Distinct Regions of the slo Subunit Determine Differential BK_Ca Channel Responses to Ethanol

Abstract

Background:

Ethanol at clinically relevant concentrations increases BK_Ca channel activity in dorsal root ganglia neurons, GH3 cells, and neurohypophysial terminals, leading to decreases in cell excitability and peptide release. In contrast, ethanol inhibits BK_Ca channels from aortic myocytes, which likely contributes to alcohol-induced aortic constriction. The mechanisms that determine differential BK_Ca channel responses to ethanol are unknown. We hypothesized that nonconserved regions in the BK_Ca channel-forming subunit (slo) are major contributors to the differential alcohol responses of different BK_Ca channel phenotypes.

Methods:

We constructed chimeras by interchanging the core and the tail domains of two BK_Ca channel-forming subunits (mslo and bslo) that, after expression, differentially respond to ethanol (activation and inhibition, respectively), and studied ethanol action on these mbslo and bmslo chimeric channels using single-channel, patch-clamp techniques.

Results and Conclusion:

Data from cell-free membranes patches demonstrate that the activity of channels that share a mslo-type core-linker (wt mslo and the mbslo chimera) is consistently and significantly potentiated by acute exposure to ethanol. Thus, a mslo tail is not necessary for ethanol potentiation of slo channels. In contrast, the activity of channels that share a bslo-type core-linker (wt bslo and the bmslo chimera) display heterogenous responses to ethanol: inhibition (in the majority of cases), refractoriness, or activation. Overall, our data indicate that the slo core-linker is a critical region likely contributing to the differential responses of BK_Ca channels to ethanol.

Acute exposure to clinically relevant concentrations of ethanol (EtOH) reversibly increases BK_Ca channel activity in neurohypophysial terminals, where drug action likely contributes to inhibition of vasopressin release and, thus, diuresis (Dopico et al., 1996). A similar potentiation of BK_Ca channels by acute EtOH was reported in GH3 cells (Jakab et al., 1997), and neurons in dorsal root ganglia, where alcohol action on BK_Ca channels effectively reduces neuronal excitability (Gruss et al., 2001). On the other hand, intoxicating [EtOH] usually decrease the activity of BK_Ca channels from aortic myocytes, which is thought to contribute to alcohol constriction of aortic smooth muscle (Dopico, 2003; Walters et al., 2000). While advance was made in linking EtOH action on BK_Ca channels to alcohol effects in the body, the molecular determinants of differential BK_Ca responses to EtOH are unknown.

In supraoptic neurons, EtOH increases nerve ending BK_Ca channel activity but fails to modulate cell body BK_Ca channels. Data were obtained in cell-free, membrane patches, indicating that the continuous presence of cytosolic messengers is not required for EtOH to evoke differential responses from native BK_Ca channels (Dopico et al., 1999).

Native BK_Ca channels consist of pore-forming (α, encoded by the slo1 or KCNMA1 gene), and accessory (β) subunits (Fig. 1A). Following expression in Xenopus oocytes, mslo (from mouse brain; Butler et al., 1993) channel activity is consistently potentiated by 10–200 mM EtOH (Dopico et al., 1998). In contrast, EtOH action on bslo channels (from bovine aortic smooth muscle; Moss et al., 1996) expressed in the same system and under similar conditions, primarily decreases activity. This distinct action of EtOH on bslo channels is not modified by the presence of modulatory β₁-subunits (Dopico, 2003).

Fig. 1.

Depiction of the BK_Ca channel and construction of chimeric slo subunits. (A) BK_Ca channel heterodimer with its pore-forming α (slo) and modulatory β subunits. The slo pore region (P), RCK structure, Ca⁺⁺ bowl, and core and tail domains joined by the linker are indicated. Arrows point to regions with nonconserved residues between mslo and bslo. The C-terminus of slo is much longer in mslo than in bslo. The dotted line indicates the point at which mslo and bslo protein domains were swapped to construct chimeric channels; (B) Construction of chimeric slo by swapping the C terminal region of mslo and bslo. Thus, mbslo contains the mslo sequence from amino acid 1 to 676 (core and linker) and bslo from 677 to 1116 (tail), while bmslo contains the bslo sequence from amino acid 1 to 676 and mslo from 677 to 1169.

Since differential responses to EtOH were found with slo expressed in a similar proteolipid environment, we hypothesized that nonconserved regions between mslo and bslo primarily determine the differential channel responses to EtOH.

The slo subunit has seven putative transmembrane segments (S0–6), a long C-terminus, which includes four hydrophobic regions (S7–10), and a “linker” between S8 and S9 (Fig. 1A) (Weiger et al. 2002). This linker connects two functional domains that determine distinct functional properties of slo channels (Wei et al., 1994). The “core” domain (S0–8) includes the voltage-sensor (in S4), the pore (between S5 and S6), and the RCK structure, which contributes to divalent sensitivity (Jiang et al., 2001). On the other hand, the “tail” domain (S9–10) includes the “Ca⁺⁺-bowl,” a major determinant of the channel Ca⁺⁺ sensitivity (Schreiber and Salkoff, 1997). Primary sequence alignment of mslo (mbr5) and bslo shows that nonconserved regions are scattered all around core, linker and tail (Fig. 1A). The least conserved region, however, is the C terminus, a modulatory region located in the tail: in mslo this region consists of 60 residues; the first 8 being substituted (4 are nonconserved) and the remaining absent in bslo, resulting in a shorter C-end (Butler et al., 1993; Moss et al., 1996).

Thus, to identify which region of the subunit is the primary structural determinant of channel responses to EtOH, “bmslo” and “mbslo” chimeras were constructed by swapping the mslo core-linker with the bslo tail (Fig. 1B). The chimeras were expressed in oocytes and their basic functional properties as BK_Ca channel-forming proteins were determined. Then, we evaluated mbslo and bmslo, as well as wild-type (wt) mslo and bslo, channel responses to EtOH under identical conditions after expression in the same batches of oocytes.

METHODS

Construction of Chimeras and Channel Expression

DNAs encoding mslo mbr5 variant (AAA39746.1), and bslo (AAB03663.1) subunits inserted into the pBluescript vector were cut with ClaI and NotI and reinserted into the pBscMXT vector for oocyte expression. The nucleotide regions coding for the mslo and bslo core-linker and tail (Fig. 1A) were separated by BclI and NotI, respectively. Then, the tail was interchanged after direct ligation of the mslo core-linker with the bslo tail (mbslo) and vice versa (bmslo; Fig. 1B). Wild-type and chimera DNAs were linearized with SalI and XbaI, and cRNAs were synthesized in vitro using T3 polymerase (mMESSAGE mMACHINE; Ambion). Mslo and bslo DNAs were generous gifts from Lawrence Salkoff (Washington University) and Ed Moczydlowski (Yale University).

Xenopus laevis females were maintained in artificial pond water on a 12-hr light/dark cycle. Isolation and injection of Xenopus oocytes were performed as described (Dopico et al., 1998; Dopico, 2003). Wild-type and chimeric cRNAs (0.1–0.5 μg/μl; volumes = 18.4–23 nl) were injected into oocytes with an automated microinjector (Drummond). Patch-clamp recordings were conducted 2–4 days following cRNA injection.

Single-Channel Recordings Single BK_Ca channel currents were recorded in inside-out (I/O) patches using patch-clamp techniques as described (Dopico et al., 1998). The bath solution contained (mM): 130 K⁺ gluconate, 2.63 CaCl₂, 1 MgCl₂, 5 EGTA, 15 HEPES, pH 7.4; free [Ca⁺⁺]≈100 nM. The electrode solution contained (mM): 130 K⁺ gluconate, 5.22 CaCl₂, 2.28 MgCl₂, 5 EGTA, 1.6 HEDTA, 15 HEPES, pH 7.4; free [Ca⁺⁺] = 11 ± 0.6 μM. The free [Ca⁺⁺] was calculated with Max Chelator Sliders (C. Patton, Stanford University) and, for solutions where nominal free [Ca⁺⁺] was >300 nM, determined with a pair of Ca⁺⁺-selective and reference electrodes (Corning). Pipettes were pulled from glass capillaries (Drummond), coated with Sylgard 184 (Dow Corning) to reduce capacitance and noise, and fire-polished with a microforge (WPI, MF200) to give a final tip resistance of 5–8 MΩ when filled with electrode solution. A Ag/AgCl electrode was used as ground electrode.

After excision from the oocyte, the membrane patch was exposed to a stream of bath solution containing either EtOH or urea isosmotically replacing alcohol (control) flowing from a “sewer” micropipette (1-mm diameter; WPI). Deionized (100% purity) EtOH (American Bioanalytical) was freshly diluted in bath solution immediately before each experiment. Patch excision and EtOH/urea application (≤1 min) were conducted as described elsewhere (Dopico, 2003). All experiments were carried out at 20–25°C.

Unitary currents were measured using an EPC8 amplifier (List), low-pass filtered at 1 kHz with an 8-pole Bessel filter (Frequency Devices 902), and sampled at 10 kHz. Data were acquired and analyzed using pClamp 8 software and a 1320 Digidata interface (Axon). Voltages given correspond to the potential at the intracellular side of the patch.

Data Analysis The product of the number of channels in the patch (N) and the probability that a channel is open (P_o) was used as index of channel steady-state activity. NP_o was calculated from all-points amplitude histograms as described elsewhere (Dopico et al., 1996). Channel mean open time (t_o) was calculated by weighting the time constants obtained from the open times distribution. Dwell time histograms were constructed using the half- amplitude threshold criterion, and an F Table (p < 0.01) was used to determine the minimum number of exponential components to appropriately fit dwell-time histogram data (Dopico et al., 1998). NP_o and t_o values were obtained from periods of at least 20 sec of continuous channel activity recording. Data are expressed as mean ± SEM. Since each oocyte was patched only once, n = number of membrane patches = number of oocytes. Statistical analysis of data was conducted according to Glantz (2001).

RESULTS AND DISCUSSION

Expression ofmbslo or bmslo subunits in Xenopus oocytes led to single channel events that displayed all major characteristics of BK_Ca unitary currents when recorded in I/O patches. First, the NP_o of both chimeric channels obtained at any given voltage increased similarly with increases in [Ca⁺⁺]_ic (Fig. 2). Second, at a fixed free [Ca⁺⁺]_ic(∼100 nM), the voltage-NP_o relationship of both chimeric channels could be described by a Boltzmann function, in which at low NP_o, the ln (NP_o) vs. voltage plot is linear. Thus, the inverse of the slope is the potential needed to produce an e-fold change in NP_o (Dopico et al., 1996): 13 ± 2 mV (n = 3) and 14 ± 3 mV (n = 3) per e-fold change in NP_o for mbslo and bmslo. These values are not only similar (p > 0.05), but also fall within the range previously reported for both mslo and bslo channels (Dopico et al., 1998; Dopico, 2003), other slo channels (Butler et al. 1993; Crowley et al., 2003; DiChiara and Reinhart, 1995), and native BK_Ca channels (Dopico et al., 1996; Toro et al., 1990). This similarity in the voltage-dependence of gating likely reflects the high identity in the voltage-sensing region (in S4) across slo subunits.

Fig. 2.

The steady-state activity of both bmslo (top pair) and mbslo (bottom pair) channels increases with increases in internal [Ca⁺⁺]. Representative traces of activity showing increases in channel NP_o when free [Ca⁺⁺] at the intracellular side of I/O patches is raised from 100 nM to 1 μM (Vm = −20 mV). Each pair of traces was obtained from the same patch. Arrows indicate the baseline; channel openings are shown as downward deflections. For display, records were digitally filtered at 1 kHz. NP_o values were obtained from periods of at least 20 sec of continuous recording.

Third, unitary current-voltage relationships from both chimeric constructs obtained in 130 mM K⁺ at both sides of the membrane were ohmic from −40 to +60 mV (not shown) rendering similar (p > 0.05) slope conductances: 253 ± 17 (n = 10) and 254 ± 14 (n = 10) pS for mbslo and bmslo. These values are similar to those reported with wt mslo (Dopico et al., 1998) and bslo (Dopico, 2003) studied in the same system and under similar conditions. The fact that mslo, bslo, mbslo and bmslo channels have the same conductance is a functional correlate of the total identity existing in the primary sequence of the pore-lining region across these slo subunits (Butler et al., 1993; Moss et al., 1996).

Finally, the mean open time, largely controlled by the slocore (Wei et al., 1994), was determined in both chimeras under identical conditions (V = 20 mV; free [Ca⁺⁺]_ic∼100 nM): 1.6 ± 0.3 and 1.4 ± 0.1 msec for mbslo and bmslo (n = 3; p > 0.05, unpaired Student's t test).

After establishing that expression of mbslo and bmslo elicited unitary events that fulfilled the basic characteristics of BK_Ca channel unitary currents, we evaluated changes in chimeric channel activity evoked by acute exposure to 100 mM EtOH. This [EtOH] represents the E_max for both EtOH inhibition of bslo (Dopico, 2003) and EtOH activation of mslo channels (Dopico et al., 1998). Chimeric channel responses to EtOH were also compared to those from wt mslo and bslo BK_Ca channels expressed in the same batches of oocytes.

Acute exposure of the cytosolic side of I/O patches to 100 mM EtOH reversibly increased mslo channel NP_o in 11 out of 11 patches. In contrast, EtOH evoked varied responses in bslo activity: decrease in 5 out of 7 patches (72% of cases), increase in 1 (14%) and no change in another (14%) patch (Fig. 4A). Despite high interpatch variability, data from all samples passed the Kolmogorov-Smirnov normality test (p > 0.1). Thus, individual means could be statistically tested with parametric methods, which established that the average responses to EtOH from mslo versus bslo were markedly different (p < 0.001, ANOVA and Tukey-Kramer's test) (Fig. 4B). These findings with wt slo channels are identical to those obtained using a wide range of [EtOH] (10–200 mM) with mslo, in which NP_o was potentiated in 97% of cases (Dopico et al., 1998), and bslo, which displayed varied responses to 3–200 mM EtOH (Dopico, 2003). Since differences between bslo and mslo were obtained in the same batches of oocytes, current data indicate that differential responses to EtOH from mslo and bslo channels cannot be explained by differences among batches of oocytes. Rather, since the differential responses were evoked in a similar proteolipid environment, they are likely determined by the existence of nonconserved regions between mslo and bslo. Ethanol responses of chimeric channels, as shown below, support this hypothesis.

Fig. 4.

Changes in the steady-state activity of wild type and chimeric channels in response to acute ethanol (100 mM). (A) Acute exposure of the cytosolic side of I/O patches to EtOH evoked varied responses in bslo and bmslo, the majority of channels being inhibited (see text), while routinely potentiating mslo and mbslo activity after expression in the same batches of oocytes. Ratios of NP_o values obtained in the presence and absence of EtOH (x100) are shown in a scatter graph, where each data point represents an individual patch/oocyte; (B) Average changes in channel NP_o in response to EtOH are shown as mean ± SEM, where n (number of patches/oocytes) is shown in parentheses at the bottom of each bar graph. In the presence of alcohol, NP_o reaches 67 ± 10 (bslo), 105 ± 22 (bmslo), 155 ± 15 (mbslo), and 196 ± 17% (mslo) of pre-EtOH values. *Significantly different when compared to bslo (p < 0.05); ^#Significantly different when compared to bslo (p < 0.001); ¶Significantly different when compared to bmslo (p < 0.01). Multicomparisons were performed with one-way ANOVA, followed by Tukey-Kramer's test. In both A and B, a dotted line highlights the point at which NP_o is unchanged by EtOH.

As found with mslo, EtOH exposure of I/O patches expressing mbslo channels resulted in a fast and reversible increase in channel NP_o (Fig. 3). This response was observed in 93% of cases (13 out of 14 patches; Fig. 4A). Thus, for both mslo and mbslo we found an increase in channel NP_o in response to EtOH in almost every patch/oocyte. In addition, not only the frequency, but also the average EtOH potentiation of NP_o was statistically similar in mbslo and mslo: 155 ± 15 and 196 ± 17% of pre-EtOH controls (p > 0.05; ANOVA followed by Tukey-Kramer test) (Fig. 4B). Therefore, the long C-terminus existing in mslo but missing in mbslo is not necessary for EtOH potentiation of slo channel activity.

Fig. 3.

As found with mslo, ethanol routinely increases mbslo channel NP_o. Representative single channel recordings obtained before and during the application of 100 mM EtOH to the cytosolic surface of the same I/O patch. Records were obtained >10 min after patch excision (Vm = 40 mV). Arrows indicate the baseline; channel openings are shown as upward deflections. For display, records were digitally filtered at 1 kHz. NP_o values were obtained from periods of at least 20 sec of continuous recording.

On the other hand, bmslo activity in response to EtOH was decreased in 8 (57%), increased in 4 (29%) and basically unchanged in 2 out of 14 (14%) patches (Fig. 4A). This heterogeneous pattern of responses to EtOH, the predominant being channel inhibition, is similar to that observed in wild-type bslo. In fact, there was no statistical difference (p > 0.05) in the average changes in NP_o induced by EtOH between bmslo and bslo (Fig. 4B). Thus, inhibitory responses evoked by EtOH (Fig. 4A) are only obtained with slo constructs that share a bslo-type core domain-linker region (Fig. 1A). In contrast, potentiation by EtOH is the only response observed in slo constructs sharing a mslo-type core-linker region. However, since EtOH-induced channel activation could also be evoked in bmslo (Fig. 4A), we conclude that, while the mslo core-linker is not necessary for EtOH potentiation of slo channel activity, the presence of a mslo core-linker “locks” (or shifts) the channel population into one state(s) that is activatable by alcohol.

Alignment of the primary sequences of mbslo and bmslo in the core-linker shows several, scattered, nonconserved residues: one in the loop between S0 and S1, and five more in the S8–9 linker. In addition, the mslo linker includes an IYF insert that is absent in bslo. Interestingly, a double S in the bslo linker constitutes a consensus site for protein kinase C phosphorylation, a process that modulates both basal BK_Ca channel activity (Schubert and Nelson, 2001) and alcohol action on native BK_Ca channels (Jakab et al., 1997). Different degrees of phosphorylation from cell to cell might contribute to the variant responses to EtOH observed in slo channels having a bslo core-linker (bslo and bmslo). An ongoing systematic study using pinpoint mutagenesis and pharmacology will determine the role of each of these nonconserved residues in the slo core-linker and , eventually, their posttranslational modification, in EtOH action on BK_Ca channel function.

We previously showed that the distinct EtOH response of bslo channels, when compared to mslo, is unmodified by coexpression of β₁ subunits (Dopico, 2003), and hypothesized that the differential alcohol responses of mslo versus bslo should be primarily attributed to non-conserved regions in these slo subunits. Now, we demonstrate that the primary structural determinant underlying the differential EtOH responses of mslo versus bslo channels is not their long, nonconserved C-terminus, but the core domain-linker region. Thus, the primary sequence of the slo subunit is a critical factor determining BK_Ca channel responses to alcohol.