Molecular mechanism of pharmacological activation of BK channels

Cym04 Opens Slo1 BK Channels in a Splice Variant-Specific Manner.

The structural element linking the channel's pore domain and the cytosolic domain, the S6/RCK linker, plays an important role in allosteric gating of the Slo1 BK channel (17, 24, 25). This linker segment is commonly encoded by exon 9 of the human Slo1 gene. However, we recently identified a splice variant of slo1 using a mutually exclusive alternative exon 9. Slo1 BK channels with such alternative exon 9, here referred to as Slo1_9a, are primarily found in neuronal tissue and display an altered dependence on the membrane potential and the [Ca²⁺]_i (24), consistent with the postulated importance of the S6/RCK linker in the allosteric gating mechanism (25). Because the BK opener Cym04 likely alters the gating machinery, we compared how Cym04 affected channels with the common S6/RCK linker (Slo1) or the distinct linker coded by the alternative exon 9 (Slo1_9a). Slo1 or Slo1_9a channels were expressed in HEK 293 cells, and modulation by Cym04 was analyzed in the inside-out configuration of the patch-clamp method. Experiments were initially performed in the virtual absence of Ca²⁺_i to simplify channel gating.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Characterization of BK channel activation by Cym04. (A) Slo1 current traces showing channel activation (Upper) and deactivation (Lower) at the indicated voltages before (Left) and after application of 10 μM Cym04 (Right). The superimposed colored curves are results of single-exponential fits. (B) Current traces at 50 and −50 mV before (black) and after (red) application of 10 μM Cym04. Bars left of the current traces indicate the current levels corresponding to closed channels (“c”) or increasing numbers of open channels (o₁, o₂). (C) Whole-cell gating currents (Left) upon depolarization to the indicated voltages before (black) and after (red) application of 10 μM Cym04. Corresponding gating charge (Q) was normalized to the maximal value (Q_max), averaged (n = 9) and plotted against voltage (Right). Continuous curves are Boltzmann fits. All data in A–C were recorded in the absence of intracellular Ca²⁺. (D) Normalized conductance-voltage plots for Slo1 currents from inside-out patches measured in the presence of 100 μM Ca²⁺ (squares) or 10 mM Mg²⁺ (circles) before (open symbols) and after (red filled symbols) application of 10 μM Cym04. Data points are mean ± SEM (n = 6), and continuous curves are data fits according to Eq. 1. (E) Cym04-induced shifts in half-maximal activation voltage (ΔV_0.5) and slope factor (ΔV_e) for experiments as shown in D at different [Ca²⁺]_i and [Mg²⁺]_i, as well as in the absence of divalent cations for Slo1 coexpressed with Slo β1 subunit and for BK channels in LNCaP cells. Vertical red lines indicate control values for Slo1 channels in the absence of divalent cations. Data points are mean ± SEM, n indicated in parentheses. *P < 0.05; **P < 0.01, comparing with Slo1 in zero [Ca²⁺]_i.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

The S6/RCK linker is a molecular determinant of BK channel activation by Cym04. (A) A plausible 3D structure of a tetrameric Slo1 BK channel. The transmembrane segment (green) is based on the homology model based on the Kv channel structure according to Yuan et al. (20) (Protein Data Bank #2R9R), and the cytosolic domain (blue) is according to Wu et al. (21) (Protein Data Bank #3NAF). The image was created using MacPyMOL v0.99. The arrow points to the probable location of the S6/RCK linker segment, which is not resolved in the available crystallographic structures. For size comparison one Cym04 molecule is also shown. (B) Comparison of amino acid sequences encoded by Slo1 exon 9 (Slo1) and the alternative exon 9 (Slo1_9a). Grouping of residues (“End S6,” “Proximal linker,” “Distal linker”) indicates the three areas for chimera construction. Dots represent conserved identical residues. S6 is assumed to end at I326; +++ denotes a cluster of positive charges and ΔΔ a site of deletion. (C) Mean half-maximal activation voltage (V_0.5, Left) and slope factors (V_e, Center Left) of Slo1 BK channel variants under control conditions obtained from Boltzmann fits to normalized tail current recordings. Corresponding shifts in half-maximal activation voltage (ΔV_0.5, Center Right) and slope factors (ΔV_e, Right) induced by 10 μM Cym04. Data are mean ± SEM (n in parentheses). *P < 0.05; **P < 0.01; ***P < 0.001, comparing with Slo1 (999) channels according to two-sided unpaired Student t tests and Holm-Bonferroni correction.

Current responses of both Slo1 isoforms to 10-ms depolarizations are compared in Fig. 1A before and after application of 10 μM Cym04 to the cytosolic side. At the voltages at which channels are activated to <5%, Cym04 rapidly increased the Slo1 current amplitude by approximately fivefold in a reversible manner and did not require previous channel opening (Fig. S1). The increase in Slo1 current size was not accompanied by a noticeable change in the single-channel current size (Fig. 2B); single-channel conductance of Slo1 channels in the range of −100 to 50 mV was unaffected by 10 μM Cym04: 240 ± 3 pS (control) and 244 ± 4 pS (Cym04, n = 12, P = 0.39). The current increase is thus most probably caused by changes in gating of Slo1 by Cym04. Analysis of tail currents indeed showed that Cym04 markedly shifted the half-activation voltage (V_0.5) estimated using a modified Boltzmann equation (Eq. 1). Cym04 (10 μM) shifted V_0.5 of Slo1 channels from 213 ± 4 mV to 184 ± 7 mV (P < 0.001, n = 12). The slope of the curve was unaffected; V_e values were 35.1 ± 1.1 mV (control) and 36.3 ± 0.7 mV (P = 0.21).

In contrast with the significant current increase observed with Slo1, Cym04 failed to alter currents through Slo1_9a even when examined over a wide voltage range (Fig. 1A, Right). The V_0.5 values in Slo1_9a were 104 ± 2 mV and 107 ± 2 mV before and after application of Cym04, respectively (P = 0.20, n = 7). Cym04 therefore is a BK opener that clearly discriminates between the two splice variants that differ in the S6/RCK linker region.

We examined whether other BK openers share the specificity of Cym04 for the Slo1 isoform over Slo1_9a. NS1619 is a well-known BK opener (10) but belongs to a structural class distinct from that of Cym04 (Fig. 1B, Left). We chose 30 μM NS1619 to obtain a shift in V_0.5 comparable to that obtained with 10 μM Cym04 in Slo1 (ΔV_0.5 = −36.0 ± 2.6 mV, n = 7; for 10 μM NS1619: ΔV_0.5 = −21.6 ± 3.9 mV, n = 5). At this concentration, frequently used for in vitro tissue studies, NS1619 did not activate Slo1_9a channels (Fig. 1B, Right). The preferred action of both Cym04 and NS1619 on Slo1 over Slo1_9a suggests that, despite the structural differences, the two BK openers may activate the channels via a common mechanism hinging on the S6/RCK linker. Unlike with the splice variant-specific action of Cym04 and NS1619, mallotoxin (MTx, 2.5 μM), a naturally occurring BK opener in a distinct structural family (27), activated both Slo1 and Slo1_9a by causing a similar large shift in V_0.5 (ΔV_0.5 = −38.1 ± 14.7 mV, n = 4 in Slo1 vs. ΔV_0.5 = −29.0 ± 11.7 mV, n = 5 in Slo1_9a). Slo1 and Slo1_9a are a promising pair of BK channels suited to study the molecular mechanism and isoform specificity of BK activation by Cym04 and NS1619-like BK openers.

Characterization of Slo1 BK Channel Activation by Cym04.

Changes in electrophysiological characteristics of Slo1 channels, whether by mutations or signaling molecules, can be interpreted using the established gating model of Slo1 (“HA model”; SI Results, and SI Materials and Methods) to gain biophysical and molecular insight (28). Using the HA model as the conceptual framework, we performed an analysis of the changes in Slo1 ionic currents caused by 10 μM Cym04. The concentration dependence of Slo1 channel activation by Cym04 as measured by the shift in V_0.5 showed that the half-maximal activation was obtained at 5.9 μM, with a Hill coefficient of 1.6 (Fig. S2). Thus, 10 μM Cym04 represents a near-saturating concentration for Slo1 and was used in most of our experiments. The electrophysiological measurements showed that Cym04 had the following two primary effects on Slo1 currents in the virtual absence of Ca²⁺. First, Cym04 decelerated deactivation at very negative voltages by ≈2.5-fold but had no effect on activation at very positive voltages (Fig. 2A and Fig. S3). Second, the effectiveness of Cym04 on Slo1 diminished but did not completely disappear at weaker depolarization (Fig. 2B and Fig. S4). For example, at 50 mV, Cym04 induced a 27 ± 11-fold increase in open probability, compared with the 7 ± 3-fold increase found at −50 mV (P < 0.05, n = 6). These observations and others, as well as our simulations, showed that the two main effectors of Cym04 are the channel's voltage-sensing mechanism and the ion conduction gate (Fig. S4), stabilized by ≈3.0 kJ/mol and ≈1.0 kJ/mol, respectively. Because Cym04 did not significantly alter Slo1 gating currents (Fig. 2C), an impact of the channel opener on the coupling of sensor and gate is suggested.

Our analysis of the results obtained in the virtual absence of Ca²⁺ suggests that the activating effect of Cym04 may be independent of the Ca²⁺- and Mg²⁺-dependent activation pathways of the BK channel (29⇓⇓–32). Consistent with this prediction, with a functionally saturating concentration of cytosolic Ca²⁺ (100 μM) or Mg²⁺ (10 mM), Cym04 (10 μM) remained effective in shifting the voltage dependence to the negative direction (Fig. 2 D and E), indicating that the channel opener does not directly act as a mimic of these divalent cations. At intermediate [Ca²⁺]_i, Cym04 remained effective in activating BK channels; there was a trend for a decrease in ΔV_0.5 (Fig. 2E), which might be indicative for an increase in the apparent Ca²⁺ binding constant, but the results were statistically insignificant.

The action of Cym04 was also independent of the presence of the auxiliary subunit β1 (33). In hSlo1+β1 channels, Cym04 shifted V_0.5 by −40.1 ± 4.9 mV (n = 6), which is not significantly greater than that with Slo1 alone without β1 (Fig. 2E). Likewise, coassembly with LRRC26, an auxiliary protein conferring a substantial (>100 mV) left-shift of V_0.5, as for BK channels in LNCaP cells (34), was without effect: in the virtual absence of intracellular Ca²⁺ Cym04 shifted LNCaP whole-cell BK channel activation by −32.5 ± 1.9 mV (n = 6) (Fig. 2E). In summary, Cym04 is a reversible BK channel opener that affects channel gating without any requirement for intracellular divalent cations or the presence of accessory subunits.

S6/RCK Linker Is a Molecular Determinant for BK Channel Activation by Cym04.

The striking difference in Cym04 response between Slo1 and Slo1_9a channels (Fig. 1) served as a starting point for the identification of molecular determinants required for the BK opener action. The two variants differ in 13 nonconsecutive amino acid residues, ranging from the transmembrane segment S6 in the pore domain to the linker to the RCK1 domain (S6/RCK linker) (Fig. 3 A and B). To reveal which of these 13 residues are important for Cym04 action on BK channels, we generated chimeras in the background of Slo1 channels, dividing the area of interest into three regions: the distal part of S6 (which is typically considered to end at I326), a proximal linker segment, and a distal linker segment. Thus, Slo1, in which Cym04 causes a ΔV_0.5 of −29 mV, is represented here as “999,” whereas Cym04-insensitive Slo1_9a is represented as “aaa” because the three segments are coded by alternative exon 9. Chimeras were tested to identify which substitution impairs the Cym04-induced shifts in V_0.5. Although ΔV_0.5 of chimera a99 with the alternative S6 segment was not significantly different (P = 0.12) from that of Slo1 (“999”) channels, the shift was significantly smaller for chimera 9aa (P < 0.005), indicating that the proximal and/or distal linkers are crucial for Cym04 action on Slo1 channels. The high-impact area was further narrowed down to the proximal linker part because chimera 9a9 (P < 0.005) but not 99a (P = 0.63) exhibited attenuated Cym04-induced activation. Within the proximal linker segment, Slo1 and Slo1_9a differ at four amino acid positions (327, 328, 330, and 332) (Fig. 3B). The results from the single point mutations changing each position from that found in Slo1 to Slo1_9a revealed that the mutations G327L, K330N, and Y332F, but not N328S, significantly diminished—but not completely abolished—the left-shift in ΔV_0.5 by Cym04.

The particularly diminished effect of Cym04 in K330N (Fig. 3C) led us to examine whether substitutions of other amino acids at this position also impaired the activating effect of Cym04. Neither the charge reversal mutation K330E nor the charge neutralization mutation K330A diminished the left-shift in V_0.5 by Cym04 (Fig. 3C).

Many of the channel mutants with diminished sensitivity to Cym04 identified above including Slo1_9a exhibited a left-shifted V_0.5 under control conditions (Fig. 3C, Left). Because Cym04 causes a left shift in the voltage dependence, the diminished effectiveness of Cym04 in these mutants may be a universal feature associated with any left-shifted mutant. We therefore tested whether other mutations known to induce a left shift in BK channel gating also attenuated BK channel activation by Cym04. Four single-site mutations—M442A in the cytosolic RCK1 intersubunit interface, G733D in the RCK2 intrasubunit interface, F315Y in the transmembrane segment S6, and R207G in the transmembrane segment S4—induced shifts in V_0.5 by approximately −50 mV under control conditions but failed to attenuate the Cym04 effect (Fig. 4B). Because of the slow activation time course of mutant F315Y, the use of 10-ms depolarizations will overestimate V_0.5 values; however, even under strongly extended pulse protocols Cym04 still left-shifted V_0.5 (Fig. S5). Therefore, the diminished effectiveness of Cym04 is not a common feature of every left-shifted mutant.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

The S6/RCK linker specifically determines whether Cym04 activates BK channels. (A) Current recordings of the indicated Slo1 mutants with altered linker length before (black) and after application of 10 μM Cym04 (gray). (B) Half-maximal activation voltages (V_0.5, Left) and slope factors (V_e, Center Left) for the indicated constructs obtained from Boltzmann fits to normalized tail current recordings. Corresponding shifts in V_0.5 (Center Right) and V_e (Right) upon application of 10 μM Cym04. Data are mean ± SEM (n in parentheses). *P < 0.05; ***P < 0.001, comparing with Slo1 channels according to two-sided unpaired Student t tests and Holm-Bonferroni correction.

The S6/RCK linker has also been implicated to play a pivotal role in voltage-dependent activation of BK channels, in particular because shortening the linker left-shifts the voltage dependence (25). Therefore, Cym04 may trigger the same conformational change induced by shortening of the S6/RCK linker. This postulate in turn suggests that shortening of the linker renders Cym04 less effective. We found that a deletion of two amino acids, Y336 and either S335 or S337, in the distal linker segment of Slo1 (Fig. 3B; “LL−2” in Fig. 4) shifted the voltage dependence of activation by −33 mV under control conditions and diminished BK channel activation by Cym04 (Fig. 4). In contrast, an insertion of two alanines between S337 and A338 in Slo1 in the distal linker segment (Fig. 3B; “LL+2” in Fig. 4) caused a 39-mV right shift of V_0.5 under control conditions but did not affect BK activation by Cym04 (Fig. 4B).

The reduced potency of Cym04 in mutant LL−2 may either result from a reduction in linker length per se or from structural changes at the deleted amino acid positions. The first possibility seems unlikely because alternative double-residue deletions at neighboring sites did not reduce the left-shift in V_0.5 upon Cym04 application (Fig. S6). Single-site mutations at positions S335, Y336, and S337 did not noticeably diminish the impact of Cym04 with the exception of Y336F (Fig. S6). Moreover, mutations Y336G, -D, and -E strongly impaired functional channel expression.