Short- and Long-Term Functional Alterations of the Skeletal Muscle Calcium Release Channel (Ryanodine Receptor) by Suramin: Apparent Dissociation of Single Channel Current Recording and [H]Ryanodine Binding

The present study demonstrates the following characteristic suramin actions on the purified skeletal muscle calcium release channel in single-channel current recordings and [H]ryanodine binding to HSR: 1) Suramin (0.3–0.9 mM) induced a concentration-dependent increase in the open probability (Po > 0.9) at 20 to 100 mM Ca and an almost fully open channel at 1 mM Ca (Po 5 0.95) with a marked shift to longer open states (to3/to4). Suramin increased the apparent calcium affinity to the activating high-affinity calcium binding sites and reduced the apparent magnesium affinity to the inhibitory low affinity Ca/ Mg binding sites. 2) Channel activation by suramin and sulfhydryl oxidation was additive. 3) Suramin (0.9 mM) reversed the Ca-calmodulin–induced channel inhibition at 0.1 or 1 to 5 mM Ca-calmodulin. 4) The open probability of the suramin activated channel was almost completely inhibited by 10 mM Mg or Ca on short suramin exposure. Prolonged suramin exposure (30–60 min) resulted in a time-dependent, slow increase in Po, with long open states of low frequency in the presence of 10 to 20 mM Mg or Ca. 5) Magnesium induced inhibition of Po (IC50 5 0.38 mM) and equilibrium [ H]ryanodine binding (IC50 5 0.30 mM) agreed well in control channels, but were dissociated in the presence of 0.9 to 1.0 mM suramin (IC50 5 0.82 mM versus 83 mM). [H]ryanodine binding seemed to monitor predominantly the long-term alteration in channel function. 6) The multiple effects of suramin on channel function suggest an allosteric mechanism and no direct effects on binding of endogenous ligands involved in channel gating. Calcium release in skeletal muscle occurs via the calcium release channel (ryanodine receptor) located in the terminal cisternae of the sarcoplasmic reticulum (Imagawa et al., 1987; Inui et al., 1987; Lai et al., 1988; Smith et al., 1988). The skeletal muscle calcium release channel (RyR-1) is a homotetramer with a molecular mass of 2260 kDa (Takeshima et al., 1989; Zorzato et al., 1990). Calcium release and the gating properties of the calcium release channel at a single channel level are regulated by endogenous effectors including calcium, magnesium, adenine nucleotides, the calcium binding proteins calmodulin and sorcin, the immunophillin FK506-binding protein, phosphorylation by protein kinases, sulfhydryl oxidation by nitric oxide, and various exogenous effectors (see reviews: Coronado et al., 1994; Meissner, 1994; Melzer et al., 1995; Zucchi and RoncaTestoni, 1997). The trypanocidal drug suramin, a polysulfonated napthylurea, is a potent activator of the ligand-gated calcium release channel of sarcoplasmic reticulum. Suramin released calcium from passively loaded skeletal muscle sarcoplasmic reticulum vesicles (Emmick et al., 1994) as well as from cells containing the RyR-3 isoform, such as Jurkat T-lymphocytes (Hohenegger et al., 1999). The gating of the calcium release channel in skeletal (RyR-1) and cardiac (RyR-2) muscle is markedly influenced by suramin or suramin analogs. Characteristic for the activating effect of suramin in single-channel current recordings was an increase in longer open states, especially with the cardiac calcium release channel and the requirement of lower suramin concentrations to increase the open probability of the cardiac calcium release channel (Sitsapesan and Williams, 1996; Sitsapesan, 1999) compared with that from the skeletal muscle channel (Hohenegger et al., 1996; Sitsapesan and Williams, 1996; Klinger et al., 1999). Furthermore, [H]ryanodine binding to the calcium release channel of heavy sarcoplasmic reticulum (HSR) of skeletal muscle was activated by suramin, but only marginABBREVIATIONS: RyR, ryanodine receptor; HSR, heavy sarcoplasmic reticulum; 4-CMPS, 4-(chloromercuri)phenyl-sulfonic acid; PMSF, phenylmethylsulfonyl fluoride; DTT, 1,4-dithiothreitol; Mops, 4-morpholinepropanesulfonic acid; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]1-propanesulfonic acid; PAGE, polyacrylamide gel electrophoresis; CaM, calmodulin; DIDS, 4,49-diisothiocyanostilbene-2,29-disulfonic acid; Po, open probability; ANOVA, analysis of variance. 0026-895X/01/5903-543–556$3.00 MOLECULAR PHARMACOLOGY Vol. 59, No. 3 Copyright © 2001 The American Society for Pharmacology and Experimental Therapeutics 293/884049 Mol Pharmacol 59:543–556, 2001 Printed in U.S.A. 543 at A PE T Jornals on Jauary 6, 2018 m oharm .aspeurnals.org D ow nladed from ally inhibited by very large concentrations of calcium or magnesium (Emmick et al., 1994; Hohenegger et al., 1996; Klinger et al., 1999). Suramin released the RyR-1 bound to a calmodulin Sepharose (Klinger et al., 1999) and inhibited [I]calmodulin binding to HSR (Suko et al., 2000). Because of the effects of suramin on the skeletal muscle calcium release channel found in ligand binding studies to HSR (reduction in calmodulin binding, marginal inhibition of [H]ryanodine binding by high concentrations of magnesium or calcium), the present investigation was carried out to determine the effect of the inhibitory endogenous ligands (calcium-calmodulin, magnesium, and millimolar calcium) on the suramin-induced activation at a single channel level. Ryanodine is known as a very valuable probe to monitor functional states of the calcium release channel and a good correlation between open probability in single-channel current recordings and the amount of [H]ryanodine bound to the channel proteins was usually observed (Coronado et al., 1994; Meissner, 1994). Recent studies showed that a single point mutation of the mouse RyR-2 (alanine 4812 to glycine), the suggested pore-forming region of the calcium release channel (Zhao et al., 1999), or mutation of amino acids in the luminal loop linking the M3/M4 region of the RyR-1 (Gao et al., 1999) reduced single-channel current fluctuation and [H]ryanodine binding (Gao et al., 1999; Zhao et al., 1999). For the above reason, [H]ryanodine binding to HSR was also carried out in the absence or presence of suramin under similar conditions as used in single-channel current recordings. The results demonstrate both short and long-term functional changes in single-channel current recordings, alterations in the gating of the calcium release channel by calcium and magnesium, and an apparent dissociation between single-channel current recordings and [H]ryanodine binding in the presence of suramin. Experimental Procedures Materials. Suramin, calmodulin, 4-(chloromercuri)phenyl-sulfonic acid (4-CMPS), MOPS, HEPES, Tris, histidine, CsCl (ultra pure), NaCl (ultra pure), ruthenium red, leupeptin, pepstatin, antipain, phenylmethylsulfonyl fluoride, tetracaine, and neomycin were purchased from Sigma-Aldrich GmbH (Vienna, Austria); [H]ryanodine was purchased from DuPont-New England Nuclear (Boston, MA); ryanodine was from Agrisystems International (Wind Gap, PA); phosphatidyl serine, phosphatidylethanolamine, and phosphatidylcholine were from Avanti Polar Lipids, Inc. (Alabaster, AL); Delrin bilayer chambers (CD22–200; CD13–200) were from Warner Instrument Corp. (Hamden, NJ). Aprotinin was a generous gift from Bayer Austria AG (Vienna, Austria). All reagents and agents (suramin, calmodulin, 4-CMPS, [H]ryanodine) were dissolved in MilliQ deionized water. Preparation of Sarcoplasmic Reticulum. Heavy sarcoplasmic reticulum vesicles (HSR) from rabbit skeletal muscle were prepared as described previously (Suko and Hellmann, 1998). Briefly, white back muscle (fast twitch muscle) was homogenized in a Waring Blender for 1.5 min in a medium containing 10 mM histidine buffer, pH 7.0, and 100 mM NaCl, and centrifuged for 35 min at 4,000g. The supernatant was filtered through cheese cloth and centrifuged for 30 min at 30,000g. The pellet was resuspended in 10 mM histidine buffer, pH 7.0, 0.6 M KCl, and 250 mM sucrose and centrifuged for 35 min at 100,000g. The pellet was washed once in a medium containing 10 mM histidine buffer, pH 7.0, 100 mM NaCl, and 200 mM sucrose, centrifuged again for 35 min at 100,000g and stored at 280°C or used immediately for the purification of the ryanodine receptor-calcium release channel. All buffers used for the preparation and resuspension of HSR contained 0.5 mg/ml leupeptin, 1 mg/ml antipain, 1.4 mg/ml aprotinin, 1 mM pepstatin, 0.1 mM PMSF, 1 mM benzamidine. Preparation of Calcium Release Channel (Ryanodine Receptor). The calcium-release channel of the terminal cisternae of sarcoplasmic reticulum vesicles was prepared as described previously (Suko and Hellmann, 1998), which was a slight modification of the preparation used before that (Suko et al., 1993). Briefly, heavy sarcoplasmic reticulum vesicles from rabbit skeletal muscle (prepared as above) were solubilized with CHAPS (medium, 40 mM Mops/Tris, pH 7.0, 1 M NaCl, 2 mM DTT, 1% CHAPS, 0.25% or 0.5% phosphatidylcholine, 0.5 mg/ml leupeptin, 1 mg/ml antipain, 1.4 mg/ml aprotinin, 1 mM pepstatin, 0.1 mM PMSF, 1 mM benzamidine, 15 mg/ml HSR; incubation, 60 min at 3–4°C), followed by centrifugation twice for 35 min at 103,000g (Beckman 65 rotor). The supernatant was centrifuged through a linear 7.5 to 20% sucrose gradient equilibrated in 40 mM Mops/Tris, pH 7.0, 300 mM NaCl, 2 mM DTT, 0.5% CHAPS, 0.25% or 0.5% phosphatidyl-choline, 0.5 mg/ml leupeptin, 1 mg/ml antipain, 1.4 mg/ml aprotinin, 1 mM pepstatin, 0.1 mM PMSF, and 1 mM benzamidine for 14 h at 2°C (Beckman SW28 rotor; 38 ml tubes). Fractions containing the ryanodine receptor (determined by SDS-PAGE) were pooled and dialysed for 24 h in a medium containing 20 mM Mops/Tris, pH 7.0, 100 mM NaCl, 2 mM DTT, 0.15 mM CaCl2, 0.1 mM EGTA, 0.5 mg/ml leupeptin, 1 mg/ml antipain, 1.4 mg/ml aprotinin, 1 mM pepstatin, 0.1 mM PMSF, and 1 mM benzamidine. Sucrose (200 mM final concentration) was added to the proteoliposomes before storage at 278°C. Preparation and dialysis were carried out at 2 to 4°C. SDS-Polyacrylamide Gel Electrophoresis (PAGE). SDSPAGE was performed in 5% polyacrylamide gels (0.75 mm thickness) with 3% stacking gels as described previously (Suko et al., 1993). Sucrose gradient fractions were added to a medium containing 10 mM Tris/HCl, pH 6.8, 2% SDS, 2% mercaptoethanol, and 10% glycerol and boiled for 2 min. Gels were stained with 0.05% Coomassie blue in 10% acetic acid. Molecular mass standards were run on two separate lanes of the same gel: Ferritin (440 kDa), thyroglobulin (330 kDa), and myosin (212 kDa). Gradient fractions with the highest content of ryanodine receptor were pooled and used for the preparation of proteoliposomes. Single Channel Recordings. Single-channel recordings were carried out after incorporation of purified calcium release channels (ryanodine receptors) into planar lipid bilayers, essentially as described previously (Suko and Hellmann, 1998). Planar lipid bilayers were formed from phosphatidylserine (10 mg/ml) and phosphatidylethanolamine (10 mg/ml) in decane (Avanti Polar Lipids). The lipid solution was spread over a 200-mm diameter aperture in a Delrin cup (Warner Instrument Corp.) separating two aqueous compartments. The cis bath solution (2.6 ml) and the trans bath solution (4 ml) were connected to the head stage input of a model EPC-9 amplifier (Heka Elektronik, Lambrecht, Germany) via Ag/AgCl electrodes and CsCl-agar bridges. The trans bath was held at virtual ground. Cs was used as the charge carrier through the calcium release channel to increase the conductance of the channel (Coronado et al., 1992). The cis solution was composed of 10 mM HEPES/ Tris, pH 7.4, 480 mM CsCl, and 50 to 100 mM CaCl2 or 100 mM CaCl2 plus 80 mM EGTA (free calcium, 20 mM). The trans solution was composed of 10 mM HEPES/Tris, pH 7.4, and 50 mM CsCl without added calcium or plus calcium in concentrations as used in the cis bath. Unless stated otherwise, purified calcium release channels and other reagents were added to the cis chamber. Recordings were filtered at 4 kHz with a low-pass Bessel filter, digitized at 40 kHz (sampling rate 25 ms) and stored on the hard disc of a Apple Macintosh (Apple, Cupertino, CA). Single channel events were identified using TAC software (ver 2.5; Skalar Instruments, Inc., Seattle, WA). Mean open probability (Po) of channels were identified by a 50% threshold analysis. The life times of open and closed events were 544 Suko et al. at A PE T Jornals on Jauary 6, 2018 m oharm .aspeurnals.org D ow nladed from determined by the method of maximum likelihood (TACFit software; Skalar Instruments). [H]Ryanodine Binding. [H]Ryanodine binding was measured under equilibrium conditions as described previously (Suko and Hellmann, 1998). Unless stated otherwise, controls and test samples were assayed in duplicate or triplicate for 90 min at 37°C in 0.2 ml of solution containing 40 mM Mops/Tris, pH 7.0, 0.5 M CsCl (or 1 M NaCl), 0.1 mg HSR, 0.5 mg/ml leupeptin, 1.4 mg/ml aprotinin, 0.1 mM PMSF, and 10 nM [H]ryanodine, 100 mM Ca without or with suramin (0.1 to 10 mM). In a few experiments, the above control medium contained, in addition, 5 to 10 mM EGTA, 10 to 20 mM ruthenium red, 1 mM tetracaine, or 1 mM neomycin. In the calcium dependence experiments, free calcium was varied between 20 nM and 50 mM. In the magnesium inhibition experiments, magnesium was varied between 0.25 and 50 mM MgCl2. Nonspecific [ H]ryanodine binding was measured in the presence of 100 mM unlabeled ryanodine. Samples were filtered on glass-fiber filters (presoaked in 1% polyethylene imine) and washed with 10 ml of 20 mM Mops/Tris,