This paper characterizes coupling between BK (K_Ca1.1, mSlo1, KCNMA1) and Ca_V1.3 (α1D, CACNA1D) channels. We present evidence for multi-channel complexes that these two channels form. In nerve and muscle, depolarization of the membrane activates voltage-gated calcium channels (Ca_V) leading to calcium influx. This increase in calcium induces the activation of BK channels. The activation of BK channels serves as negative feedback, tending to repolarize the membrane and to close Ca_V channels. Therefore, understanding the functional and physical relation between BK and Ca_V channels is critical to elucidate the biophysical and physiological actions of this feedback mechanism.

BK channels are expressed in nearly every excitable cell and their activation depends on both voltage and intracellular calcium (Cui et al., 2009). Without calcium ([Ca²⁺]_i = 0), BK channels activate only at very depolarized voltages with a V_1/2 >100 mV (Yan and Aldrich, 2010). The affinity of BK channels for calcium ions ranges between 1 and 10 μM (Contreras et al., 2013). Therefore, a rise in calcium concentration from ~100 nM resting levels to the micromolar range is needed to increase the open-channel probability of BK channels at −50 to 0 mV (Barrett et al., 1982).

Opening of a single calcium channel raises local calcium to more than 100 μM within a few tens of nanometers from the inner mouth of the channel, but most of these ions are quickly buffered--within microseconds (Simon and Llinás, 1985; Roberts, 1994). BK channels have been found in close proximity to all subfamilies of Ca_V channels, suggesting a molecular interaction (Roberts et al., 1990; Roberts, 1994; Marrion and Tavalin, 1998; Berkefeld and Fakler, 2008; Müller et al., 2010; Rehak et al., 2013). This association can be recapitulated in heterologous systems and has been shown to reconstitute functional nanodomains (Berkefeld et al., 2006). The physical arrangement and channel stoichiometry of these putative complexes in the nanodomain are still not clear. Some models depict a 1:1 relation between Ca_V and BK channels arranged at distances as close as 10 nm (Fakler and Adelman, 2008; Cox, 2014). Other models are similarly effective in reproducing the functional coupling without a 1:1 stoichiometry (Prakriya and Lingle, 2000; Montefusco et al., 2017). Thus, a comprehensive characterization of Ca_V-BK interactions within functional nanodomains remains to be completed.

The activation threshold of BK channels tends to reflect the activation threshold of associated Ca_V partners. Dihydropyridine-sensitive, L-type Ca_V1.3 channels are unique in their activation profile; they can start activating at voltages as negative as −65 mV (Platzer et al., 2000; Lipscombe et al., 2004). These channels play essential roles in the auditory system (Koschak et al., 2001; Michna et al., 2003; Brandt et al., 2003, 2005), in the pacemaker cells of the heart (Mangoni et al., 2003; Christel et al., 2012; Torrente et al., 2016), in chromaffin cells (Marcantoni et al., 2010), and in neurons (Putzier et al., 2009; Liu et al., 2014; Cooper et al., 2015; Lu et al., 2015; Stanika et al., 2016). Deficiency of Ca_V1.3 channels leads to deafness and bradycardia (Platzer et al., 2000).

Despite the potential relevance for many excitable cells, functional and structural coupling between BK and Ca_V1.3 channels is poorly defined. Here, we tested the hypothesis that coupling between these channels can activate BK channels at very negative voltages. This would promote a negative feedback mechanism whereby BK-channel opening would oppose further opening of Ca_V1.3 channels even before the membrane potential reaches the firing threshold for action potentials. In order to activate BK channels at these low voltages, the Ca_V1.3 and BK channels must be organized in close proximity (<20 nm), so that nanodomains of calcium forming near the inner mouth of the calcium channel overlap with the calcium sensors of the BK channel.

To test the concept of spatial proximity, we (i) compared the activation properties of BK channels when co-expressed with two different Ca_V isoforms, (ii) used classical biophysical and biochemical approaches to test for coupling and direct interaction, and (iii) imaged BK and Ca_V channels with super-resolution microscopy to characterize the spatial distribution of these channels.

Expression of Ca_V1.3 channels shifts the voltage dependence of activation of BK channels to more negative potentials

We tested the hypothesis that BK and Ca_V1.3 channels are functionally coupled. A prediction of this hypothesis is that the voltage dependence of activation of the BK channels should follow the activation curve of the Ca_V1.3 channels (Berkefeld et al., 2006). Therefore, we compared activation of BK channels in a cell line (tsA-201 cells) co-expressing BK with either low-voltage activated Ca_V1.3 channels or high-voltage activated Ca_V2.2 channels. These cells do not express BK or any of the isoforms of Ca_V channels endogenously (Berjukow et al., 1996; Zhu et al., 1998; Avila et al., 2004).

As expected, the inward calcium currents from cells expressing only Ca_V2.2 or Ca_V1.3 (Figure 1A) are blocked by cadmium, a general blocker of Ca_V channels, but they differ in their gating kinetics. Figure 1B shows the voltage dependence of activation gating as conductance-voltage (G-V) curves fitted with Boltzmann functions. The midpoint of activation of the Ca_V1.3 channels is 35 mV more negative than that for the Ca_V2.2 channels. Outward K⁺ currents from cells expressing only BK channels are insignificant in the voltage range that we used (Figure 1—figure supplement 1). When Ca_V channels are co-expressed with the BK channels, the same voltage steps elicit an outward current that is blocked by cadmium (Figure 1C and D). These results confirm that calcium influx through Ca_V isoforms is required for the activation of BK channels in the expression system. As anticipated, the mid-point voltage for activation of BK channels depends on the co-expressed Ca_V channel isoform (Figure 1E). For a quantitative comparison, we use the fitted Boltzmann functions in Figure 1 to interpolate the voltage for 5% activation of the channels. Loosely, we will call this the activation threshold, although activation is actually continuously graded at all voltages. In this terminology, the activation threshold for Ca_V2.2 channels was −13.8 mV and for the co-expressed BK channels it was −20.2 mV, whereas for Ca_V1.3 channels the activation threshold was −48.8 mV and for the co-expressed BK channels it was −52.3 mV. In short, there is quantitative agreement with the hypothesis that the Ca_V channel with the more negative activation voltage allows BK channels to activate at correspondingly more negative potentials. The activation kinetics of BK channels has also been shown to change when coupled with different Ca_V isoforms (Berkefeld and Fakler, 2008). In agreement, the time constant of BK channel activation at 0 mV is shorter when co-expressed with Ca_V1.3 (8 ± 2 ms, n = 7) than when is co-expressed with Ca_V2.2 channels (18 ± 5, n = 9), resembling the difference in activation kinetics between Ca_V1.3 (0.7 ± 0.1 ms, n = 9) and Ca_V2.2 (4.5 ± 0.7 ms, n = 8) channels (Figure 1—figure supplement 2).

Figure 1 with 2 supplements see all

Download asset Open asset

Ca_V1.3 channels are in close proximity to BK channels

The activation of BK channels at −40 mV when co-expressed with Ca_V1.3 channels suggests that these two channels are in close proximity. We tested this hypothesis several ways. As a crude first approximation, we recorded from membrane patches on tsA-201 cells transfected with both BK and Ca_V1.3 channels with the hypothesis that the channels should be close enough to allow us to observe single-channel events. The pipette solution was similar to Ringer’s solution except that the calcium concentration was increased from 2 to 20 mM. A 1 s voltage step from −80 mV to 0 mV was applied to activate Ca_V1.3 channels and to let calcium in only at the membrane patch. If nearby BK channels were present, they would then open. BK channels outside the membrane patch are not expected to open given the decay in calcium concentration. The solution in the bath, outside the membrane patch, contained high potassium to maintain the cytoplasmic potential close to 0 mV, helping to prevent channel openings outside the pipette. To prevent the activation of BK channels inside the patch by calcium coming from Ca_V1.3 channels outside the patch, 2 mM EGTA was added to the bath solution. Indeed, unitary outward K⁺ currents with multiple levels were recorded from these patches (Figure 2A, Figure 2—figure supplement 1 for raw trace), reflecting the activation of Ca_V1.3 channels and subsequent activation of BK channels. As a measure of BK channel activity, we calculated NP_o the product of number of channels times probability of opening. NP_o in these patches was 3.7 ± 0.7 (n = 8, Figure 2B). NP_o decreased to 0.8 ± 0.4 (n = 7, p value = 0.002) after bath application of the dihydropyridine blocker nifedipine (10 μM) suggesting that Ca_V1.3 channels need to be activated to induce these outward current events. Openings were abolished and NP_o decreased to 0.2 ± 0.1 (n = 7, p value = 0.0004) after the sequential application of a BK channel blocker, 500 nM paxilline, confirming the identity of BK channels coupled with Ca_V1.3 channels (Figure 2B).

Figure 2 with 1 supplement see all

Download asset Open asset

Ca_V1.3 and BK channels have been previously shown to co-immunoprecipitate from brain extracts (Grunnet and Kaufmann, 2004), suggesting a direct interaction. However, expression of these channels in a heterologous system might not recapitulate a direct interaction, perhaps requiring additional proteins present uniquely in brain tissue. Thus, we investigated whether, as in neurons, BK and Ca_V1.3 channel are physically associated in tsA-201 cells using co-IP approaches. A comparison between protein extractions from tsA-201 cells transfected with both BK and Ca_V1.3 channels, only one channel (BK or Ca_V1.3), or none, shows that the BK antibody stains the input blot when BK channels are expressed alone or together with Ca_V1.3 channels (Figure 2C, left gel). The BK antibody also stains immunoprecipitates pulled down with Ca_V1.3 antibodies (n = 3, Figure 2C, right gel), suggesting that some level of interaction between these channels is reconstituted when expressed in a heterologous system and that no other proteins specific to brain tissue are required for this interaction.

A third biophysical approach tested whether BK channels are distributed within nanometer distances from Ca_V1.3 channels. We recorded potassium currents in whole-cell configuration and dialyzed calcium buffers with different binding rate constants into the cytoplasm (through the patch pipette) (Naraghi and Neher, 1997). The buffers limited calcium diffusion to specific distances from the calcium source. The equation of Prakriya and Lingle, 2000, predicts that dialysis of 10 mM EGTA restricts local elevations of intracellular calcium to a radius of 100 nm from the calcium source, whereas dialysis of 10 mM BAPTA restricts diffusion to a radius of 10 nm (Figure 3A). Tellingly, potassium currents activated by Ca_V1.3 channels were reduced by dialysis with BAPTA but not EGTA (Figure 3B). Interestingly, some current remains even with the dialysis of BAPTA, suggesting that some BK channels lie closer than 10 nm to Ca_V1.3 channels. In contrast, potassium currents were reduced by dialysis of either BAPTA or EGTA in tsA-201 cells co-expressing BK and Ca_V2.2 channels (Figure 3C), suggesting that BK channels are distributed farther than 100 nm from Ca_V2.2 channels. These observations imply that BK and Ca_V1.3 channels organize in close proximity, creating a functional unit with preferential co-localization in a range of less than 100 nm.

Figure 3

Download asset Open asset

Clusters of Ca_V1.3 channels surround BK channels

We proceeded to determine the spatial organization of BK and Ca_V1.3 channels with nanometer resolution. Super-resolution microscopy can resolve subcellular structures with a localization precision of approximately 20 nm that varies with the fluorophore used (Dempsey et al., 2011). We first checked how well our microscope could resolve previously studied structures. Figure 4—figure supplement 1 shows localization maps from tsA-201 cells stained for α-tubulin. In agreement with previous reports (Burnette et al., 2011), these single microtubules exhibit a full-width-at-half-maximum of 33 ± 1 nm. This result validates our experimental approach.

tsA-201 cells co-expressing BK and Ca_V1.3 channels were fixed and stained for super-resolution imaging. Secondary antibodies conjugated to Alexa-647 or Alexa-568, were used to detect single molecules in two-color microscopy in the same cell. To determine bleed-through between spectral channels, we obtained super-resolution images of microtubules in tsA-201 cells labeled with secondary Alexa-647 or 568. We followed the same imaging settings and imaging sequence normally used in all our experiments (imaging first the 647 channel and then the 568 channel). We calculated the fraction of bleed-through by dividing the total number of localization events detected in the ‘wrong’ channel by the total number of localization events detected in the ‘correct’ channel for both fluorophores (Figure 4—figure supplement 2). The bleed-through from the emission of Alexa-568 into the 647 channel was 0.092 ± 0.032 (n = 5) and the bleed-through from the emission of Alexa-647 into the 568 channel was 0.008 ± 0.001 (n = 3), suggesting that the bleed-through between these two fluorophores is minimal.

In Figure 4A, pixels positive for BK channels (red) formed multi-pixel aggregates, with a median area of 1600 nm² (n = 5 cells). They are much larger than the expected size for a single ion channel, which has known dimensions of 13 by 13 nm (Wang and Sigworth, 2009; Tao et al., 2017). Therefore, we interpreted these aggregates as clusters of multiple BK channels. In numerous cases, the clusters of BK channels (red) were surrounded by variable numbers of clusters of Ca_V1.3 channels (green) rather than showing any fixed stoichiometry or geometry (Figure 4A). Only 18 ± 5% of the BK clusters were not surrounded by Ca_V1.3 clusters within a radius of 200 nm (n = 5 cells).

Figure 4 with 5 supplements see all

Download asset Open asset

To provide a more quantitative description of the super-resolution maps, we measured the distance between BK and Ca_V1.3 clusters from edge to edge. The frequency distribution for the distance between Ca_V1.3 and BK-positive pixels (Figure 4C, green bars) was empirically fitted by a bell-shaped curve with a peak at 20 nm, our resolution limit, and a half width of 60 nm. This distribution fits with the distances suggested by the use of calcium buffers (Figure 3B), where BK channels are closer than 100 nm and even BAPTA could not block all the current, meaning that some BK channels probably lie at distances less than 10 nm from Ca_V1.3 channels. In another approach to describe the distribution of Ca_V1.3 channels around BK channels, we defined concentric areas of 20 nm width around the irregular BK clusters and measured the percentage of area occupied by Ca_V1.3 positive pixels in each concentric area (Figure 4D, green squares). To avoid bias in the analysis, data were analyzed from the whole cell instead of selected regions of interest (ROIs). The first point corresponds to Ca_V1.3-positive pixels that actually overlap with BK clusters. This plot shows a skewed distribution of Ca_V1.3 channels towards areas within 50 nm from BK clusters, indicating again that Ca_V1.3 clusters are preferentially concentrated in close proximity to BK clusters. Importantly, only 17 ± 2% of Ca_V1.3 clusters are at <200 nm from BK channels while the other 83% of Ca_V1.3 clusters are not localized close to any BK cluster (n = 5 cells).

To rule out that this distribution was a fixation artifact, we imaged the clusters from live tsA-201 cells in an AiryScan microscope which doubles the resolution achieved with a diffraction-limited confocal microscope. For these experiments we expressed BK channels tagged with GFP and Ca_V1.3 channels tagged with mRuby. Similar to what we observed in fixed cells, live cells exhibited BK channel puncta surrounded by Ca_V1.3 channel puncta (Figure 4—figure supplement 4).

We repeated the super-resolution study in cells co-expressing BK channels with Ca_V2.2 channels. Figure 4B shows representative ROIs from these cells. The median BK cluster size is not significantly different when BK channels are co-expressed with Ca_V2.2 channels (2000 nm², n = 4 cells). In comparison to the results with Ca_V1.3 channels, there is much less evidence for an association of Ca_V2.2 with BK channels. The frequency distribution for the distance between Ca_V2.2 and BK-positive pixels is roughly described by a bell-shaped curve centered at 100 nm from BK-positive pixels (Figure 4C, cyan bars). This distribution also fits with the distances suggested by the use of calcium buffers in Figure 3C. In addition, Ca_V2.2-positive pixels are more randomly distributed around BK channels (Figure 4D, cyan circles). Altogether, these results suggest that Ca_V1.3 channels are more likely to be close to BK channels than are Ca_V2.2 channels in our heterologous expression system. These results do not contradict the observed coupling between BK and Ca_V2.2 channels in neurons (Marrion and Tavalin, 1998). Rather they suggest some differences in the coupling mechanism and its ability to reconstitute in an expression system.

In the super-resolution maps, there were 2.5 times more Ca_V1.3 clusters (1843 ± 587) with a median cluster size of 1600 nm² than Ca_V2.2 clusters (741 ± 273) with a median cluster size of 400 nm². Presumably a higher density of Ca_V1.3 channels is expressed in these cells. We confirmed this observation by comparing the calcium current densities. The peak inward current density was −38 ± 11 pA/pF (n = 11) in cells co-transfected with Ca_V1.3 and only −10 ± 3 pA/pF (n = 9) in cells co-transfected with Ca_V2.2 (p=0.04). Because a higher expression of Ca_V1.3 channels might explain the closer average proximity of Ca_V1.3 and BK channels, we compared the distribution of BK channels to a random distribution of Ca_V1.3-positive pixels. To generate the random distribution from our data while keeping the number, size, and shape of each cluster, we binarized the maps of Ca_V1.3-positive pixels and randomized them with a custom macro written in MATLAB (Source code 1). Then, we merged the original map of BK-positive pixels unchanged with the new random-distribution map of Ca_V1.3-positive pixels for each cell (Figure 4—figure supplement 5) and analyzed the distribution around BK clusters. The distribution curve of randomized Ca_V1.3 pixels did not depend on the position of BK clusters (Figure 4D, orange circles), being more similar to the distribution observed for Ca_V2.2 channels. This result rejects the possibility that a higher expression level of Ca_V1.3 channels alone accounts for the greater coupling with BK channels.

BK and Ca_V1.3 channels form clusters in sympathetic and hippocampal neurons

Does this clustered distribution require overexpression of channels? Hippocampal and sympathetic neurons abundantly express BK and Ca_V1.3 channels, where they are known to serve important physiological functions (Storm, 1987; Lancaster and Nicoll, 1987; Gu et al., 2007). Thus, we investigated the spatial distribution of BK and Ca_V1.3 channels in these neurons using super-resolution imaging of immunostained neurons. As neurons express both Ca_V1.3 and Ca_V1.2 channels, it is important to be certain of the specificity of the Ca_V1.3 antibody. The antibody used in this work recognizes a highly variable site within the intracellular loop between domains II and III with no homology to the Ca_V1.2 channel (Figure 5—figure supplement 1A) and its specificity has been tested previously (Hell et al., 1993). To further assess the antibody’s specificity in our super-resolution experiments, we co-transfected tsA-201 cells with Ca_V1.3 and Ca_V1.2-EGFP channels and co-immunostained with anti-Ca_V1.3 and anti-GFP. Figure 5—figure supplement 1B shows that the two antibodies label different structures. In fact, the overlap between the two labels is only 2.2 ± 0.5% (n = 4), supporting that the antibody used against Ca_V1.3 channels does not recognize Ca_V1.2 channels.

Similar to the distribution observed in the heterologous expression system, BK channels formed clusters surrounded by clusters of Ca_V1.3 channels (Figure 5A,B). Specific secondary antibodies conjugated to Alexa-647 or Alexa-568 were used to detect BK and Ca_V1.3 channels, respectively. Reciprocal immunostaining using Alexa-647 for the Ca_V1.3 and Alexa-568 for the BK channels was also performed. The BK-Ca_V1.3 clustering organization was not dependent on the secondary antibody used (Figure 5—figure supplement 2). Independent of the secondary antibody used, all the cells were included in the distribution analysis. The median BK cluster size is 2800 nm² in SCG neurons (n = 7), significantly larger than the median BK cluster size in hippocampal neurons (2000 nm², n = 6) or the expression system (1600 nm², n = 5). The distribution analysis in both types of neurons also showed a skewed distribution of Ca_V1.3 clusters occupying areas adjacent to BK clusters (Figure 5C,E, green squares), indicating that Ca_V1.3 clusters are preferentially concentrated in close proximity to BK clusters, similar to the distribution observed in the heterologous expression system. Interestingly, whereas in hippocampal neurons 3% of the pixels inside BK clusters also included Ca_V1.3 channels, in SCG neurons as many as 10% of the BK cluster pixels included Ca_V1.3 channels, suggesting more intimate coupling of these interacting channels in SCG neurons. Randomized Ca_V1.3 channel clusters were distributed independently of the position of BK channels in the neurons (Figure 5C,E, orange circles). Such results show that endogenous BK and Ca_V1.3 channels in neurons exhibit a distribution and coupling similar to those in the expression system and suggest that clustering and proximity are fundamental properties of endogenous BK and Ca_V1.3 channels and not an artifact of overexpression.

Figure 5 with 3 supplements see all

Download asset Open asset

Our original hypothesis was that low-voltage activated Ca_V1.3 channels can confer on BK channels the ability to be activated at quite negative voltages, and we have found that these channels are distributed in a spatial pattern consistent with functional coupling. Given that BK and Ca_V1.3 channels display a similar coordinated distribution in neurons, we tested whether BK channels are activated at −40 mV in these cells. We recorded potassium currents in dissociated hippocampal and sympathetic neurons before and after the application of 500 nM paxilline and analyzed subtracted currents for evidence of BK channel activity. Figure 5D and F show representative potassium currents activated at −40 mV. On average, we observed 200 ± 20 pA (n = 3) of paxilline-sensitive potassium currents at −40 mV in hippocampal neurons and 100 ± 20 pA (n = 3) in sympathetic neurons. Calcium current at −40 mV is minimal or undetectable in whole-cell recordings in hippocampal and sympathetic neurons (Regan et al., 1991; Vivas et al., 2012). Nonetheless, it is possible that activation of a few channels, or clusters, is enough to begin to activate BK channels at low voltages. To determine the activity of individual calcium channels at negative voltages, we recorded spontaneous calcium events in total internal reflection fluorescence (TIRF) microscopy. Sympathetic neurons were patched in whole-cell mode. The calcium indicator Rhod2 (200 μM) and the calcium chelator EGTA (10 mM) were included in the patch pipette to restrict calcium elevations to the plasma membrane. Membrane potential was held at −80 mV. Localized bursts of fluorescent events were detected (Figure 5—figure supplement 3). These events, termed sparklets, have been characterized in tsA-201 cells expressing Ca_V1.3 channels and are known to reflect the opening of single or clustered channels (Navedo et al., 2007). Importantly, these events increased in number and amplitude after application of 500 nM Bay K 8644 (Figure 5—figure supplement 3D), suggesting that they are mediated by L-type calcium channels. Altogether, these results show that SCG neurons exhibit spontaneous activity of single and coupled L-type calcium channels at negative potentials that can contribute to the activation of nearby BK channels even in the absence of detectable whole-cell calcium current.

We found that Ca_V1.3 channels couple to BK channels in multi-channel complexes, enabling BK channels to open at voltages as negative as −50 mV. Our single-molecule localization experiments did not support proposed models of strict stoichiometry such as a 1:1 ratio of Ca_V and BK channels. Instead, we found that BK channels form clusters that are surrounded by clusters of Ca_V1.3 channels with no apparent fixed stoichiometry.

We provide several lines of evidence that support the hypothesis that these two channels congregate and are functionally coupled: (i) the functional and spatial coordination can be reconstituted in a heterologous system, (ii) the activation profile of BK channels changes with the expression of Ca_V1.3 channels, (iii) functional interaction persists when isolated in a patch pipette, (iv) coordinated gating is abolished by BAPTA, but not by EGTA in the cell, (v) the two channels co-immunoprecipitate, and (vi) they are physically close when observed under super-resolution microscopy. These findings are consistent with previous studies that show that BK channels co-precipitate with Ca_V1.3 channels from rat brain extracts (Grunnet and Kaufmann, 2004) and that chromaffin cells from mice lacking Ca_V1.3 channels exhibit smaller BK currents (Marcantoni et al., 2010).

Berkefeld and colleagues have previously shown that the activation profile of BK channels changes with the expression of different Ca_V isoforms. For instance, in their work BK channels start activating at around −10 mV when co-expressed with Ca_V2.1 channels but start activating at around +10 mV when co-expressed with Ca_V1.2 channels (Berkefeld et al., 2006). In agreement, our results show that the BK channel activation curve corresponds closely to the activation of co-expressed Ca_V2.2 or Ca_V1.3 channels, respectively, and with the latter can start activating at voltages as negative as −50 mV. Our experiments differ, however, in that we did not co-express the auxiliary β subunit of BK channels, suggesting that this auxiliary subunit is not necessary to reconstitute the functional coupling between BK and Ca_V channels. Our results add information about the composition of these complexes by using imaging at nanometer resolution, arguing against a 1:1 channel stoichiometry with a 10 nm distance between them. Instead, BK channels form homotypic clusters and then form higher order heterotypic clusters with clusters of Ca_V isoforms.

Clustering is a characteristic of several ion channels, including potassium and calcium channels, (Choi, 2014; King et al., 2014; Dixon et al., 2015). Recently, it was shown that Ca_V1.3 channels form clusters that are functionally coupled to each other and gate cooperatively (Moreno et al., 2016). Here we show that BK channels arrange in clusters as well and that these clusters of BK channels arrange in higher order distributions surrounded by clusters of Ca_V1.3 channels. Multi-channel complexes between Ca_V and BK channels have been proposed before. For example, in frog hair cells, it has been proposed that 85 calcium channels and 40 BK channels organize in the presynaptic active zone (Roberts, 1994). An important conclusion of that work is that, assuming the simultaneous opening of 10 Ca_V channels at any given time, the calcium concentration could reach >100 μM inside the active zone, decaying steeply at the edge. Others have proposed complexes with a ratio of 5:1 Ca_V to BK channels (Prakriya and Lingle, 2000). In this case, calcium channels can be at distances as far as 50 nm from BK channels and would still be able to accumulate enough calcium to activate BK channels, if more than one calcium channel opened at the same time. The association we see between BK and Ca_V1.3 channels appears to have neither fixed stoichiometry nor fixed geometry. Rather it can be described as a statistical bias for proximity of clusters, a bias that would not be present with a random distribution. We describe the most common composition of this mega complex. Homotypic BK clusters have various sizes, with a median of 2000 nm² and appear to be nucleation centers for Ca_V1.3 clusters to come very close, forming a heterotypic cluster. By counting the number of Ca_V1.3 clusters within 50 nm from BK clusters, we estimated that there are on average 4 Ca_V1.3 clusters around a single BK cluster (n = 128 ROIs). However, the super-resolution data show considerable lack of symmetry, with some BK clusters lacking nearby Ca_V1.3 clusters, Ca_V1.3 clusters not touching BK clusters, and Ca_V1.3 clusters in close proximity to BK clusters at various distances and numbers. The mechanisms favoring the formation of higher order cluster between BK and Ca_V1.3 clusters might be (1) direct interactions between potassium and calcium channels, (2) indirect interactions mediated by scaffolding proteins, (3) membrane domains that restrict and concentrate their distribution, or (4) weak interactions between BK and Ca_V1.3 channels, that by lateral diffusion and mass action, result in aggregations around each other. The later alternative might explain why not all BK clusters are surrounded by Ca_V1.3 clusters, given an equilibrium where some clusters are in a non-interacting state while others are in an interacting state.

What concentration of intracellular calcium might be found in the center of this multi channel mega-complex? The amount of calcium reaching a BK channel cluster will depend on: (i) the number of Ca_V1.3 clusters around it (a mean of 4 according to our estimations), (ii) the number of Ca_V1.3 channels open at any given time, and (iii) their distance from the target BK channel. Interestingly, clustering of Ca_V1.3 channels leads to cooperative opening of Ca_V1.3 channels within the cluster (Moreno et al., 2016). It remains uncertain, however, whether a nearby second cluster of Ca_V1.3 channels could open at the same time as the first. Calcium entering from two equidistant Ca_V1.3 clusters around a BK cluster could sum linearly and therefore contribute equally to the activation of BK channels.

To conclude, we propose that the coupling of BK channels with low-voltage activated Ca_V1.3 channels in a multi-channel complex enables this feedback mechanism to activate BK channels at potentials near the threshold, preventing neurons from firing action potentials as we observed recently in sympathetic neurons (Vivas et al., 2014).

1. 1. Avila G
   2. Sandoval A
   3. Felix R

   (2004) Intramembrane charge movement associated with endogenous K+ channel activity in HEK-293 cells

   Cellular and Molecular Neurobiology 24:317–330.

   https://doi.org/10.1023/B:CEMN.0000022765.52109.26

   • PubMed
   • Google Scholar
2. 1. Barrett JN
   2. Magleby KL
   3. Pallotta BS

   (1982) Properties of single calcium-activated potassium channels in cultured rat muscle

   The Journal of Physiology 331:211–230.

   https://doi.org/10.1113/jphysiol.1982.sp014370

   • PubMed
   • Google Scholar
3. 1. Berjukow S
   2. Döring F
   3. Froschmayr M
   4. Grabner M
   5. Glossmann H
   6. Hering S

   (1996) Endogenous calcium channels in human embryonic kidney (HEK293) cells

   British Journal of Pharmacology 118:748–754.

   https://doi.org/10.1111/j.1476-5381.1996.tb15463.x

   • PubMed
   • Google Scholar
4. 1. Berkefeld H
   2. Sailer CA
   3. Bildl W
   4. Rohde V
   5. Thumfart JO
   6. Eble S
   7. Klugbauer N
   8. Reisinger E
   9. Bischofberger J
   10. Oliver D
   11. Knaus HG
   12. Schulte U
   13. Fakler B

   (2006) BKCa-Cav channel complexes mediate rapid and localized ^Ca2+-activated K⁺ signaling

   Science 314:615–620.

   https://doi.org/10.1126/science.1132915

   • PubMed
   • Google Scholar
5. 1. Berkefeld H
   2. Fakler B

   (2008) Repolarizing responses of BKCa-Cav complexes are distinctly shaped by their Cav subunits

   Journal of Neuroscience 28:8238–8245.

   https://doi.org/10.1523/JNEUROSCI.2274-08.2008

   • PubMed
   • Google Scholar
6. 1. Brandt A
   2. Striessnig J
   3. Moser T

   (2003) CaV1.3 channels are essential for development and presynaptic activity of cochlear inner hair cells

   Journal of Neuroscience 23:10832–10840.

   https://doi.org/10.3410/f.1016742.201257

   • PubMed
   • Google Scholar
7. 1. Brandt A
   2. Khimich D
   3. Moser T

   (2005) Few CaV1.3 channels regulate the exocytosis of a synaptic vesicle at the hair cell ribbon synapse

   Journal of Neuroscience 25:11577–11585.

   https://doi.org/10.1523/JNEUROSCI.3411-05.2005

   • PubMed
   • Google Scholar
8. 1. Burnette DT
   2. Sengupta P
   3. Dai Y
   4. Lippincott-Schwartz J
   5. Kachar B

   (2011) Bleaching/blinking assisted localization microscopy for superresolution imaging using standard fluorescent molecules

   PNAS 108:21081–21086.

   https://doi.org/10.1073/pnas.1117430109

   • PubMed
   • Google Scholar
9. 1. Choi KH

   (2014) Cooperative gating between ion channels

   General Physiology and Biophysics 33:1–12.

   https://doi.org/10.4149/gpb_2013076

   • PubMed
   • Google Scholar
10. 1. Christel CJ
    2. Cardona N
    3. Mesirca P
    4. Herrmann S
    5. Hofmann F
    6. Striessnig J
    7. Ludwig A
    8. Mangoni ME
    9. Lee A

    (2012) Distinct localization and modulation of Cav1.2 and Cav1.3 L-type ^Ca2+ channels in mouse sinoatrial node

    The Journal of Physiology 590:6327–6342.

    https://doi.org/10.1113/jphysiol.2012.239954

    • PubMed
    • Google Scholar
11. 1. Contreras GF
    2. Castillo K
    3. Enrique N
    4. Carrasquel-Ursulaez W
    5. Castillo JP
    6. Milesi V
    7. Neely A
    8. Alvarez O
    9. Ferreira G
    10. González C
    11. Latorre R

    (2013) A BK (Slo1) channel journey from molecule to physiology

    Channels 7:442–458.

    https://doi.org/10.4161/chan.26242

    • PubMed
    • Google Scholar
12. 1. Cooper G
    2. Lasser-Katz E
    3. Simchovitz A
    4. Sharon R
    5. Soreq H
    6. Surmeier DJ
    7. Goldberg JA

    (2015) Functional segregation of voltage-activated calcium channels in motoneurons of the dorsal motor nucleus of the vagus

    Journal of Neurophysiology 114:1513–1520.

    https://doi.org/10.1152/jn.00432.2014

    • PubMed
    • Google Scholar
13. 1. Cox DH

    (2014) Modeling a Ca²⁺ channel/BKCa channel complex at the single-complex level

    Biophysical Journal 107:2797–2814.

    https://doi.org/10.1016/j.bpj.2014.10.069

    • PubMed
    • Google Scholar
14. 1. Cui J
    2. Yang H
    3. Lee US

    (2009) Molecular mechanisms of BK channel activation

    Cellular and Molecular Life Sciences 66:852–875.

    https://doi.org/10.1007/s00018-008-8609-x

    • PubMed
    • Google Scholar
15. 1. Dempsey GT
    2. Vaughan JC
    3. Chen KH
    4. Bates M
    5. Zhuang X

    (2011) Evaluation of fluorophores for optimal performance in localization-based super-resolution imaging

    Nature Methods 8:1027–1036.

    https://doi.org/10.1038/nmeth.1768

    • PubMed
    • Google Scholar
16. 1. Dixon RE
    2. Moreno CM
    3. Yuan C
    4. Opitz-Araya X
    5. Binder MD
    6. Navedo MF
    7. Santana LF

    (2015) Graded Ca²⁺/calmodulin-dependent coupling of voltage-gated CaV1.2 channels

    eLife 4:e05608.

    https://doi.org/10.7554/eLife.05608

    • PubMed
    • Google Scholar
17. 1. Fakler B
    2. Adelman JP

    (2008) Control of K_Ca channels by calcium nano/microdomains

    Neuron 59:873–881.

    https://doi.org/10.1016/j.neuron.2008.09.001

    • PubMed
    • Google Scholar
18. 1. Grunnet M
    2. Kaufmann WA

    (2004) Coassembly of big conductance ^Ca2+-activated K⁺ channels and L-type voltage-gated ^Ca2+ channels in rat brain

    Journal of Biological Chemistry 279:36445–36453.

    https://doi.org/10.1074/jbc.M402254200

    • PubMed
    • Google Scholar
19. 1. Gu N
    2. Vervaeke K
    3. Storm JF

    (2007) BK potassium channels facilitate high-frequency firing and cause early spike frequency adaptation in rat CA1 hippocampal pyramidal cells

    The Journal of Physiology 580:859–882.

    https://doi.org/10.1113/jphysiol.2006.126367

    • PubMed
    • Google Scholar
20. 1. Hell JW
    2. Westenbroek RE
    3. Warner C
    4. Ahlijanian MK
    5. Prystay W
    6. Gilbert MM
    7. Snutch TP
    8. Catterall WA

    (1993) Identification and differential subcellular localization of the neuronal class C and class D L-type calcium channel alpha 1 subunits

    The Journal of Cell Biology 123:949–962.

    https://doi.org/10.1083/jcb.123.4.949

    • PubMed
    • Google Scholar
21. 1. King AN
    2. Manning CF
    3. Trimmer JS

    (2014) A unique ion channel clustering domain on the axon initial segment of mammalian neurons

    Journal of Comparative Neurology 522:2594–2608.

    https://doi.org/10.1002/cne.23551

    • PubMed
    • Google Scholar
22. 1. Koschak A
    2. Reimer D
    3. Huber I
    4. Grabner M
    5. Glossmann H
    6. Engel J
    7. Striessnig J

    (2001) Alpha 1D (Cav1.3) subunits can form l-type ^Ca2+ channels activating at negative voltages

    Journal of Biological Chemistry 276:22100–22106.

    https://doi.org/10.1074/jbc.M101469200

    • PubMed
    • Google Scholar
23. 1. Lancaster B
    2. Nicoll RA

    (1987) Properties of two calcium-activated hyperpolarizations in rat hippocampal neurones

    The Journal of Physiology 389:187–203.

    https://doi.org/10.1113/jphysiol.1987.sp016653

    • PubMed
    • Google Scholar
24. 1. Lipscombe D
    2. Helton TD
    3. Xu W

    (2004) L-type calcium channels: the low down

    Journal of Neurophysiology 92:2633–2641.

    https://doi.org/10.1152/jn.00486.2004

    • PubMed
    • Google Scholar
25. 1. Liu Y
    2. Harding M
    3. Pittman A
    4. Dore J
    5. Striessnig J
    6. Rajadhyaksha A
    7. Chen X

    (2014) Cav1.2 and Cav1.3 L-type calcium channels regulate dopaminergic firing activity in the mouse ventral tegmental area

    Journal of Neurophysiology 112:1119–1130.

    https://doi.org/10.1152/jn.00757.2013

    • PubMed
    • Google Scholar
26. 1. Lu L
    2. Sirish P
    3. Zhang Z
    4. Woltz RL
    5. Li N
    6. Timofeyev V
    7. Knowlton AA
    8. Zhang XD
    9. Yamoah EN
    10. Chiamvimonvat N

    (2015) Regulation of gene transcription by voltage-gated L-type calcium channel, Cav1.3

    Journal of Biological Chemistry 290:4663–4676.

    https://doi.org/10.1074/jbc.M114.586883

    • PubMed
    • Google Scholar
27. 1. Mangoni ME
    2. Couette B
    3. Bourinet E
    4. Platzer J
    5. Reimer D
    6. Striessnig J
    7. Nargeot J

    (2003) Functional role of L-type Cav1.3 ^Ca2+ channels in cardiac pacemaker activity

    PNAS 100:5543–5548.

    https://doi.org/10.1073/pnas.0935295100

    • PubMed
    • Google Scholar
28. 1. Marcantoni A
    2. Vandael DH
    3. Mahapatra S
    4. Carabelli V
    5. Sinnegger-Brauns MJ
    6. Striessnig J
    7. Carbone E

    (2010) Loss of Cav1.3 channels reveals the critical role of L-type and BK channel coupling in pacemaking mouse adrenal chromaffin cells

    Journal of Neuroscience 30:491–504.

    https://doi.org/10.1523/JNEUROSCI.4961-09.2010

    • PubMed
    • Google Scholar
29. 1. Marrion NV
    2. Tavalin SJ

    (1998) Selective activation of ^Ca2+-activated K⁺ channels by co-localized ^Ca2+ channels in hippocampal neurons

    Nature 395:900–905.

    https://doi.org/10.1038/27674

    • PubMed
    • Google Scholar
30. 1. Michna M
    2. Knirsch M
    3. Hoda JC
    4. Muenkner S
    5. Langer P
    6. Platzer J
    7. Striessnig J
    8. Engel J

    (2003) Cav1.3 (alpha1D) ^Ca2+ currents in neonatal outer hair cells of mice

    The Journal of Physiology 553:747–758.

    https://doi.org/10.1113/jphysiol.2003.053256

    • PubMed
    • Google Scholar
31. 1. Montefusco F
    2. Tagliavini A
    3. Ferrante M
    4. Pedersen MG

    (2017) Concise Whole-Cell modeling of BKCa-CaV activity controlled by local coupling and stoichiometry

    Biophysical Journal 112:2387–2396.

    https://doi.org/10.1016/j.bpj.2017.04.035

    • PubMed
    • Google Scholar
32. 1. Moreno CM
    2. Dixon RE
    3. Tajada S
    4. Yuan C
    5. Opitz-Araya X
    6. Binder MD
    7. Santana LF

    (2016) Ca²⁺ entry into neurons is facilitated by cooperative gating of clustered CaV1.3 channels

    eLife 5:e15744.

    https://doi.org/10.7554/eLife.15744

    • PubMed
    • Google Scholar
33. 1. Müller CS
    2. Haupt A
    3. Bildl W
    4. Schindler J
    5. Knaus HG
    6. Meissner M
    7. Rammner B
    8. Striessnig J
    9. Flockerzi V
    10. Fakler B
    11. Schulte U

    (2010) Quantitative proteomics of the Cav2 channel nano-environments in the mammalian brain

    PNAS 107:14950–14957.

    https://doi.org/10.1073/pnas.1005940107

    • PubMed
    • Google Scholar
34. 1. Naraghi M
    2. Neher E

    (1997)

    Linearized buffered ^Ca2+ diffusion in microdomains and its implications for calculation of [^Ca2+] at the mouth of a calcium channel

    Journal of Neuroscience 17:6961–6973.

    • PubMed
    • Google Scholar
35. 1. Navedo MF
    2. Amberg GC
    3. Westenbroek RE
    4. Sinnegger-Brauns MJ
    5. Catterall WA
    6. Striessnig J
    7. Santana LF

    (2007) Ca(v)1.3 channels produce persistent calcium sparklets, but ca(v)1.2 channels are responsible for sparklets in mouse arterial smooth muscle

    AJP: Heart and Circulatory Physiology 293:H1359–H1370.

    https://doi.org/10.1152/ajpheart.00450.2007

    • PubMed
    • Google Scholar
36. 1. Platzer J
    2. Engel J
    3. Schrott-Fischer A
    4. Stephan K
    5. Bova S
    6. Chen H
    7. Zheng H
    8. Striessnig J

    (2000) Congenital deafness and sinoatrial node dysfunction in mice lacking class D L-type ^Ca2+ channels

    Cell 102:89–97.

    https://doi.org/10.1016/S0092-8674(00)00013-1

    • PubMed
    • Google Scholar
37. 1. Prakriya M
    2. Lingle CJ

    (2000)

    Activation of BK channels in rat chromaffin cells requires summation of Ca²⁺ influx from multiple Ca²⁺ channels

    Journal of Neurophysiology 84:1123–1135.

    • PubMed
    • Google Scholar
38. 1. Putzier I
    2. Kullmann PH
    3. Horn JP
    4. Levitan ES

    (2009) Cav1.3 channel voltage dependence, not ^Ca2+ selectivity, drives pacemaker activity and amplifies bursts in nigral dopamine neurons

    Journal of Neuroscience 29:15414–15419.

    https://doi.org/10.1523/JNEUROSCI.4742-09.2009

    • PubMed
    • Google Scholar
39. 1. Regan LJ
    2. Sah DW
    3. Bean BP

    (1991) ^Ca2+ channels in rat central and peripheral neurons: high-threshold current resistant to dihydropyridine blockers and omega-conotoxin

    Neuron 6:269–280.

    https://doi.org/10.1016/0896-6273(91)90362-4

    • PubMed
    • Google Scholar
40. 1. Rehak R
    2. Bartoletti TM
    3. Engbers JD
    4. Berecki G
    5. Turner RW
    6. Zamponi GW

    (2013) Low voltage activation of KCa1.1 current by Cav3-KCa1.1 complexes

    PLoS One 8:e61844.

    https://doi.org/10.1371/journal.pone.0061844

    • PubMed
    • Google Scholar
41. 1. Roberts WM
    2. Jacobs RA
    3. Hudspeth AJ

    (1990)

    Colocalization of ion channels involved in frequency selectivity and synaptic transmission at presynaptic active zones of hair cells

    Journal of Neuroscience 10:3664–3684.

    • PubMed
    • Google Scholar
42. 1. Roberts WM

    (1994)

    Localization of calcium signals by a mobile calcium buffer in frog saccular hair cells

    Journal of Neuroscience 14:3246–3262.

    • PubMed
    • Google Scholar
43. 1. Simon SM
    2. Llinás RR

    (1985) Compartmentalization of the submembrane calcium activity during calcium influx and its significance in transmitter release

    Biophysical Journal 48:485–498.

    https://doi.org/10.1016/S0006-3495(85)83804-2

    • PubMed
    • Google Scholar
44. 1. Stanika R
    2. Campiglio M
    3. Pinggera A
    4. Lee A
    5. Striessnig J
    6. Flucher BE
    7. Obermair GJ

    (2016) Splice variants of the CaV1.3 L-type calcium channel regulate dendritic spine morphology

    Scientific Reports 6:34528.

    https://doi.org/10.1038/srep34528

    • PubMed
    • Google Scholar
45. 1. Storm JF

    (1987) Action potential repolarization and a fast after-hyperpolarization in rat hippocampal pyramidal cells

    The Journal of Physiology 385:733–759.

    https://doi.org/10.1113/jphysiol.1987.sp016517

    • PubMed
    • Google Scholar
46. 1. Tao X
    2. Hite RK
    3. MacKinnon R

    (2017) Cryo-EM structure of the open high-conductance Ca²⁺-activated K⁺ channel

    Nature 541:46–51.

    https://doi.org/10.1038/nature20608

    • PubMed
    • Google Scholar
47. 1. Torrente AG
    2. Mesirca P
    3. Neco P
    4. Rizzetto R
    5. Dubel S
    6. Barrere C
    7. Sinegger-Brauns M
    8. Striessnig J
    9. Richard S
    10. Nargeot J
    11. Gomez AM
    12. Mangoni ME

    (2016) L-type Cav1.3 channels regulate ryanodine receptor-dependent ^Ca2+ release during sino-atrial node pacemaker activity

    Cardiovascular Research 109:451–461.

    https://doi.org/10.1093/cvr/cvw006

    • PubMed
    • Google Scholar
48. 1. Vivas O
    2. Arenas I
    3. García DE

    (2012) Voltage-independent inhibition of ca(V)2.2 channels is delimited to a specific region of the membrane potential in rat SCG neurons

    Acta Biochimica Et Biophysica Sinica 44:544–549.

    https://doi.org/10.1093/abbs/gms025

    • PubMed
    • Google Scholar
49. 1. Vivas O
    2. Castro H
    3. Arenas I
    4. Elías-Viñas D
    5. García DE

    (2013) PIP₂ hydrolysis is responsible for voltage independent inhibition of CaV2.2 channels in sympathetic neurons

    Biochemical and Biophysical Research Communications 432:275–280.

    https://doi.org/10.1016/j.bbrc.2013.01.117

    • PubMed
    • Google Scholar
50. 1. Vivas O
    2. Kruse M
    3. Hille B

    (2014) Nerve growth factor sensitizes adult sympathetic neurons to the proinflammatory peptide bradykinin

    Journal of Neuroscience 34:11959–11971.

    https://doi.org/10.1523/JNEUROSCI.1536-14.2014

    • PubMed
    • Google Scholar
51. 1. Wang L
    2. Sigworth FJ

    (2009) Structure of the BK potassium channel in a lipid membrane from electron cryomicroscopy

    Nature 461:292–295.

    https://doi.org/10.1038/nature08291

    • PubMed
    • Google Scholar
52. 1. Yan J
    2. Aldrich RW

    (2010) LRRC26 auxiliary protein allows BK channel activation at resting voltage without calcium

    Nature 466:513–516.

    https://doi.org/10.1038/nature09162

    • PubMed
    • Google Scholar
53. 1. Zhu G
    2. Zhang Y
    3. Xu H
    4. Jiang C

    (1998) Identification of endogenous outward currents in the human embryonic kidney (HEK 293) cell line

    Journal of Neuroscience Methods 81:73–83.

    https://doi.org/10.1016/S0165-0270(98)00019-3

    • PubMed
    • Google Scholar

Author details

1. Oscar Vivas

   1. Department of Physiology and Biophysics, University of Washington, Seattle, United States
   2. Department of Physiology and Membrane Biology, University of California, Davis, Davis, United States

   Contribution

   OV, Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   ovivas@ucdavis.edu

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0002-0964-385X

2. Claudia M Moreno

   1. Department of Physiology and Biophysics, University of Washington, Seattle, United States
   2. Department of Physiology and Membrane Biology, University of California, Davis, Davis, United States

   Contribution

   CMM, Formal analysis, Investigation, Writing—original draft, Writing—review and editing

   Competing interests

   The authors declare that no competing interests exist.

3. Luis F Santana

   Department of Physiology and Membrane Biology, University of California, Davis, Davis, United States

   Contribution

   LFS, Funding acquisition, Methodology, Writing—review and editing

   Contributed equally with

   Bertil Hille

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0002-4297-8029

4. Bertil Hille

   Department of Physiology and Biophysics, University of Washington, Seattle, United States

   Contribution

   BH, Conceptualization, Supervision, Funding acquisition, Writing—original draft, Project administration, Writing—review and editing

   Contributed equally with

   Luis F Santana

   Competing interests

   The authors declare that no competing interests exist.

• 3,333

  views

• 614

  downloads

• 53

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.