A Novel Human Cl- Channel Family Related to Drosophila flightless Locus
2004; Elsevier BV; Volume: 279; Issue: 21 Linguagem: Inglês
10.1074/jbc.m313813200
ISSN1083-351X
Autores Tópico(s)Ion Channels and Receptors
ResumoLarge conductance chloride (maxi-Cl-) currents have been recorded in some cells, but there is still little information on the molecular nature of the channel underlying this conductance. We report here that tweety, a gene located in Drosophila flightless, has a structure similar to those of known channels and that human homologues of tweety (hTTYH1–3) are novel maxi-Cl- channels. hTTYH3 mRNA was found to be distributed in excitable tissues. The whole cell current of hTTYH3 was large enough to be discriminated from the control but emerged only after treatment with ionomycin. Analysis of pore mutants suggested that positively charged amino acids contributed to anion selectivity. Like a maxi-Cl- channel in situ, the hTTYH3 single channel showed 26-picosiemen linear current voltage, complex kinetics, 4,4′-diisothiocyanato-stilbene-2,2′-disulfonic acid sensitivity, subconductance, and the permeability order of I- > Br- > Cl-. Similarly, hTTYH2 encoded an ionomycin-induced maxi-Cl- channel, but TTYH1 encoded a Ca2+-independent and swelling-activated maxi-Cl- channel. Therefore, the hTTYH family encoded maxi-Cl- channels of mammals. Further studies on the hTTYH family should lead to the elucidation of physiological and pathophysiological roles of novel Cl- channel molecules. Large conductance chloride (maxi-Cl-) currents have been recorded in some cells, but there is still little information on the molecular nature of the channel underlying this conductance. We report here that tweety, a gene located in Drosophila flightless, has a structure similar to those of known channels and that human homologues of tweety (hTTYH1–3) are novel maxi-Cl- channels. hTTYH3 mRNA was found to be distributed in excitable tissues. The whole cell current of hTTYH3 was large enough to be discriminated from the control but emerged only after treatment with ionomycin. Analysis of pore mutants suggested that positively charged amino acids contributed to anion selectivity. Like a maxi-Cl- channel in situ, the hTTYH3 single channel showed 26-picosiemen linear current voltage, complex kinetics, 4,4′-diisothiocyanato-stilbene-2,2′-disulfonic acid sensitivity, subconductance, and the permeability order of I- > Br- > Cl-. Similarly, hTTYH2 encoded an ionomycin-induced maxi-Cl- channel, but TTYH1 encoded a Ca2+-independent and swelling-activated maxi-Cl- channel. Therefore, the hTTYH family encoded maxi-Cl- channels of mammals. Further studies on the hTTYH family should lead to the elucidation of physiological and pathophysiological roles of novel Cl- channel molecules. Chloride (Cl-) anion passes through the cell membrane in greatest abundance mainly by channels. Cl- channels play roles in the stabilization of membrane potential, transport, and cell volume regulation. Cl- channels are observed in many tissues and are thus variably regulated by voltage, calcium, pH, or cell volume. There are three types of Cl- channels based on their conductance: small ( pS, maxi-Cl-) conductance channels. Small and medium conductance channels are found ubiquitously, even in oocytes, whereas large conductance channels are rarely found. To our knowledge, less than 12 cells possessing maxi-Cl- channels have been found over the past 40 years. A maxi-Cl- channel in muscle or in other tissues is activated directly by cytosolic Ca2+ (1Thorn P. Martin R.J. Q. J. Exp. Physiol. 1987; 72: 31-49Crossref PubMed Scopus (30) Google Scholar, 2Fahmi M. Garcia L. Taupignon A. Dufy B. Sartor P. Am. J. Physiol. 1995; 269: E969-E976Crossref PubMed Google Scholar), whereas activity of the channel in other tissues is independent of Ca2+ (3Nilius B. Droogmans G. Physiol. Rev. 2001; 81: 1415-1549Crossref PubMed Scopus (756) Google Scholar, 4Kemp P.J. MacGregor G.G. Olver R.E. Am. J. Physiol. 1993; 265: L323-L329PubMed Google Scholar). GTP-binding protein (5Sun X.P. Supplisson S. Torres R. Sachs G. Mayer E. J. Physiol. (Lond.). 1992; 448: 355-382Crossref Scopus (50) Google Scholar), cell volume (3Nilius B. Droogmans G. Physiol. Rev. 2001; 81: 1415-1549Crossref PubMed Scopus (756) Google Scholar), and protein kinases (6Groschner K. Kukovetz W.R. Pfluegers Arch. 1992; 421: 209-217Crossref PubMed Scopus (43) Google Scholar) have been reported to modify the function of a maxi-Cl- channel. However, because of their rarity, the physiological role of maxi-Cl- channels remains obscure. The molecular structures of Cl- channels are diverse, with different transmembrane segments (TMSs) such as cystic fibrosis transmembrane regulator (7Schwiebert E.M. Benos D.J. Egan M.E. Stutts M.J. Guggino W.B. Physiol. Rev. 1999; 79: S145-S166Crossref PubMed Scopus (379) Google Scholar) and ClC (8Jentsch T.J. Stein V. Weuinreich F. Zdebik A.A. Physiol. Rev. 2002; 82: 503-568Crossref PubMed Scopus (1056) Google Scholar) with 10 or 12 TMSs. Aquaporin-6 encodes a unique acid-dependent Cl- channel with 6 TMSs (9Yasui M. Hazama A. Kwon T.H. Nielsen S. Guggino W.B. Agre P. Nature. 1999; 402: 184-187Crossref PubMed Scopus (404) Google Scholar). The ClCA family, with 5 TMSs, encodes medium conductance Ca2+-activated Cl- (CaC) channels (10Gandhi R. Elble R.C. Gruber A.D. Ji H.L. Copeland S.M. Fuller C.M. Pauli B.U. J. Biol. Chem. 1998; 273: 32096-32101Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar, 11Gruber A.D. Elble R.C. Ji H.L. Schreur K.D. Fuller C.M. Pauli B.U. Genomics. 1998; 54: 200-214Crossref PubMed Scopus (205) Google Scholar). Thus, it is possible that there exists a Cl- channel with more variable TMSs. On the other hand, cDNA encoding ion channels is generally expressed in ovary cells, such as Chinese hamster ovary (CHO) and Xenopus oocytes. They possess Cl- channels endogenously. CaC currents are frequently observed as endogenous currents of these cells. The CaC current is driven by small or medium conductance Cl- channels activated by a high concentration of intracellular Ca2+, is outwardly rectifying, and is inhibited by 4,4′-diisothiocyanato-stilbene-2,2′-disulfonic acid (DIDS) and niflumate (3Nilius B. Droogmans G. Physiol. Rev. 2001; 81: 1415-1549Crossref PubMed Scopus (756) Google Scholar). Human embryonal kidney (HEK) cells have also been used for the expression of cDNA encoding CaC (10Gandhi R. Elble R.C. Gruber A.D. Ji H.L. Copeland S.M. Fuller C.M. Pauli B.U. J. Biol. Chem. 1998; 273: 32096-32101Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar), but they possess endogenous maxi-Cl- channels regardless of the concentration of Ca2+ (12Zhu G. Zhang Y. Xu H. Jiang C. J. Neurosci. Methods. 1998; 81: 73-83Crossref PubMed Scopus (86) Google Scholar). To try to find a novel ion channel, we used the BLAST program to search for proteins that have four or more TMSs as predicted by hydrophobicity and a gene related to behavior abnormality. By using a cluster of leucine residues as a transmembrane probe, we found a human gene (httyh3) that has homology to Drosophila tweety (dttyh1) located in the flightless locus (13Maleszka R. Hanes S.D. Hackett R.L. DeCouet H.G. Gabormiklos G.L. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 447-451Crossref PubMed Scopus (97) Google Scholar), which is related to behavior abnormality. flightless is a mutant of Drosophila melanogaster. The flightless gene is molecularly characterized by four transcription units within it, which are named tweety (twe), fli, dodo (dod), and penguin (pen). These genes are required for normal flight. We therefore investigated the localization and expressed currents of hTTYH3 in the present study. Isolation and Detection of hTTYH3—Human TTYH3 cDNA (AL530584) was purchased from Incyte Genomics. Mutants were made by using PCR (QuickChange, Stratagene). Northern blot analysis was performed with a membrane (human multiple tissues Northern blot, Clontech, Tokyo, Japan) according to the manufacturer's protocol using a PstI fragment as a probe. Mouse RNA was isolated by using the guanidine thiocyanate method with organic extraction (Trizol, Invitrogen). A mouse TTYH3 (mTTYH3) fragment was cloned by PCR followed by reverse transcriptase (RT) using a primer set (5′-AAGCTGTCGGGCAGCCACAA-3′ and 5′-CGATGGAGCTGAACATGAGG-3′) that detects an 84-amino acid fragment from S positions 310–394. RT-PCR detections of human, hamster (Cricetulus griseus), and mouse TTYH3 (hmcTTYH3) exons in the cell lines and tissues were performed by conventional methods. Human endothelial cells (number 375) and smooth muscle cells (number 716) were purchased from Applied Cell Biology Research Institute. Neuronal cells (PC12 and N2A) were gifts from K. Shimazaki (Department of Neurophysiology, Jichi Medical School, Japan). The mTTYH3 protein was detected in mouse tissues. Specific antibodies were raised against the C-terminal region of MRAKYLATSQ, which was common in human, hamster, and mouse clones. An antigen (1 mg/ml distilled water) conjugated with 1 ml of Freund's adjuvant was intramuscularly injected into New Zealand White rabbits, and this was followed by biweekly booster injections of the same dose of the antigen in Freund's incomplete adjuvant. Serum in which the titer was over 10,000 times higher than that of the control was obtained 8 weeks later. The immunized rabbit serum was passed through a column (Amersham Biosciences) and affinity-purified by a kit (Pierce) according to the manufacturer's protocol. Extracts of subconfluent HEK293 or CHO cells (in 60-mm dishes) with or without transfection of hTTYH3 cDNA were lysed by gentle sonication in 50 mm NaH2PO4, 300 mm NaCl, 10 mm imidazole, 0.05% Tween 20 (pH 8.0), and protease inhibitors. Digestion was performed with the proteins at a concentration of 20 mg/20 ml and incubated with N-glycanase (0.5 units/20 μl) for 18 h at 37 °C. Western blot analysis with a blocking test by an excess of antigen (1 μg/ml) was performed to evaluate specificity. Immune complexes were detected by horseradish peroxidase-labeled anti-rabbit IgG by using an ECL detection system (PerkinElmer Life Sciences). Histologic staining was performed by conventional methods and detected by fluorescent isothiocyanate (FITC)-labeled anti-rabbit IgG (Dako, Kyoto, Japan). The tissues were viewed through a microscope (BX-5, Olympus, Tokyo, Japan), and photographs were stored in a computer. Expression of cDNA—A plasmid expressing green fluorescence protein (pEGFP-N1, Clontech, Tokyo, Japan) or red fluorescence protein (pDsRed1-N1, Clontech) was used as a marker for transfection. 0.5 μg of the marker cDNA and 1 μg of hTTYH3 cDNA or its mutants were transfected into mammalian cells grown on coverslips using FuGENE6™ (Roche Applied Science). Growth and transfection were performed in 10% fetal calf serum-containing media. The cells were used without any further treatment for patch clamp samples. The experiments were performed at 48–72 h after the transfection. To make stable transformed cells, we inserted hTTYH3 cDNA into a plasmid (pIREShygr3, Clontech). The CHO cells were transfected by electroporation, and the stable transformant was cloned using the immunoreactivity as a marker. Electrophysiology—Patch clamp recordings were carried out according to conventional methods (14Suzuki M. Morita T. Hanaoka K. Kawaguchi Y. Sakai O. J. Clin. Investig. 1991; 83: 735-742Crossref Scopus (41) Google Scholar). The experiments were performed at room temperature (20–25 °C). Micropipettes were made from capillary tubes (Drummond Scientific, Broomall, PA) that were coated by bees-wax before fire-polishing. They had a resistance of <10 megohms. Series resistance in a whole cell configuration was around 300 megohms. The GFP-positive cells were visualized by fluorescence measurement (CAM 2000 System, Jasco, Tokyo, Japan) with an emission of 490 nm. Currents were recorded with an EPC-9 patch clamp amplifier (HEKA, Pfalz, West Germany). Applied voltage and sampling (10 kHz) were controlled by a computer system (HEKA, Pfalz, West German). Samples were filtered through 1 kHz, and analysis was performed using the software Patch Analyst Pro. The bath solution contained 140 mm NaCl (or TEA-Cl or sodium gluconate), 1.0 mm MgCl2, and 3 mm HEPES (pH 7.4). The whole cell patch pipette contained a filtered solution of 140 mm CsCl, 1.0 mm MgCl2,and 3 mm HEPES (pH 7.2) with or without 0.5 mm CaCl2. To calculate the ratio of X- to Cl- permeability in a whole cell configuration, the current-clamp mode was used to measure reversal potential. Then the reversal potential was compensated by liquid junction potential, and the ratio was calculated by using an equation (15Hille B. Ionic Channels of Excitable Membranes. 2nd Ed. Sinauer Associates, Inc., Sunderland, MA1992: 337-361Google Scholar). For single channel analysis, the pipette solution was changed to 140 mm NaX, 1.0 mm MgCl2, and 3 mm HEPES (pH 7.4). Single channel recording was performed in inside-out patches, and then reversal potential was calculated to obtain the selective permeability. To obtain the desired Ca2+ concentration, Ca/EGTA concentration in the solution for the cytoplasmic side was adjusted with appropriate amounts of KCl, KOH, MgCl2, and HEPES, keeping a constant ionic strength (Σ ions2) of 1.5. A hypotonic solution was made by addition of distilled water to the bath solution. Osmolality was measured using an osmometer (One-Ten Osmometer, Fiske, MA). The data were analyzed using one-way analysis of variance, and the significance was calculated using Bonferroni's analysis. p < 0.05 was considered statistically significant. Isolation and Characterization of hTTYH3 Protein—The sequence of the purchased cDNA (AL530584) had significant homology to ttyh1 and ttyh2 (NM020659 and NM032646) and was named ttyh3. The gene ttyh3 is located in 7p22.3. Other family members of ttyh3 were found by using the BLAST program; one was found in Drosophila (CG3638) and two were found in Homo sapiens (ttyh1 and ttyh2). hTTYH3 mRNA encodes 4800 nucleotides and 523 amino acids. hTTYH1 is thought to have 5 TMSs (16Campbell H.D. Kamei M. Claudianos C. Woollatt E. Sutherland G.R. Suzuki Y. Hida M. Sugano S. Young I.G. Genomics. 2000; 68: 89-92Crossref PubMed Scopus (32) Google Scholar), whereas hTTYH3 is thought to possess 6 TMSs. Both were predicted by the Sosui program (GenomNet, Kyoto University) (Fig. 1A). The third TMS (TMS3) was predicted in hTTYH3 but not in hTTYH1. The cluster of leucine used for the search was found in TMS4. To determine the membrane topology, a marker (His6) was inserted between TMS3 and TMS4 (amino acids 199–209) and between TMS5 and TMS6, which were expressed in mammalian cells and then detected with an FITC-labeled antibody against His6 (data not shown). The mutants containing the marker in the given positions (amino acids 202, 205, 208, and 257, Fig. 1A, green triangles) were positive, and the other mutants (red triangles) were negative for binding when the antibody was incubated outside the cells. However, addition of His6 may change the transmembrane topology. Detection of an ttyh3 product and glycosylation scanning were performed with an antibody raised against C-terminal peptides. The amino acid sequence of the C-terminal region used for the antibody is a common sequence among hamsters, mice, and humans. Western blot analysis showed positive signal in expressed CHO cells (Fig. 1B), whereas immunoreactivity was not detected in control CHO cells. On the other hand, immunoreactivity was faintly detected by Western blotting and clearly detected by immunohistochemistry in HEK cells (Fig. 1C), suggesting that hTTYH3 is an endogenous Cl- channel. Incubation of the same membrane with an excess of antigen (1 μg/ml) abolished the signals, indicating that the antibody is specific for hTTYH3 detection. Possible glycosylation sites (Asn-Xaa-Thr/Ser, red circle) were found at the positions 128, 144, and 353. Glycosylation scanning was performed by using N-glycanase digestion. The protein of hTTYH3 expressed in CHO was incubated with N-glycanase, resulting in digestion of the upper band, but the sequence-predicted 58-kDa band remained. Thus, the upper band in CHO was a glycosylated form of hTTYH3. When threonine was substituted by alanine at the three sites, Thr at position 353 was the only site glycosylated. Thr at 353 was therefore located in an extracellular domain. Therefore, the structure of hTTYH3 is like that illustrated in Fig. 1A (lower panel), regardless of whether TMS3 (white column) penetrates the membrane or not. The putative structure of hTTYH3 may be comparable with that of the Ca2+-activated large conductance potassium channel (17Piskorowski R. Aldrich R.W. Nature. 2002; 420: 500-503Crossref Scopus (60) Google Scholar, 18Schreiber M. Yuan A. Salkoff L. Nat. Neurosci. 1999; 2: 416-421Crossref PubMed Scopus (126) Google Scholar) (maxi-K+, Fig. 1A, lower). The pore region in maxi-K+ determines potassium ion selectivity located between TMS5 and TMS6. The position of the pore region in hTTYH3 is predicted to be similar to that of maxi-K+, where positively charged amino acids, arginine and histidine, are commonly conserved in the TTYH family. Maxi-K+ has a unique aspartate cluster below the last transmembrane domain. This negatively charged cluster is called the “Ca2+ bowl” and is related to direct activation by Ca2+ (18Schreiber M. Yuan A. Salkoff L. Nat. Neurosci. 1999; 2: 416-421Crossref PubMed Scopus (126) Google Scholar). A Ca2+ bowl-like sequence was recognized in the TTYH family as a glutamate- or Asp-rich domain. Knockout of the maxi-K+ channel gene resulted in a “slowpoke,” whereas that of the tweety gene resulted in a “flightless” in Drosophila; both genes are related to behavior abnormality. Localization of hTTYH and mTTYH3—hTTYH3 and mTTYH3 are distributed mainly in excitable tissues. hTTYH3 mRNA of 4.8 kb was found by Northern blotting in the brain, heart, skeletal muscle, colon, spleen, kidney, and peripheral blood leukocytes (Fig. 2A). RT-PCR using mouse tissues revealed positive signals also in fat, the pancreas, thymus, and uterus (Fig. 2B). mTTYH3 was also detected in an endothelial cell line, kidney epithelial cells (opossum kidney and HEK), and a neuronal cell line (PC12 and N2A). To confirm the absence of a hamster cTTYH3 product in CHO cells, we cloned a partial sequence of hamster cTTYH3 (AB158504), 84 amino acids, from genomic DNA of CHO cells by PCR (arrows in Fig. 1A). A comparison of the hamster sequence with those of other species showed that one and nine amino acids differed from those in mouse and human, respectively. By using species-common primers, cTTYH3 cDNA was not found in CHO by RT-PCR. Immunohistological analysis was performed on mouse tissues (Fig. 3). Thalamic neurons and ventricular cells of the heart and skeletal muscle membrane were positively stained. A negative control was obtained with an excess of antigen. In the kidney, cells in glomeruli, presumably mesangium cells, were positively stained. Although opossum kidney cells or HEK cells showed positive hmcTTYH3 signals by RT-PCR, renal proximal tubules were not positively stained. The signal may be indistinguishable because the level of autofluorescence of proximal tubules is high and/or the level of expression of mTTYH3 is low in this segment. Therefore, mTTYH3 is located mainly in excitable membranes. Whole Cell Recordings of hTTYH3-transfected CHO Cells— We first selected HEK293 cells for the expression of hTTYH3, because HEK293 cells were used for discovery of the ClCA family (10Gandhi R. Elble R.C. Gruber A.D. Ji H.L. Copeland S.M. Fuller C.M. Pauli B.U. J. Biol. Chem. 1998; 273: 32096-32101Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar). When transiently expressed in HEK293 cells, hT-TYH3 was associated with a prominent Cl- current that was activated by ionomycin (10-6m). The current was much less marked in GFP vector (mock)-transfected cells. However, a similar large conductance Cl- single channel (12Zhu G. Zhang Y. Xu H. Jiang C. J. Neurosci. Methods. 1998; 81: 73-83Crossref PubMed Scopus (86) Google Scholar) was detected in the following investigation. Therefore, we used CHO cells. The basal current in hTTYH3-transfected CHO cells was not different from that in untransfected control cells. Although ionomycin induced an outward-rectified current in mock-transfected cells, it induced an overt linear current in hTTYH3-transfected cells (Fig. 4A). The current was not altered by substitution of sodium with TEA but was altered by substitution of Cl- with gluconate in the bath solution, suggesting that the current is a current driven by an anion. The permeability ratio of Cs+/Cl- was nominally zero and that of gluconate/Cl- was 0.33 (15Hille B. Ionic Channels of Excitable Membranes. 2nd Ed. Sinauer Associates, Inc., Sunderland, MA1992: 337-361Google Scholar). To investigate the ion selectivity, we used whole cell patches with a pipette solution containing 0.5 mm CaCl2, because this large amount of Ca2+ was required to stabilize the current-voltage relation. The ionic selectivity was altered in hTTYH3 pore mutants by exchanges of charged amino acids (Fig. 4B). Mutant H370D showed a different selectivity, in which gluconate/Cl- decreased from 0.3 to 0.12 and Cs+/Cl- increased to 0.1. Mutant R366Q was more permeable to cations. The reversal potential with isosmotic NaCl/CsCl shifted to 12.5 mV (mean of eight), suggesting significant permeability to Na+. Assuming that TEA and gluconate are impermeable ions, the Na+/Cs+/Cl- permeability ratio was calculated to be 0.7:0.25:1. Thus, positively charged amino acids at positions 366 and 370 composed the pore of hTTYH3, playing a role in anion selectivity. Halide selectivity of hTTYH3 was precisely measured by single channel analysis. Anions in the pipette solution were changed in each experiment, and reversal potential was measured (n = 5 in each). The resultant order of permeability (P X/Cl-) was I- (2.2), Br- (1.8), SCN- (1), NO3− (0.83), gluconate- (0.3), aspartate- (0.2), and Na+ (<0.1). Effects of Blockers on the hTTYH3 Channel—The effects of blockers were investigated next. The evoked current in non-transfected cells was blocked by niflumate, whereas the evoked current in hTTYH3-transfected cells was not inhibited by niflumate (300 μg/ml). The CaC current was not inhibited by 10 μm dithiothreitol, 10 μm ZnCl2, or 10 μm SITS but was completely inhibited by 10 μm DIDS (Fig. 4C, left). Effects of reagents on the single hTTYH3 channel in an inside-out configuration were also examined. Niflumate had no effect (Po = 0.8 to 0.82, n = 4). SITS induced a flickering opening but had no significant inhibitory effect (Po = 0.58, n = 4), whereas DIDS induced a blockade (Po = 0.12, n = 0.4, p < 0.01) within a few seconds (Fig. 4C, right). Therefore, the hTTYH3 current was only sensitive to DIDS. Single Channel Analysis of hTTYH3—hTTYH3 was characterized by single channel analysis. Maxi-Cl- channel activity was observed within 30 s after detachment of the membrane when bath solution contained 0.1 mm of Ca2+ (Fig. 5A). Such a large conductance was observed in hTTYH3-transfected cells but not in non-transfected cells. The current-voltage relation was linear in symmetrical Cl- solution (n = 12) with slope conductance of 260 pS (Fig. 5B). Open probability was not dependent on voltage of less than 0 mV but decreased at voltage over 60 mV (Fig. 5C). The distribution of current amplitudes fitted well to results of Gausian's analysis, indicating a subconductance of 50 pS (Fig. 5D). The channel had a long open lifetime as was observed in the trace at +40 mV, whereas it also had a fluctuating rapid opening as was observed in the trace at -60 mV. Long opening and rapid (but transient) fluctuations were observed in the same patch membrane. The hTTYH3 channel therefore involved a multiopening mechanism. Open-closed lifetimes fit better to two kinetic states (τo1 = 52, τo 2 = 510 ms, τc1 = 60 and τc2 = 250 ms, mean of 12) rather than 1. As in the case of a maxi-K+ channel (17Piskorowski R. Aldrich R.W. Nature. 2002; 420: 500-503Crossref Scopus (60) Google Scholar) and ClCA channel (10Gandhi R. Elble R.C. Gruber A.D. Ji H.L. Copeland S.M. Fuller C.M. Pauli B.U. J. Biol. Chem. 1998; 273: 32096-32101Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar), a high concentration of cytosolic Ca2+ was essential for hTTYH3 activation. The open probability was plotted against Ca2+ concentration, resulting in an ED50 of about 2 μm (Fig. 5E). The range of Ca2+ concentrations is similar to that for maxi-K+ activation (18Schreiber M. Yuan A. Salkoff L. Nat. Neurosci. 1999; 2: 416-421Crossref PubMed Scopus (126) Google Scholar). The value 2 μm is within the physiologic range of Ca2+ in excitable cells in which hTTYH3 is detected. Thus, like the maxi-K+ channel, hTTYH3 may play a role in Ca2+ signal transduction. However, it took around 10 s for activation when the cytosolic side of the isolated membrane was exposed to high Ca2+ solution. Thus, protein phosphorylation or another complex mechanism might be involved in the activation. To search for activators, activity of hTTYH3 was observed in a cell-attached configuration with stably transformed CHO cells, and the presence of hTTYH3 channels in the attached membrane was determined later in an inside-out configuration in high Ca2+ solution. Although protein kinase sites on the amino acid sequences were predicted, direct opening of the hTTYH3 channel by protein kinase A (dibutyryl cAMP, 10-5 m) or kinase C (phorbol ester, 10-8 m) was not observed. Addition of stauro-sporine (10-6 m), GTPγS (10-5 m), or okadaic acid (10-5 m) to the bath solution did not activate the hTTYH3 channel (data not shown). Therefore, the physiological role of hTTYH3 remains obscure and awaits in vivo study. Expression and Activation of hTTYH1 and hTTYH2 Channels—To elucidate the physiological role of the TTYH family, we examined the functional expression of other members, hTTYH1 and hTTYH2. Based on the above results, we tested whether hTTYH1 and hTTYH2 were also expressed as evoked Cl- channels (Fig. 6). Activity of the hTTYH1 single channel was sometimes observed in cell-attached patches, showing slope conductance of 280 pS. However, the hTTYH2-induced channel appeared only after treatment with ionomycin. hTTYH2 was expressed as a Ca2+-dependent inward-rectified Cl- channel of 120 pS in an inward direction and of 45 pS in an outward direction. In a cell-attached configuration, reversal potential of hTTYH1 and hTTYH2 was around -30 mV, which is similar to the Cl- equilibrium potential. By substitution of NaCl with TEA-Cl in a pipette solution, the permeability ratio, PNa/PCl, was calculated to be less than 0.1 for hTTYH1 and 0.1 for hTTYH2. We next tested the influence of a hypotonic solution on hTTYH1 and hTTYH3 channels. In a whole cell configuration with CsCl in the pipette solution without Ca2+ (pCa = 7.2), the current by hTTYH1 was usually silent but remarkably activated when the bath osmolarity was changed to a hypotonic solution of 250 mOsm (Fig. 7A). The current was further activated in a hypotonic solution of 220 mOsm. A whole cellular current-voltage relation of hTTYH1 in a hypotonic solution was not obtained because the magnitude of the current was unstably high. Because the current at 0 mV was stable even in a hypotonic solution, this prominent current was not due to leakage but due to fluctuated opening and closing of the large channel. The current magnitude at 100 mV was measured (n = 6, Fig. 7B). The evoked current at 220 mOsm was diminished by 50 μm GdCl3 and abolished by 10 μm DIDS. On the other hand, the current of hTTYH3-expressed cells was significantly activated only when exposed to a solution of 220 mOsm. The evoked current was blocked by 50 μm GdCl3. This magnitude of activation in a hypotonic solution was, however, observed in GFP-transfected control CHO cells (19Suzuki M. Mizuno A. Kodaira K. Imai M. J. Biol. Chem. 2003; 278: 22664-22668Abstract Full Text Full Text PDF PubMed Scopus (584) Google Scholar). Thus, we conclude that hTTYH1 is involved in a swelling-activated Cl- channel, whereas hTTYH3 may not be involved. Thus, the common feature of these three channels is large conductance. In this study, we showed that the hTTYH family possesses 5 or 6 transmembrane segments encoding a large conductance Cl- channel. The structure of hTTYH3 was estimated by computer-based hydrophobic analysis, addition of tag alignment, and glycosylation scanning. The results of a mutation study suggested that the structure of this novel maxi-Cl- channel is similar to the structure of known cation channels, in which a selective pore is located between TMS5 and TMS6 and a regulatory domain is located in the C terminus. An ion channel with the pore domain between TMS5 and TMS6 is a prototype structure of cation-permeable channels, such as the K+ channel encoded by the Kv family and the Ca2+ channel encoded by the Trp family (20Minke B. Cook B. Physiol. Rev. 2002; 82: 429-472Crossr
Referência(s)