Functional and molecular identification of a TASK-1 potassium channel regulating chloride secretion through CFTR channels in the shark rectal gland: implications for cystic fibrosis

Abstract

In the shark rectal gland (SRG), apical chloride secretion through CFTR channels is electrically coupled to a basolateral K⁺ conductance whose type and molecular identity are unknown. We performed studies in the perfused SRG with 17 K⁺ channel inhibitors to begin this search. Maximal chloride secretion was markedly inhibited by low-perfusate pH, bupivicaine, anandamide, zinc, quinidine, and quinine, consistent with the properties of an acid-sensitive, four-transmembrane, two-pore-domain K⁺ channel (4TM-K2P). Using PCR with degenerate primers to this family, we identified a TASK-1 fragment in shark rectal gland, brain, gill, and kidney. Using 5′ and 3′ rapid amplification of cDNA ends PCR and genomic walking, we cloned the full-length shark gene (1,282 bp), whose open reading frame encodes a protein of 375 amino acids that was 80% identical to the human TASK-1 protein. We expressed shark and human TASK-1 cRNA in Xenopus oocytes and characterized these channels using two-electrode voltage clamping. Both channels had identical current-voltage relationships (outward rectifying) and a reversal potential of −90 mV. Both were inhibited by quinine, bupivicaine, and acidic pH. The pKa for current inhibition was 7.75 for shark TASK-1 vs. 7.37 for human TASK-1, values similar to the arterial pH for each species. We identified this protein in SRG by Western blot and confocal immunofluorescent microscopy and detected the protein in SRG and human airway cells. Shark TASK-1 is the major K⁺ channel coupled to chloride secretion in the SRG, is the oldest 4TM 2P family member identified, and is the first TASK-1 channel identified to play a role in setting the driving force for chloride secretion in epithelia. The detection of this potassium channel in mammalian lung tissue has implications for human biology and disease.

the rectal gland of the spiny dogfish shark (Squalus acanthias) is an ancient tubular epithelial organ that is highly specialized for a single function, sodium chloride secretion. As such, it is a powerful model for the study of this process in higher vertebrates. Basolateral membranes of the rectal gland contain record amounts of the Na/K/ATPase pump (55) and the Na/K/2Cl cotransporter (19, 65), whereas apical membranes contain large amounts of the CFTR chloride channel protein (35). Stimulatory hormones elicit 30–40-fold rapid increases in chloride secretion (20) that is accompanied by upregulation of both the Na/K/2Cl cotransporter and Na/K/ATPase activity (18, 19, 54). When secretion is activated, potassium ions that enter the cell through these two transport proteins must exit though basolateral membrane potassium channels because the apical membrane of the SRG is impermeant to this ion (24).

The importance of basolateral potassium channels in chloride secretion was demonstrated in experiments with barium, a nonspecific K channel inhibitor (22, 25, 53). When the exit of potassium from rectal gland tubules is reversibly blocked by barium, chloride secretion is markedly reduced (53) because blocking exit to K depolarizes the cell, reducing the intracellular voltage that drives apical chloride exit through CFTR (25). Thus, in this system, basolateral potassium channels function to maintain a favorable gradient for apical chloride secretion by maintaining an apical membrane potential more negative than the Nernst potential value for Cl.

Despite their significance, the molecular identity of potassium channels in the shark rectal gland is unknown. Two potassium channels have been cloned from this tissue. The ortholog of the human K⁺ channel KCNQ1 was cloned from the SRG and characterized in the Xenopus oocyte expression system (32, 63). However, investigators were unable to correlate early patch-clamp measurements performed using isolated tubules (22, 24) with inhibitor studies in perfused SRGs. When electrophysiological results were compared with inhibitor studies in the intact SRG perfused with chromanol 293 B, this channel accounted for only a minor part of the total K⁺ conductance (24). An inward-rectifying KIR6.1 potassium channel was also cloned from the rectal gland, but expression studies were not successful (64).

A family of background or “leak” potassium channels, characterized by four transmembrane domains and two channel pores (4TM-2P channels) has been identified in excitable tissues where they function to set the membrane potential (14, 45). By quantitative PCR, these channels are also expressed in nephron segments of the mammalian kidney (38), although their role in specific ion transport functions is unknown.

In the present study, we provide evidence that the TASK-1 channel, a member of the 4TM-2P superfamily, is the dominant potassium channel in the rectal gland tubule and functions to regulate the membrane potential that drives chloride secretion though CFTR channels. Our finding of this channel in human airway cells has implications for mammalian physiology and disease.

MATERIALS AND METHODS

In Vitro Perfusion of Shark Rectal Glands

Rectal glands were obtained from dogfish sharks weighing 1–3 kg, which were caught by gill nets in Frenchman Bay, Maine. They were kept in tanks with flow-through seawater until use, usually within 3 days of capture. Sharks were killed by pithing of the spinal cord [Mount Desert Island Biological Laboratory Institutional Animal Care and Use Committee (MDIBL IACUC) protocol no. 10-03]. Rectal glands were excised, and cannulas were placed in the artery, vein, and duct, as previously described (31, 35). Glands were placed in a glass perfusion chamber, maintained at 15°C with running sea water, and perfused with shark Ringer solution containing (in mM) 270 NaCl, 4 KCl, 3 MgCl₂, 2.5 CaCl₂, 1 KH₂PO₄, 8 NaHCO₃, 350 urea, 5 glucose, and 0.5 Na₂SO₄, and this solution was equilibrated to pH 7.6 by bubbling with 99% O₂ and 1% CO₂. All glands were perfused for 30 min with Ringer solution to achieve basal rates of chloride secretion. Chloride secretion was then activated by continuous perfusion of forskolin (1 μM) and IBMX (100 μM) from 30 min to the end of the experiment at 90 min. Individual potassium channel inhibitors were perfused from 50 min until the 70-min time point. The maximum level of inhibition achieved was compared with the 50- to 70-min controls without the inhibitor. Inhibitors were then removed for the remainder of the experiment to assess reversibility of the inhibitory effect. Inhibitors were chosen on the basis of their specificity in blocking specific families of potassium channels (1, 2, 8, 28, 30, 58, 68), and perfusate concentrations were at or above published K_i values. Results are expressed as microequivalents of chloride secreted per hour per gram wet weight (μeq·h⁻¹·g⁻¹) ± SE. Unless otherwise specified, all reagents were obtained from Sigma-Aldrich (St. Louis, MO).

Molecular Biology, DNA Sequencing, and Sequence Analysis

Preparation of total RNA from rectal gland tissue, degenerate primer design, PCR, and cloning of PCR fragments.

Total RNA was extracted with TRIzol reagents (Invitrogen, Carlsbad, CA) from 100 mg of shark tissue that had been excised and immediately flash frozen in liquid nitrogen. First-strand cDNA was synthesized using Invitrogen SuperScript synthesis system for RT-PCR. ClustalW (27, 61) was used to identify regions of high amino acid homology in available primate, rodent, and teleost 4TM 2P family potassium channel subtypes: TWIK, THIK, TASK, TREK, and TRAAK.

Fifteen degenerate primer pairs were designed using Consensus-Degenerate Hybird Oligonucleotide Primers (51) that targeted specific two-pore channel subtypes. Of the fifteen primer pairs, one pair targeting the TASK subtype successfully amplified target sequence in shark rectal gland, kidney, brain, and gill. These TASK primers were designed from regions of high conservation in nine sequences from four species: Fugu rubripes, Danio rerio, Rattus norvegicus, and Mus musculus. The successful pair was the eight-fold degenerate forward oligonucleotide primer [5′-ATCCCCCTGACCCTGGT(A/C/G/T) ATGTT(C/T)CA-3′] that codes for the Pro-Leu-Thr-Leu-Val-Met-Phe-Gln sequence and the 48-fold degenerate reverse oligonucleotide primer [5′-GCACCACCAGGT TCAGGAA(A/C/G/T)GC(A/C/G/T)CC(A/G/T)AT-3′] that codes for the Val-Iso-Gly-Ala-Phe-Leu-Asn-Leu. PCR amplifications were performed with cDNA template using Taq DNA Polymerase (Promega, Madison, WI). The touch-down PCR amplifications were performed in a MasterCycler Gradient 2 (Eppendorf, Hauppauge, NY) with the following parameters: hot start for 3 min at 94°C; followed by 40 cycles of 95°C for 45 s, 65°C annealing for 1 min that was lowered by 0.5°C each cycle, and 72°C for 1.5 min; and an extension period of 72°C for 10 min. The resulting 394-bp PCR product was purified from a 2% agarose gel and ligated into the pCR II TOPO TA cloning vector (Invitrogen). Positive transformants were screened on LB plates with 100 μg ampicillin ml-1 and by colony PCR. Positive transformant plasmid DNA was then replicated with QIAprep Spin Miniprep Kit (Qiagen, Valencia, CA). Clones were bidirectionally sequenced, and the 394-bp sequence was analyzed against the GenBank database with a BLAST alignment to identify the fragment as having high similarity with human TASK proteins.

Rapid Amplification of cDNA Ends-PCR to Obtain 5′ and 3′ Ends of Shark TASK and Cloning of Rapid Amplification of cDNA Ends-PCR Fragment

3′ rapid amplification of cDNA ends-PCR.

Shark rectal gland total RNA was adaptor ligated, according to the manufacturer's protocol (SMART RACE cDNA Amplification, BD Biosciences, San Jose, CA). Two TASK-specific primers were designed for 3′ rapid amplification of cDNA ends (RACE)-PCR (external 5′-TGGGGCTGCTGCATTCTC ATATTA CG-3′ and an internally nested primer 5′-TTGCTTCATCACACTCACCACCATT-3′) that were used in subsequent PCRs. External TASK-specific primer and external cassette-specific primer (Universal Primer A Mix, BD Biosciences) were used in a PCR using Advantage 2 PCR kit (BD Biosciences) with an adaptor-ligated shark rectal gland 3′ SMART RACE-ready cDNA template. Fifty-microliter reaction mixtures were cycled in a MasterCycler Gradient 2 (San Jose, CA) with the following parameters: 95°C for a 2-min hot start; followed by 32 cycles of 94°C for 30 s, 58 °C for 45 s, and 72°C for 2.5 min; and an extension period of 72°C for 10 min. The products were diluted 50-fold and used as a template for another RACE-PCR amplification with nested primers for both the adaptor-ligated cDNA ends and the TASK gene using the same cycling parameters.

5′RACE-PCR.

The degenerate-antisense primer targeting the TASK subtype from the initial amplifications and the external cassette-specific primer (Universal Primer A Mix, BD Biosciences) were used in an amplification using Advantage 2 PCR kit (BD Biosciences) with adaptor-ligated shark rectal gland 5′ SMART RACE-ready cDNA template. Fifty-microliter reaction mixtures were cycled in a Mastercycler Gradient 2 (Eppendorpf) with the following parameters: hot start for 3 min at 94°C; followed by 35 cycles of 95°C for 45 s, 65°C annealing for 1 min that was lowered by 0.5°C each cycle, and 72°C for 1.5 min; and an extension period of 72°C for 10 min. The products were diluted 50-fold and used as a template for another RACE-PCR amplification with nested primers for both the adaptor-ligated cDNA ends and the TASK gene (5′-GCTAAATGCCACA TAGTGAGGGTTC-3′ or 5′-GCAGAGCCACATAGTCCCCGAA-3′) using the cycling parameters used in both 3′ RACE-PCR amplifications. The 694-bp 3′ product and a partial 592-bp 5′ product were extracted from a 2% agarose gel of the nested RACE PCR amplifications. Genome walking (BD Biosciences GenomeWalker universal kit) with shark genomic DNA was necessary to obtain the remainder of the 5′ sequence (428 bp band extracted from 2% gel), including 335 bp of untranslated region sequence upstream of the methionine start codon. The PCR products were then cloned in TOPO TA cloning vectors (Invitrogen), as previously described (5). Clones were sequenced bidirectionally, and complete sequence information, including untranslated regions above the methionine start codon and below the stop codon for the shark gene was assembled. Shark-specific primers were designed to amplify full-length shark TASK-1 cDNA. The sense primer used was, 5′-CCGCCGCGATCTCCAGTCTTC TTCT-3′ The antisense primer used was 5′-GGATATGGTCCAATTGACTTACTAACTTCTA-3′. This full-length product was amplified and cloned, and the sequence was confirmed by bidirectional sequencing using vector-specific primers M13F and M13R (Invitrogen). This shark sequence was compared with the human isoform using ClustalW.

Generation of expression construct for oocyte injection.

The 1,188-bp start-to-stop PCR product was cloned in a pRAT vector and used as the template for cRNA synthesis using T7 in vitro transcription (Ambion, Austin, TX). Human TASK-1 plasmid cDNA in the pRAT vector was kindly provided by Dr. Steve N. Goldstein (University of Chicago). Capped hTASK-1 cRNA was synthesized using T7 RNA polymerase and in vitro transcription.

Oocyte preparation and expression of shark and human TASK-1.

Mature female Xenopus laevis were anesthetized in a 0.15% cold solution of tricaine for 20 min, and several ovarian lobules were removed under sterile conditions through an abdominal incision (MDIBL IACUC protocol no. 10-03). Oocytes were incubated in a 2.5 mg/ml solution of type I collagenase for 1 h to effect defolliculation. Mature stage V and VI oocytes were selected and stored. After 12–24 h, the oocytes were injected with 1 ng shark TASK cRNA/50 nl, 1 ng of human TASK, or an equivalent volume of water. Oocytes were then stored at 18–20°C in modified Barth solution holding medium (MBSH) containing (in mM) 88 NaCl, 1 KCl, 2.4 NaHCO₃, 0.82 MgSO₄, 0.33 Ca(NO₃)2·4H₂O, 10 HEPES (5 Na salt, 5 acid), buffered to pH 7.4, and 150 mg/l gentamicin sulfate.

Two electrode voltage clamping.

Electrophysiological recordings were performed 18–30 h after injection. Electrodes were pulled on a micropipette puller (Sutter Instruments, Novato, CA) and had input resistances between 0.5 and 2.5 MΩ. Electrodes were filled with 3M KCl. I-V curves were obtained by clamping the voltage over a series of steps from −140 to +60 mV at 20-mV increments (each increment 250 ms), using a Dagan TEV-200 amplifier and Axon 1320 digidata interface (Dagan Instruments, Axon Instruments).

Oocytes were perifused with frog Ringer solution containing (in mM) 98 NaCl, 2 KCl, 1.8 CaCl₂, 1 MgCl₂, and 10 HEPES, buffered to pH 7.4. ND96 containing (in mM) 93.5 NaCl, 2 KCl, 1.8 CaCl₂, 2 MgCl₂, and 5 HEPES was the perifusion solution for experiments with increasing potassium concentrations with KCl. (Fig. 6).

Fig. 6.

Effect of increasing K⁺ concentration on shark and human TASK-1 expressed in Xenopus oocytes. A: shark TASK-1. B: human TASK-1. Oocytes were perifused with ND96 (K⁺ concentration 2 mM) to which additional potassium concentrations were added as KCl. The legend indicates the extracellular potassium concentration to which oocytes expressing the channels were exposed.

I-V ramps were taken under basal conditions approximately every 8 min, with multiple readings in the basal state before drug addition to the Ringer solution and for at least three readings when a steady state was reached after the drug. Electrophysiological data were acquired and analyzed with pCLAMP software (version 9.0; Axon Instruments).

Antibody design and generation.

Two antibodies to shark TASK-1 were used in this study, the first to detect the protein in Western blots and the second for immunochemistry. The first antibody was made and purified by Pocono Rabbit Farms (Canadensis, PA). This antibody was raised in a host rabbit to residues 39–59 of shark TASK-1. It detected a band in SRG of the expected size (44–50 kDa) in shark rectal gland and other tissues. The epitope was KNLEERRFALMTKYNLSEKKY. The epitope localization was intracellular; the identity of the antigen peptide was confirmed by mass spectroscopy and amino acid analysis.

For immunochemistry, we purchased an antibody from Alomone Labs (Jerusalem BioPark, Hadassah Ein Kerem, Jerusalem, Israel). The antibody was made to epitope EDEKRDAE HRALLTRNGQ, corresponding to amino acid residues 252–269 of human TASK-1. The epitope localization was intracellular in the COOH terminal part. The homology with other species, including shark, was 17/18 residues identical. The control antigen was also from Alomone Labs.

Western Blot Analysis and Confocal Immunofluorescence Imaging

Shark tissues were homogenized in lysis buffer containing 1% NP-40, 150 mM NaCl, 20 mM Tris base, pH 8.0, 5 mM EDTA, 10 mM NaF, 10 mM Na₄P₂O₇, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 10 mM iodocetimide, 1 mM PMSF, and 1 mM sodium vanadate. The lysate was spun at 15,000 rpm for 15 min, and the protein concentration of the supernatant was determined using a BCA assay. The lysate was diluted 1:1 with sample buffer containing 62.5 mM Tris·HCl, pH 6.8, 2% SDS, 10% glycerol, 0.01% bromophenol blue, and 5% β-mercaptoethanol. Lysates were subjected to SDS-PAGE, electrotransferred to Immobilon transfer membranes and probed with shark-specific anti-TASK-1 primary antibody (1:10), donkey anti-rabbit secondary antibody (1:20,000) (Jackson Immunoresearch), and visualized with the enhanced chemiluminescence (ECL) detection system (Amersham Biosciences, Piscataway, NJ).

For immunofluorescent imaging, stimulated perfused rectal glands were snap frozen and fixed with paraformaldehyde, and 4-μm sections were obtained from the gland 2–3 mm from the distal tip. Specimens were stored at −20°C. and rehydrated in PBS 3× for 5 min and blocked for 20 min with 1% BSA in PBS. Slides were washed 3× for 5 min with PBS and probed with a sheep antishark TASK-1 antibody at a 1:10 dilution. The primary antibodies were visualized with FITC-conjugated antisheep secondary antibody (shark rectal gland slices and human bronchial epithelia). Preparations were treated with actin and nuclear labels and viewed with an Olympus Fluoview confocal microscope (Center Valley, PA).

Statisitics and data analysis.

Statistical significance was determined using Excel (Microsoft Office) and GraphPad Prism statistical functions. P < 0.05 was considered statistically significant. Inhibitor dose-responses were fit using GraphPad Prism software.

RESULTS

In Vitro Perfusion of Shark Rectal Glands

From 30 min until the end of each experiment, chloride secretion was activated by perfusion of forskolin (1 μM) and IBMX (100 μM) (Fig. 1). Seventeen potassium channel inhibitors (Table 1) were added individually to the perfusate from 50 to 70 min during forskolin + IMBX stimulation in 123 shark rectal gland perfusion experiments and then were removed from 70 to 90 min to access reversibility. The maximum level of inhibition achieved with each inhibitor was compared with the 50–70-min controls without the inhibitor. The doses perfused, the number of experiments with each channel blocker, and a scale of percent inhibition with statistics are summarized in Table 1.

Fig. 1.

In vitro SRG perfusions illustrating the effects of BaCl₂, acidic conditions, bupivicaine, and anandamide on stimulated Cl⁻ secretion. All glands were perfused to basal levels for 30 min, then the secretagogues forskolin (1 μM) + IBMX (100 μM) were added to the perfusate for the remainder of the experiment. In experimental perfusions, K⁺ channel inhibitors were added to the perfusate from minute 50 to minute 70; then they were removed for the remainder of the experiment. A: effect of BaCl₂, a universal K⁺ channel inhibitor, on stimulated chloride secretion (n = 6, P < 0.001 vs. controls at each time point between 52 and 70 min). B: effect of low pH (addition of HCl to perfusate buffer) on stimulated chloride secretion (n = 6 at each pH, P < 0.01 at pH 6.0 at each time point between 53 and 70 min). C: effects of bupivacaine (n = 6, P < 0.001 vs. controls at each time point between 53 and 70 min). D: effects of anandamide (n = 9, P < 0.01 vs. controls at each time point between 55 and 70 min). A, C, and D: n = 10 for controls. Values are mean Cl secretion (μeq·h⁻¹·g⁻¹) ± SE.

Table 1.

Effects of K⁺ channel inhibitors on chloride secretion in the perfused SRG

Drug	Concentration perfused based on Ki values	n	Scale of Inhibition of Cl− Secretion
BaCl2	5 mM	6	++++
Acidosis	pH 6	6	++++
Bupivicaine	2.5 mM	6	++++
Quinidine	200 μM	6	++++
Anandamide	150 μM	9	+++
Zinc acetate	200 μM	9	+++
Quinine	1 mM	6	+++
Glybenclamine	1 μM	6	++
Lidocaine	1 mM	6	++
Chromanol 293b	200 μM	6	+
Tolbutamide	100 μM	15	0
Phentolamine	200 μM	8	0
Charybdotoxin	50 nM	6	0
TEA	10 μM	8	0
3,4 DAP	1 μM	9	0
Chlortrimazole	10 μM	8	0
Sotalol	100 μM	3	0

Explanation of scale and P values. SRG, shark rectal gland; TEA, tetraethylammonium; 3,4 DAP, 3,4-diaminopyridine. 0, no inhibition; +0–20% inhibition; ++20–40% inhibition; +++40–80% inhibition; ++++80–100% inhibition; 0 vs. controls, P = NS; +vs. controls, P < 0.05; ++vs. controls, P < 0.05; +++vs. controls P < 0.01; and ++++vs. controls P < 0.001.

Figure 1, A–D, illustrates the dramatic response to inhibitors that implicate a 4TM-2P K⁺ channel as a contributor to the secretory response. BaCl₂ (Fig. 1A), a potent but nonselective K⁺ channel blocker, was used as a positive control for comparison, and it inhibited chloride secretion from 2,700 to <300 meq·h⁻¹·g⁻¹. Lowering the perfusion solution pH to 6.0 (Fig. 1B) also inhibited chloride secretion dramatically, as did bupivacaine (2.5 mM) (Fig. 1C) and the TASK-1-specific inhibitor anandamide (100 μM) (Fig. 1D), Lowering of pH, bupivicaine, and anandamide are each inhibitors of TASK channels (6, 11, 14, 47). Each had striking barium-like inhibitory effects and reduced chloride secretion by 50–100%. Anandamide had slower kinetics at its onset, which may fit with its lipid solubility. Anandamide inhibited by 50% in perfusion studies (Fig. 1D) compared with 54% in Xenopus oocytes Figure inhibition by bupivicaine was entirely reversible, while anandamide was partially reversible when the solution was switched to forskolin (1 μM) and IBMX (100 μM) without the inhibitor. Zinc acetate (100 μM) (Table 1), a specific inhibitor of the TASK-1 channel (12), showed strong reversible inhibition.

As shown in Table 1, quinidine, an inhibitor of several K⁺ channels, including 4 TM-2P channels, 6 transmembrane (TM) voltage-sensitive K⁺ channels [including KvLQT1, K_A, and K_V(r)], and volume-sensitive K⁺ channels (9, 10, 13, 50, 56, 59), showed a reversible barium-like effect in the perfused SRG, as did quinine. Glibenclamide, an inhibitor of KATP channels (15, 39, 60) had only modest inhibitory effects, possibly due to inhibition of CFTR (46, 52, 57). Lidocaine, a blocker of 4TM, 2P K⁺ channels and cell-volume sensitive K⁺ channels (28, 34) showed moderate and reversible inhibition. Chromanol 293b, an inhibitor of slow delayed rectifier K⁺ current (IKs) (3, 16) had minimal effect. Tolbutamide and phentolamine, inhibitors of K_ATP channels (21), and charybdotoxin, which acts on BK_CA and K_V channels (48, 66), had no effects. Although the published K_i of charybdotoxin is 0.5 nM and the concentration we used was 50 nM, 100-fold above the K_i, this concentration of charybdotoxin would be insufficient to inhibit BK channels in renal epithelial cells. TEA and 3,4 DAP, blockers of K_A channels (67), also had no effect on secretion. Clotrimazole an inhibitor that acts on IK_CA channels (30, 42), and Sotalol, a blocker of rapid delayed rectifier K⁺ channels (41, 62), had no effects on chloride secretion in the SRG. Although our results point to the TASK-1 potassium channel as the dominant K channel in the SRG, there could possibly be other potassium channels present in this tissue.

Cloning of shark TASK-1 from the shark rectal gland.

The results of perfusion studies with K⁺ channel inhibitors and low-pH perfusate narrowed our search to the two-pore-domain (4TM 2P/KCNK/K2P) family of potassium channels. Our strategy to clone the shark-specific potassium channel and PCR gel images are illustrated in Fig. 2. Using degenerate PCR and shark cDNA template reverse transcribed from RNA extracted from shark tissues, one primer pair amplified a putative TASK-1 fragment (394 bp) in the SRG, brain, gill, and kidney (Fig. 2A, lanes 7–10). Using the initial fragment sequence, we constructed primers for 3′ and 5′ RACE PCR, which yielded a 694-bp 3′ product that included the stop codon, and a partial 5′ product of 592 bp that did not include the start codon. (Fig. 2C) To obtain the remainder of the 5′ sequence, genome walking was used to obtain a 428-bp product, including the start codon and 334 bp of the untranslated region (Fig. 2C).

Fig. 2.

Cloning of shark TASK-1. A: degenerate PCR fragments (394 bp) of shark TASK-1 with cDNA template from SRG (lane 7), brain (lane 8), kidney (lane 9), gill (lane 10). Lanes 2–5 are degenerate primer pairs that did not amplify template cDNA. cDNA was omitted in the PCR reaction in lanes 6 and 11. B: full-length clone, confirmed with nested PCR. Lane 2, 1,386-bp primary product; lane 3, 1,282-bp nested product. One-kilobit plus ladder was run in lane 1 of these gels. C: cloning strategy for obtaining full-length sequence of shark TASK-1. 394-bp fragment from degenerate PCR was identified as TASK-1. Using this sequence, we constructed primers for 5′ and 3′ RACE, which yielded a 694 bp 3′ product that included the stop codon, and a partial 5′ product of 592 bp. To obtain the remainder of the 5′ sequence Genome Walking (GW) was used to obtain a 428-bp product, which included the start codon and 334 bp of upstream untranslated region. Using the complete sequence, primers were designed to amplify a 1,282 bp full-length clone. *Position of the start and stop sequence of shark TASK-1.

With the complete shark sequence identified, PCR primers were designed to generate a full-length clone. An agarose gel showing the full-length clone confirmed with nested PCR is seen in Fig. 2B. The full-length clone (1282 bp) had an 1,128 bp open reading frame encoding a protein of 375 amino acids. The sequence was confirmed with bidirectional sequencing, and the full-length clone was compared against the genomic database using BLAST (basic local alignment search tool) with the highest homology to the human TASK-1 channel (1,188 bp, 394 amino acids). The nucleotide sequence was 71% conserved between shark and human isoforms, while conservation was 80% at the amino acid level. Fig. 3 shows a ClustalW alignment of the shark and human TASK-1 amino acid sequences. Major structural features of the human protein are conserved in the shark ortholog, including the four transmembrane segments (% identity M1–M4: 100%, 90.5%, 95%, and 100%), the 2P domains (P1 and P2: 95.8% and 95.8%), the short NH₂-terminal (100%), the middle (70%), and the long COOH-terminal cytoplasmic element (66.9%). The predicted molecular mass of sTASK-1 is 42.4 kDa.

Fig. 3.

Amino acid alignment of shark and human TASK-1 proteins. Transmembrane domains (TM1-4) are identified by red boxes, pore domains (P1-2) by green boxes, and cytoplasmic domains by blue text. The GYG/GFG K⁺ selectivity motif conserved in each pore region is highlighted in green. Human amino acid residues that are identical to shark TASK-1 are represented as dots. Dashes indicate gaps between the two sequences.

Two-electrode voltage-clamping electrophysiology.

Shark TASK-1 and human TASK-1 were then expressed in Xenopus oocytes. Similar current-voltage relationships of human and shark TASK-1 are shown in Fig. 4A. The dose-dependent inhibition of the 4TM 2P family blocker quinine (0.1–1 mM), and the TASK-1 blocker bupivacaine (0.1–2.5 mM) are shown in Fig. 4, B and C. At the concentrations used, both blockers markedly inhibited (75–80%) the channels from the two species.

Fig. 4.

Expression of shark TASK-1 (sTASK) and human TASK-1 (hTASK) in Xenopus oocytes. A: I-V relationship of sTASK-1 and hTASK-1 in oocytes demonstrating identical reversal potentials, current, and outward rectification compared with water-injected controls. n = 32 for hTASK-1, n = 44 for sTASK-1, and n = 3 for water-injected controls. B: dose-dependent inhibition of shark vs. human TASK-1 by quinine. n = 8 for sTASK-1, and n = 6 for hTASK-1. IC_50s are 0.17 mM for human TASK-1 and 0.18 mM for shark TASK-1. C: dose-dependent inhibition of shark vs. human TASK-1 by bupivicaine. n = 11 for sTASK-1, and n = 6 for hTASK-1. IC_50s are 0.3 mM for human TASK-1 and 0.4 mM for shark TASK-1 Bupivicaine and quinine are relatively specific blockers of TASK channels and show a similar inhibition in both shark and human TASK-1.

The effects of pH on the shark and human orthologs are seen in Fig. 5A. Shark and human TASK-1 channels are inhibited greater than 90% at low pH but have different pK_as (7.74 ± 0.02 for shark vs 7.36 ± 0.03 for human). These pK_as are remarkably similar to the arterial pH of each species where the channels function. These differences are highly significant (P < 0.001 at pH 7.4).

Fig. 5.

Shark and human TASK-1 channels expressed in Xenopus oocytes have different sensitivities to pH and zinc acetate. A: different pH sensitivity in shark and human TASK-1. The pKa for shark is 7.74 ± 0.02, and the pKa for human is 7.36 ±. 0.03. pKa was determined as the pH at the midpoint between maximum and minimum values. These differences are highly significant (P < 0.001 at pH 7.4). n = 9 for shark TASK-1, and 11 for human TASK-1. B: shark TASK-1 is more sensitive to zinc than human TASK-1. n = 12 for shark TASK-1, and 6 for human TASK-1. (P < 0.01 sTASK vs. hTASK at all concentrations of zinc).

Zinc is a channel modulator that blocks select members of the TASK family. Zinc is 10 times more effective against TASK-1 than TASK-3, and is ineffective against TASK-2 (17, 23, 33, 36, 49). We found that shark TASK-1 is much more sensitive to zinc than human TASK-1 (P < 0.01 sTASK vs. hTASK at all concentrations of zinc). (Fig. 5B). Figure 6 depicts the effect of increasing K⁺ concentrations on shark (A) and human (B) TASK-1. Because KCl was added to increase the K⁺ concentration, the observed depolarizing shift in the reversal potentials with increased K concentration (Fig. 6) may be related in part to cell shrinkage due to the osmolality.

Finally, we examined the effect endocannabinoid anandamide, the amide derivative of arachidonic acid, on shark TASK-1 expressed in oocytes (Fig. 7). Anandamide has been shown to be a selective inhibitor of human TASK-1 channels (44). Figure 7 shows substantial inhibition (>54%) of the shark TASK-1 channel expressed in oocytes with anandamide, with inhibition at all concentrations above 20 μM (P < 0.01 compared with controls). This inhibition by anandamide is comparable to its effect in the perfused gland (Fig. 1D), where anandamide inhibition was 50%.

Fig. 7.

Effect of anandamide on shark TASK-1 expressed in Xenopus oocytes. Anandamide is a specific inhibitor of TASK-1 channels (44) Anandamide markedly inhibited sTASK-1 at all concentrations above 20 μM (P < 0.01 compared with controls). IC₅₀ is 99.6 μM for anandamide.

Western blot analysis.

Using an anti-shark antibody prepared by Pocono Rabbit Farms (Canadensis, PA), we detected the TASK-1 protein in Western blots of shark tissue lysates. The antibody was raised in a host rabbit to residues 39–59 of shark TASK-1. It detected (Fig. 8) a band in SRG of the expected size (44–50 kDa) (4, 29) and confirmed the presence of the TASK-1 protein. Similar-size bands were also detected in brain, muscle, and heart. Preincubation with 3× antigen peptide completely eliminated all bands (Fig. 8, bottom).

Fig. 8.

Detection of TASK-1 in shark tissue lysates by Western blot analyses. Top: an antibody raised to residues 39–59 of shark TASK-1 detected a band in SRG of the expected size (44–50 kDa) (4). Similar size bands were also detected in brain, muscle, and heart. Bottom: preincubation with 3× antigen peptide completely eliminated all bands. Scale: 25–100 kDa.

Using an anti-shark TASK-1 antibody, we obtained TASK-1 bands in lysates of several shark tissues, including rectal gland, brain, muscle, and heart (Fig. 8, top). Bands are ∼44–50 kDa and were similar to hTASK-1 (4, 29), confirming the presence of the protein. In Fig. 8, the lower graph is a control Western blot preincubating the TASK-1 antibody with 3× antigenic peptide; no bands were detected.

Immunohistochemistry of shark rectal gland tubules and human bronchial epithelia.

With the anti-shark TASK-1 antibody, intense specific staining (green) of the basolateral membrane of SRG tubules was evident. (Fig. 9, A–C). In SRG tubules, red staining is the nuclear dye propidium iodide. (TASK-1 staining is indicated by red arrows, Fig. 9, A–C). A control longitudinal SRG tubule incubated with a five-fold TASK antigen excess incubated with the anti-shark TASK-1 antibody is shown in Fig. 9D. Using an antibody to mammalian TASK 1, we visualized this protein in human bronchial epithelial cells, which showed intense basal membrane staining (Fig. 9E; TASK-1 staining is indicated by the red arrow in E). In these cells, some binding of antibody to TASK-1 was observed in a subapical location beneath the actin array. In these human airway cells (E), red is actin stained with rhodamine phalloidin, and blue is a nuclear DNA dye.

Fig. 9.

Confocal images showing membrane localization of the TASK-1 protein in SRG tubules and human bronchial epithelia. Intense specific staining of TASK-1 in the basolateral membrane of tubules was seen in SRG with minimal signal present in the apical membrane domain. In SRG tubules red staining indicates nuclei stained with propidium iodide, and green staining indicates TASK-1 (TASK 1 is indicated by red arrows, A, B, and C). A: cross section of a SRG tubule. B and C: longitudinal sections of SRG tubules. D: longitudinal SRG tubule preincubated with 5× antigen peptide blocking TASK-1 staining. E: human bronchial epithelial cells showing intense basal membrane staining (TASK-1 indicated by red arrow, D) and some binding in a subapical location beneath the actin array. In these cells, red is actin stained with rhodamine phalloidin, and blue is a nuclear DNA dye.

DISCUSSION

Our studies in the SRG provide the first evidence that TASK-1 channels play a critical role in establishing the membrane potential that drives chloride secretion through CFTR in the SRG.

A role for 4 TM-2P potassium channels has been emphasized in the biology of excitable tissues, but their function in epithelial tissues has not been clarified. In the heart, central nervous system, smooth muscle, and adrenal cortex, 4 TM-2P K channels contribute to setting the membrane voltage and maintaining cell volume (4, 14, 40, 43, 45). In the SRG model, despite extensive electrophysiological studies (8, 25, 27–29) and the previous cloning of two potassium channels (32, 63, 64), the identity of the functionally dominant potassium channel remained unknown.

We first carried out inhibitor fingerprints (Table 1) to narrow the search for the dominant channel in the SRG. Inhibitors of potassium channels, if the channel is present, should inhibit chloride secretion by depolarizing the membrane potential. These studies indicated that in the SRG, Ca²⁺-activated K⁺ channels, ATP-dependent and voltage-sensitive K⁺ channels, delayed rectifiers, and K_A channels do not play important roles in the regulation of chloride secretion. In contrast, the TASK-1 inhibitors, anandamide, acidic pH, zinc, and quinidine markedly inhibit chloride secretion in the SRG, indicating the importance of this channel in the single function of this model.

We identified the full-length clone of shark TASK-1 and found it to be 80% identical in amino acids to the human TASK-1 protein. Major structural features of the human protein are conserved in the shark ortholog, including the four-transmembrane segments, the 2P domains, the short NH₂- and long COOH-termini, and an extended extracellular loop between M1 and P1.

In the kidney, TASK-1 and TASK-2 channels are sensitive to pH changes in a narrow range near the physiological pH (37). Our findings that the pK_as of shark and human TASK-1 differ significantly (7.74 ± 0.02 vs. 7.36 ± 0.03) and approximate the arterial pH of each species (7.7 in shark, 7.4 in humans) indicate that these proteins are excellent sensors of pH.

The TASK-1 protein was previously identified by mRNA expression studies in human pancreas, placenta, brain, heart, lung, and kidney (45). TASK-1 has segment-specific expression in the human nephron, being present in the glomerulus and distal nephron segments (38).

This work has direct implications for secretory epithelia in mammals and human cystic fibrosis, in particular. Many chloride secretory epithelia, including bronchi, airway cells, pancreas, small intestine, and the vas deferens in males are defective in the human disease cystic fibrosis. The first human Northern blots indicate that TASK 1 channels are highly expressed in lung, pancreas, small intestine, and colon (17). Each of these organs have abnormal chloride transport in patients with cystic fibrosis. Therefore, drug openers of the TASK-1 potassium channel, which should hyperpolarize the cellular membrane, could enhance chloride secretion through defective CFTR in patients with cystic fibrosis.

Our perfusion studies in the SRG indicate that acidic pH perfusate markedly inhibits chloride secretion. When shark TASK-1 and human TASK-1 protein are expressed in Xenopus oocytes, they are both inhibited (>90%) by lowering the pH. The pKa for shark is 7.74 ± 0.02 and the pKa for human is 7.36 ± 0.03. Given our findings that the pK_as approximate the arterial pH of each species (7.4 in human and 7.7 in shark), we speculate that there is a physiological advantage to control chloride secretion through inhibition by acidic pH and activation by alkaline pH.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.J.T., S.E.D., W.W.M., A.W.P., A.P.M., R.A.F., and J.N.F. conception and design of research; C.J.T., S.E.D., W.W.M., A.W.P., A.P.M., and J.N.F. performed experiments; C.J.T., S.E.D., W.W.M., A.W.P., A.P.M., R.A.F., and J.N.F. analyzed data; C.J.T., S.E.D., W.W.M., A.W.P., A.P.M., R.A.F., and J.N.F. interpreted results of experiments; C.J.T., S.E.D., W.W.M., A.W.P., A.P.M., R.A.F., and J.N.F. prepared figures; C.J.T., S.E.D., W.W.M., A.W.P., A.P.M., R.A.F., and J.N.F. drafted manuscript; C.J.T., S.E.D., W.W.M., A.W.P., A.P.M., R.A.F., and J.N.F. edited and revised manuscript; C.J.T., S.E.D., W.W.M., A.W.P., A.P.M., R.A.F., and J.N.F. approved final version of manuscript.