The methods were very similar to the ones used before (Levic et al., 2007, 2011).

To investigate the presence of I_K(Ca) in developing longitudinal middle section (midsection) chick, THC whole-cell currents were elicited by 250 ms depolarizing voltage steps in 10-mV increments from a holding potential of −90 mV. Currents were recorded without Ca²⁺ added to the bath solution (Figure 1A, top left panel) and after the addition of 2 mM [Ca²⁺]_ext, which produced an immediate increase in outward current (Figure 1A, top middle panel). The addition of 300 nM apamin abolished the effect of Ca²⁺ on whole-cell current (Figure 1A, top left panel). Indeed, the profile of I_K(Ca) was comparable with apamin-sensitive currents (Figure 1A, bottom panels). The plot of the current–voltage relationship (I–V) for before and after the addition of 2 mM Ca and apamin is shown in the upper panel of Figure 1B, and the I–V plot comparing Ca²⁺ and apamin-sensitive components is shown in the lower panel of Figure 1B (lower panel) for E10 cells (n = 15). While the SK currents are exclusively sensitive to intracellular Ca²⁺ availability (Martin and Fuchs, 1992; Kennedy, 2002), increased depolarization could lead to activation of voltage-sensitive potassium currents present in chick HCs (Levic et al., 2011) and accumulation of intracellular Ca²⁺, thereby coupling intracellular Ca²⁺ levels and membrane potential and producing the observed increase in current amplitude (Yamashita et al., 1990).

The average peak current density doubled in the presence of Ca²⁺ [Figure 1B, measured in nominally Ca²⁺-free solution at + 50 mV had an amplitude of 96 ± 16 pA/pF, which increased to 210 ± 24 pA/pF in the presence of 2 mM (Ca²⁺)_ext]. This current was abolished in the presence of 300 nM apamin (reducing to 89 ± 15 pA/pF). Similar experiments were performed over the developmental range covering the period after HC terminal differentiation (E8) until maturation (P2) (E8, n = 15; E10, n = 15; E12, n = 23; E16, n = 36; E18, n = 9; P2, n = 10). The similarity in Ca²⁺ and apamin-sensitive currents could be generalized in all cells between E8 and E16. The apamin-sensitive component showed developmental upregulation [measured in the% of total current density at + 10 mV, apamin-sensitive component was (in% of total): E7: 32 ± 15; E10: 50 ± 20; E12: 53 ± 21; E16: 42 ± 18; E18:10 ± 3; P2: 2 ± 1]. The currents were compared at + 10 mV step potentials, as this is the peak potential often observed during spontaneous action potentials in developing chick HCs (Levic et al., 2007, 2011). An ANOVA was used to show significant differences in the presence of the SK current across different developmental ages, F (5, 103) = 362.6, p < 0.001. The post hoc Tukey test showed that there is a significant difference between all ages (p < 0.001), except between E8 and E16 (p = 0.6), E10 and E12 (p = 0.6), and E18 and P2 (p = 0.9). Lower concentrations of apamin did not produce an effect on whole-cell currents (Figure 1D, n = 10). This suggested that of the three identified SK channel isomers, only the SK2 channels are present, as the SK2 channels are sensitive to apamin concentrations in the hundreds of nM range (Figure 1, Marcotti et al., 2004; Stocker, 2004). Together, these results indicate the developmental upregulation of I_K(Ca) which is likely to be of the SK2 type in developing chick HCs.

Next, the role of the ISK in spontaneous electrical activity was examined. As reported previously (Levic et al., 2007), action potentials are driven by Ca²⁺. Spontaneous activity is not possible in the absence of external Ca²⁺ (Figure 2A). However, raising the external Ca²⁺ [(Ca²⁺)_ext] to 2 mM initiated spontaneous action potentials (Figure 2A). Subsequent perfusion with 300 nM apamin reduced this Ca²⁺-dependent electrical activity (Figure 2A). Washing out of apamin for about 5 min allowed for the spontaneous action potentials to resume (Figure 2A right) indicating that the observed phenomenon is due to SK current block by apamin. The interspike interval (ISI) distribution histogram for spontaneous action potentials before and after the addition of apamin indicates that the ISI time is prolonged with the addition of apamin until eventually all activities cease (Figure 2B). The inhibition of spontaneous action potentials results from diminished repolarization after the upstroke of action potentials, leading to depolarization of cells, which is consistent with observations in developing mammalian IHCs (Marcotti et al., 2004). Together, these results suggested that SK currents determine the patterning of spontaneous action potentials in developing THCs.

The expression of SK currents coincides with the upregulation of T-type VGCC in developing HCs (Levic et al., 2007). Thus, any contribution of T-type VGCC on the SK current activation was then assessed. Kurtoxin, a T-type VGCC-specific blocker (Levic et al., 2007), was used to determine the sensitivity of the SK currents to T-type VGCC. The currents were elicited as described in Figure 1 and measured before (Figure 3A left panel) and after the addition of kurtoxin (Figure 3A middle panel), and subsequent addition of apamin (Figure 3A right panel) was used to further block the remaining SK currents. The current density–voltage relationship for this cell is shown in Figure 3B. Measured at +10 mV, currents showed very little sensitivity to kurtoxin compared with apamin (Figure 3C). The kurtoxin-sensitive component of the current also decreased with development (measured at +10 mV,% kurtoxin-sensitive current: E10: 10 ± 3; E12: 12 ± 5; E16: 5 ± 2; E18: 1 ± 1). Thus, T-type currents contribute little to activating the ISK, suggesting the possibility that these channels are loosely coupled and most likely spatially distant.

While the dominant VGCC in developing chick HCs is of T-type (Levic et al., 2007), the experiments were also conducted to investigate the contributions of L-type VGCC, which are dominant channels in mature chick HCs (Fuchs et al., 1990). Nifedipine, as an established blocker of L-type VGCC in chick (Fuchs et al., 1990; Spassova et al., 2001; Levic and Dulon, 2012) and mouse HCs (Marcotti et al., 2003b), was used to test the contributions of L-type VGCC to SK current activation. The currents were elicited as described in Figure 1 and measured before (Figure 4A left panel) and after the addition of nifedipine (Figure 4A middle panel), and subsequent addition of apamin (Figure 4A right panel) was used to further block any remaining SK currents. While it is evident that nifedipine completely eliminates the apamin-sensitive component, it is worth noting that it also affects the delayed rectifier channels (Valmier et al., 1991; Fagni et al., 1994; Zhang and Gold, 2009; Levic et al., 2011). However, using TEA to eliminate delayed rectifier channels also affected the ISK (data not shown, Lang and Ritchie, 1990; Bond et al., 1999; Blanchet and Dulon, 2001; Dilly et al., 2013). The remaining currents were completely inhibited by 4AP (data not shown, Levic et al., 2011). The average current density–voltage relationship for E12 cells is shown in Figure 4B (n = 11). In addition to blocking L-type VGCC, nifedipine (and nimodipine, data not shown) also had non-specific effects on voltage-gated potassium currents. Thus, using these pharmacological manipulations did not help in determining the contribution of L-type VGCC on the activation of ISK. However, the results allow the conclusion that the L-type VGCC has a profound effect on the activation of ISK current. The same effect was observed across all developmental stages.

Since nifedipine and nimodipine (not shown) affected not only L-type VGCC but also delayed rectifier potassium currents, further experiments were conducted to isolate the role of L-type VGCC on ISK activation. Bay K8644, an agonist of L-type VGCC (Fuchs et al., 1990), was used to directly test the effects of L-type VGCC on the ISK current activation (Figure 5). These experiments involved first determining the amount of the ISK in an individual cell by applying apamin after stimulating ISK by 2 mM Ca²⁺ (Figure 5A top row). After this test, apamin was washed off, and Bay K8644 was applied to measure any further increase in ISK, which was subsequently confirmed by applying apamin to test if the Bay K8644-elicited current were ISK, based on its sensitivity to apamin block (Figure 5A bottom row). Bay K8644 elicited a further increase of 15% in ISK current amplitude as measured at +10 mV in E12 THCs [mean current density (pA/PF): in 2 mM Ca²⁺ 114.8 ± 9.6, after the application of Bay K8644 130.2 ± 10.4]. The experiments revealed that the current elicited by Bay K8644 was indeed fully sensitive to apamin as illustrated in Figure 5A (bottom row), as well as in current density plots (Figure 5B). Therefore, L-type currents contribute substantially to ISK activation, suggesting that these channels are functionally (and spatially) tightly coupled (Zhang et al., 2018).

The next step was to evaluate the sensitivity of ISK to intracellular Ca²⁺availability (Figure 6). The two well-established Ca²⁺ chelators are BAPTA and EGTA, with BAPTA being ∼150 times faster in binding Ca²⁺ than EGTA (Naraghi and Neher, 1997). Cellular processes that show sensitivity to both BAPTA and EGTA are thought to be within Ca²⁺ microdomains (from ∼50nm to several hundred nm). However, processes sensitive only to BAPTA are localized within “Ca²⁺nanodomains” (within ∼20–50 nm of the Ca²⁺ source) (Naraghi and Neher, 1997; Yu et al., 2014). Thus, a comparison of the effects of BAPTA and EGTA on the activation of SK currents was carried out. Examples of currents recorded before and after Ca²⁺ and subsequently apamin using the intracellular solution containing 1 mM EGTA are shown in Figure 5A and using 10 mM BAPTA-AM are shown in Figure 6B. Ca²⁺ and apamin-sensitive components were similar under all conditions; using BAPTA-AM significantly reduced the ISK (Figure 6C). Measured at + 10 mV at E12, apamin-sensitive components were (% total current): 1 mM EGTA, 53 ± 21; 10 mM EGTA, 36 ± 18; 1 mM BAPTA-AM, 24 ± 11; 10 mM BAPTA-AM, 8 ± 3 (n = 11 for each concentration). An ANOVA was used to show significant differences in the presence of SK current across different intracellular buffering conditions, F (4, 44) = 327.8, p < 0.001. The post hoc Tukey test showed that there is a significant difference between EGTA and BAPTA buffering conditions (p < 0.001), as well as between 1mM and 10 mM BAPTA (p < 0.001). Thus, intracellular BAPTA buffering significantly reduced the SK currents in a concentration-dependent manner (Figure 6).

These findings suggest nanodomain proximity of SK current to the sources of Ca²⁺, creating the possibility that intracellular Ca²⁺ is an important source of the SK current activation. In the cells, a universally important source of intracellular Ca²⁺ is the sarcoplasmic reticulum (SR), where the sarcoendoplasmic reticulum (SR) Ca²⁺ transport ATPase (SERCA) is a pump that transports Ca²⁺ ions from the cytoplasm into the SR.

Thus, the sensitivity of ISK to SR Ca²⁺ was probed using thapsigargin, a specific inhibitor of SERCA pumps that leads to intracellular store depletion (e.g., Bian et al., 1991). Prolonged pretreatment with 10 μM thapsigargin (> 20 min) effectively abolished 30% of the total outward current in response to depolarizing voltage step to + 10 mV in E12 THCs (2 mM Ca²⁺ 85.4 ± 8.7, thapsigargin 59.7 ± 5.3). However, the subsequent application of L-type VGCC agonist Bay K8644 increased the magnitude of ISK by 10%, as measured at +10 mV in E12 THCs [mean current density (pA/PF): in the presence of thapsigargin 59.7 ± 5.3; Bay K8644 65.7 ± 7.8] (Figure 5). Thus, these results suggest the importance of intracellular Ca²⁺ availability from SR stores in activating the ISK (Figure 7). The results also demonstrate that Ca²⁺ influx through LTCCs alone is not sufficient to activate SK currents and that SR Ca²⁺ release is important in activating ISK.

Spontaneous action potentials initiated by the T-type VGCC are hallmarks of developing and regenerating chick HCs (Levic et al., 2007). The functional changes seen in regeneration seem to share similarity with those seen in development. Thus, the functional presence of SK currents was also assessed in regenerating midsection THCs between 7 and 40 days posttreatment (PT) after inducing HC damage with a single treatment of gentamicin, which causes the loss of HCs, as confirmed with scanning electron microscopy (data not shown, Stone and Rubel, 2000; Levic et al., 2007). During the experiments, it was possible to visually identify the HCs that were newly regenerated, as the cellular appearance markedly differed from the HCs normally seen in the untreated animals. One of the main features was the very immature stereocilia which are starting to form on the apical side. In addition, the cells were tested for the presence of spontaneous action potentials in the whole-cell patch-clamp configuration, which has been identified as one of the markers for regenerating HCs. Also, adjacent cells of similar appearance were tested for the presence of T-type VGCC, another marker for regenerating cells (Levic et al., 2007).

During regeneration, spontaneous electrical activity also showed sensitivity to apamin (Figure 8A, left panel). ISK currents also contribute to the patterning of spontaneous action potentials, as indicated by prolonged ISI (Figure 8A, right). Together, these findings indicate that SK current reappeared again during the process of regeneration. SK currents recorded at PT7 from midsection THCs are shown in Figure 8B. Currents were elicited by 250 ms depolarizing voltage steps in 10-mV increments from a holding potential of −90 mV. Traces are shown before (left) and after (middle) the application of 300 nM apamin. The current density–voltage relationship for the control, apamin, and apamin-sensitive current is shown in Figure 8B. Moreover, since T-type VGCC are also present in regenerating HCs (Levic et al., 2007), their contribution to ISK was determined as shown in Figures 3, 8C. The mean current density plots of these currents are shown in Figure 8D.

Comparing the percentage of apamin- and kurtoxin-sensitive current during the process of regeneration revealed that both kurtoxin-sensitive current and the SK current are transiently upregulated between PT7 and PT25 (Figure 8D right). Measured as the% of total current density at + 10 mV, the apamin-sensitive component was (% of total): (PT7: 27 ± 13; PT15: 53 ± 20; PT25: 42 ± 19; PT40: 3 ± 1). An ANOVA was used to show significant differences in the presence of SK current across different posttreatment days, F (4, 21) = 559.3, p < 0.001. The post hoc Tukey test showed that there is a significant difference between all groups tested (p < 0.001). Thus, the SK current transient upregulation is recapitulated during the regeneration of THCs.