β1 subunits facilitate gating of BK channels by acting through the Ca²⁺, but not the Mg²⁺, activating mechanisms

Abstract

The β1 subunit of BK (large conductance Ca²⁺ and voltage-activated K⁺) channels is essential for many key physiological processes, such as controlling the contraction of smooth muscle and the tuning of hair cells in the cochlea. Although it is known that the β1 subunit greatly increases the open probability of BK channels, little is known about its mechanism of action. We now explore this mechanism by using channels in which the Ca²⁺- and Mg²⁺-dependent activating mechanisms have been disrupted by mutating three sites to remove the Ca²⁺ and Mg²⁺ sensitivity. We find that the presence of the β1 subunit partially restores Ca²⁺ sensitivity to the triply mutated channels, but not the Mg²⁺ sensitivity. We also find that the β1 subunit has no effect on the Mg²⁺ sensitivity of WT BK channels, in contrast to its pronounced effect of increasing the apparent Ca²⁺ sensitivity. These observations suggest that the β1 subunit increases open probability by working through the Ca²⁺-dependent, rather than Mg²⁺-dependent, activating mechanisms, and that the action of the β1 subunit is not directly on the Ca²⁺ binding sites, but on the allosteric machinery coupling the sites to the gate. The differential effects of the β1 subunit on the Ca²⁺ and Mg²⁺ activation of the channel suggest that these processes act separately. Finally, we show that inhibits, rather than activates, BK channels in the presence of the β1 subunit for intermediate levels of . This Mg²⁺ inhibition in the presence of the β1 subunit provides an additional regulatory mechanism of BK channel activity.

BK channels (Slo1) are large conductance K⁺ channels that are activated in a highly synergistic manner by membrane depolarization, intracellular Ca²⁺ (), and intracellular Mg²⁺ () (1–12). When BK channels are open, K⁺ flows out of cells through the open channels, driving the membrane potential more negative. This change in membrane potential allows BK channels to modulate many important physiological processes, such as smooth muscle contraction (13–15), frequency tuning of sensory hair cells (16), and regulation of neurotransmitter release (17, 18). BK channels can be comprised of four α subunits alone, as in skeletal muscle, or four α subunits with four associated β subunits, as in smooth muscle and neurons (19–21). Of the four types of β subunits that have been identified, the β1 subunit is predominantly expressed in smooth muscle (13, 20). When coexpressed with α subunits, the β1 subunit greatly increases the open probability (P_o) of BK channels over the physiological range of and voltage (22–27). This increase in P_o gives an apparent increase in the Ca²⁺ sensitivity of the channel, because less is required for 50% activation (22–27). Decreasing the activation of BK channels in smooth muscle by knocking out the β1 subunit results in chronic hypertension from increased contraction of the arterial smooth muscle (13, 14).

Despite the importance of the β1 subunit, the underlying mechanism by which the β1 subunit increases P_o of the BK channel remains unclear, although progress has been made (22–24, 26, 27). Previous studies at both the single-channel and macrocurrent level have suggested that the major (80%) facilitating effect of the β1 subunit is independent of Ca²⁺, with a smaller (20%) Ca²⁺-dependent contribution (23, 24, 26). Consistent with the idea that the major action of the β1 subunit is not through changes in Ca²⁺ affinity (23, 24, 26), the Ca²⁺ bowl, a key site located on the C terminus associated with Ca²⁺ activation of BK channels (28, 29), is not required for the β1 subunit to increase the apparent Ca²⁺ sensitivity of the BK channel, but other structural features of the C terminus are required (27). Thus, the β1 subunit may act on the allosteric machinery coupling the Ca²⁺ binding sites to the gate rather than directly on the sites.

To explore this possibility, we take advantage of recent studies that suggest that there are at least two major classes of divalent binding sites located on the intracellular C terminus of the α subunit of the BK channel. One of these classes is of low affinity (millimolar) and activates the channel through the binding of Ca²⁺ or Mg²⁺, and the other is of high affinity (micromolar) and activates the channel through the binding of Ca²⁺. For the high-affinity class, combined mutations at two separate sites, the Ca²⁺ bowl and D362/D367, remove the high-affinity Ca²⁺ sensitivity (11). For the low-affinity class, mutation of E399 removes the activation by millimolar concentrations of Ca²⁺ or Mg²⁺ (10, 12). Mutation of these three sites together produces a triple mutant (TM) channel with no Ca²⁺ or Mg²⁺ sensitivity for concentrations of and <10 mM (11). Because the TM channel has no Ca²⁺ and Mg²⁺ sensitivity, it provides a unique platform to examine to what extent the β1 subunit may require functional Ca²⁺ and Mg²⁺ activating mechanisms to modulate the BK channel.

We found that the β1 subunit increased P_o of the TM channel in the absence of and , just as it does for WT channels, indicating that the triple mutation does not destroy the gating machinery on which the β1 subunit acts. Surprisingly, we also found that the β1 subunit restored some Ca²⁺ sensitivity to the TM Ca²⁺-insensitive channel. This restoration was still present after removing the only intracellular negative charge on the β1 subunit, suggesting that the β1 subunit was not directly adding an additional Ca²⁺ binding site. In contrast to the restoration of Ca²⁺ sensitivity to the TM channels, the β1 subunit did not restore any Mg²⁺ sensitivity. In addition, the β1 subunit had no effect on the Mg²⁺ sensitivity of WT BK channels, compared with its pronounced effect of increasing the apparent Ca²⁺ sensitivity. Taken together, these results suggest that the β1 subunit couples Ca²⁺, but not Mg²⁺, activating mechanisms to the gating of BK channels, and that the action of the β1 subunit is not directly on the Ca²⁺ binding sites, but rather on the allosteric machinery coupling the sites to the gate. Finally, we made the observation that Mg²⁺ can inhibit channel activity in the presence of the β1 subunit for physiological levels of , in contrast to its activation in the absence of the β1 subunit.

Materials and Methods

Clones, Mutagenesis, and Channel Expression. The WT Slo1 α subunit of the mouse BK channel (GenBank accession no. L16912) was provided by L. Salkoff (30), and the human β1 subunit (GenBank accession no. NM_004137) was provided by Merck Research Laboratories. The α subunit of the BK channel with mutations to remove three regulatory mechanisms (11), 5D5N in the Ca²⁺ bowl, D362A/D367A, and E399A (TM channel), and the human β1 subunit with mutations E13K and T14R were provided by X. Xia and C. Lingle (Washington University, St. Louis). Starting with the construct of the TM channel, we made two additional types of mutated channels: one by making a mutation in the pore region to decrease the tetraethylammonium (TEA) sensitivity (GYGDVYAKT to GYGDVVAKT) and the other by making a mutation on the C terminus by changing M513 to I513, using Stratagene's QuikChange Site-Directed Mutagenesis Kit. The mutations were checked by sequencing. The cDNA was transcribed in vitro by using the mMESSAGE mMACHINE Kit (Ambion, Austin, TX) to obtain cRNA for injection into the Xenopus oocytes. The oocytes were microinjected with 0.5–50 ng of cRNA 2–8 days before recording. For coexpression of α and β1 subunits, the cRNA for the β1 subunit was in 4–6 M excess over that for the α subunit to drive coassembly (22).

Electrophysiology and Solutions. Single-channel currents were recorded with the patch clamp technique (31) from inside-out patches of membrane excised from Xenopus oocytes, except for two experiments when TEA was applied to outside-out patches. (TEA was in the pipette for seven additional experiments with TEA.) The pipette solution contained 158 mM KCl, 5 mM N-[tris(hydroxymethyl)methyl]-2-aminoethanesulfonic acid (TES), and usually 10 μM GdCl₃ to block the endogenous stretch-activated channels (32). The bath solution contained 158 mM KCl, 5 mM TES, 1 mM EGTA, 1 mM N-(2-hydroxyethyl) ethylenediamine-N,N′,N′-triacetic acid (HEDTA), and sufficient added Ca²⁺ and Mg²⁺ to bring the free Ca²⁺ and Mg²⁺ levels to those indicated (23, 33). All solutions were adjusted to pH 7.0. Solutions with no added Ca²⁺ had a calculated free Ca²⁺ of <10^–8 M and will be referred to as 0 Ca²⁺ solutions because at these concentrations has essentially no effect on the gating of the channel (3, 24, 34). Voltages refer to the intracellular potential. Experiments were performed at room temperature (20–23°C).

Data Analysis. Single-channel currents were typically filtered at 3–10 kHz and sampled at 200 kHz directly to disk by using pclamp8 (Axon Instuments, Union City, CA). P_o and burst duration were then estimated from the digitized records by using custom programs (33, 35). P_o vs. V plots derived from the single-channel data were fitted with a Boltzmann function (Eq. 1):

[1]

where V_0.5 is the voltage of half-maximal activation of the channel and K_v is the voltage dependence of the activation process in mV per e-fold change. P_o vs. plots derived from the single-channel data were fitted with a Hill equation (Eq. 2):

[2]

where P_max is the maximum P_o for the channels, K_d is the required for half activation, and n is the Hill coefficient that reflects the slope of the dose–response relationship. Data are expressed as the mean ± SE.

Results and Discussion

The β1 Subunit Restores Ca²⁺ Sensitivity to the TM Channel. We first examined whether the TM channel is Ca²⁺-insensitive by using single-channel recording techniques. Increasing from 0 to 10 mM was found to have no effect on the activity of single BK channels in seven experiments (example in Fig. 1A), confirming the Ca²⁺ insensitivity of the TM channel (11).

Fig. 1.

The β1 subunit partially restores Ca²⁺ sensitivity to TM Ca²⁺-insensitive BK channels. (A and B) Single-channel currents recorded from a TM channel in the absence (A) and presence (B)ofthe β1 subunit at the indicated and +50 mV. (C) Plots of P_o vs. from 10 representative experiments to show the range of restoration of Ca²⁺ sensitivity to the TM channel by the β1 subunit. (D) Plots of P_o vs. V at 0 for TM channels with and without the β1 subunit (data from six patches for each point). The standard errors are often obscured by the symbol.

To examine the effect of the β1 subunit on the TM channel, we then coexpressed β1 subunits with TM α subunits. Surprisingly, in the presence of the β1 subunit the same change in now greatly increased the single-channel activity of the TM channel (Fig. 1B). The degree of restoration of Ca²⁺ sensitivity to the TM channel by the β1 subunit was consistent for each channel but variable among channels (data from 10 different experiments at +50 mV in Fig. 1C), with maximum P_os of 0.44 ± 0.10 (range 0.002–0.9), typical Hill coefficients of 1.1 ± 0.1 (range 0.5–1.8), and K_ds of 38.1 ± 11.2 μM (range 14.7–108.2). In contrast, such a wide variability in the response is not observed for WT channels with β1 subunits, which typically have maximum P_os of >0.9, Hill coefficients of 3.5 or 4.5, and K_ds of 1.2 or 2.6 μM for data obtained at either +60 or +30 mV, respectively (23, 26). Thus, the β1 subunit only partially restored Ca²⁺ sensitivity to the TM Ca²⁺-insensitive channel, as reflected by the decreased Hill coefficient, greater K_d values, and inability to consistently drive the P_o to the maximal level of >0.9 observed for WT channels with β1 subunits (23).

The reason for the variability in the restoration of Ca²⁺ sensitivity for the TM channels is not known, but Wang et al. (36) have shown for WT channels that decreasing the number of β1 subunits per channel results in less activation for a given . Perhaps TM channels are less likely than WT channels to have a full complement of four β1 subunits, giving rise to graded effects. The TM channels typically gate in normal mode activity, so the variability in response is not likely to arise from entry into other modes of gating (see ref. 33 for discussion of modes). Although the reason for the variability is not known, the important point is that the β1 subunit typically restored some Ca²⁺ sensitivity to the TM channels. The following sections explore possible mechanisms for the restoration of Ca²⁺ sensitivity.

The Mutations Producing the TM Channel Do Not Facilitate the Action of the β1 Subunit. Nimigean and Magleby (23, 24) showed that the β1 subunit acts by increasing burst duration and P_o by a relatively fixed multiplicative amount in the absence and presence of . If this multiplicative effect were greatly facilitated in some manner by the three mutations producing the TM channel, then an imperceptible Ca²⁺ sensitivity in the TM channel might become pronounced in the presence of the β1 subunit. To test for this possibility we examined whether the β1 subunit in the absence of had a greater effect on burst duration and P_o on the TM channel than on the WT channel. We found that the β1 subunit increased burst duration (5.2± 0.7-fold) and P_o (8.7± 1.3-fold at +50 mV; n = 8) in the TM channel (example in Fig. 1B, upper trace, and results over a range of voltage in Fig. 1D), and that these effects were somewhat less than reported for WT channels (24). These results indicate that the mutations in the TM channel do not facilitate the action of the β1 subunit on burst duration and P_o. If anything, the action of the β1 subunit on burst duration and P_o is reduced for the TM channel. Consequently, the restoration of Ca²⁺ sensitivity to the TM channel by the β1 subunit must arise by some other mechanism.

TM Channels with Restored Ca²⁺ Sensitivity Do Not Incorporate Subunits from Endogenous BK Channels. Endogenous BK channels have been reported at very low levels in Xenopus oocytes (37). Consequently, it is possible in our experiments where the β1 subunit restored Ca²⁺ sensitivity to the TM channels that either endogenous channels or endogenous subunits mixed with TM subunits contribute to the Ca²⁺ sensitivity. To check this possibility, we introduced a point mutation to the TEA binding site located at the outer mouth of the TM subunit. The subunit composition of the channels was then identified from the single-channel current amplitude in the presence of 1.5 mM external TEA (TEAo) (38, 39). For the conditions of these experiments (+50 mV, symmetrical 158 mM KCl), the single-channel current amplitude should be ≈13 pA without TEAo, and should be reduced to ≈9.3, 7.5, 4.0, 1.5, and 0 pA in the presence of 1.5 mM TEAo for channels with 4, 3, 2, 1, and 0 subunits having their TEA binding sites mutated (39).

When oocytes were injected with cRNA for TM channels with the TEA binding sites intact, the single-channel currents were 13 pA in the absence of TEAo and then were completely blocked with 1.5 mM TEAo (Fig. 2A), just as expected for channels with four intact TEA binding sites (38, 39). After injecting cRNA for TM channels with the TEA binding site mutated, the observed single-channel currents were ≈13 pA in the absence of TEAo and ≈9 pA with 1.5 mM TEAo (Fig. 2B), just what would be expected if all four subunits had the TEA binding site mutated (38, 39). Similar results were found in eight additional experiments with TEAo, where the single-channel amplitudes of the TM channels were ≈9 pA with 1.5 mM TEAo. This indicates that the channels being recorded from were homotetrameric channels in which each subunit had the pore mutation to reduce TEA sensitivity and consequently also had the TM. If one or more of the subunits were endogenous (without mutated TEA sites), then the single-channel currents should have been considerably less than 9 pA, as described above. Thus, the β1 subunit restores Ca²⁺ sensitivity to homotetrameric TM channels.

Fig. 2.

The β1 subunit restores the Ca²⁺ sensitivity to channels in which the four α subunits all have the TM, such that no endogenous α subunits from BK channels are present. (A) Single-channel currents from a TM channel in the presence of the β1 subunit are fully blocked by 1.5 mM TEAo, as indicated in the current amplitude histogram. Data are from the same outside-out patch. (B) Single-channel currents in the presence of the β1 subunit from a TM channel with the TEA binding sites mutated are partially blocked by 1.5 mM TEAo, reducing the current from ≈13 pA to ≈9 pA. Data are from the same outside-out patch. Similar results were obtained in eight additional experiments. If one or more of the subunits did not have the TEA site mutation, then the current would have been ≈7.5 pA or less (39).

The β1 Subunit Does Not Directly Provide a Ca²⁺ Binding Site. Another possible mechanism for the restoration of Ca²⁺ sensitivity to the TM channel is that the β1 subunit directly adds a novel Ca²⁺ binding site to the channel. To explore this possibility, we examined whether β1 subunits with the double mutation E13K and T14R to replace the only negative intracellular charge with two positive charges still restored Ca²⁺ sensitivity, because such charge alteration should destroy any potential Ca²⁺ binding site. After the double mutation E13K and T14R to the N terminus, we found that the β1 subunit still restored the Ca²⁺ sensitivity to the TM channel (Fig. 3). Thus, it is unlikely that the β1 subunit is directly adding a Ca²⁺ binding site to the TM channel.

Fig. 3.

The β1 subunit still restores Ca²⁺ sensitivity to the TM channel after replacing the only intracellular negative charge on the β1 subunit with a positive charge (E13K) and adding another adjacent positive charge (T14R). (A) Representative single-channel currents recorded from a TM channel in the presence of the doubly mutated β1 subunit. (B) Plots of P_o vs. for TM channels with the doubly mutated β1 subunit.

M513 Is Not Responsible for Restoring Ca²⁺ Sensitivity to the TM Channel by the β1 Subunit. In addition to the three sites (Ca²⁺ bowl, D362/D367, and E399) in the C terminus of the TM channel associated with Ca²⁺ sensitivity, an additional site, M513, has also been identified in the C terminus that greatly reduces Ca²⁺ sensitivity when mutated (8). The M513 site provides no Ca²⁺ sensitivity to the TM channel even though it has not been mutated, because the TM channel is Ca²⁺-insensitive (Fig. 1 A). Nevertheless, the possibility arises that the M513 site may be involved in the restoration of the Ca²⁺ sensitivity in the presence of the β1 subunit. To explore this possibility, we mutated the M513 site in the TM channel to obtain a quadra mutant channel and then examined whether the β1 subunit still restored Ca²⁺ sensitivity. The quadra mutant channel was found to be Ca²⁺-insensitive in the absence of the β1 subunit (Fig. 4A), just as the TM channel was (Fig. 1 A). The β1 subunit then restored Ca²⁺ sensitivity to the quadra mutant channel (Fig. 4B), just as it did to the TM channel (Fig. 1B). Thus, the M513 site is not responsible for restoring Ca²⁺ sensitivity to the TM channel by the β1 subunit.

Fig. 4.

The M513 site is not responsible for the restoration of the Ca²⁺ sensitivity to the TM channel by the β1 subunit. (A) Single-channel currents recorded from a Ca²⁺-insensitive quadra mutant channel made by adding the mutation M513I to the TM channel. (B) The β1 subunit restores Ca²⁺ sensitivity to the quadra mutant channel, +50 mV.

The β1 Subunit May Restore the Ca²⁺ Sensitivity by Restoring Coupling of Ca²⁺ Binding Sites to the Gating of the Channel. The mutations to specific sites in the TM and quadra mutant channels could remove the Ca²⁺ sensitivity by disrupting the Ca²⁺ binding sites or by disrupting the coupling between the Ca²⁺ sites and the gating machinery. If the mutations disrupt the Ca²⁺ binding sites, then the β1 subunit could act by directly restoring one or more of the mutated Ca²⁺ binding sites. If the mutations disrupt the coupling, then the β1 subunit could restore Ca²⁺ sensitivity by restoring the coupling between the Ca²⁺ binding sites and the gating machinery. It is unlikely that β1 subunits act directly on the Ca²⁺ binding sites, because the β1 subunit has little effect on either the K_d for high-affinity Ca²⁺ binding or the critical levels of Ca²⁺ required to first activate the WT channel (23, 26).

Consistent with the idea that the β1 subunit is unlikely to act through the Ca²⁺ binding sites, we found that the β1 subunit still increased burst duration and P_o for WT channels in which each of the four sites in the quadra mutant channel were mutated individually (data not shown), similar to what we observed previously for a BK channel with a mutated Ca²⁺ bowl (27). Thus, the β1 subunit does not specifically require any one of the four sites identified on the C terminus to either increase the apparent Ca²⁺ sensitivity of the WT channel or to restore Ca²⁺ sensitivity to the triple or quadra mutant channel, suggesting that it may be acting on the coupling between the sites and the gating mechanism.

Nevertheless, it cannot be ruled out that the β1 subunit restores Ca²⁺ sensitivity to the triple and quadra mutant channels by unmasking additional Ca²⁺ binding sites on either S0–S6 (40, 41) or the C terminus that are silent in the triple and quadra mutant channels. A recent study (41) has reported that BK channels with the entire C terminus removed beyond S6 are still gated by . In this case, then, the triple and quadra mutations would have to disrupt the coupling between the proposed Ca²⁺ sites on S0–S6 and the gating, and the β1 subunit would then have to restore the coupling.

The β1 Subunit Does Not Restore Ca²⁺ Sensitivity Through the Low-Affinity Ca²⁺/Mg²⁺ Site. To examine whether the β1 subunit might restore Ca²⁺ sensitivity to the TM channel by recoupling the low-affinity Ca²⁺/Mg²⁺ site to the gating, we took advantage of the observation that activates both the high- and low-affinity sites, whereas activates only the low-affinity E399 site (10, 12). If the restoration of the Ca²⁺ sensitivity to the TM channel by the β1 subunit were through the low-affinity site, then the β1 subunit would be expected to also restore Mg²⁺ sensitivity. If the restoration of Ca²⁺ sensitivity were through a Ca²⁺-specific high-affinity site, then the β1 subunit would not be expected to restore Mg²⁺ sensitivity.

We first verified that activates WT channels (10, 12). Increasing from 1 to 5 to 10 mM increased P_o from 0.05 to 0.2 to 0.4 at 100 mV (Fig. 5A). We also verified (Fig. 5B) that the P_o of the TM channel is not sensitive to for ≤ 10 mM (11). We then examined whether the β1 subunit restored the lost Mg²⁺ sensitivity to the TM channel. In contrast to the restoration of Ca²⁺ sensitivity by the β1 subunit to the TM channel (Fig. 1 B and C), the β1 subunit did not restore the lost Mg²⁺ sensitivity to the TM channel (Fig. 5 C and D). Hence, the β1 subunit does not restore Ca²⁺ sensitivity through the low-affinity Ca²⁺/Mg²⁺ activating mechanism.

Fig. 5.

The β1 subunit does not restore the Mg²⁺ sensitivity to the TM channel, and it has little effect on the Mg²⁺ sensitivity of WT BK channels. (A) Single-channel currents recorded from WT channels at the indicated at +100 mV. (B and C) Currents from TM channels in the absence (B) and presence (C)ofthe β1 subunit. (D) Plots of P_o vs. for TM and WT channels with and without the β1 subunit at +100 mV. (E) Currents from WT channels in the absence and presence of the β1 subunit with 5.7 μM at –50 mV. (F) Currents from WT channels in the absence and presence of the β1 subunit with 0 and 10 mM at +50 mV. (G) Plots of P_o vs. V from experiments like those in E and F (n ≥ 5 for each point in D and G). The open triangles typically obscure the filled triangles.

The β1 Subunit Has No Effect on Activation of the BK Channel by Mg²⁺. If the β1 subunit does not act through the low-affinity Ca²⁺/Mg²⁺ activating mechanism, as suggested in the previous section, then it might be expected that the β1 subunit would not amplify the increased channel activity induced by for WT channels, because Mg²⁺ activates only through the low-affinity Ca²⁺/Mg²⁺ site (9, 10, 12). This was found to be the case. In contrast to the pronounced (22–27) large increase in burst duration and P_o induced by the β1 subunit for WT channels (example for 5.7 μM in Fig. 5E), the β1 subunit had no effect on the burst duration and P_o for WT channels activated by 10 mM (Fig. 5F). This differential effect of the β1 subunit on facilitating the activation of BK channels by , but not , over a range of voltages is shown in Fig. 5G. With 5.7 μM , the β1 subunit shifted the P_o vs. membrane potential curve 75 mV to the left (Fig. 5G, from filled to open diamonds) so that less depolarization was required to activate the channel. In contrast, with 10 mM the β1 subunit had no effect on the P_o vs. V plots (Fig. 5G, triangles). A lack of effect of the β1 subunit on the P_o vs. V plots was also observed with 1, 5, 30, and 100 mM added (Fig. 5D). Hence, in contrast to the facilitating effect of the β1 subunit on burst duration and P_o in the presence of (22–27), the β1 subunit had no effect on burst duration and P_o in the presence of (Fig. 5 D, F, and G).

These results suggest that the β1 subunit does not act through the low-affinity Ca²⁺/Mg²⁺ activating mechanism, but rather acts through the high-affinity Ca²⁺ activating mechanisms. This differential effect of the β1 subunit on the high- and low-affinity activating mechanisms supports the previous proposal that these mechanisms are separate and act relatively independently to facilitate channel activity (10, 12). This differential effect also rules out the possibility that the β1 subunit acts at a common step in the gating mechanism shared by the high- and low-affinity sites, for if it did, the β1 subunit should have enhanced activation by both and , rather than just . Since previous studies have shown that the major action of the β1 subunit is not through changes in the binding affinity of (23, 26), these results when taken together with our observations suggest that the β1 subunit acts on the allosteric linker at a location between the high-affinity Ca²⁺ binding sites and any common elements in the Ca²⁺ and Mg²⁺ gating mechanisms.

Inhibits, Rather than Activates, BK Channels in the Presence of the β1 Subunit for Intermediate Levels of . Previous studies in the absence of the β1 subunit have shown that Mg²⁺ activates BK channels, with the maximal activation at 0 and saturating and the minimal activation at intermediate levels of (10, 12). Since we found in the experiments delineated in previous sections that the β1 subunit has differential effects on the Ca²⁺ and Mg²⁺ activating systems, we examined whether Mg²⁺ still activates BK channels in the presence of the β1 subunit.

The key observation is shown in the single-channel records in Fig. 6 A and B. In the absence of the β1 subunit, adding 10 mM to 5.7 μM had minimal activating effect on the channel. In the presence of the β1 subunit the activity with 5.7 μM was greatly increased (compare Fig. 6 B, upper trace, with A, upper trace), and adding 10 mM to the 5.7 μM now greatly decreased P_o (Fig. 6B, lower trace). Results from experiments of this type over a range of voltages and levels are presented in Fig. 6 C and D. In the absence of the β1 subunit with 5.7 μM,10mM had a minimal activating effect on channel activity (Fig. 6C, squares). In the presence of the β1 subunit with 5.7 μM , 10 mM now inhibited channel activity, as indicated by the 40 mV rightward shift in the P_o vs. V curves (Fig. 6D, from filled to open squares). In contrast to this inhibition by 10 mM Mg²⁺ in the presence of the β1 subunit with 5.7 μM , greatly increased channel activity with 0 or 100 μM , shifting the P_o vs. V plots 50–60 mV to the left, and this was the case in the presence (Fig. 6D) and absence (Fig. 6C) of the β1 subunit.

Fig. 6.

In the presence of the β1 subunit, inhibits the activity of WT channels for intermediate levels of . (A) Single-channel currents recorded from WT channels in the absence of the β1 subunit at 5.7 μM without and with 10 mM at –30 mV. (B) As in A, except for the presence of the β1 subunit. (C) Plots of P_o vs. V for WT channels with the β1 subunit with 0, 5.7, and 100 μM_, each with and without 10 mM . (D)Asin C, except for the presence of the β1 subunit. (E) Plots of the voltage for half activation, V_0.5, vs. for WT channels with and without the β1 subunit, as indicated. (n ≥ 3 for each point in C–E.)

To define the range of levels over which Mg²⁺ activates and inhibits the BK channel in the absence and presence of the β1 subunit, we performed experiments like those shown in Fig. 6 A–D for additional , and we plotted the results as V_0.5, the voltage required for half activation of the channel, vs. . The more negative the V_0.5, the greater the activity of the channel at a fixed voltage. In the absence of the β1 subunit, 10 mM Mg²⁺ always activated the channel, but with a negligible effect at 5.7 μM (Fig. 6E). In the presence of the β1 subunit, 10 mM Mg²⁺ inhibited channel activity for ranging from ≈2 to 30 μM (Fig. 6E, shaded area) and activated the channel for Ca²⁺ <2 μM or >30 μM. This plot clearly shows that the β1 subunit has differential effects on the Ca²⁺ and Mg²⁺ activating mechanisms.

The reason for the inhibition by in the presence of the β1 subunit is not clear, but it has been proposed that Mg²⁺ has two effects on channel activity: activation of the channel through the low-affinity Mg²⁺ sites, and antagonism of the Ca²⁺-dependent activation of the channel by displacement of Ca²⁺ from the high-affinity Ca²⁺ sites (10, 12). Whether the net result of Mg²⁺ is activation or inhibition would depend on the relative contributions of the low-affinity Mg²⁺ site and high-affinity Ca²⁺ sites to the activation. On this basis, our observations can be explained by this hypothesis as follows. At very low or very high , Mg²⁺ would not displace significant Ca²⁺ from the high-affinity sites, so that the activation by Mg²⁺ on the Mg²⁺ sites would be readily apparent (Figs. 5 and 6). At intermediate levels of , Mg²⁺ would displace Ca²⁺ from the high-affinity sites, decreasing the activation by . In the absence of the β1 subunit, adding 10 mM Mg²⁺ to 5.7 μM would have little effect because the reduced activation by due to the displacement of the from the high-affinity sites by Mg²⁺ would be about equal to the increased activation by (Fig. 6 A, C, and E) (10, 12). In the presence of the β1 subunit, which greatly facilitates the activation of the channel by but not by (Fig. 5), adding 10 mM to 5.7 μM would now have an inhibitory effect (Fig. 6 B, D, and E), because the reduced activation by due to displacement of the from the high-affinity sites by would now be greater than the activation by .

Because intracellular free Mg²⁺ can increase to 2–4 mM with ischemia (42, 43), such an increased in the presence of the β1 subunit could decrease the activation of BK channels for intermediate levels of , decreasing the protective ability of BK channels to prevent depolarization. This inhibitory effect of Mg²⁺ on BK channels in the presence of the β1 subunit would act through changes in gating and would be in addition to the inhibitory effect of Mg²⁺ on the conductance described previously (1, 42, 44).

Conclusions. (i) Our results suggest that Ca²⁺ and Mg²⁺ activate BK channels through separate mechanisms, with the β1 subunit acting through the Ca²⁺ activating mechanism rather than the Mg²⁺ activating mechanism. The action of the β1 subunit appears not to be on the Ca²⁺ binding sites, but on the allosteric machinery that couples the Ca²⁺ site to the opening–closing transition. (ii) It has previously been shown that only facilitates the action of BK channels (10, 12). Here we show that, in the presence of the β1 subunit, can also inhibit the activity of BK channels for physiologically relevant levels of . This inhibition results from the fact that the β1 subunit has differential effects on the Ca²⁺ and Mg²⁺ activating mechanisms.