The large-conductance Ca2+-activated voltage-dependent K+ channel (maxi-K channel) induces a significant repolarizing current that buffers cell excitability. This channel can derive its diversity by alternative splicing of its transcript-producing isoforms that differ in their sensitivity to voltage and intracellular Ca2+. We have identified a novel 132-bp exon of the maxi-K channel from human myometrial cells that encodes 44 amino acids within the first intracellular loop of the channel protein. Distribution analysis reveals that this exon is expressed predominantly in human smooth muscle tissues with the highest abundance in the uterus and aorta and resembles the previously reported distribution of the total maxi-K channel transcript. Single-channel K+ current measurements in fibroblasts transfected with the maxi-K channel containing this novel 132-bp exon demonstrate that the presence of this insert attenuates the sensitivity to voltage and intracellular Ca2+. Alternative splicing to introduce this 132-bp exon into the maxi-K channel may elicit another mode to modulate cell excitability.
- potassium channel
- calcium activated
- splice varient
the large-conductance Ca2+-activated voltage-dependent K+ channel (maxi-K channel) functions to provide a potent repolarizing current in multiple cell types. The maxi-K channel possesses various mechanisms to regulate cell excitability. Recently, the role of auxiliary β-subunits in modulating maxi-K channel function has been explored (2, 12), and at least one subunit, β1, which is abundant in smooth muscle tissues, is known to increase the sensitivity of the channel to voltage and intracellular Ca2+ (10, 11, 20). A second mechanism to modulate maxi-K channel behavior is alternative splicing of the channel transcript to produce various protein isoforms that differ in their sensitivity to voltage, intracellular Ca2+levels, and potential susceptibility to posttranslational modification (1, 7, 21). Multiple splice variants of the maxi-K channel are present in humans (4, 21); however, the tissue distribution of the isoforms and their ability to regulate cell excitability are currently unknown. Multimeric association of maxi-K channel isoforms produced by differential splicing with varied sensitivities to regulators allows for fine tuning of cell excitability in various tissue types (9, 18). In addition, splicing may underlie differences in posttranslational modification of maxi-K channels via phosphorylation by endogenous protein kinases, which, in turn, alters channel behavior (5, 23).
We have identified a novel 132-bp alternatively spliced exon of the maxi-K channel. Its transcript is prevalent in human smooth muscle tissues, similar to that reported for the maxi-K channel. The 132-bp exon encodes a 44-amino acid segment that is introduced into the first intracellular loop of the channel. Channels that contain this exon, when heterologously expressed in mouse Ltk− fibroblasts, elicit a current similar to the maxi-K channel current but possess a diminished sensitivity to increasing concentrations of intracellular Ca2+ and depolarizing voltages. These studies indicate that the novel exon within the maxi-K channel modulates channel function and may be a regulator of cell excitability.
MATERIALS AND METHODS
Human myometrial tissue from the lower uterine segment was collected from patients undergoing elective or emergency cesarean section under spinal anesthesia at 31–32 and 38–40 wk of gestation either in the presence or absence of spontaneous or induced labor contractions. All patients signed written consent approved by Internal Review Board 199809066. Tissue was placed in Hanks' balanced salt solution or phosphate-buffered saline on ice and used for culturing within 1–3 h after collection.
Cell isolation and culturing.
Freshly isolated human myometrial tissue was disaggregated with 1 mg/ml of collagenase (type 1; Worthington Biochemicals) at 37°C with mild agitation for 1 h. Partially digested explants were plated in DMEM/F-12 medium supplemented with 5% fetal bovine serum, 4 ng/ml basic fibroblast growth factor (Sigma), 1 ng/ml epithelial growth factor (Sigma), 50 μg/ml gentamicin (GIBCO BRL), and 5 μg/ml fungizone (GIBCO BRL) and grown in a humidified incubator with 5% CO2 at 37°C for 2 days until smooth muscle cell outgrowth was observed. At that point, the explants were removed, and adherent cells were trypsinized, plated in fresh dishes, and grown for two additional days in the medium described above (fungizone removed) until they reached 70–80% confluency, at which point they were used experimentally.
Total RNA isolation, RT-PCR, and Southern blotting.
Total RNA from cell cultures was isolated by the guanidinium isothiocyanate method, ethanol precipitated, and resuspended in 10 mM Tris and 1 mM EDTA, pH 6.5. The concentrations were determined by spectrophotometry at a wavelength of 260 nm. Total RNA was reverse transcribed using random primers (ProStar First Strand RT-PCR kit; Stratagene), and cDNA was PCR amplified using primers flanking human Slo (hSlo) bp 1–351 of the coding region (forward: 5′ ATGGATGCGCTCATCATC, reverse: 5′ ACAACCAGGACTCTGCCAGT). The PCR products were visualized with ethidium bromide on 1% agarose gels. The gels were denatured in 1.5 M NaCl/0.5 M NaOH solution, neutralized in 1 M Tris/1.5 M NaCl solution (pH 8.0), and transferred to nitrocellulose by capillary action. After transfer, the blots were hybridized in QuickHyb solution (Stratagene) with a radiolabeled cDNA probe (Prime-It II kit; Stratagene). The probe was designed to recognize bp 1–351 of hSlo. Autoradiography was performed at −90°C overnight. To prepare for sequencing, the band representing an ∼500-bp fragment was gel purified (QIAEX II Gel Extraction kit; Qiagen), ligated into the pCRII vector, and transformed into INVαF′ cells (TA cloning kit; Invitrogen). The DNA was then sequenced and analyzed. The accession number of 132-bp exon is AF349445.
3′ RACE PCR.
Human myometrial total RNA was reverse transcribed using oligo(dT) primers (ProStar First Strand RT-PCR kit; Stratagene). The subsequent cDNA was PCR amplified with the forward primer 5′ GGGAGGAAACCAAGACTC (nt 1–18 of the 132-bp exon) and an oligo(dT) reverse primer (Stratagene). The PCR product obtained was amplified using the forward primer 5′ GGGATTCGTATGTTTCACCTG (nt 38–51 of the 132-bp exon) and the same oligo(dT) reverse primer. The resultant DNA was separated on a 1% agarose gel, visualized by ethidium bromide, and transferred to nitrocellulose, and the blot was hybridized with a 5′-end radiolabeled cDNA probe (Klenow Fill-in kit; Stratagene) for bp 38–132 of the 132-bp exon. Bands identified by autoradiography were subsequently gel purified, subcloned, and sequenced.
Genomic DNA from National Institute of General Medical Sciences human/rodent cell hybrid mapping panel no. 1 (BIOSMAP somatic cell hybrid PCRable DNAs, BIOS Labs) was used for PCR amplification of the 132-bp exon, with primers flanking the exon (Southern blotting method) as previously described (3). PCR samples were separated on a 1% agarose gel, and the banding pattern was recorded and analyzed.
Tissue distribution analysis.
A human Multiple Tissue Expression (MTE) array (Clontech) was hybridized according to the manufacturer's instructions with a radiolabeled cDNA probe (Prime-It II kit; Stratagene) against the 132-bp exon (1.0 × 106 counts per minute/ml). Autoradiography was performed at −90°C in the presence of an intensifying screen for 72 h. Transcript expression levels were assessed by dot densitometry (LabWorks 3.0; UVP) and normalized to the dot with the most intense signal.
In vitro expression and electrophysiological characterization.
The 132-bp exon-containing and -lacking forms of the hSlo α-subunit cDNA were isolated from human myometrial cDNA by PCR with the forward primer 5′ TTGCGGCCGCATGGATGCGCTCATC and the reverse primer 5′ GGTCTAGATCAAAGCCGCTCTTCC. The insertless hSlo sequence found in myometrium corresponded to a previously described clone (13). The resulting PCR fragments were digested withNotI and XbaI (sites within primers underlined) and subcloned into the pTracer-CMV2 plasmid (Invitrogen). Ltk− mouse fibroblasts grown to 50–80% confluency in 60-mm dishes in DMEM supplemented with 10% horse serum and 0.1% penicillin and streptomycin or 50 μg/ml of gentamicin were transfected with 10 μg of construct DNA (Lipofectamine Plus reagent kit; GIBCO) and incubated one to three additional days. Cells expressing green fluorescent protein (pTracer-CMV2 plasmid reporter gene) were used for patch clamping.
All patch-clamp experiments were performed at room temperature (∼22°C). Single-channel currents were elicited in symmetrical K+ at constant holding potentials. Transfected Ltk− fibroblasts were placed in a pH 7.2 bath solution containing (in mM) 145 KCl, 1 MgCl2, 10 HEPES, and 1 EGTA. The bath solution was supplemented with the appropriate amount of CaCl2 to produce free Ca2+ concentrations of ∼10−7, 10−6, and 10−5 M, based on a free intracellular Ca2+ concentration ([Ca2+]) estimation program (17). Borosilicate glass pipettes of 8–20 MΩ were filled with a pH 7.2 solution containing (in mM) 145 KCl, 2 CaCl2, 1 MgCl2, and 5 HEPES. High-resistance inside-out patch seals (3–30 GΩ) were achieved for single-channel measurements. Constant membrane potentials at 20-mV increments from −60 mV up to as high as +160 mV were applied to membrane patches in the three different Ca2+ concentrations for up to 1 min using an AxoPatch 200B voltage-current amplifier. The elicited currents were recorded using pCLAMP 6.0 software (Axon Instruments). The open state was defined as 50% of the single-channel current level. The number of open events was measured using the Fetchan program of pCLAMP 6.0. These data were analyzed relative to the total number of open levels seen at maximal stimulation using the pStat program within pCLAMP 6.0 to determine the open-state probability of a patch at each potential and Ca2+ concentration. Half-activation potential (V ½) values were determined only from those patches that were raised as nearly as possible to maximally stimulating potentials. These values ranged between +100 and +160 mV, since voltages higher than this altered seal integrity. The currents were fit to normalized Boltzmann curves using Origin 6.0 software, and the potentials at which half-maximal activation was achieved, based on the Boltzmann curves, were taken as theV ½ values. TheV ½ was not determined for hSlo containing the 132-bp exon in 10−7 M Ca2+because the channels could not be opened at +160 mV at this concentration. Additionally, at high potentials at 10−5 M Ca2+, many cells containing the hSlo lacking the 132-bp exon exhibited significant rundown at high potentials, and thusV ½ values were determined from those cells that did not strongly demonstrate this effect.
Graphical results are displayed as means ± SE. Statistical significance of differences of means was evaluated by a two-tailed Student's t-test for unpaired observations. Differences were considered significant at P < 0.05.
To identify the splice variants of the maxi-K channel in human myometrial cells, first-strand cDNA from primary cell cultures was PCR amplified using primers that flank the previously described splice site A (13, 16, 21). The expected 351-bp insertless variant (Fig. 1 A, open arrow) was evident consistently, as was a second fragment of ∼500 bp (Fig.1 A, closed arrow). Southern blot analysis was used to demonstrate that the 351-bp probe could hybridize to the ∼500-bp band (Fig. 1 B, closed arrow), suggesting that the larger band represents either an alternative transcript of the hSlo mRNA or a transcript from a homologous gene. To identify this band, the fragment was purified, subcloned, and sequenced. Sequence analysis indicated that a portion of this fragment represents a novel 132-bp exon located between bp 183 and 184 of a previously reported human maxi-K channel coding region (13). Recovery of the nucleotide sequence downstream from the insert by 3′ RACE PCR confirmed that this insert is a novel exon within the human maxi-K channel. Chromosomal localization using a somatic cell hybrid mapping panel proved that the exon is localized to human chromosome 10 (data not shown), similar to the gene encoding the human maxi-K channel (21), but the exon was not detected in either control mouse or hamster genomic DNA by PCR.
The amino acids encoded by the 132-bp exon are located in the first intracellular loop of the maxi-K channel (Fig.2). Analysis of the deduced amino acid sequence of the exon revealed an abundance of hydrophilic residues. A motif search indicated that the exon introduces four potential consensus sites for posttranslational modification within the first intracellular loop: two CK2 phosphorylation sites, one myristylation site, and one amidation site (Fig. 2). The amidation site is a combination of three amino acids that are within the insert and the residue immediately 5′ of the novel exon.
To determine the distribution of the exon among human tissues, a commercially available MTE array blot of poly(A)+ RNA derived from multiple human tissues was hybridized with a radiolabeled probe against the 132-bp exon sequence. Our results are largely consistent with what is currently known about tissue and organ distribution of the maxi-K channel. The exon was most abundant in smooth muscle-containing and lymphoid organs, in skeletal muscle, and throughout the central nervous system but was absent in the heart (Fig.3 A). The uterus (closed arrow) and aorta (open arrow) presented the strongest signals by densitometry, but signals were also prevalent in other smooth muscle-containing organs that were present on the blot. In addition, this transcript was very prevalent in liver (A9). Densitometric quantitation of the entire blot was performed to determine the distribution of the 132-bp exon transcript (Fig. 3 B). The 132-bp exon transcript expression demonstrates a similar pattern to that reported for the maxi-K channel, with a greater abundance of the transcript in smooth muscle-containing organs. However, significant differences are apparent in the preferential expression of the 132-bp exon transcript in liver (A9) and salivary gland (E9), which has not been reported for the maxi-K channel.
To determine the single-channel properties of the exon-containing isoform of this channel, inside-out patch recordings were performed from heterologously expressed channels. At +40 mV membrane potential, an outward current of large amplitude was measured in Ltk−cells transfected with either hSlo or hSlo containing the exon (hSlo+132). Initial observations suggested that when patches containing equal numbers of detectable channels were compared, hSlo+132 currents showed a lower open-state probability than channels lacking the novel exon (Fig. 4 A). The single-channel conductance was similar in the absence (241 ± 3.0 pS) or presence of the exon (243 ± 3.0 pS, Fig. 4 B). The exon-containing channel demonstrated a much lower open-state probability at any given Ca2+ concentration. At 10−7 M Ca2+, hSlo+132 channel opening was not detectable up to voltages of +160 mV, while hSlo lacking the exon could be measured at potentials greater than +80 mV. At higher [Ca2+] of 10−6 and 10−5 M, the exon-containing channel isoform demonstrated a much lower open-state probability (Fig. 4 C). V ½, as determined from a Boltzmann curve, was shifted to the right by ∼50 mV in 10−6 M Ca2+ (from ∼83 to ∼137 mV), indicating a greatly reduced Ca2+ sensitivity in the 132-bp exon-containing isoform (Fig. 4 D). In addition, at 10−5 M Ca2+, theV ½ of the channels showed a similar trend (∼59 to ∼97 mV); however, significant channel rundown and seal integrity at this concentration decreased the number of cells that could be measured, which prevented obtaining statistical significance.
The maxi-K channel is believed to provide a potent repolarizing current to attenuate cell excitability and thereby temper smooth muscle contractility and neuronal firing. Regulation of this channel occurs via multiple mechanisms, including posttranslational modification (14) and alternative splicing, to introduce novel exons that can modulate intracellular regulation, and, consequently, channel function. We have isolated a novel splice variant of the maxi-K channel containing a 132-bp exon that encodes for an additional 44 amino acids into the first intracellular domain of the channel. This is the second splice site identified within the NH2-terminal portion of the maxi-K channel (13, 16, 21). Most of the currently known splice sites are confined to the COOH-terminal intracellular “tail” region (21). This novel exon possesses similar characteristics to the COOH-terminal inserts, such as the ability to modulate channel responses to voltage and Ca2+.
The amino acid composition encoded by the 132-bp exon suggests that it is a functionally active part of the channel protein. The two potential CK2 consensus sites, which may allow protein phosphorylation independent of cyclic nucleotides or intracellular Ca2+, could produce variability in posttranslational modifications in the first intracellular loop and thus a differential effect on the channel's function, as it does for other channels (8). The addition of a myristylation site may facilitate membrane targeting and/or membrane assembly of the maxi-K channel protein (15). It is possible that this channel is alternatively targeted to, or away from, the plasma membrane to alter cellular excitability by changing the number of functional channels. The potential significance of an amidation site in this protein is difficult to surmise. The effect of amidation on ion channels has been studied in relation to the ClC-2G Cl− channel, where it causes a significant increase in the open channel probability (19). Whether the novel 132-bp exon-containing maxi-K channel could alter biological activity of a K+ channel protein by posttranslational processing is unknown, but it is an interesting question to address.
The location of the novel exon within the first intracellular loop indicates that other regions within the channel protein besides the COOH-terminal region may be able to alter the maxi-K channel's sensitivity to intracellular Ca2+ and voltage. Our single-channel data indicate that the channel containing this exon is less sensitive to Ca2+ and voltage. Conformational changes in the maxi-K channel may alter β1 binding (22); however, experiments to coexpress α- and β-subunits have not yet been performed to assess this point. Our findings also may result from an interaction of the first intracellular loop with the COOH-terminal domain (6). The addition of 44 amino acids may allow for interactions between other domains such as the S4-S5 domains to alter voltage sensing (6). An additional possibility is that intracellular signaling pathways present in our heterologous expression system affect maxi-K channel activity. Further studies could address interacting domains important for channel function, since maxi-K channel isoforms that contain this 132-bp exon likely form heteromeric channels with isoforms that do not include this novel exon. This will help explain the differences in the sensitivity of this channel to voltage and intracellular Ca2+ in various tissues.
The abundance of the 132-bp exon transcript may indicate the functional significance of this particular isoform in different organs. While transcript levels of the 132-bp exon are prevalent in smooth muscle similar to the entire maxi-K channel, other tissues also contain this isoform. The preferential use of this exon in liver and salivary glands is interesting and may indicate an important function in these tissues. Heterologous expression system studies demonstrate that the novel exon alters the maxi-K channel by conferring a lower sensitivity to changes in voltage and intracellular Ca2+. Therefore, on the basis of electrophysiological data, the function of isoforms containing this novel exon may be to provide a repolarizing current following strong depolarization or intracellular Ca2+release. The presence of this exon may also modulate the voltage and Ca2+ sensitivity of other isoforms with which it is associated.
It is clear that the 132-bp exon discovered in the human myometrium represents a novel splice variant of the maxi-K channel. Insertion of this exon into the hSlo transcript 3′ of nt 183 places 44 additional amino acids in the first intracellular loop of the channel. To date, this is the second splice variant described near the NH2terminus. Upon stimulation, it generates currents similar to those of the exon-lacking isoform of the maxi-K channel; however, this channel is less sensitive to voltage and intracellular Ca2+. The 132-bp exon-containing isoform is found predominantly in smooth muscle-containing organs and liver, with the highest levels of expression in smooth muscle tissues, similar to the whole channel. The tissue expression pattern may indicate a functional significance of the exon for these organs. Thus it is possible that this splice variant is subject to regulation by local and/or systemic factors.
We thank Nancy A. Benkusky for critical review of the manuscript.
This work was supported in part by National Institute of Child Health and Human Development Grant HD-37831 (to S. K. England), National Science Foundation Grant IBN 98-19339 (to S. K. England), and National Institute of Diabetes and Digestive and Kidney Diseases Postdoctoral Training Grant Fellowship DK-07759 (to V. P. Korovkina). Reagents and facilities to conduct a portion of these studies were provided by the Diabetes and Endocrinology Research Center at the University of Iowa (DK-25295).
Address for reprint requests and other correspondence: S. K. England, Dept. of Physiology and Biophysics, 5-660 Bowen Science Bldg., Univ. of Iowa, Iowa City, IA 52242 (E-mail:).
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.
- Copyright © 2001 the American Physiological Society