HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 30, 27045-27052, July 26, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/30/27045    most recent M203087200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Erxleben, C. Articles by Armstrong, D. L. Search for Related Content PubMed PubMed Citation Articles by Erxleben, C. Articles by Armstrong, D. L. Social Bookmarking                      What's this?

Interacting Effects of N-terminal Variation and Strex Exon Splicing on slo Potassium Channel Regulation by Calcium, Phosphorylation, and Oxidation^*

Christian Erxleben, Angela L. Everhart, Charles Romeo, Hannah Florance^§, Mary Beth Bauer, David A. Alcorta, Sandra Rossie^¶, Michael J. Shipston^§, and David L. Armstrong

From the  Laboratory of Signal Transduction, NIEHS, National Institutes of Health, Research Triangle Park, North Carolina 27709, the ^§ Membrane Biology Group, University of Edinburgh, Edinburgh EH8 9XD, Scotland, and the ^¶ Department of Biochemistry, Purdue University, West Lafayette, Indiana 47907

Received for publication, April 1, 2002, and in revised form, May 10, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We have investigated the structural basis for the phenotype of a native rat Slo (rSlo) potassium channel (BK_Ca; KCNMA1) in a rat pituitary cell line, GH₄C₁. Opposing regulation of these calcium- and voltage-activated potassium channels by cAMP- and cGMP-dependent protein kinases requires an alternatively spliced exon (strex) of 59 amino acids in the cytoplasmic C terminus of the pore-forming  subunit encoded by rslo. However, inclusion of this cysteine-rich exon produces a 10-fold increase in the sensitivity of the channels to inhibition by oxidation. Inclusion of the strex exon also increases channel sensitivity to stimulation by calcium, but responses in the physiological ranges of calcium and voltage require coassembly with ₁ subunits. With strex present, however, ₁ subunits only stimulated channels assembled from rSlo  subunits with a truncated N terminus beginning MDALI- . Thus N-terminal variation and strex exon splicing in rSlo interact to produce BK_Ca channels with a physiologically relevant phenotype.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Only one family of potassium channels (BK_Ca; KCNMA) responds directly to both depolarization and intracellular calcium, which allows them to integrate cellular excitability and provide critical feedback inhibition (1-3). In mammalian tissues, a single gene, slo (KCNMA1), encodes these big conductance, Ca²⁺-activated potassium channels (4). Nevertheless, the phenotype of BK_Ca channels varies widely between tissues, between cells of the same tissue (5, 6), and in the same cell types under different hormonal environments (7, 8). Several factors that contribute individually to this diversity have been identified, including alternative splicing (9, 10), regulatory  subunits (11, 12), and protein phosphorylation (13, 14), but the influence of these factors on one another has not been studied in detail. For example, most recombinant forms of BK_Ca channels are stimulated by the cAMP-dependent protein kinase (PKA)¹ (15), but the response of BK_Ca channels to PKA in a mouse pituitary cell line, AtT20 (16, 17), is reversed by insertion of a 59-amino acid exon (strex) at splice site 2 in the C terminus (18). However, most studies of recombinant BK_Ca channel regulation by protein phosphorylation have employed a cDNA that begins at the third potential initiator methionine (MDALI-) encoded by the slo gene (19), omitting up to 65 amino acids in the N terminus. Because this extracellular N-terminal sequence of the pore-forming  subunit interacts with the  subunit (20), which also increases calcium sensitivity, we postulated that these structural changes might interact with one another. In addition, we wondered whether some of the variability in the regulation of BK_Ca channels by exogenous protein kinases might result from nonspecific effects associated with stimulation of the channel by the sulfhydryl reducing agents (21, 22), which are routinely included in purified preparations of the enzymes.

We have previously characterized the regulation of native BK_Ca channels by protein phosphorylation in a rat pituitary cell line, GH₄C₁ (23-25). These channels are stimulated by the cGMP-dependent protein kinase (PKG) and inhibited by PKA, which is consistent with effects of hypothalamic neuropeptides on pituitary cell excitability and secretion (26) and with BK_Ca channel behavior in cell-free patches from GH₄C₁ cells (27). However, to date no one has reported a recombinant form of BK_Ca that displays all these properties. Therefore, we believed it would be instructive to identify the structural basis for BK_Ca regulation in the GH₄C₁ cells. The data reveal new interactions between the extracellular N terminus and the cytoplasmic C terminus of the BK_Ca  subunit with potential physiological significance for channel regulation by calcium, oxidation, and phosphorylation.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture and Transfection-- HEK293 cells, a human embryonic kidney cell line (ATCC CRL-1573), were cultured on glass coverslips (Deutsche Spiegelglas, Carolina Biological) in Dulbecco's modified Eagle's medium (H-21) supplemented with penicillin and streptomycin and 10% fetal bovine serum (Hyclone). Channel constructs in the pCI-neo vector were coexpressed transiently with a plasmid-encoding green fluorescent protein (GFP) at a DNA ratio of 8:1 using LipofectAMINE 2000 (Invitrogen) as described by the manufacturer. In less than 24 h, excised patches from weakly fluorescent cells contained hundreds of BK_Ca channels. Stable cell lines for single-channel recordings were generated by transfection of Ecr293 cells (Invitrogen), a HEK cell line stably expressing the ecdysone receptor, with constructs in the ecdysone-inducible pIND vector (Invitrogen) and selected in G418. When these cells were grown in the absence of ecdysone, basal transcription from the plasmid was sufficient to produce just enough channels for reliable single-channel recording from excised patches of membrane. Native BK_Ca channels were studied in GH₄C₁ cells, a rat pituitary cell line, which were cultured in Ham's F-10 medium supplemented with penicillin, streptomycin, 15% equine serum, 2.5% fetal bovine serum, and 10 mg/liter phenol red as described previously (27). It should be noted that GH₄C₁ cells cultured without phenol red had a much lower incidence of BK_Ca modulation by PKA (19%, n = 3/16) than cells in phenol red, which showed inhibition by PKA in 85% of patches (n = 22/26).

Molecular Biology-- Both rSlo and ₁ subunits of the BK_Ca channel were cloned by RT-PCR from total GH₄C₁ cell RNA using primers based on the GenBank^TM sequences U55995 isolated from adult female rat myometrium and U79661 from rat cerebral vascular smooth muscle. An EcoRI-MluI oligonucleotide adapter designed to encode an N-terminal FLAG epitope with a short linker region (MVDYKDDDKLGGGATR) was inserted into a multiple cloning site of pCI (Promega) to generate pCI-FLAG. The forward primer covered U55995 nucleotides 64-87 (5'-ATG AGC AAT ATC CAC GCG AAC CAT), whereas the reverse primer was from nucleotides 3574-3597 (the reverse complement of 5'-CCA GAA ACA TTC AAA TCA AGC CCA taaagcggccgc) and included a TAA stop codon and NotI cloning site. An endogenous MluI site (position 98 in U55995) was used as the 5'-cloning site for the PCR product. This necessitated the insertion of an adapter oligonucleotide sequence between the MluI site in the pCI-FLAG vector and the internal rSlo MluI site to encode the N-terminal 12 amino acids of the mature peptide reported in U55995. The resulting clone expressed rSlo-FLAG. The N-terminal FLAG sequences were removed from rSlo-FLAG by swapping out an EcoRI-MluI fragment with the oligonucleotide sequence 5'-GAA TTC CCC ATG AGC AAT ATC CAC GCG AAC CAT CTC AGC CTA GAC GCG T-3' to generate rSlo. Five full-length clones of the rSlo  subunit were isolated independently in separate PCR reactions and sequenced on both strands. All clones contained identical N and C termini (Fig. 1) with no inserts at known sites of alternative splicing. The GH₄C₁ clone of rSlo is identical to the rat smooth muscle sequence (U55995) up to Gln-1140, but then it encodes only seven additional amino acids before terminating in a stop codon.

View larger version (32K): [in this window] [in a new window]   Fig. 1.   Primary structure of rSlo isolated from GH4C1 cells. A, N-terminal alignment of rSlo with slo gene products isolated from rat smooth muscle (U55995) and human brain (U11717). B, the C terminus of the same clones indicating putative phosphorylation sites for PKG (15) and PKC (36). C, the strex exon of 59 amino acids that is inserted at Leu-668 and contains a putative phosphorylation site for PKA (17). D, diagram of rSlo topology and the variants used in this study.

The strex insert was isolated from total GH₄C₁ cell RNA by RT-PCR using an anchored deoxyoligo (dT) primer (5'-TTCTAgAATTCAgCggCCgC(T)₃₀N₁N₂) and Superscript II reverse transcriptase (Invitrogen) for first-strand cDNA synthesis and forward (5'-gTC CTT CCC TAC TgT TTg) and reverse (5'-gTg TTT gAg CTC ATg ATA gT) PCR primers spanning splice site 2. Inserts containing rat strex were sequenced on both strands and subcloned into unique BlpI and PacI restriction sites of rSlo to give the rSlo-strex construct (Fig. 1). N-terminal truncations (N) of both rSlo and rSlo-strex were generated by site-directed mutagenesis of the initiator methionine (MSNIH-) to alanine, forcing transcription to begin at the next potential initiator methionine, MDALI- (Fig. 1).

Electrophysiology-- Slo channel currents were recorded in the excised inside-out configuration using an EPC9 patch-clamp amplifier/interface (HEKA) and an Apple Power Macintosh computer running Pulse software (HEKA) for generation of pulses and data acquisition. All experiments were performed at room temperature (20-24 °C). Whenever BK channel modulators were applied to the patch, the recording chamber (0.2 ml of effective volume) was perfused with at least 10× the bath volume using a gravity-driven system at a flow rate of 1-2 ml/min. Patch pipettes were manufactured from Corning 7052 glass with resistances of 2.5-4.5 M for single-channel recordings and 1.5-2.5 M for pipettes used to record macroscopic currents from patches with high channel densities. To minimize series resistance errors, the access resistance was partially compensated (~60%), and analysis was limited to patches with currents 3 nA at 150 mV.

Reagents and Solutions-- Dehydrosoyasaponin-I was prepared as a crude extract as published (28). ATP was added as the Mg-salt and cyclic nucleotides cGMP and cAMP as Na-salts (Calbiochem). Okadaic acid, PKG (recombinant bovine isoform I), and the PKA inhibitor (6-22 amide) were obtained from Calbiochem, and the PKA catalytic subunit from Biolab. An anti-FLAG monoclonal antibody (M2, F-3165) was purchased from Sigma. Unfixed cells were stained in complete medium with serum for 20 min, washed 3× with complete medium, and exposed to 1 µg/ml of a second antibody, Alexa FluorÆ488 goat anti-mouse IgG (Molecular Probes), in complete medium for 15 min.

For cell-free recordings from excised inside-out patches of membrane, the patch pipette, which contacts the extracellular side of the membrane, contained either physiological saline (in mM): 135, NaCl; 10 HEPES; 1, MgCl₂; 0.1, CaCl₂ with the pH adjusted to 7.4 with NaOH, or symmetrical K⁺ solution (in mM): 135, KCl; 10, HEPES; 1, MgCl₂; 0.1, CaCl₂ with the pH adjusted to 7.4 with KOH. The standard bath solution for the former cytoplasmic side of the patch was (in mM): 145, KCl; 20, HEPES; 1, MgCl₂; and 0.1, dithiothreitol with the pH adjusted to 7.4 with KOH. Bath solutions were passed over a Chelex 100 (Bio-Rad) ion-exchange column, prior to adding Ca²⁺ buffers and divalent ions in order to remove potential metal contaminants. To buffer calcium reliably without saturation at physiological concentrations between 0.1 and 10 µM, we used the Mg- and pH-insensitive buffer, BAPTA, and its lower affinity dibromo derivative (Molecular Probes). Free Ca²⁺ concentrations were calculated with WinMAXC (www.stanford.edu/~spatton/maxc.html). The actual free Ca²⁺ concentrations were independently verified with a calcium electrode (Orion, Model 97-20 ionoplus), which was calibrated with buffered calcium solutions (Molecular Probes, Buffer kit 2 with 1 mM Mg).

Data Analysis-- Single-channel open probability (P_o) was derived either from single-channel analysis using TAC and TACFit (Bruxton) for patches with 1-3 channels, or in the case of patches with 4-15 channels, by the integration-over-baseline technique using Igor Pro (WaveMetrix). In the latter case N*P_o values were determined as follows. All-point histograms were plotted to obtain the offset, i.e. leak current, as well as the single-channel current amplitude from the peak intervals. After subtraction of the offset from the traces these were integrated over 10-60-s segments. The integral divided by integration time and single-channel current amplitude gives N*P_o. Statistical significance of differences between treatments was evaluated using an unpaired Student's t test on the N*P_o or P_o averages obtained from several minutes of recordings. The voltage-dependence of Slo currents was either measured from P_o versus membrane potential plots (patches with 1-3 channels), or in the case of patches with macroscopic currents, from normalized plots of the tail currents recorded in symmetrical K⁺ solution at 70 mV. Plots of P_o or normalized tail current amplitudes were fitted with a Boltzmann curve and the voltage of half-maximal activation (V_1/2), derived from 1/(1 + exp(V_mV_1/2)/k).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The BK_Ca  subunit channel protein encoded by the rSlo cDNA isolated from GH₄C₁ RNA (Fig. 1) is identical to the rat smooth muscle sequence (U55995) up to Gln-1140 and almost as similar to clones from mouse and human brain (29, 30). The GH₄C₁ rSlo cDNA was expressed efficiently in HEK293 cells and produced robust currents activated by depolarization and cytoplasmic calcium (Fig. 2). Expression of the entire cDNA was confirmed by addition of a FLAG epitope to the N terminus (see "Experimental Procedures"), which was detected immunologically at the surface of intact cells (Fig. 2A), as predicted (31). Excised patches of membrane expressing rSlo produced currents in the nanoampere range (Fig. 2B), corresponding to hundreds of functional BK_Ca channels in each square micrometer of membrane. With or without the FLAG epitope, the channels in such patches showed relative sensitivities to depolarization and cytoplasmic calcium (Fig. 2C) comparable to native channels from GH₄C₁ cells (27). Thus, on average, 17-mV depolarization produced an e-fold increase in activity at any fixed calcium. Similarly a 10-fold increase in the calcium concentration at the cytoplasmic surface shifted the voltage required for half-maximal stimulation by 50 mV toward more negative potentials. However, the absolute sensitivity of the channels to calcium was 10-fold lower than native channels, which are stimulated half-maximally by 1 µM calcium at approximately +20 mV (27).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   The rSlo gene encodes a functional BKCa channel. A, confocal image shows surface staining with anti-FLAG antibody of intact HEK293 cells expressing rSlo with a FLAG tag at the N terminus. B, currents recorded from excised, inside-out patches of HEK293 cells, which had been transfected the night before with FLAG-tagged rSlo. Recording in symmetrical potassium solution; voltage protocol illustrated under the current traces. C, plots of normalized, quasi-macroscopic currents (G/Gmax) as a function of the step potential from tail currents at 70 mV show the voltage and calcium dependence of rSlo currents (without the FLAG tag).

The rSlo channels also differed from native BK_Ca channels in GH₄C₁ cells by their response to adenine nucleotides on the cytoplasmic side of the patch (Fig. 3). In patches from GH₄C₁ cells, addition of Mg-ATP inhibits channel activity through protein phosphorylation by PKA (27), which is reversed by a peptide inhibitor of PKA (Fig. 3A; n = 5 patches). In contrast, rSlo inhibition by 0.5 mM ATP in patches from HEK293 cells is not blocked by the PKA inhibitor peptide (n = 0/3) and is mimicked by a nonhydrolysable analogue of ATP (AMP-PNP) (Fig. 3B, n = 4/4). This direct inhibitory effect of adenine nucleotides on recombinant BK_Ca channels is similar to the effect reported previously on a cDNA from mouse brain, mb2 (32), which has identical N and C termini as the rSlo cDNA from GH₄C₁ cells (29). As reported by Clark et al. (32) for mb2 channels, coexpression with the ₁ subunit eliminated the direct effect of ATP on the channels (Fig. 3C), but the rSlo channels remained insensitive to regulation by PKA, showing neither inhibition nor stimulation, even in the presence of 1 µM okadaic acid to inhibit endogenous protein phosphatases (Fig. 3C, n = 0/10).

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Inhibition of rSlo single-channel activity by ATP is not mediated by PKA-dependent phosphorylation. A, native BK channels from GH4C1 cells are inhibited by PKA-dependent phosphorylation. Channel activity is inhibited in the presence of 0.5 mM Mg-ATP and inhibition is partially reversed upon application of 500 nM PKA-I peptide. Recording from an inside-out patch with physiological saline in the pipette at 0 mV patch potential, 2 µM free cytoplasmic Ca2+. B, rSlo channel activity (N*Po) at +50 mV and 2 µM Ca2+ versus time from an inside-out multichannel patch shows inhibition by 0.5 mM ATP, which is not reversed by 500 nM PKA-inhibitor peptide but is mimicked by 0.2 mM AMP-PNP. C, rSlo channels coexpressed with the 1 subunit do not show direct inhibition by ATP and do not respond to PKA (10 units/ml catalytic subunit), even in the presence of 0.1 mM cAMP and 100 nM okadaic acid (OA). Recording from an inside-out patch at +50 mV with physiological saline in the pipette and 2 µM free cytoplasmic Ca2+.

The Strex Exon Partially Restores the Native Phenotype to rSlo Channels-- Previous studies of BK_Ca channels in a mouse pituitary cell line AtT20 (16) have identified a 59-amino acid exon, strex, in the C terminus near the calcium-binding site (Fig. 1, C and D), which is required for BK_Ca channel inhibition by PKA (17). Expression of rSlo with the strex exon inserted in place of L668 (rSlo-strex) in HEK cells produced BK_Ca channels that were inhibited by PKA (Fig. 4). Even in patches with quasi-macroscopic currents, addition of ATP and cAMP reduced the activity of all the channels, increasing the voltage required to produce half-maximal activity by +20 mV (Fig. 4A), which is quantitatively very similar to the effect of PKA on native BK_Ca channels in GH₄C₁ cells (27). At the single-channel level, ATP/cAMP reduced activity in 86% of the patches (n = 12/14) producing on average a 49% decrease (Fig. 4B). Unlike the effect of ATP on rSlo channels, inhibition of rSlo-strex channels by ATP/cAMP was reversed by a peptide inhibitor of PKA, PKA-I, which increased channel activity in the presence of ATP/cAMP by 67% on average (n = 6, Fig. 4C). Thus, endogenous PKA and phosphatase molecules appear to be associated closely with the BK_Ca channel in cell-free patches of membrane.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   rSlo channels with incorporated strex exon are inhibited by PKA phosphorylation. A, plot of normalized, quasi-macroscopic currents (G/Gmax) as a function of the step potential from tail currents at 70 mV recorded in symmetrical K+ solution and 2.5 µM free cytoplasmic Ca2+. After application of 0.5 mM Mg-ATP and 0.1 mM cAMP (filled circles, V0.5 = 60 mV) the Po versus membrane potential relationship is shifted some 20 mV toward positive membrane potentials compared with the control (open circles, V0.5 = 41 mV) and after washout (open diamonds, V0.5 = 46 mV). The insets show typical current records during steps to 60 to +40 mV in 20-mV increments. B, representative single-channel records before and 5 min after application of 0.5 mM Mg-ATP and 0.1 mM cAMP. Inside-out patch held at +30 mV in solution with 0.5 µM free Ca2+. C, all-point amplitude histogram from a multichannel patch shows reversal of ATP/cAMP inhibition (0.5 and 0.1 mM, respectively) by 500 nM of the PKA inhibitor peptide. Recorded from an inside-out patch at +40 mV in symmetrical K+ solution and 1 µM free cytoplasmic Ca2+.

Inclusion of the strex exon also increased the sensitivity of recombinant BK_Ca channels to inhibition by oxidation (Fig. 5). Native BK_Ca channels in GH₄C₁ cells showed a wide range of sensitivity to oxidation (Fig. 5C), which was measured by replacing 1 mM dithiothreitol with 0.1 mM thimerosal (22). Some native channels were inhibited 2-3-fold by thimerosal (Fig. 5A), whereas other native BK_Ca channels were much more sensitive to inhibition (Fig. 5B). Recombinant rSlo channels in HEK cells were uniformly less sensitive to inhibition by oxidation (Fig. 5C), which is comparable to the sensitivity reported previously for other recombinant channels (22). However, rSlo-strex channels were an order of magnitude more sensitive to inhibition by thimerosal (Fig. 5C). Although thimerosal inhibited the activity of rSlo channels to 40% of their control activity in dithiothreitol (n = 7), the same concentration of thimerosal inhibited 95% of the activity of rSlo-strex channels (n = 6). The increased sensitivity of rSlo-strex channels to inhibition by oxidation is consistent with the high cysteine content of the strex exon (18), and provides additional evidence that native BK_Ca channels in GH₄C₁ cells, like the BK_Ca channels in mouse pituitary cells (16), also contain the strex exon. Nevertheless, rSlo-strex channels also differed from the native channels in GH₄C₁ cells in one essential feature; their absolute sensitivity to stimulation by calcium.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   rSlo-strex channels are highly sensitive to inhibition by oxidation. Native BKCa channels from GH4 cells differ widely in their sensitivity to oxidation, and recombinant rSlo channels show low sensitivity, whereas channels with strex insert are highly sensitive to inhibition by oxidation. A and B, typical single-channel activity from native BK channels in patches from two different GH4C1 cells to show activity under reduced (1 mM dithiothreitol) and oxidized (0.1 mM thimerosal) conditions. A, patch potential +20 mV. B, 0 mV, 0.5 µM free cytoplasmic Ca2+. C, scatter plot of N*Po reduced to N*Po oxidized ratio for native BK channels from GH4C1 cells and recombinant rSlo channels with and without strex insert.

rSlo-strex Is Insensitive to Physiological Calcium-- Although inclusion of the strex exon shifted the calcium sensitivity of rSlo by ~20 mV toward more negative potentials (Fig. 6), as reported previously for other recombinant channels (8, 18), membrane potentials more positive than +75 mV were required to produce half-maximal activation of rSlo-strex in 2 µM calcium (Fig. 6A). In contrast, half-maximal activation of native BK_Ca channels under similar conditions occurs at voltages around +20 mV (27), which are routinely observed during spontaneous spiking. Surprisingly, coexpression of the rat ₁ subunit from GH₄C₁ cells did not alter the calcium sensitivity of rSlo-strex (Fig. 6A) even though ₁ did alter the kinetics of gating (Fig. 6B) and the sensitivity to stimulation by dehydrosoyasaponin-I (Fig. 6C), indicating efficient expression of ₁ and interaction with the  subunit (33). The insensitivity to calcium was also not an artifact of the HEK cell expression system. Mutation of tyrosine 294 to valine at the external mouth of the pore (34) eliminated the sensitivity of rSlo-strex to inhibition by external tetraethylammonium (not shown) and allowed us to study its calcium sensitivity in GH₄C₁ cells without contamination by native channels. Nevertheless, Y294V-strex channels in GH₄C₁ cells were just as insensitive to calcium as rSlo-strex channels in HEK293 cells (Fig. 6A).

View larger version (23K): [in this window] [in a new window]   Fig. 6.   The strex exon but not 1 subunit increases the Ca2+ sensitivity of rSlo channels. A, membrane potential of half-maximal activation (V0.5) determined from normalized tail currents at 70 mV in symmetrical K+ solution with 2 µM free cytoplasmic Ca2+ for recombinant channels expressed in HEK293 cells and for TEA-insensitive rSlo-strex channels with the Y294V mutation expressed in GH4C1 cells. *, p < 0.05 versus rSlo. B, representative traces of currents (steps ranging from 30 to +170 mV) from HEK293 cells expressing rSlo-strex without (top trace) or with (lower trace) 1 subunit. Note in particular the slowing of the deactivation time constant of the tail currents, indicating that the 1 subunits associate with . C, DHS-I (extract, 1:1000) increases single-channel activity of recombinant rSlo-strex channels when coexpressed with 1 subunits. Recording at +30 mV in solution with 2 µM free cytoplasmic Ca2+ and normal saline in the pipette.

rSloN-strex Behaves Like the Native BK_Ca Channel in GH₄C₁ Cells-- In other studies of recombinant BK_Ca channels (33, 35), the ₁ subunit is highly effective at increasing calcium sensitivity, shifting the voltage required for half-maximal activation at a given concentration of calcium toward more negative potentials by 50 mV or more. Because many of the studies used clones that began MDALI- in the N terminus, we also expressed truncated forms of rSlo and rSlo-strex that began at MDALI, which we labeled rSloN and rSloN-strex, respectively (Fig. 1). The calcium sensitivity of rSloN is slightly higher than rSlo (Fig. 7A), but strex inclusion at splice site 2 in rSloN shifted the voltage required for half-maximal activation toward more negative potentials by >40 mV (Fig. 7A). In addition, coexpression of ₁ shifted rSloN-strex by a further 25 mV toward more negative voltages (Fig. 7A). Thus, the absolute calcium sensitivity of rSloN-strex is very similar to native BK_Ca channels in GH₄C₁ cells. Furthermore, rSloN-strex retains the native property of simultaneous sensitivity to inhibition by PKA and stimulation by PKG (Fig. 7, B and C). On average, cAMP-dependent phosphorylation inhibited rSloN-strex channel activity by 45.5% (n = 8), whereas cGMP-dependent phosphorylation stimulated activity in 13 of 20 patches by 89%. In both cases, regulation was mediated by endogenous protein kinases, which remained associated with the channels in cell-free patches of membrane.

View larger version (22K): [in this window] [in a new window]   Fig. 7.   N-terminal truncation restores the native phenotype to rSlo-strex channels. A, membrane potential of half-maximal activation (V0.5) determined from normalized tail currents at 70 mV in symmetrical K+ solution with 2 µM free cytoplasmic Ca2+ for recombinant rSloN channels expressed in HEK293 cells. *, p < 0.05. B, in rSloN channels with strex insert, the inhibition by ATP/cAMP (0.5 and 0.1 mM, respectively) is reversed by 500 nM of the PKA inhibitor peptide. Recording from an inside-out patch at +40 mV in symmetrical K+ solution and 2 µM free cytoplasmic Ca2+ show channel activity in ATP/cAMP, 5 min after application of PKA-I and 7 min after washout of the inhibitor peptide. C, rSloN-strex channels can be stimulated by cGMP-dependent protein kinase. Single-channel traces in 0.5 mM ATP and 10 min after stimulation with 0.1 mM cGMP. Recording from an inside-out patch at +20 mV in symmetrical K+ solution and 2 µM free cytoplasmic Ca2+.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The big-conductance, calcium- and voltage-activated potassium channels BK_Ca encoded by the slo gene (KCNMA1) play a powerful integrative role in the regulation of electrical excitability in many tissues, so it is not surprising that they are also regulated by a complex network of signal transduction pathways. Many of these regulatory mechanisms have been identified, but this is the first study that specifically investigates the interactions between them on recombinant channels. We previously characterized the regulation of native BK_Ca channels by reversible protein phosphorylation in a rat pituitary cell line, GH₄C₁ (23-25, 27). In this study we have identified a structural variant of the rSlo gene product, rSloN-strex (Fig. 1), which recapitulates that regulation (Fig. 7). Furthermore, when it is coexpressed with ₁ subunits, rSloN-strex is just as sensitive to calcium as native channels (Fig. 7). The requirement of strex inclusion in Slo channels for inhibition by PKA has been observed previously in murine BK_Ca channels cloned from a mouse pituitary cell line (16), where it was shown that PKA produces inhibition by phosphorylating a serine in the strex exon (17). Without strex, however, we never observed stimulation of rSlo by PKA, unlike many other recombinant forms of the BK_Ca channel (15, 17, 36). We have not ruled out the possibility that background PKC-dependent phosphorylation of our clone occludes its stimulation by PKA as reported for a full-length bovine Slo channel (36). However, unlike the bovine Slo construct, our rSloN-strex construct can be simultaneously inhibited by PKA and stimulated by PKG (Fig. 7), as reported previously for native channels from GH₄C₁ cells (24).

In designing our experimental solutions, we took four uncommon precautions to ensure reproducible recordings of physiological significance. All pipette solutions were passed over an ion-exchange column to remove potential metal contaminants prior to adding divalent ions. Metal ions are known to block BK_Ca channel pores, particularly at strongly depolarized membrane potentials, where they might artificially reduce estimates of G_max. To maintain the activity of kinases and phosphatases associated with the cell-free patches, we added 1 mM Mg²⁺ and 0.1 mM dithiothreitol to all the solutions that bathed the former cytoplasmic side of the cell-free patches. In addition ATP was always added as a magnesium salt, so free Mg²⁺, which has direct effects on BK_Ca gating (37, 38), remained constant. To buffer calcium reliably without saturation at physiological concentrations between 0.1 and 10 µM, we used the Mg- and pH-insensitive buffer, BAPTA, and its lower affinity dibromo derivative (39). With these precautions we have also identified two novel properties of BK_Ca regulation of potential physiological significance: the effect of N-terminal variation on interactions with the  subunit and the unusually high sensitivity of the strex form to inhibition by oxidation.

There have been very few studies of recombinant BK_Ca channels that begin at the first methionine encoded by the slo gene (19, 40), and this is the first study to specifically compare the behavior of channels with different initiator methionines under identical experimental conditions. Such a study is important because the initiation site for translation has not been identified unambiguously for any native BK_Ca channel, and there are no antibodies available to distinguish N-terminal variations in native BK_Ca channels at the cell surface. The rSlo constructs beginning MSNIH- produce BK_Ca channels with the same relative sensitivity to calcium, but a much lower absolute sensitivity than native BK_Ca channels in GH₄C₁ cells (27). This difference cannot be attributed to the longer N-terminal sequence because the rSloN construct beginning MDALI- produces channels with very similar sensitivity as rSlo channels (Fig. 7) and other recombinant Slo channels (29, 30). The rSlo construct from GH₄C₁ cells also differs from native channels in that the channels are inhibited directly by ATP at concentrations that are predicted to occur in vivo (Fig. 3). This effect does not involve protein phosphorylation because it is also produced by AMP-PNP, a nonhydrolysable analogue of ATP. A similar phosphorylation-independent effect of ATP on a BK_Ca clone from mouse brain, mb2 (29), with identical N and C termini has been reported previously (32). Other Slo constructs, such as mbr5 (41) with different N and/or C termini do not show such inhibition by ATP (32). Coexpression of the ₁ subunit with rSlo prevents direct inhibition by ATP (not shown) and increases their sensitivity to calcium (see below), as reported for mb2 (32, 33). However, based on their low sensitivity to calcium and their inhibition by physiological concentrations of ATP, it is unlikely that BK_Ca channels comprised solely of rSlo  subunits from GH₄C₁ cells contribute much current under normal conditions. In ischemic conditions however, with calcium rising and ATP falling, the rSlo subunits might provide a reservoir of emergency K⁺ channels that could limit further excitotoxicity.

The calcium sensitivity of other recombinant Slo channels is increased by several factors, including alternative mRNA splicing (8, 18),  subunits (19, 33, 40), and protein phosphorylation (14). The calcium sensitivity of BK_Ca channels encoded by the rSlo cDNA from GH₄C₁ cells is also increased by inclusion of the strex exon (8) at splice site 2 (Leu-668 in rSlo) near the calcium-binding site in the cytoplasmic C terminus of the protein (Figs. 1 and 6). The strex exon was originally identified as a cysteine-rich insert at splice site 2 in BK_Ca channels from chromaffin and pituitary cells, which increased the channels' sensitivity to stimulation by calcium (18). Subsequently, the strex exon was shown to determine the response of BK_Ca channels in mouse pituitary AtT20 cells to protein phosphorylation by the cAMP-dependent protein kinase (16, 17). Inclusion of the strex exon in rSlo from rat pituitary GH₄C₁ cells also produced BK_Ca channels that were inhibited by PKA (Fig. 4), like the native channels from those cells (23, 27). However, strex inclusion in rSlo had two additional effects of potential physiological significance, which have not been described before.

Firstly, strex inclusion in rSlo makes the resulting BK_Ca channels extremely vulnerable to inhibition by oxidation. For example, the activity of fully reduced channels in dithiothreitol, which is presumed to be comparable to the redox conditions in vivo, is inhibited greater than 95% by exposure to chemical oxidants such as thimerosal (Fig. 3). In contrast recombinant Slo channels without strex are inhibited less than 50% by exposure to comparable oxidants (21, 22). Expressed as a fold reduction in activity, strex-containing channels are 10× more sensitive to oxidation.

Secondly, we discovered that the incorporation of strex near the calcium-binding site in the C terminus (Fig. 1) alters the effectiveness of the ₁ subunit at increasing calcium sensitivity, and this effect depends on the sequence at the extracellular N terminus of rSlo that interacts with ₁ (20). ₁ subunits have been reported to produce large increases in the calcium sensitivity of many recombinant forms of Slo (19, 33, 40), although the effectiveness of the  subunit varies with the absolute calcium concentration, becoming less effective at concentrations in the physiological range near 1 µM and below (35, 42). For example, the ₁ subunit that we isolated from GH₄C₁ cells increased the activity of both our rSlo and rSlo-strex channels at calcium concentrations >100 µM (not shown), but the ₁ subunit did not stimulate either rSlo or rSlo-strex channels in the physiological range of 1-10 µM calcium (Fig. 6). We confirmed that the ₁ subunits interact with both rSlo and rSlo-strex channels under these conditions because the currents had slower kinetics and channels were stimulated by dehydrosoyasaponin when the ₁ subunits were expressed (Fig. 6). In contrast, inclusion of strex in the truncated rSloN channels that start at MDALI- extends the effectiveness of ₁ on calcium sensitivity to calcium concentrations in the physiological range (Fig. 7), making rSloN-strex as sensitive to calcium as native channels in GH₄C₁ cells. Thus, the polyserine domain between the first and second methionines of rSlo (Fig. 1) inhibits the ability of ₁ subunits to alter calcium sensitivity, which is consistent with the evidence that the transmembrane ₁ subunit interacts with Slo channels at their extracellular N terminus (20). Consequently, the regulation of BK_Ca structure through alternative splicing of rSlo by hormones (8, 10) and activity-generated calcium transients (43) could have dramatic effects on endocrine excitability in the pituitary (44).

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Laboratory of Signal Transduction (F2-05), NIEHS, National Institutes of Health, 111 Alexander Dr., Research Triangle Park, NC 27709. Tel.: 919-541-0062; E-mail: armstro3@niehs.nih.gov.

Published, JBC Papers in Press, May 16, 2002, DOI 10.1074/jbc.M203087200

	ABBREVIATIONS

The abbreviations used are: PKA, cAMP-dependent protein kinase; PKG, cGMP-dependent protein kinase; r, rat; HEK, human embryonic kidney; N, N-terminal truncations; BAPTA, 1,2-bis-(2-aminophenoxy)ethane-N,N,N,N-tetraacetate; AMP-PNP, adenosine 5'-(,-imino)triphosphate.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				B. Wang and R. Brenner An S6 Mutation in BK Channels Reveals {beta}1 Subunit Effects on Intrinsic and Voltage-dependent Gating J. Gen. Physiol., December 1, 2006; 128(6): 731 - 744. [Abstract] [Full Text] [PDF]

				I. Semenov, B. Wang, J. T. Herlihy, and R. Brenner BK channel beta1-subunit regulation of calcium handling and constriction in tracheal smooth muscle Am J Physiol Lung Cell Mol Physiol, October 1, 2006; 291(4): L802 - L810. [Abstract] [Full Text] [PDF]

				G.-J. Lai and D. P. McCobb Regulation of Alternative Splicing of Slo K+ Channels in Adrenal and Pituitary during the Stress-Hyporesponsive Period of Rat Development Endocrinology, August 1, 2006; 147(8): 3961 - 3967. [Abstract] [Full Text] [PDF]

				H. Sade, K. Muraki, S. Ohya, N. Hatano, and Y. Imaizumi Activation of large-conductance, Ca2+-activated K+ channels by cannabinoids Am J Physiol Cell Physiol, January 1, 2006; 290(1): C77 - C86. [Abstract] [Full Text] [PDF]

C. E. McCartney, H. McClafferty, J.-M. Huibant, E. G. Rowan, M. J. Shipston, and I. C. M