amplitude of the sine wave was chosen to be small enough not to activate any significant current since membrane capacitance measurements require a high and constant membrane resistance. The capacitance signal from the Optopatch was amplified (×50), filtered at 250 Hz, and sampled at 5 kHz. ΔC_m was measured by averaging the C_m trace over a 200 ms period following the voltage step and subtracting from prepulse baseline. For these experiments the intracellular solution contained (mm): 110 caesium glutamate, 20 CsCl, 3 MgCl₂, 1 EGTA-CsOH, 5 Na₂ATP, 0.3 Na₂GTP, 5 Hepes-CsOH, 10 Na₂-phosphocreatine (pH 7.3, 294 mmol kg⁻¹). ΔC_m was recorded while applying 30 mm TEA (Fluka, UK), and additionally 300 nm apamin in immature IHCs or 60–100 μm linopirdine (Tocris, UK) in adult cells, to reduce K⁺ currents. In mature IHCs the Ca²⁺ dependence of ΔC_m was studied by superfusing a Ca²⁺-free solution (containing 0.5 mm EGTA) or different Ca²⁺ concentrations (1.3, 5 and 10 mm). When the concentration of the blockers or Ca²⁺ that were added to or removed from the solution was > 1 mm, NaCl was adjusted to keep the osmolality constant. Membrane potentials were corrected for R_s (4.9 ± 0.1 MΩ, n = 55, P6–P50) and the LJP (−11 mV).

SK2 current is required for sustaining action potential activity in immature IHCs

Before the onset of hearing, mouse IHCs generate Ca²⁺ APs spontaneously (Kros et al. 1998; Beutner & Moser, 2001; Marcotti et al. 2003a, 2003b). To determine whether I_SK2, which is transiently expressed during immature stages, exerts a critical role in maintaining repetitive spontaneous APs, we investigated voltage responses from SK2 knockout mice (Bond et al. 2004). For these experiments early postnatal-day mice (P2–P4) were used to increase the likelihood of finding IHCs that fired spontaneous APs (Marcotti et al. 2003a). Figure 1A shows an example of spontaneous APs from an immature control IHC. Since there were no apparent differences between wild-type (+/+) and heterozygous (+/Δ) IHCs, the results from the two cell populations were pooled. Control IHCs fired spontaneous APs with a frequency of 4.2 ± 0.5 Hz (n = 12, P2–P4), similar to that previously found in IHCs from CD-1 mice (Marcotti et al. 2003b). The presence of I_SK2 in control cells was confirmed by the slowly activating, apamin-sensitive current (Marcotti et al. 2004b) recorded in voltage-clamp (Fig. 1B, +/+).

Despite the absence of I_SK2 (Figs 1B, Δ/Δ), mutant IHCs were able to generate spontaneous APs (Fig. 1C) for several seconds (48 ± 12 s, n = 9, P2–P4) with a frequency of 3.4 ± 0.3 Hz. However, with time APs gradually became wider until a sustained depolarization occurred (Fig. 1C, red inset) in all P3–P4 IHCs investigated (n = 8) but in only 1 out of 7 P2 cells. This indicates that from around P2–P3 I_SK is essential for sustaining continuous and rapid firing activity by providing a more robust AP repolarization. This also agrees with previous findings indicating that I_SK2 becomes expressed in IHCs from around P2–P3 (Katz et al. 2004; Marcotti et al. 2004b). In addition to differences in width, mutant APs were significantly (P < 0.04) smaller in amplitude (32.6 ± 2.2 mV, n = 9, P2–P4) compared to those of control cells (39.1 ± 1.9 mV, n = 12, P2–P4). The sustained depolarization observed in mutant IHCs (which lasted for 32 ± 6 s, n = 7, P2–P4) was not permanent since they were able to repolarize completely within a few seconds and fired spontaneously again (Fig. 1C and green inset) after an additional 53 ± 10 s with a frequency of 2.3 ± 0.4 Hz. This pattern was repeated three times in one of the longest recordings made. A similar behaviour was also observed when a fully blocking concentration of apamin (300 nm), a selective blocker of SK channels (Sah & Faber, 2002), was superfused onto control IHCs (Fig. 1D). The block of I_SK2 was confirmed by voltage-clamp recordings (Fig. 1B, +/+ and Apa). The change in the AP waveform with time, recorded in the three different situations (Fig. 1A, C and D), is shown in Fig. 1E. The gradual widening of APs in both mutant and control (in the presence of apamin) IHCs was evident when the width of the APs (at 20% from baseline) was plotted as a function of time (Fig. 1F). Qualitatively similar results to those shown in Fig. 1 were obtained for all cells investigated (+/+ and +/Δ: n = 13; +/+ and Apa: n = 3; Δ/Δ: n = 12). To ensure that possible differences in resting membrane potential between control and mutant IHCs did not bias our observations we studied evoked APs. In control IHCs (Fig. 2A), the frequency of APs increased with the amount of depolarizing current injection applied as previously described in CD-1 mice (Marcotti et al. 2003a). In contrast a gradual disappearance of the firing activity in mutant IHCs (Fig. 2B) and control cells superfused with apamin (Fig. 2C) was seen when current injections were applied. It is also worth noting that in control IHCs the contribution of I_SK was evident by the more robust AP repolarization occurring towards the beginning of each current step (Fig. 2A, arrowheads). The absence/block of I_SK, together with the progressive inactivation of I_K,neo, is likely to be responsible for the gradual increase in the input resistance during large current steps (e.g. +60 pA and +70 pA in Fig. 2B and C).

To determine whether the whole-cell configuration, a condition that could modify the intracellular milieu, had an effect on action potential activity (Figs 1 and 2) we recorded from a few IHCs using the perforated-patch technique, a condition closer to the in vivo situation. The results shown in Fig. 3 clearly indicate that spontaneous action potentials from control and mutant IHCs (Fig. 3A and B, respectively) were comparable to those recorded in whole-cell conditions. Mutant IHCs exhibited difficulties in repolarizing completely following an AP upstroke and with time a sustained depolarization occurred (Fig. 3B). Interestingly, APs with a long-lasting ‘plateau’, as seen in whole-cell experiments (Fig. 1C), could be recorded following membrane perforation (Fig. 3B), indicating that mutant IHCs are likely to fire prolonged APs and therefore spend a considerable amount of time depolarized in vivo. It is worth noting that a persistent depolarization was never observed at the beginning of current clamp recordings. This could occur if the membrane potential cycle observed in mutant IHCs is initially reset by voltage clamping the cells near their resting potential for patch break-in/perforation. Evoked APs were also similar between whole-cell (Fig. 2A and B) and perforated patch (Fig. 3C and D) with mutant IHCs losing the ability to fire APs with time and current injection.

Potassium currents in SK2 mutant IHCs develop normally

It has been suggested that AP activity in pre-hearing IHCs could serve as a developmental signal by influencing the expression of ion channels and other proteins within the cells themselves (Beutner & Moser, 2001; Marcotti et al. 2003a, b). Since the absence of I_SK2 from mutant IHCs disrupted the normal spontaneous and evoked firing activity of immature cells (Figs 1–3), we investigated some of the cell's biophysical properties in order to verify whether they matured normally. Typical examples of outward (delayed rectifier I_K,neo: Marcotti et al. 2003a) and inward (inward rectifier I_K1: Marcotti et al. 1999) K⁺ currents characteristic of immature IHC are shown in Fig. 4. The size of both K⁺ currents was found to be indistinguishable between control and mutant IHCs (Fig. 4C and F and see Table 1). Although the inward Na⁺ current was not studied in isolation, in some recordings (e.g. Fig. 4A and B) it was seen to precede the much slower outward K⁺ current of both control and mutant IHCs. Table 1 also shows that the basolateral membrane properties of immature IHCs were not significantly different between control and mutant cells. Furthermore, the K⁺ currents characteristic of adult IHCs (Fig. 5A–C, F and G; I_K,s and I_K,n: Marcotti et al. 2003a, 2004a; Oliver et al. 2003; I_K,f: Kros et al. 1998), voltage responses (Fig. 5D and E) and resting membrane potentials (Fig. 5H) did not differ significantly between control and mutant adult cells, indicating that the normal development of K⁺ currents and the general biophysical properties of IHCs were unlikely to be affected by the mutation.

Exocytosis in IHCs from SK2 mutant mice fails to mature

Since the maturation of K⁺ currents was normal in mutant IHCs, we investigated whether the development of I_Ca and the induced exocytosis was altered by the mutation. Exocytosis was estimated by measuring increases in cell membrane capacitance (ΔC_m) following depolarizing voltage steps, which is generally interpreted as a sign of neurotransmitter release from presynaptic cells (Parsons et al. 1994; von Gersdorff et al. 1998; Moser & Beutner, 2000; Schnee et al. 2005). Figure 6A and B shows I_Ca and the corresponding ΔC_m recorded in response to a 100 ms depolarizing voltage step from control and mutant immature IHCs, respectively. The I_Ca in pre-hearing mutant IHCs was significantly (P < 0.0005) smaller than that recorded in control cells (Fig. 6C, current–voltage (I–V) relation) and is likely to be responsible for the reduced AP amplitude observed in these cells (Fig. 1). As a consequence of the smaller I_Ca a significant reduction (P < 0.005) in the Ca²⁺-induced ΔC_m (Fig. 6D, capacitance–voltage (ΔC_m–V) relation) was also seen in mutant IHCs (10.4 ± 1.0 fF, n = 8, P6–P8) compared to that in control cells (37.2 ± 6.9 fF, n = 11, P6–P8).

Figure 6E and F shows synaptic transfer functions that show the relation between the peak I_Ca and exocytosis for control and mutant IHCs, at different membrane potentials (Augustine et al. 1985; Johnson et al. 2005). For the above comparison, peak I_Ca was preferred to the time integral since I_Ca is likely to be contaminated by the presence of residual unblocked K⁺ currents (I_K,neo in immature IHCs and I_K,s and I_K,Ca in adult cells), which are known to be difficult to block completely (Marcotti et al. 2003a, 2004a). While slowly activating K⁺ currents are absent or relatively small at the peak I_Ca, their interference would become more prominent towards the end of the voltage steps causing variability in the apparent I_Ca inactivation (Fig. 6A and B; see also Johnson et al. 2005). This is particularly true when working near body temperature since the activation kinetics of the contaminating outward K⁺ currents become faster.

The single data points shown in Fig. 6E and F were approximated using a power function:

(1)

where c is a scaling coefficient and the exponent n is the power. In immature IHCs, the power obtained from fitting the single data points from each control (Fig. 6E; 2.7 ± 0.1, n = 11) and mutant (Fig. 6F; 2.7 ± 0.4, n = 8) IHC indicates that each release event requires around three Ca²⁺ binding steps to occur (Dodge & Rahamimoff, 1967; Johnson et al. 2005). This suggests that mutant IHCs, despite the reduced ΔC_m, show a normal Ca²⁺ dependence of exocytosis at this stage of development (see the complete overlap of the two synaptic transfer functions: Fig. 6G).

An example of I_Ca and corresponding ΔC_m recorded in adult control and mutant IHCs is shown in Fig. 7A and B, respectively. In contrast to immature IHCs, depolarizing voltage steps elicited a similar peak I_Ca between mutant and control cells (Fig. 7C) although the former showed a significantly (P < 0.0001) smaller ΔC_m (Fig. 7D, mutant: 17.8 ± 1.4 fF, n = 10, P31–P49; control: 34.0 ± 2.2 fF, n = 12, P28–P50). The power value of the synaptic transfer functions in control adult cells (Fig. 7E) was 0.9 ± 0.1 (n = 13), suggesting a near linear relation between Ca²⁺ entry and exocytosis as previously described in normal CD-1 mice (Johnson et al. 2005). Surprisingly, the mutant transfer function (Fig. 7F) was best approximated using a power of 2.3 ± 0.2 (n = 9), significantly larger (P < 0.0001) than that found in the controls but similar to that of immature IHCs. A direct comparison of the synaptic transfer function between control and mutant IHCs is shown in Fig. 7G.

Since the Ca²⁺ dependence of exocytosis was altered in adult mutant IHCs (using physiological 1.3 mm extracellular Ca²⁺), we investigated this finding in greater detail by recording responses while superfusing cells with different Ca²⁺ concentrations (Augustine & Charlton, 1986). Changing the extracellular Ca²⁺ concentration over the range from zero to 10 mm gradually increased the size of both I_Ca and ΔC_m in control (Fig. 8A, +/+) and mutant (Figs 8B, Δ/Δ) IHCs. However, in mutant IHCs the increase in ΔC_m was less linear than that in control cells. Figure 8C and D shows the peak I–V (bottom panel) and ΔC_m–V (top panel) curves from control and mutant IHCs, respectively. The direct relation between the size of I_Ca and elicited ΔC_m (Fig. 8E and F) can be obtained by plotting the maximum responses in different Ca²⁺ concentrations and membrane potentials from each individual cell tested, similar to the synaptic transfer functions shown in Figs 6 and 7, where I_Ca and ΔC_m were varied by membrane potential in only 1.3 mm extracellular Ca²⁺. The power values obtained by fitting the data using eqn (1) for control (+/+ and +/Δ,Fig. 8E) and mutant (Δ/Δ, Fig. 8F) adult IHCs were 1.2 ± 0.1 (n = 5) and 2.5 ± 0.1 (n = 4, significant at P < 0.0001), respectively, which are similar to those obtained in Fig. 7E and F. The direct comparison between the two fits is shown in Fig. 8G.

The altered Ca²⁺ dependence of exocytosis in adult SK2 mutant IHCs would theoretically affect the release kinetics of their different vesicle pools. In order to investigate this hypothesis, the rate of neurotransmitter release in mutant and control IHCs was studied by measuring ΔC_m in response to depolarizing voltage steps, from 2 ms to 3 s in duration, to −11 mV from the holding potential of −81 mV (Fig. 9). This allowed us to study the emptying of different synaptic vesicle pools (Moser & Beutner, 2000; Johnson et al. 2005). While relatively short stimuli reveal the number of vesicles docked at the active zones (readily releasable pool: RRP), longer steps induce the release of vesicles from a secondarily releasable pool (SRP) that is located further away from the Ca²⁺ channels (von Gersdorff et al. 1996; Moser & Beutner, 2000; Beutner & Moser, 2001). Figure 9A and B shows ΔC_m responses from a control and mutant adult IHC, respectively, in response to depolarizing voltage steps of varying duration using physiological 1.3 mm extracellular Ca²⁺. The average ΔC_m responses to voltage steps of varying duration from control (P28–P31, n = 5) and mutant (P31–P32, n = 6) IHCs are shown in Fig. 9C. Similar to previous findings in normal CD-1 mice using similar recording conditions (body temperature and 1.3 mm Ca²⁺: Johnson et al. 2005) voltage steps of up to about 100 ms are likely to recruit mainly the RRP in both control and mutant IHCs since the increase of capacitance responses (Fig. 9D) could be approximated with a single exponential. The SRP (evident from the slower additional increase of ΔC_m > 100 ms, Fig. 9C) was isolated (Fig. 9E) by subtracting the RRP. The cross-over in the size of control and mutant ΔC_m responses associated with the SRP (Fig. 9C and E) is likely to reflect the higher Ca²⁺ dependence in mutant IHCs. The fact that mutant IHCs showed large ΔC_m responses following prolonged Ca²⁺ entry, similar to those of control cells, indicates that the number of available vesicles and/or release sites is unlikely to be affected by the mutation in adult cells. This might also suggest that the smaller ΔC_m responses observed for the RRP in mutant IHCs (Fig. 9D) could simply be due to a reduced Ca²⁺ efficiency of exocytosis (Figs 7 and 8) rather than a difference in the relative number of active zones and/or docked vesicles.

SK2 channels are essential for sustaining continuous repetitive spontaneous action potentials in pre-hearing IHCs

Small conductance Ca²⁺-activated K⁺ (SK) channels have a crucial role in modulating the excitability of neuronal cells (Bond et al. 2005). This is achieved by the ability of these channels to hyperpolarize the cell membrane potential when activated by transient and small increases in cytosolic Ca²⁺ resulting from the activation of voltage-gated Ca²⁺ channels (Tucker & Fettiplace, 1996; Marcotti et al. 2004b) or α9α10 nAChRs (Glowatzki & Fuchs, 2000; Oliver et al. 2000; Elgoyhen et al. 2001). In IHCs the SK current, which is carried by SK2 channels (Nie et al. 2004), is only transiently expressed during pre-hearing stages of development (Katz et al. 2004; Marcotti et al. 2004b). Using SK2 knockout mice (Bond et al. 2004) we have shown that repetitive spontaneous and evoked AP activity was gradually abolished when the SK2 current was absent (Figs 1C, 2B and 3B and D) or blocked by apamin (Figs 1D and 2C) leading to a sustained depolarization. Similar results, with apamin, have also been described on evoked APs from normal CD-1 mice, which was found to result from the direct coupling between SK2 and Ca²⁺ channels (Marcotti et al. 2004b). However, the disappearance of APs (Figs 1C and 3B) was not permanent since mutant IHCs were able to repolarize completely and fire spontaneously again. Therefore, the membrane potential of IHCs appears to cycle between periods of AP activity and maintained depolarization, probably causing periodical prolonged Ca²⁺ influx into these cells. Although we did not look for the presence of nAChRs in immature mutant IHCs, any efferent transmission onto these cells in vivo would, if anything, cause the cells to depolarize because of the absence of SK2 channels and most likely facilitate the appearance of wide APs and sustained depolarizations (Marcotti et al. 2004b).

Immature IHCs fire spontaneous APs from just before birth up to the end of the first postnatal week (Marcotti et al. 2003a). The generation and modulation of these APs depends on a variety of ion channels expressed in their basolateral membrane, most of which are only present transiently during pre-hearing stages of development (Housley et al. 2006). Although both Ca²⁺ and Na⁺ channels are expressed in IHCs, only the former is essential for the generation of APs (Marcotti et al. 2003b). Moreover, I_Ca together with a delayed rectifier K⁺ current (I_K,neo) determines the rates of rise and fall of APs, while I_Na reduces the interspike interval by speeding up the time required for the membrane potential to reach spike threshold (Marcotti et al. 2003a, b). The inward rectifier K⁺ current I_K1 is mainly involved in setting the cell's resting membrane potential (Marcotti et al. 1999). I_SK2 first appears in immature IHCs at around P2–P3 (Katz et al. 2004; Marcotti et al. 2004b), a period when APs change from wide and small to become narrower and much taller (Marcotti et al. 2003a). This suggests that soon after birth the presence of I_SK could be required for sustaining continuous and rapid firing activity by providing a more robust AP repolarization. Although I_K,neo is initially sufficient to sustain repetitive APs, I_SK2 (when activated by Ca²⁺ channels: Marcotti et al. 2004b) ensures IHCs repolarize completely following each AP and consequently prevents the progressive inactivation of I_K,neo. A transient role for SK2 in development is not unique to IHCs since I_SK2 is also known to be involved in modulating spontaneous firing activity in both retinal and cerebellar neurons where its expression is confined to early developmental stages (Klöcker et al. 2001; Cingolani et al. 2002).

Physiological consequences of the absence of SK2 channels in IHC development

Although the function of spontaneous APs in pre-hearing IHCs in currently unknown, the influx of Ca²⁺ during AP activity (Beutner & Moser, 2001) is likely to regulate a variety of cellular responses involved in their functional maturation (Berridge et al. 2000). Therefore, any alteration in the frequency, amplitude and duration of APs, as observed in SK2 mutants (Figs 1–3), could have major consequences on the normal developmental maturation of IHCs. In the developing nervous system spontaneous firing has been implicated in the refinement of neural circuits before the onset of externally driven activity (Zhang & Poo, 2001). In addition to this extrinsic role, spontaneous activity could also influence intrinsic cell development by modulating the expression of ion channels or other proteins implicated in normal cell maturation (Moody & Bosma, 2005). Indirect evidence for the latter has come from experiments on IHCs implying that the expression of the BK current I_K,f is induced by elevations in [Ca²⁺]_i from APs (Brandt et al. 2003). Although immature SK2 mutant IHCs expressed a smaller I_Ca, this is likely to be of minor relevance compared to the prolonged Ca²⁺ influx during sustained depolarizations (Figs 1–3). However, this Ca²⁺ elevation had no significant effects on the normal developmental expression of K⁺ currents (Figs 4 and 5) characteristic of immature (I_K,neo and I_K1: Marcotti et al. 2003a, 2003b) and adult (I_K,n: Marcotti et al. 2003a; Oliver et al. 2003 and I_K,f: Kros et al. 1998; Marcotti et al. 2004a) IHCs.

Although most of the biophysical properties were indistinguishable between control and mutant IHCs, the absence of I_SK was accompanied by the abnormal development of Ca²⁺-induced exocytosis. In immature IHCs the similar Ca²⁺ dependence of exocytosis between control and mutant cells (Fig. 6) suggests that the smaller I_Ca and ΔC_m observed in the latter is simply due to the expression of fewer Ca²⁺ channels compared to control cells. This is likely to result from fewer presynaptic active zones since the disrupted AP activity in these cells may not facilitate the survival of as many newly formed synaptic connections in the immature organ (Zhang & Poo, 2001; Moody & Bosma, 2005). If this is the case then the Ca²⁺ elevation at each active zone in pre-hearing IHCs is unlikely to be affected by the mutation, explaining the similar Ca²⁺ dependence observed between control and mutant IHCs (Fig. 6G). In adult IHCs while the size of I_Ca was similar between controls and mutants (Fig. 7) the induced ΔC_m was smaller in the latter, indicating a reduction in the Ca²⁺ efficiency of exocytosis in mutant cells. This was also evident from the release kinetics of the readily releasable pool of vesicles (Fig. 9D). The reduced efficiency (for I_Ca smaller than about 150 pA; Figs 7G and 8G) in mutant IHCs resulted from a higher order Ca²⁺ dependence of exocytosis (power N≈ 2.3–2.5) compared to control cells (near linear: N≈ 0.9–1.2) (Figs 7 and 8). This higher order Ca²⁺ dependence caused exocytosis in mutant IHCs to become more Ca²⁺ efficient than control cells following large Ca²⁺ currents (> 150 pA: Fig. 8G) or prolonged stimulation (> 1.5 s: Fig. 9E). The Ca²⁺ dependence of exocytosis in control IHCs changes with development from a third order relation in pre-hearing IHCs (Fig. 6E) to be approximately linear in adult cells (Fig. 7E), similar to previous studies in mice (Johnson et al. 2005). A linear relation of the stimulus–secretion coupling has also been confirmed in adult mouse IHCs (Brandt et al. 2005), turtle auditory hair cells (Schnee et al. 2005) and bullfrog papillae (Keen & Hudspeth, 2006). The higher-order Ca²⁺ dependence of exocytosis and the lower Ca²⁺ efficiency in adult mutant IHCs is again indicative of a failure in the normal functional development of the exocytotic machinery, most likely caused by the disrupted AP activity (rather than the smaller I_Ca) in immature cells. A recent investigation has shown that the complete absence of APs in pre-hearing IHCs of Ca_v1.3 knockout mice did not prevent robust exocytotic responses to flash photolysis of caged Ca²⁺ (Brandt et al. 2003). Although these findings suggest that the presence of synaptic machinery in IHCs is unlikely to be affected by a reduced number of Ca²⁺ channels and/or AP activity, the Ca²⁺-sensitivity of exocytosis in Ca_v1.3 mutant cells was not determined. While the factors that lead to the developmental linearization in the Ca²⁺ dependence of exocytosis in normal IHCs remains unknown, the Ca²⁺ nanodomain control of the stimulus–secretion coupling (Brandt et al. 2005; for a recent review see Moser et al. 2006) as well as the specific expression of Ca²⁺ sensing molecules such as otoferlin (Roux et al. 2006), are likely to play a crucial role.

The adult cochlea is specialized to encode acoustic signals with high temporal precision over a wide dynamic range of stimuli. In order to achieve these tasks, IHCs have developed synaptic specializations named ribbon synapses (Sterling & Matthews, 2005) that allow continuous release of neurotransmitter both spontaneously and in response to changes in membrane potential. Although SK2 mutant mice are not deaf (they showed a positive Preyer's reflex, authors' personal observations), the higher Ca²⁺ dependence of exocytosis in IHCs would elevate the threshold for neurotransmitter release, such that they would respond less efficiently for small stimuli and thus narrow their dynamic range. The direct influence of the altered Ca²⁺ dependence of exocytosis on the animals hearing sensitivity has yet to be determined. A high order Ca²⁺ dependence of exocytosis has recently been described in adult IHCs from Beethoven and deafness mutant mice (Marcotti et al. 2006) in which the transmembrane cochlear-expressed gene 1 (tmc1) is mutated. Both mutants exhibit absent (dn) or elevated (Bth) compound action potentials (CAP), although in this case the concomitant failure in the normal development of K⁺ currents characteristic of adult IHCs (I_K,f and I_K,n) is likely to have a prominent effect on the CAP.

The data presented here indicate that SK2 channels are essential for supporting continuous repetitive spontaneous Ca²⁺-dependent AP activity in pre-hearing IHCs. In the absence of SK2 channels, immature IHCs expressed fewer Ca²⁺ channels although the Ca²⁺ dependence of exocytosis was normal. One possible explanation for this finding is the presence of fewer afferent synaptic contacts in immature mutant IHCs. Despite having a normal I_Ca, adult IHCs from SK2 mutant mice exhibited an altered (higher-order) Ca²⁺ dependence of exocytosis. Since SK2 channels can only have a functional influence during immature stages of development and are not directly involved in neurotransmitter release, our results suggest that the disrupted spontaneous APs in mutants could affect the normal maturation of the exocytotic machinery. The results presented here give the first indication of a possible direct link between the spontaneous firing activity of pre-hearing IHCs and their normal intrinsic development.