Are Ca_v1.3 pacemaker channels in chromaffin cells?

Abstract

Mouse and rat chromaffin cells (MCCs, RCCs) fire spontaneously at rest and their activity is mainly supported by the two L-type Ca²⁺ channels expressed in these cells (Ca_v1.2 and Ca_v1.3). Using Ca_v1.3^−/− KO MCCs we have shown that Ca_v1.3 possess all the prerequisites for carrying subthreshold currents that sustain low frequency cell firing near resting (0.5 to 2 Hz at −50 mV):¹ low-threshold and steep voltage dependence of activation, slow and incomplete inactivation during pulses of several hundreds of milliseconds. Ca_v1.2 contributes also to pacemaking MCCs and possibly even Na⁺ channels may participate in the firing of a small percentage of cells. We now show that at potentials near resting (−50 mV), Ca_v1.3 carries equal amounts of Ca²⁺ current to Ca_v1.2 but activates at 9 mV more negative potentials. MCCs express only TTX-sensitive Na_v1 channels that activate at 24 mV more positive potentials than Ca_v1.3 and are fully inactivating. Their blockade prevents the firing only in a small percentage of cells (13%). This suggests that the order of importance with regard to pacemaking MCCs is: Ca_v1.3, Ca_v1.2 and Na_v1. The above conclusions, however, rely on the proper use of DHPs, whose blocking potency is strongly holding potential dependent. We also show that small increases of KCl concentration steadily depolarize the MCCs causing abnormally increased firing frequencies, lowered and broadened AP waveforms and an increased facility of switching “non-firing” into “firing” cells that may lead to erroneous conclusions about the role of Ca_v1.3 and Ca_v1.2 as pacemaker channels in MCCs.²

Introduction

L-type calcium channels (LTCCs, Ca_v1) contribute to the pacemaker current of central neurons,^3–6 neuroendocrine^1,7,8 and cardiac sino-atrial node cells⁹ generating spontaneous action potentials of low frequency (0.5 to 2 Hz) (reviewed in refs. ¹⁰ and ¹¹). Between the two LTCCs expressed in these tissues (Ca_v1.2, Ca_v1.3), Ca_v1.3 appears the most suitable isoform for pacemaking in excitable cells: (1) it activates at relatively more negative potentials^12–15 and (2) has faster activation but slower and less complete voltage-dependent inactivation.^13,14 Using Ca_v1.3^−/− KO mice,¹² we found that these peculiarities are well preserved in mouse adrenal chromaffin cells and that Ca_v1.3 and Ca_v1.2 contribute equally to the total Ca²⁺ current (∼25% each).¹ Given this, it appears evident that near resting conditions (−50 mV), Ca_v1.3 is privileged for controlling AP firings with interspike intervals of 0.5 to 2 s (2-0.5 Hz). As suggested by the blocking action of nifedipine on Ca_v1.3^−/− MCCs, we also concluded that Ca_v1.2 also contributes to pacemaking, but at a lower degree given its more positive voltage range of activation and faster inactivation. A small contribution to the pacemaker current is expected also from TTX-sensitive Na⁺ channels, which can account for a small fraction of Ca_v1.3^−/− MCCs that preserves their ability to fire in the presence of nifedipine, i.e., when Ca_v1.2 is pharmacologically blocked and Ca_v1.3 is absent.¹

In apparent contrast with us, a recent paper by Pérez-Alvarez et al.² reports that Ca_v1.2 is likely to be the dominant pacemaking channel in MCCs and that Ca_v1.3 contributes only marginally to the spontaneous activity. The conclusions are based on experiments in WT and Ca_v1.3^−/− KO-MCCs bathed in solutions with higher KCl concentration than ours (5.5 vs. 4 mM) which steadily depolarize the cells by ∼6 mV. Moreover, the authors evaluated the contribution of Ca_v1.2 and Ca_v1.3 by the block of DHPs at Vh = −80 mV assuming that the same is valid at −40 mV, without taking into consideration that block of LTCCs by DHPs is strongly holding potential-dependent.^13,15–17 We show here that the findings by Pérez-Alvarez et al.² are not in contrast with ours if properly corrected for the different KCl concentration and if DHP blocking potency is suitably tested at the interspike potential of spontaneous firing (à50 mV in 4 mM KCl). Thus, with the proper corrections their data confirm our previous proposal that, near resting physiological conditions the order of importance for pacemaking is: Ca_v1.3, Ca_v1.2 and Na_v1 channels.

To clarify these critical points, we show here that: (1) full block of Ca_v1.3 in MCCs requires high doses of nifedipine at V_h = −80 mV (∼30 µM) and 10-fold lower concentrations at V_h = −50 mV, making this channel very sensitive to DHPs near resting conditions, (2) the predominance of Ca_v1.3 in pacemaking derives from its steep voltage range of activation which is 9 and 24 mV more negative than Ca_v1.2 and Na_v1 channels, respectively, (3) MCCs express only fast and fully inactivating TTX-sensitive Na_v1 channels that can contribute little to pacemaker currents lasting 0.5 to 2 s. Saturating doses of TTX block the firing in a small fraction of MCCs (13%), (4) steady depolarizations of ∼6 mV using higher KCl concentrations or few pA of current injection cause a marked reduction and broadening of AP waveform, a 3-fold increase of firing frequency and an increased capability of turning “non-firing” into “firing” cells. These conditions favor the contribution of fast inactivating channels, which activate at more positive potentials (Ca_v1.2 and Na_v1 channels) and introduce serious bias to the true estimate of Ca_v1.3 role in pacemaking MCCs.

Results

Ca_v1.3 and Ca_v1.2 contribute equally to the total Ca²⁺ current in WT MCCs.

MCCs express both Ca_v1.2 and Ca_v1.3 and the two channels together contribute about 50% of the total Ca²⁺ currents.^1,8 It is thus critical to establish the percentage of current carried by each of the two Ca_v1 channels at the potentials where cells spontaneously fire (−50 mV). We have previously shown that WT and Ca_v1.3^−/− express equal Ca²⁺ current amplitudes (Fig. 1A top-right inset) and that the block of LTCCs by nifedipine at V_h −50 mV saturates at about 3 µM to give maximal inhibitions of 52.8% (WT) and 22.4% (Ca_v1.3^−/−) (Fig. 1A bottom-right inset). Given the importance of this issue, we show here two examples of L-type current block with increasing doses of nifedipine in WT and KO MCCs held at −50 mV (Fig. 1A). Since WT currents are carried by Ca_v1.3 and Ca_v1.2, and KO currents by Ca_v1.2 alone, it is evident that the DHP effectively blocks both Ca_v1.2 and Ca_v1.3 and that Ca_v1.3 contributes to the total L-type current equally to Ca_v1.2. The same is not valid when V_h is lowered to −80 mV, where LTCCs sensitivity to DHPs decreases,^16,17 requiring higher doses of nifedipine to fully block the channels. Figure 1B shows that after blocking N-, P/Q- and R-type channels with ω-CTx-GVIA (3.2 µM), ω-CTx-MVIIC (10 µM) and SNX-482 (0.4 µM), full block of the remaining LTCCs requires nifedipine concentrations >10 µM. Comparison of the dose response curve of nifedipine action on LTCCs at V_h −80, −50 and −40 mV shows that nifedipine concentrations that mediate complete block of L-type currents at −50 mV or −40 mV can only block a small fraction of current at −80 mV (Fig. 1C). This suggests that using DHP concentrations tested on whole cell Ca²⁺ currents at −80 mV is not an appropriate basis for arguing on their action on AP firing at interspike potentials near −50 mV or −40 mV, as in the case of reference ². This procedure would lead to an obvious underestimation of Ca_v1.3 with respect to Ca_v1.2.

Figure 1.

Ca²⁺ current block by nifedipine at different holding potentials (V_h) in WT and Ca_v1.3^−/− MCCs. (A) Time course of normalized peak Ca²⁺ currents recorded from a WT (open squares) and a Ca_v1.3^−/− MCC (filled squares) before, during and after sequential applications of 0.3, 1 and 3 µM nifedipine. Step depolarization to +10 mV from V_h = −50 mV was applied every 10 s. The insets to the bottom show the current traces recorded at the time indicated by the letters. To the top-right are shown the mean Ca²⁺ current amplitudes at +10 mV of WT (n = 29) and KO-MCCs (n = 26).¹ To the bottom-right are given the percentage of Ca²⁺ current blocked by 3 µM nifedipine in WT (mean 52.8%; n = 7) and KO MCCs (mean 22.4%; n = 5) (see also ref. 1). (B) Time course of peak Ca²⁺ currents at +10 mV from V_h = −80 mV recorded from a WT MCC during sequential applications of 3.2 µM ω-CTx-GVIA + 10 µM ω-CTx-MVII C (GVIA + MVII) and 1, 3, 10 and 30 µM nifedipine. SNX-482 (0.4 µM) was constantly applied throughout the experiment to block Ca_v2.3 channels. The inset shows the current traces recorded at the time indicated by the letters. Notice that nearly full block is obtained only after applying 30 µM nifedipine. (C) Blocking potency of nifedipine Ca²⁺ currents recorded at V_h = −80 mV and −50 mV in WT MCCs. The dose-response curve at V_h = −50 mV is taken from (C) in reference ¹ (IC₅₀ = 0.21 µM; Hill slope 1.2) while the curve at V_h = −80 mV was obtained from ten WT-MCCs (IC₅₀ = 1.31 µM; Hill slope 1.2). Cells were pre-treated with ω-CTx-GVIA (3.2 µM), ω-CTx-MVII C (10 µM) to block N and P/Q-type channels as in (B). SNX-482 (0.4 µM) was added to all solutions containing nifedipine to block also Ca_v2.3 channels. Step depolarizations were to +10 mV for 20 ms. The dashed curve to the left is an hypothetical dose-response curve at −40 mV drawn just to underline the increased blocking potency of nifedipine at this V_h.

Ca_v1.3 activates at more negative potentials than Ca_v1.2 and Na_v1 channels.

Given that Ca_v1.3 and Ca_v1.2 contribute equally to L-type currents, our next issue was to determine the true voltage-dependence of Ca_v1.3 activation using the conductance vs. voltage curves of LTCCs in WT and KO MCCs previously determined (Fig. 2B in ref. 1). We assumed that the WT conductance curve is a linear combination of the Ca_v1.3 and Ca_v1.2 curves with equal weight (50%) and that the KO curve is representative of Ca_v1.2. The resulting Ca_v1.3 curve is shown in Figure 2A (dashed line) and appears shifted by about −9 mV from Ca_v1.2 curve (see legend). This finding together with the observation that Ca_v1.3 and Ca_v1.2 contribute equally to the total Ca²⁺ current and that Ca_v1.3 inactivates slowly and incompletely during long depolarizations suggest that Ca_v1.3 outweighs Ca_v1.2 in pacemaking MCCs.

Figure 2.

Voltage-dependence of Ca_v1.3, Ca_v1.2 and Na_v1 channel conductance derived from WT and Ca_v1.3^−/− KO-MCCs. The two solid curves with no data points represent the voltage-dependence of WT and KO MCCs Ca²⁺ conductance taken from reference ¹. The dashed curve to the left is the voltage-dependence of Ca_v1.3 channels obtained by assuming that the WT curve is a linear combination of Ca_v1.3 and Ca_v1.2 channel conductance with equal contribution from the two channel isoforms (WT = 0.5 Ca_v1.3 + 0.5 Ca_v1.2) and that the KO curve is representative of Ca_v1.2. The Ca_v1.3 curve is a Boltzmann function with V_1/2 = −26.8 mV and slope factor k = 6.1 mV. The Ca_v1.2 (KO) curve has V_1/2 = −18.1 mV (ref. ¹) and is shifted about 9 mV from Ca_v1.3. The voltage dependence of Na_v1 channels was obtained by fitting the mean values of the conductance vs. voltage obtained from families of Na_v1 current recordings shown in (B). The Na_v1 conductance curve was calculated as I_peak/(V − V_rev) with V_rev = +70 mV and normalizing to the maximum value at +10 mV (n = 12). The solid curve is a Boltzmann curve with V_1/2 = −2.9 mV and slope factor k = 7.7 mV. (B) Overlapped TTX-sensitive Na⁺ currents recorded from a WT MCC after subtraction of the currents remaining after adding TTX (0.3 µM). K⁺ and Ca²⁺ channels were blocked by adding 200 µM Cd²⁺ to the bath solution and replacing K⁺ with Cs⁺ inside the patch pipette. The sequential depolarizations were on step of 10 mV from −40 to +70 mV. They were preceded by a 50 ms depolarization to −50 mV from V_h = −70 mV to partially inactivate the available Na⁺ channels and attenuate the capacitative artifact. The inset shows a Na⁺ current trace at +10 mV before and after addition of TT X (0.3 µM). Sustained (non-inactivating) Na⁺ currents were never observed during pulses of 100 ms duration (n = 12). (C) AP recordings from a WT MCC whose spontaneous activity is fully blocked by TTX (0.3 µM). Notice the slight hyperpolarization and full block of APs during TTX application and the complete recovery after wash out. (D) AP recordings from a WT MCC whose spontaneous activity is altered, but not blocked, by TTX (0.3 µM). The toxin depolarizes by ∼5 mV the firing cell and changes the shape of APs into small bursts of 3 to 5 spikes. This may be explained by a decreased number of active BK channels (due to the reduced Na⁺-dependent upstroke), accumulation of Ca²⁺ during the burst and activation of SK channels that terminate the burst. The insets show single or trains of APs recorded at the time indicated on a faster time scale.

Once we established the voltage-dependence of Ca_v1.3 and Ca_v1.2 we next studied the gating properties and the range of activation of Na_v1 channels that support the upstroke of APs and may contribute to the pacemaker current in MCCs. We found that MCCs express mainly Na⁺ TTX-sensitive channels that activate at ∼24 mV more positive potentials than Ca_v1.3 (Fig. 2A) and inactivate fully within tens of milliseconds at all potentials (maximal peak current −635 ± 46 pA at +10 mV, n = 12; Fig. 2B). In twelve MCCs we found no evidence of slowly inactivating “persistent” Na⁺ channels, which are able to support neuronal pacemaker currents.¹⁸ This confirms that Na_v1 channels contribute little to the pacemaker current in MCCs (Fig. 3C in ref. 11). Interestingly, when applied to spontaneously firing cells, saturating doses of TTX (300 nM) caused full block only in two out of 17 cells (13%; Fig. 2A). In the remaining 15 cells (87%), TTX caused a 5–6 mV depolarization followed by a persistent firing of smaller APs, often characterized by unusually rapid bursts of three to five spikes (Fig. 2D and legend). In conclusion, fast inactivating Na_v1 channels activate at much more positive potentials than Ca_v1.2 (+15 mV) and Ca_v1.3 (+24 mV) and contribute little to pacemaking MCCs near resting physiological conditions.

Figure 3.

Higher KCl increases the firing frequency, lowers and broadens the AP and turns into “firing” a “non firing” MCCs. (A) Spontaneous firing of a WT MCC in control solution (4 mM KCl) and during a transient increase of KCl to 5.5 mM. Notice the sizeable cell depolarization causing the increased frequency and change of AP shape (lowered spike height and broadened half-width). The insets show single or trains of APs recorded at the time indicated on a faster time scale. (B) A “non-firing” WT cell bathed in control solution (4 mM KCl) generates burst firings when exposed to 5.5 mM KCl. The increased KCl depolarizes the cell by ∼7 mV and favor AP firings. (C) Mean values (±SEM) of firing frequency, spike height and half-width from MCCs in control conditions (4 KCl, empty bar; n = 15) and after applying 5.5 mM KCl (n = 7) or injecting 1–3 pA of current to depolarize the cells by ∼6 mV (n = 8) (5.5 KCl; filled bar). Mean interspike potential was −44.9 mV (4 KCl) and −38.4 mV (5.5 KCl + depo).

Higher KCl solutions or steady depolarizations increase the firing frequency and lowers the AP size.

To assay the true role of Ca_v1.3 on spontaneous firing, it is important that the cells are maintained near resting physiological conditions, avoiding unnecessary biased potentials which may alter the firing frequency and AP waveform. Either a depolarization induced by 1–3 pA of current injection or bathing the cell in higher than normal KCl solutions changes cell firing and AP waveforms. For instance, ∼6 mV cell depolarization induced by passing 1–3 pA or raising external KCl from 4 to 5.5 mM produces a marked increase of firing frequency, marked reduction of the AP amplitude and pronounced broadening of the half-width (Fig. 3A). In 15 cells (eight depolarized by passing current and seven by raising KCl from 4 to 5.5 mM) we found that AP frequency increased three times (from 0.8 ± 0.1 to 2.4 ± 0.2 Hz; p < 0.001), the height decreased by ∼20% (from 83 ± 3 to 69 ± 4 mV; p < 0.01) and the half-width increased by ∼40% (from 2.4 ± 0.1 to 3.4 ± 0.2 ms; p < 0.01) (Fig. 3C). These altered parameters are comparable to those reported by reference ² and uncover the different experimental conditions in which this set of data is collected.

Figure 3A also suggests that steadily depolarized MCCs, with higher firing frequencies (3–4 Hz), overrate the role of channels with faster inactivation and higher threshold of activation (Ca_v1.2 and Na_v1) with respect to channels like Ca_v1.3 that activate at more negative potentials and inactivate slowly. In addition, more positive resting potentials overestimate the percentage of “firing” versus “non-firing” cells, which is a critical parameter to determine whether deletion of Ca_v1.3 affects MCCs firing. Figure 3B shows an example out of seven in which a “nonfiring” cell in 4 mM KCl is turned into “firing” by simply raising KCl to 5.5 mM. Under these conditions, more positive resting potentials favor the firing of Ca_v1.3^−/− cells, which express only Ca_v1.2 and possess smaller pacemaker currents than WT MCCs.¹ This may explain the larger number of firing Ca_v1.3^−/− KO-MCCs estimated by reference ² (57%) with respect to us (30%)1 and suggests that most of the different results can be simply explained by methodological differences rather than postulating pathological states of KO mice (see Discussion in ref. 2).

Discussion and Conclusions

The present study extends the work of Marcantoni et al.¹ concerning the role of Ca_v1.3 as pacemaker channel in MCCs. Here we establish unequivocally that Ca_v1.3 contributes to about half of the L-type current which sustains AP firings near −50 mV and that Ca_v1.3 activates at 9 mV and 24 mV more negative potentials than Ca_v1.2 and Na_v1 channels. Since TTX-sensitive Na_v1 channels are fully inactivating within tens of milliseconds and Ca_v1.2 inactivates faster and more completely than Ca_v1.3,^13,15 it is evident that Ca_v1.3 is the most suitable channel for carrying most of the subthreshold pacemaker currents in slow firing MCCs. This does not exclude the possibility that Ca_v1.2 also contributes to pacemaking these cells. Indeed, KO MCCs lacking Ca_v1.3 are still able to fire and nifedipine is able to reduce or block the spontaneous firing in most, but not all the KO MCCs.¹ This also suggests that besides Ca_v1.3 and Ca_v1.2, Na_v1 or other ion channels carrying subthreshold inward currents may also contribute to the pacemaker current in a small fraction of cells. Figure 2C clearly shows that this occurs in 13% of cells in which TTX causes a slight hyperpolarization and a block of MCCs firing.

The current study gives also a partial explanation to the contrasting results reported by Pérez-Alvarez et al.² regarding the marginal role of Ca_v1.3 as pacemaker channel in MCCs. We showed that this might derive from two anomalous methodological approaches. The first one is the improper use of DHPs that, when tested at V_h −80 mV, underestimate the true contribution of Ca_v1.3 channels near resting potentials (−50 mV). DHPs have very lowaffinity for Ca_v1.3 at very negative holding potentials.^13–16 Regarding this issue, the use of 0.3 µM nifedipine to identify the contribution of Ca_v1.2 on AP firing, as proposed by reference ², is unjustified because this DHP concentration blocks equal fractions of Ca²⁺ currents in WT and KO-MCCs at −50 mV (see the dose-response curves of Fig. 1C in ref. 1). If 0.3 µM nifedipine inhibits equally Ca_v1.2 and Ca_v1.3 a similar block potency of WT and KO firing cells by this DHP concentration cannot be taken as proof that Ca_v1.2 is the only LTCC pacemaking MCCs. The second one is related to the use of KCl solutions, which depolarize the cell by 5–6 mV, increases the firing and changes the AP shape with marked alteration of cell functioning. As shown in Figure 3, the fraction of “firing” vs. “non-firing” cells can be significantly increased at more depolarized resting potentials. We also cannot exclude that cell firing at higher frequencies (2–3 Hz or 5 Hz as in the case of ref. 2) will increase Ca²⁺ entry and stimulate Ca²⁺-dependent pathways that upregulate LTTCs openings^19,20 and further contributes to switch “non-firing” into “firing” cells.

In conclusion, to answer the question raised in the title, Ca_v1.3 is unequivocally a pacemaker channel in MCCs, not secondary to Ca_v1.2. Most likely its role is not limited to chromaffin cells, dopaminergic neurons⁶ and cardiac sinoatrial node cells.⁹ Ca_v1.3 may even soon result a potential pacemaker channel in other neurons and neuroendocrine cells (reviewed in ref. 10). We also recall that Ca_v1.3 is particularly coupled to fast inactivating BK channels¹ and drives a considerable fraction of SK channels at resting potentials (Vandael DHF, Marcantoni A and Carbone E, unpublished results). Thus, their loss in Ca_v1.3^−/− MCCs is not expected to necessarily reduce the firing frequency as postulated by reference ², but depending on their coupling to Ca²⁺-activated K⁺ channels^21,22 it could paradoxically increase it.¹

Finally, we would underline the new frontiers on Ca_v1.3 gating properties associated with the existence of the newly identified short splice variant Ca_v1.3_42A,²³ whose degree of expression and role in different tissues still needs to be clarified. An important clarification is still expected about the role of Ca_v1.2 and Ca_v1.3 on catecholamine secretion and vesicles retrieval in which LTCCs are shown to play a critical role (reviewed in ref. 11). Recent data suggest that Ca_v1.3 contributes more to the exocytosis at low membrane potentials,^2,11 in a way resembling T-type Ca_v3.2 channels.²⁴

Materials and Methods

Isolation and culture of WT and Ca_v1.3^−/− mouse adrenal medulla chromaffin cells.

All experiments were performed in accordance with the guidelines established by the National Council on Animal Care and were approved by the local Animal Care Committee of Turin University. Chromaffin cells were obtained from young (13 months) male C57BL/6N mice. Like for our previous work (ref. ¹), Ca_v1.3^−/− mice¹² were obtained from the animal house of the Eberhard Karls Universität Tübingen (Germany) which breeds animals under SPF conditions and provided us with the respective health certificates that are on file. Animals were killed by cervical dislocation and chromaffin cells were isolated and cultured following the method described previously.⁸

Voltage-clamp and current-clamp recordings and solutions.

Current-clamp and voltage-clamp recordings were made in perforated-patch conditions using an Axopatch 200-A amplifier and pClamp 10.0 software programs (Axon Instruments Inc., USA). Details on patch pipette fabrication, perforated-patch conditions, current and voltage recordings, filtering and capacitative transient cancellation are given elsewhere in references 8 and 20. The indicated voltages were not corrected for the liquid junction potential. For Ca²⁺ and Na⁺ currents the pipette contained (mM): 135 CsMeSO₃, 8 NaCl, 2 MgCl₂, 20 HEPES, pH 7.3 (CsOH) plus amphotericin B. For Ca²⁺ currents the external bath contained (mM): 135 TEACl, 2 CaCl₂, 2 MgCl₂, 10 glucose, 10 HEPES, pH 7.4 (with CsOH) and for Na⁺ currents (mM): 104 NaCl, 30 TEACl, 4 KCl, 2 CaCl₂, 2 MgCl₂, 10 glucose, 10 HEPES, pH 7.4 (NaOH) plus 200 µM CdCl₂. Action potentials (APs) were recorded in current-clamp mode at resting conditions without injecting any current. The intracellular solution contained (mM): 135 KAsp, 8 NaCl, 20 HEPES, 2 MgCl₂, 5 EGTA pH 7.3 (KOH) and the external bath contained (mM): 137 NaCl, 4 KCl (or 5.5 KCl), 2 CaCl₂, 2 MgCl₂, 10 glucose, 10 HEPES, pH 7.4 (NaOH). The cell perfusion system, all the drugs and toxins were purchased and used as described in references 1 and 24. All the experiments were performed at room temperature (22–24°C). Data analysis and statistical significance were carried out as described previously.¹

Addendum to: Marcantoni A, Vandael DHF, Mahapatra S, Carabelli V, Sinneger-Brauns MJ, Striessnig J, et al. Loss of Cav1.3 channels reveals the critical role of L-type and BK channel coupling in pacemaking mouse adrenal chromaffin cells. J Neurosci. 2010;30:491–504. doi: 10.1523/JNEUROSCI.496.