Amphibian ryano ine receptor isoforms are related to those of mammalian s eletal or cardiac muscle F. ANTHONY LAI, &I-Y1 LIU, LE XU, ANITA EL-HASHEM, NEAL R. KRAMARCY, ROBERT SEALOCK, AND GERHARD MEISSNER Departments of Biochemistry and Biophysics, and Physiology, University of North Carolina, Chapel Hill, North Carolina 27599- 7260 Lai, F. Anthony, Qi-Yi Liu, Le Xu, Anita El-Hashem, Neal R. Kramarcy, Robert Sealock, and Gerhard Meissner. Amphibian ryanodine receptor isoforms are related to those of mammalian skeletal or cardiac muscle. Am. J. Physiol. 263 (Cell PhysioZ. 32): C365C372, 1992.-The ryanodine receptor (RyR)-Ca2+ release channels of frog skeletal muscle have been purified as 30s protein complexes comprised of two high molecular weight polypeptides. The upper and lower bands of the frog doublet comigrated on sodium dodecyl sulfate polyacylamide gels with the mammalian skeletal and cardiac RyR polypeptides, respectively. Immunoblot analysis showed that a polyclonal antiserum to the rat skeletal RyR preferentially cross-reacted with the upper band, whereas monoclonal antibodies to the canine cardiac RyR preferentially cross-reacted with the lower band of the frog receptor doublet. Immunoprecipitation studies indicated the presence of two homooligomer 30s RyR complexes comprised of either the lower or upper polypeptide band of the frog doublet, and immunocytochemical staining revealed their colocalization in frog gastrocnemius muscle. After planar lipid bilayer reconstitution of the 30s frog RyR, single-channel currents were observed that exhibited a Na+ and Ca” + conductance and pharmacological characteristics similar to those of the mammalian skeletal and cardiac Ca’+ release channels. These results suggest that amphibian skeletal muscle expresses two distinct RyR isoforms that share epitopes in common with the mammalian skeletal or cardiac RyR. excitation-contraction skeletal muscle

coupling;

calcium

release channel;

frog

AND CARDIAC MUSCLE, an action potential propagated along the surface membrane and its infoldings, the transverse tubule (T-tubule), triggers the rapid release of Ca2+ ions from an intracellular compartment, the sarcoplasmic reticulum (SR) (7, 8). This process, commonly referred to as excitation-contraction (E-C) coupling, is believed to be mediated by several integral membrane proteins located in the T-tubule and SR membranes. A dihydropyridine (DHP) receptor in the cardiac surface and T-tubule membranes has been shown to function as a voltage-dependent Ca’-)+channel (19), which, by facilitating Ca2+ influx during a surface membrane action potential, is thought to induce the opening of a Ca2+-activated Ca2+ release channel in the SR of cardiac muscle (5, 20). In contrast, the skeletal muscle DHP receptor is believed to function as a voltage-sensing molecule, which effects skeletal muscle SR Ca2+ release through a direct physical interaction with a SR Ca2+ release channel (3, 25, 30). The mammalian cardiac and skeletal muscle DHP receptor and Ca”+ release channel proteins have been biochemically purified and their primary structures determined by cDNA cloning and sequencing. These studies have indicated that the skeletal and cardiac DHP receptors are related oligomeric complexes (6, 27), which, however, exhibit differences in amino acid sequence and IN SKELETAL

0363-6143/92

$2.00

Copyright

function of one of their major subunits (19). Similarly, the mammalian skeletal and cardiac Ca2+ release channels are homologous channels comprised of four molecular weight (M,) -560,000 polypeptides and have also been shown to be distinct with respect to their Ca”+ activation (17), mobility on sodium dodecyl sulfate (SDS) gels (9, 13), and amino acid sequence (21, 23). Although biochemical studies have focused on mammalian tissues, frog skeletal muscle has been the preferred system to study the physiological mechanisms of E-C coupling and SR Ca’)+ release. In this study, we have sought to correlate biochemical and physiological studies by purifying the frog skeletal SR Ca’* release channel. Our studies indicate that the frog Ca’+ release channel can be isolated as a large 30s complex possessing a high-affinity ryanodine binding site, as has been observed for mammalian skeletal and cardiac ryanodine receptors (2, 11). SDS-polyacrylamide gel electrophoresis (PAGE) of the frog ryanodine receptor indicated the presence of two high molecular weight proteins that corresponded to the mammalian skeletal and cardiac isoforms in electrophoretic mobility. Two recent reports have shown that in avian, bullfrog, and toad fish skeletal muscle, there appears to be a coexpression of two electrophoretically and immunologically distinct ryanodine receptor isoforms (1, 22). Here, we present immunological data indicating that the frog ryanodine receptor isoforms can be distinguished using antibodies raised against the rat skeletal and canine cardiac ryanodine receptors. Some of the results of this study have been presented in abstract form (12, 15). EXPERIMENTAL

PROCEDURES

Isolation of SR vesicles and X8 Ca2+ release channel complex. SR vesicles enriched in [“Hlryanodine binding activity were isolated from the leg muscles of Rana pipiens as a 2,600- to 100,000-g pellet (16) in the presence of 0.5 mM ethylene glycolbis(p-aminoethyl ether)-~,N,X’,N’-tetraacetic acid (EGTA) and protease inhibitors [lo0 nM aprotinin, 1 ,uM leupeptin, 1 PM pepstatin A, 1.0 mM benzamidine, 1.0 mM iodoacetamide, and 0.2 mM phenylmethylsulfonyl flouride (PMSF)]. The pellets were resuspended in medium A (0.6 M KCI, 100 PM EGTA, 50 FM Ca”+, 0.2 mM PMSF, 1 PM leupeptin), incubated for 1 h at 4°C and layered at the top of a discominuous sucrose gradient consisting of 20 ml of 25% and 10 ml of 40% sucrose in medium A. After centrifugation for 3 h at 150,000 g in a Beckman Ti 45 rotor, membranes at the 2540% sucrose interface were collected, diluted with two volumes of water, and sedimented. Pellets were resuspended in 0.3 M sucrose and 5 mM K-piperazine-NJ’-bis(2-ethanesulfonic acid) (PIPES), pH 7, rapidly frozen, and stored at -75°C before use. The 3- [ (3-cholamidopropyl)dimethvlammonio] -1 -propanev sulfonate (CHAPS) solubilized 30s ryanodine receptor-Cay+ release channel complex was isolated by rat/e density gradient

0 1992 the American

Physiological

Society

C365

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 11, 2019.

C366

FROG

RYANODINE

centrifugation in the presence of protease inhibitors as described by Lai et al. (11). p?Yjryanodine binding. [“HI ryanodine binding was determined as described previously (14). Membranes were incubated for 3 h at 37°C in media containing 1.0 M NaCl, 100 PM EGTA, 200 PM Ca”+, 5 mM AMP, 0.1 mM diisopropyl fluorophosphate, 5 PM leupeptin, variable concentrations of [‘jH]ryanodine (15,000 nM), and 20 mM Na-PIPES, pH 7.4. Nonspecific binding was estimated using an l,OOO-fold excess of unlabeled ryanodine. Antibody production. Polyclonal antibodies against the purified rat skeletal muscle ryanodine receptor were produced in rabbits as described using standard methods (18). Sera were reacted with rabbit skeletal SR proteins (depleted of ryanodine receptor) before use (18) and shown by immunoblot analysis to recognize the ryanodine receptor from rat, pig, dog, and rabbit skeletal muscle. Monoclonal antibodies were produced in BALB/c mice immunized by serial injection of cardiac SR vesicles and the purified canine cardiac ryanodine receptor. Spleen cells of immunized mice were fused with P3X 63 AG 8.653 myeloma cells at a 1:l ratio, plated, and grown using standard conditions (10). Hybridomas were screened by enzyme-linked immunosorbent assay and immunoblot analysis for reactivity with the purified ryanodine and canine cardiac SR membrane proteins, respectively, and classified using a Boehringer Mannheim subtyping kit. Specificity of monoclonal antibodies was determined by immunoblot analysis as described in Fig. 3. Immunoprecipitation and immunolocalization of ryanodine receptor complexes. Agarose beads with covalently attached goat anti-mouse immunoglobulin G, (IgG,) antibodies were washed once in solution B (0.5 M NaCl, IO mM Na phosphate, pH 7.4), then incubated at 4°C overnight with an equal volume of 0.15 M NaCl, 10 mM Na phosphate, pH 7.4, containing 0.010-0.5 mg/ml affinity-purified cardiac ryanodine-receptor monoclonal antibody C3-33. Beads were washed three times with 1 ml of solution B, resuspended in 0.1 ml of solution B, and then incubated with 0.2 ml of sucrose gradient fractions containing CHAPS-solubilized, purified 30s ryanodine receptor complexes (-35 pg protein/ml) of rabbit or frog skeletal or canine cardiac muscle. After incubation for 36 h at 4°C beads were washed three times with 1 ml of solution B containing 0.1% CHAPS and then treated for 30 min at room temperature with 0.1 ml SDS-PAGE sample buffer [O.l M tris(hydroxymethyl)aminomethane (Tris) . HCl, pH 6.8, 2% SDS, 3% P-mercaptoethanol, and 10% glycerol]. After pelleting of beads, 90 ~1 of supernatant was loaded onto polyacrylamide gels. Immunofluorescent localization of ryanodine receptors was done on 4-pm-thick cryostat sections of unfixed frog gastrocnemius muscle using methods described elsewhere (26). Controls included use of all antibodies over a broad concentration range, substitution of the specific antibodies by nonspecific monoclonal antibodies of the IgG, isotype or polyclonal rabbit IgG, and extensive testing of the second antibodies for cross-reactivities (26). SDS gel electrophoresis and immunoblot staining. SDSPAGE was performed in the Laemmli buffer system using 3-12% linear polyacrylamide gradient gels and 3% stacking gels (18). Unless otherwise indicated, samples were denatured for 3 min at 95°C in 0.1 M Tris HCl, pH 6.8, containing 2% SDS, 3% P-mercaptoethanol, and 10% glycerol. After electrophoresis, gels were stained with 0.1% Coomassie Brilliant Blue R-250 in 50% methanol and 10% acetic acid and destained with 10% methanol and 15% acetic acid. For immunoblots, the separated proteins from SDS-PAGE were electrophoretically transferred onto Immobilon polyvinylidene difluoride membranes (Millipore) at 15°C for 1 h at 400 mA and then for 12-15 h at 1,500 mV. Transfer membranes were blocked with 5% nonfat dried milk proteins and then incubated either with a rabbit anti-rat

RECEPTOR

ISOFORMS

skeletal muscle ryanodine receptor antiserum (l:l,OOO) (18) or with canine cardiac muscle ryanodine receptor monoclonal antibodies (1:lO). After incubation for 1 h at room temperature, blots were developed with peroxidase-conjugated secondary antibodies (1:2,000) using 3,3’-diaminobenzidine and H,O,. Single-channel recordings. Single-channel measurements were performed by incorporating the CHAPS-solubilized 30s Ca2+ release channel complex, purified in the absence of [“HIryanodine, into Mueller-Rudin-type bilayers containing phosphatidylethanolamine, phosphatidylserine, and phosphatidylcholine in the ratio 5:3:2 (50 mg/ml total phospholipid in n-decane) (11). Unless otherwise indicated, a symmetrical solution of buffered NaCl (0.5 M NaCl, 20 mM Na-PIPES, pH 7) was used to record channel currents. Small aliquots of the 30s gradient fractions were added to the cis-chamber (defined as the SR cytoplasmic side; Ref. 11) of the planar lipid bilayer apparatus. Bilayer currents were recorded and analyzed as described previously (28). Materials. [:‘H]ryanodine (54.7 Ci/mmol) was obtained from Du Pont-New England Nuclear and ryanodine from AgriSysterns International (Wind Gap, PA). Phospholipids were purchased from Avanti Polar Lipids (Birmingham, AL), CHAPS from Boehringer Mannheim, ruthenium red from Fluka, SDS gel molecular weight markers and anti-mouse IgG1 antibodies linked to beaded agarose from Sigma, and peroxidase-conjugated secondary antibodies from Calbiochem. Fluorescein- and Texas Red-conjugated Affini-Pure donkey anti-rabbit and antimouse secondary antibodies were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA). All other chemicals were of analytical grade. RESULTS

The neutral plant alkaloid, ryanodine, has been shown to bind with high and low affinity and in a Ca”+-dependent manner to the mammalian skeletal and cardiac SR Ca2+ release channels (for review see Ref. 13). Figure 1 shows specific [:jH]ryanodine binding to frog skeletal SR vesicles at free ryanodine concentrations ranging from 0.1 to 5,000 nM. Scatchard analysis of binding data (Fig. 1, inset) indicated the presence of a high-affinity site with a

g k

1.5

5

1.0

2

0.5

ii 0.0

0 0-O

.,

I

I

1o-g

Free

10-O

lo-7

[3~]~y

anodine

1

1o-6

(M)

Fig. 1. Dependence of [:‘H]ryanodine binding on [:‘H]ryanodine concentration. Frog sarcoplasmic reticulum (SR) vesicles (0.5-2.0 mg protein/ml) were incubated for 3 h at 37°C in 1.0 M NaCl, 100 FM EGTA, 200 PM Ca”‘, 5 mM AMP, 1 mM diisopropyl fluorophosphate, 5 FM leupeptin, 20 mM Na-PIPES, pH 7.0, and l-5,000 nM [:‘H]ryanodine. Specific [:‘H] ryanodine binding was determined as described by Lai et al. (14). Scatchard plot analysis (inset) indicates presence of a high-affinity site with a B,,,, of ,- 8 pmol of [ :lH ] ryanodine bound/mg protein and a K,, of 4.4 nM.

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 11, 2019.

FROG

RY’ANODINE

RECEPTOR

maximum b inding (B ma ,) value of 8.1 t 4.2 pmol/mg protein and dissoci ation constan t (KD) of 4.1 t 1.4 nM (n = 3). Under identical binding conditions, rabbit skeletal and canine cardiac SR vesicles bound [“Hlryanodine with B,,, values of 14.2 t 2.7 (n = 3) and 6.2 t 1.9 (n = 3) pmol/mg protein and Kp of 5.0 t 1.5 (n = 3) and 2.2 & 0.7 (n = 3) nM, respectively. Data in Fig. 1 further indicate that, for each high-affinity [“Hlryanodine binding site, there were present two or more sites of lower affinity. Ca2+ dependence of [“HI ryanodine binding to frog and rabbit skeletal and canine cardiac SR membranes is compared in Fig. 2. At micromolar Ca”+ concentrationsJ3H]ryanodine binding to frog SR membranes showed a Ca2+ dependence similar to that to rabbit SR membranes, whereas at millimolar Ca2+ concentrations [“Hlryanodine binding to frog SR membranes resembled more closely that to canine cardiac SR membranes. SDS-PAGE of frog skeletal SR vesicles indicated the presence of two high-M, protein bands (Lanef, Fig. 3A). The upper and lower bands of the frog doublet comigrated with the 1M, --560,000 band of rabbit skeletal (lane s, Fig. 3A) and canine cardiac (lane c, Fig. 3A) SR, respectively (21, 23). The rabbit skeletal and canine cardiac M, -560,000 bands have been previously identified to form tetrameric 30s complexes, which comprise the SR ryanodine receptor-Ca”+ release channel and are identical with protein bridges (feet) that span the gap between T-membranes and SR membranes (9, 13). The SR ryanodine receptor has been shown to be highly susceptible to attack by endogenous and exogenous proteases (Ref. 18 and cited references). It was therefore conceivable that the lower band of the frog ryanodine receptor doublet resulted from partial degradation of the

I

-

-8

0

lf7

10

-8

lo-6

1

1o-4

I

1o-3

I

1o-2

[Ca2+l (M) Fig . 2. Ca”+ dependence of [:iH]ryanodine binding to frog and rabbit skeletal and canine cardiac SR vesicles. Specific [:‘H]ryanodine binding to frog (n), rabbit (o), and canine (A) SR vesicles was determined at indicated Ca2+ concentrations as described in Fig. 1, using a protein concentration of 80, 100, and 120 pg/ml, respectively, and [:‘H]ryanodine concentration of 5 nM. Values (100%) of [:%H]ryanodine binding correspond to 9, 7.5, and 5.5 pmol bound [‘3H]ryanodine/mg of frog, skeletal, and canine SR membrane protein, respectively. Data points are average of 2 different experiments cArried out in triplicate (SD, 210% or less).

ISOFORMS

C367

upper polypeptide. An alternative explanation was that frog skeletal muscle contains two distinct ryanodine receptor polypeptides, as initially observed for chicken pectoral muscle (1). To distinguish between these two possibilities, rabbit and frog skeletal and canine cardiac muscle SR proteins were separated on 3-12s SDS polyacrylamide gradient gels, electrophoretically transferred onto Immobilon membranes, and probed with an antiserum against the purified rat skeletal muscle ryanodine receptor (Fig. 3L3)and two monoclonal antibodies against the canine cardiac muscle ryanodine receptor (Fig. 3, C and 0). The immunoblot analysis showed that polyclonal antiserum to the skeletal ryanodine receptor preferentially cross-reacted with the upper band of the frog receptor doublet. Conversely, two monoclonal antibodies to canine cardiac ryanodine receptor (C3-24 and C3-33) preferentially cross-reacted with the lower band of the frog receptor. These observations disfavored proteolysis as an explanation for the presence of the lower frog band and suggested a possible coexpression in amphibian muscle of two isoforms having some properties in common with the mammalian skeletal and cardiac muscle isoforms. The CHAPS-solubilized ryanodine receptor-Ca2+ release channels of mammalian skeletal and cardiac muscle SR have been purified on linear sucrose gradients as homotetrameric 30s protein complexes (2, 11, 14). With the use of an identical procedure for purifying the frog skeletal muscle, ryanodine receptor also resulted in identification of a 30s ryanodine-binding protein complex (Fig. 4A), which comigrated with two distinct high molecular weight polypeptides (Fig. 4B). In Coomassiestained gels, the two protein bands in the 30s fraction were present with about equal intensity in gradient fraction 15 (Fig. 4B) and the two neighboring fractions 14 and 16 (not shown). Immunoblot analysis of sucrose gradient fractions 15 (Fig. 4, C and D) and 14 and 16 (not shown) with a skeletal muscle ryanodine receptor antiserum and a third cardiac muscle ryanodine receptor monoclonal antibody (C2-997) showed a specific comigration of the immunoreactive high A& bands and the 30s ryanodine receptor peak fractions. Further, as observed in immunoblots of frog SR membranes (Fig. 3), the skeletal antiserum and cardiac monoclonal antibody C2-997 preferentially recognized the upper and lower bands of the doublet, respectively. Immunoprecipitation studies with the 30s ryanodinebinding peak fractions of sucrose gradients showed that, at a low antibody-to-ryanodine receptor protein ratio, cardiac monoclonal antibody C3-33 selectively bound the lower band of the frog ryanodine receptor doublet in addition to the canine cardiac receptor protein (Fig. 5, A-C). Under these conditions, the antibody also immunoprecipitated, to a lesser extent, the high molecular weight polypeptide of the rabbit skeletal 3OS receptor complex. However, at a higher antibody-to-receptor ratio, we observed a partial immunoprecipitation of the upper band of the frog doublet together with a more complete precipitation of the lower band of the frog doublet as well as the rabbit skeletal polypeptide (Fig. 5D). These results indicated that cardiac monoclonal antibody C3-33 recognized

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 11, 2019.

C368

FROG

RYANODINE

the two frog M, -560,000 polypeptides with differing affinity and suggested that two distinct 30s ryanodine receptor complexes comprised of either the upper or lower band are present in frog muscle. The two ryanodine receptor complexes were localized in frog muscle by double-label immunofluorescence. Cryostat sections of unfixed frog muscle were labeled with the polyclonal antibody against rat skeletal muscle ryanodine receptor and monoclonal antibody C3-33 followed by fluorescein- and Texas Red-conjugated secondary antibodies, respectively. The monoclonal antibody was used over a concentration range giving from barely detectable to quite strong staining; preferential labeling of the complex corresponding to the lower band of the doublet would be expected at low concentrations (see Fig. 5C). At all concentrations, the two labels were entirely coextensive (Fig. 5, E and F). No sites were labeled by one antibody but not the other, and all labeling was consistent with the simultaneous presence of both complexes in triads throughout the muscle. The presence of a Ca2+ release channel activity intrinsic to the frog skeletal ryanodine receptor was tested by incorporating purified, CHAPS-solubilized receptor fractions into Mueller-Rudin planar lipid bilayers. It should

A

B

RECEPTOR

ISOFORMS

be noted that the two frog ryanodine receptor complexes could not be sufficiently separated using immunological approaches to allow their specific properties to be determined in these reconstitution experiments. A symmetrical NaCl medium was used because the reconstituted rabbit skeletal SR Ca2+ release channel has been shown to be impermeant to anions such as Cl- but to conduct monovalent cations more efficiently than it does Ca2+ ions (11, 29). In Fig. 6, a single purified channel was recorded in the presence of a symmetric 150 mM NaCl, 20 mM NaPIPES, pH 7 medium. The channel was partially activated in the presence of 2 PM free Ca2+ and 2.4 mM ATP in the &-chamber at holding potentials ranging from +30 to -30 mV. Single-channel conductance was 490 pS with -160 mM Na+ as the current carrier. Perfusion of the trans-chamber with 50 mM CaCl, resulted in a Ca2+ conductance of 110 pS and a reversal potential of +18 mV. In Fig. 7A, a single purified frog channel was recorded in symmetric 500 mM NaCl buffer. With 500 mM Nat as the current carrier, single-channel conductance was 600 f 15 pS (n = 6). In Fig. 7, top trace, the channel was activated by 50 PM free Ca2+ in the c&-chamber. Channel activity was greatly decreased by lowering the free Ca”+ concentration to