Structural and Functional Correlation of the Trypsin-digested Ca2+ Release Channel of Skeletal Muscle Sarcoplasmic Reticulum*

The effect of trypsin digestion on the (i) fragmenta- tion pattern, (ii) activity, (iii) [‘Hlryanodine binding, and (iv) sedimentation behavior of the skeletal sarco- plasmic reticulum (SR) ryanodine receptor-Ca2+ release channel complex has been examined. Mild tryptic digestion of heavy, junctional-derived SR vesicles resulted in the rapid disappearance of the high molecular weight (Mr -400,000) Ca2+ release channel protein on sodium dodecyl sulfate gels and appearance of bands of lower M, upon immunoblot analysis, without an appreciable effect on [‘Hlryanodine binding or the ap- parent S value (30 S) of the 3-[3-cholamidopro-pyl)dimethylammonio]-1-propanesulfonate (Chaps)- solubilized channel complex. Further degradation to bands of M, > 70,000 on immunoblots correlated with a reduction of channel size from 30 S to 10-15 S and loss of high affinity [‘Hlryanodine binding to the tryp- sinized receptor, while low affinity [3H]ryanodine binding and [“Hlryanodine bound prior to digestion were retained. Parallel indicated retention of the Ca2+, Mg2+, and ATP regulatory sites, although Ca2+-induced 46Ca2+ release rates were changed. In planar lipid bilayer-single channel measurements, addition of trypsin to the cytoplasmic side of the high conductance


Structural and Functional Correlation of the Trypsin-digested Ca2+
Release Channel of Skeletal Muscle Sarcoplasmic Reticulum* (Received for publication, August 22, 1988) Gerhard MeissnerS, Eric Rousseau, and F. Anthony Lai From the Departments of Biochemistry and Physiology, University of North Carolina, Chapd Hill, North Carolina 27599-7260 The effect of trypsin digestion on the (i) fragmentation pattern, (ii) activity, (iii) ['Hlryanodine binding, and (iv) sedimentation behavior of the skeletal sarcoplasmic reticulum (SR) ryanodine receptor-Ca2+ release channel complex has been examined. Mild tryptic digestion of heavy, junctional-derived SR vesicles resulted in the rapid disappearance of the high molecular weight (Mr -400,000) Ca2+ release channel protein on sodium dodecyl sulfate gels and appearance of bands of lower M, upon immunoblot analysis, without an appreciable effect on ['Hlryanodine binding or the apparent S value (30 S) of the 3-[3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (Chaps)solubilized channel complex. Further degradation to bands of M, > 70,000 on immunoblots correlated with a reduction of channel size from 30 S to 10-15 S and loss of high affinity ['Hlryanodine binding to the trypsinized receptor, while low affinity [3H]ryanodine binding and ["Hlryanodine bound prior to digestion were retained. Parallel 4SCa2+ efflux measurements also indicated retention of the Ca2+, Mg2+, and ATP regulatory sites, although Ca2+-induced 46Ca2+ release rates were changed. In planar lipid bilayer-single channel measurements, addition of trypsin to the cytoplasmic side of the high conductance (100 pS in 50 mM Ca2+), Ca2+-activated SR Ca2+ channel initially increased the fraction of channel open time and was followed by a complete and irreversible loss of channel activity. Trypsin did not change the unitary conductance, and was without effect on single channel activity when added to the lumenal side of the channel.
Rapid release of Ca2+ from skeletal muscle sarcoplasmic reticulum (SR)' is triggered by a surface membrane action potential that is thought to be communicated to SR via the transverse (T-) tubule system at specialized areas where the *The research was supported by United States Public Health Service Grant AR18687 and fellowships from the Canadian Heart Foundation (to E. R.) and the Muscular Dystrophy Association (to F. A. L.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ To whom correspondence should be addressed Dept. of Biochemistry CB# 7260, The University of North Carolina, Chapel Hill, NC 27599.
The abbreviations used are: SR, sarcoplasmic reticulum; T, transverse; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; Pipes, 1,4-piperazinediethanesulfonic acid Hepes, N-2-hydroxyethylenepiperazine-N'-2-ethanesulfonic acid; DIFP, diisopropyl fluorophosphate; Chaps, 3-[3-cholamidopropyl)dimethylammonio]-l-propanesulfonate; AMP-PCP, adenosine 5'-(@,y-methylene) triphosphate; pS, picosiemens; SDS-PAGE, sodium dodecyl sulfatepolyacrylamide gel electrophoresis; PBS, phosphate-buffered saline. two membrane systems come in close contact (for reviews see . Ultrastructural studies have revealed distinct protein bridging structures ("feet") (5) which span the gap between T-tubule and SR and have been immunologically related to high molecular mass proteins of 3 300,000 daltons (6). Although the mechanism of signal transmission across the junctional gap remains to be determined, recent single channel recording measurements using isolated membrane fractions have suggested that SR Ca2+ release is mediated by a high conductance, Ca2+-and ATP-activated "Ca2+ release" channel (7,8). Using the Ca2+ release channel-specific probe [3H]ryanodine, a 30 S ryanodine receptor complex has been isolated (9) which is composed of polypeptides of apparent molecular mass -400,000 (9-11). Upon reconstitution of the 30 S complex into planar lipid bilayers, Ca2+ conducting channels were evident with a conductance and pharmacological behavior similar to those observed when SR Ca2+ release vesicles were fused with the bilayers (9, 12, 13). Negativestain electron microscopy further revealed the four-leaf clover structure described for the feet that span the T-tubule-SR junction (9, 14). These findings have suggested that the feet are synonymous with the high conductance, ligand-gated Ca2+ release channel of sarcoplasmic reticulum. Furthermore, they support the hypothesis that the channel directly senses the T-tubule potential with one end, while the other end regulates a Ca2+ conducting pore in SR (15).
The high molecular weight Ca2+ release channel proteins are highly susceptible to proteolysis (9,11,16). Mild digestion of "heavy" SR vesicles with trypsin resulted in partial loss of ATP-dependent Ca2+ accumulation (17) and increased ruthenium red-sensitive Ca2+ efflux from the vesicles (18,19). Ca2+ efflux was potentiated by Ca2+ and ATP and inhibited by M$+ as observed similarly in undigested vesicles. Furthermore, [3H]ryan~dine binding to trypsin-digested vesicles was reduced (18). These observations have suggested that limited tryptic digestion of heavy SR vesicles activates the SR Ca2+ release channel.
In this paper, the effects of controlled trypsin proteolysis on the SR Ca2+ release channel have been studied using a heavy SR vesicle fraction isolated from rabbit skeletal muscle. We have determined the size and proteolytic fragmentation pattern of the digested channel complex by sedimentation analysis and immunoblot staining using a rabbit anti-rat polyclonal antibody, respectively. In parallel experiments, the Ca2+ release behavior of, and [3H]ryanodine binding to, the digested vesicles was determined. In addition, we report the first direct observations of the effects of trypsin on single channels incorporated into planar lipid bilayers. Some of these results have been communicated in abstract form (20).

EXPERIMENTAL PROCEDURES
Materials-Trypsin (Type 111) from bovine pancreas, soybean trypsin inhibitor protein, diisopropyl fluorophosphate (DIFP), and the 1715 This is an Open Access article under the CC BY license.
ATP analog AMP-PCP were obtained from Sigma, ruthenium red from Fluka, 46Ca2+ from ICN Pharmaceuticals, [3H]ryanodine (54.7 Ci/mmol) from Du Pont-New England Nuclear, and unlabeled ryanodine from AgriSystems International (Wind Gap, PA). Phospholipids were purchased from Avanti Polar Lipids. All other reagents were of reagent grade.
Isolation of SR Vesicles-Heavy SR vesicles enriched in Ca2+ release activity were isolated from rabbit skeletal muscle as a 2,600-35,000 X g pellet (21) in the presence of 2.5 mM EGTA and 1 mM DIFP. The pellets were resuspended in 0.3 M sucrose, 0.6 M KCl, 0.1 mM EGTA, 0.1 mM DIFP, and 10 mM KPipes, pH 7.0, incubated for 1 h at 4 "C and sedimented by centrifugation at 100,000 X g for 30 min. After washing and resuspension in 0.3 M sucrose, 0.1 M NaC1, 100 g M EGTA, 100 g~ Ca", and 20 mM NaPipes, pH 7.0, membranes were rapidly frozen and stored at -65 "C before use. "Light" SR vesicles lacking the CaZ' release channel were prepared as described previously (21).
Trypsin Digestion of SR Vesicles-For all experiments, unless otherwise indicated, SR vesicles (2-10 mg of protein/ml) were incubated with trypsin at 22 "C in 0.1 M NaCl, 0.1 mM EGTA, 1.1 mM Ca", and 20 mM NaPipes, pH 7.0. The reaction was quenched with a 20-fold weight excess of soybean trypsin inhibitor protein and 1 mM DIFP (dissolved at a concentration of 100 mM in dimethyl sulfoxide). In control experiments trypsin was omitted, or the two protease inhibitors were added to the vesicles before the addition of trypsin.
Preparation of Ca2+ Release Channel Protein Antiserum-A rabbit anti-rat Ca2+ release channel protein antiserum was prepared by intradermal injection of an emulsion of 1 ml of Freund's adjuvant (complete) and 1 ml of a saline solution containing a small amount of Chaps and 250 gg of purified Ca2+ release channel-ryanodine receptor protein (9) prepared from rat skeletal muscle. The [3H] ryanodine-labeled receptor sedimented as a 30 S complex and comprised one major high molecular weight band of M, -400,000, as has been previously reported for the rabbit muscle ryanodine receptor (9). After two booster injections at 4-week intervals, rabbit serum was collected. Immunoblot analysis using the complete serum indicated strong immunoreactivity with the rat and rabbit heavy SR M, -400,000 protein bands. A faint band of M, -100,000 was also recognized in the immunoblot. Immunostaining of this latter band was removed by prior incubation of the complete serum for 3 h at 22 "C with SR protein coupled to cyanogen bromide-activated Sepharose 4B (1 ml of serum/ml gel containing 1 mg of SR protein). The SR protein fraction used for coupling was obtained by centrifuging Chaps-solubilized heavy SR vesicles from rabbit muscle through a linear sucrose gradient and collecting the top region of the gradient which was devoid of the 30 S channel complex (9). SDS Gel Electrophoresis and Immunoblot Staining-SDS-PAGE was performed in the Laemmli buffer system (22) using a 1.5 mm thick, 5-12% linear polyacrylamide gradient gel and 3% stacking gel. Samples were denatured for 3 min at 95 "C in 0.1 M Tris-HC1, pH 6.8, containing 2% SDS, 3% 8-mercaptoethanol, and 10% glycerol. Electrophoresis was at 15 "C and constant current (30 mA/gel) and was monitored with 0.004% bromphenol blue as tracking dye. Gels were stained with 0.1% Coomassie Brilliant Blue R-250 in 50% methanol, 10% acetic acid and destained with 10% methanol, 15% acetic acid.
For immunoblots, the separated proteins from SDS-PAGE were electrophoretically transferred onto Immobilon PVDF membranes (Millipore Corp.) for 1 h at 400 mA then 12 h at 1500 mA and 15 'C (23). Transfer membranes were washed in PBS (15 rnin), blocked with PBS/5% non-fat dried milk for 1 h at 37 "C (24), and twice washed for 20 min in PBS/O.5% Tween 20 (PBS-T), before overnight incubation at 4 "C with rabbit and anti-rat ryanodine receptor antiserum (1:10,000 dilution in blocking buffer) preabsorbed with rabbit SR protein devoid of the ryanodine receptor as described above. After 4 washes in PBS-T as above, membranes were incubated for 1 h at 23 "C with peroxidase-conjugated goat anti-rabbit IgG antiserum (Calbiochem), washed 3 times in PBS-T as above, and then the color developed using 3,3'-diaminobenzidine as substrate.
45Ca2+ Flux Measurement~--"~Ca~+ efflux rates from vesicles passively loaded with '%a2+ were determined with the use of an Update System 1000 Chemical Quench apparatus (Madison, WI) and by filtration (25). Vesicles (2-10 mg of protein/ml) were passively loaded by incubation for 2 h at 23 "C with '%a2+ in a medium containing 100 mM NaC1, 20 mM NaPipes, pH 7.0, 100 g~ EGTA, and 1.1 mM 300-fold dilution into isosmolal unlabeled release media containing 46Ca2+ . The '%a2' efflux behavior of the vesicles was measured by 5-varying concentrations of Ca2+, M%f, and adenine nucleotide. In the rapid quench experiments, "CaZ+ efflux was inhibited at time intervals ranging from 25 to 1000 ms by the addition of 10 mM Mg+, 5 mM EGTA, and 10 p M ruthenium red. Untrapped as well as released "Ca2+ was separated away by filtration using Millipore Type HA filters. Radioactivity retained by the vesicles on the filters was determined by liquid scintillation counting. '%a2+ efflux measurements were carried out at least in triplicate with three or more time points. The standard errors were +20% or less.
Planar Bilayer Measurements-Single channel recordings were performed by fusion of heavy SR vesicles into Mueller-Rudin planar bilayers containing phosphatidylethanolamine, phosphatidylserine, and phosphatidylcholine in the ratio 5:3:2 (50 mg/ml phospholipid in n-decane). Bilayer currents were measured and analyzed as described previously (8).
PHIRyanodine Binding-SR vesicles (0.5-3 mg of protein/ml) were incubated at 22 "C in a medium containing 20 mM NaPipes, pH 7.0, 1 M NaC1, 1 mM DIFP, 0.1 mM EGTA, 0.15 mM Ca2+, 2.5 mM AMP-PCP, and 2 nM to 1 mM [3H]ryanodine. [3H]Ryanodine (54.7 Ci/mmol) was added to a concentration up to 25 nM; greater concentrations were prepared as admixtures of labeled and unlabeled ryanodine. After 15 h, aliquots of the vesicle suspensions were (i) placed into a scintillation vial to determine total radioactivity, (ii) centrifuged for 30 min at 90,000 X g in a Beckman Airfuge to determine free [3H]ryanodine, and (iii) placed, after dilution with 25 volumes of ice-cold water, on a Whatman GF/B filter soaked in 1% polyethylenimine. After rinsing with three 5-ml volumes of ice-cold HzO, radioactivity remaining with the filters was determined by liquid scintillation counting to obtain bound [3H]ryanodine.
[3H]Ryanodine binding measurements were carried out in triplicate. The standard errors were 210% or less.
In the trypsin digestion studies with prelabeled membranes, vesicles were incubated in the above medium with 2 nM [3H]ryanodine for 2 h at 22 "C and then centrifuged for 45 min at 4 'C and 100,000 X g in a Beckman ultracentrifuge through a layer of 0.5 M sucrose to remove unbound [3H]ryanodine and remaining unhydrolyzed DIFP.
Sucrose Gradient Analy~is-[~H]Ryanodine-labeled membranes were solubilized at 1.5 mg of protein/ml in buffer A (1 M NaCl, 0.1 mM EGTA, 0.15 mM Ca2+, 3 mM AMP, 1 mM DIFP, 0.5% soybean phosphatidylcholine, 20 mM NaPipes, pH 7.0) containing 1.5% Chaps (9). After 2 h incubation at 22 "C and centrifugation at 4 "C for 30 min at 100,000 X g in a Beckman ultracentrifuge, the supernatant (1 ml) was loaded onto 5-20% linear sucrose gradients in buffer A containing 1% Chaps and then centrifuged at 26,000 rpm at 2 "C for 16 h in a Beckman SW-41 rotor. Gradient fractions were collected from the top and analyzed for radioactivity by liquid scintillation counting. Bound [3H]ryanodine was determined by filtration (26) after dilution of samples with 25 volumes of ice-cold water. Apparent sedimentation coefficients (uncorrected for bound detergent and lipid) of the [3H]ryanodine receptor were determined by extrapolation of an enzyme calibration curve obtained with Escherichin coli 8galactosidase (16 S), bovine catalase (11.2 S), and yeast alcohol dehydrogenase (7.6 S).
Biochemical Assays-Protein was determined by the Lowry method using bovine serum albumin as a standard (27). Free Ca2+ concentrations were calculated by a computer program using binding constants published by Fabiato (28).

SDS-Polyacrylamide Gel Electrophoresis and Immunoblot
Analysis-SDS-polyacrylamide gradient gel electrophoresis of rabbit muscle heavy SR membranes indicated the presence of several major protein bands including two bands of apparent M, (relative molecular mass) of 110,000 (110 kDa) and -400,000 (-400 kDa) (Fig. lA, lane 1). The band of 110 kDa is the Ca2+ pump or (Caz+ + M%+)-ATPase of sarcoplasmic reticulum. The -400-kDa polypeptide forms, in Chaps solution, a tetrameric 30 S complex which is identical with the SR Ca" release channel and the feet structures that span the T-tubule-SR junction (9). Proteolytic digestion of heavy SR vesicles at the high protein:trypsin ratio of 10,OOO:l for 1 min resulted in the disappearance of the -400-kDa band (Fig. M,   lune 3 ) . Under our experimental conditions, the 110-kDa Ca2+ pump protein was less sensitive to proteolytic digestion. Lane 3, plus trypsin (10,0001) with addition of soybean trypsin inhibitor protein and DIFP after 1 min at 22 "C. h f t , upper arrow denotes migration distance of the -400-kDa protein and the lower arrow that of soybean trypsin inhibitor protein (Mr 20,100). Right, the molecular mass standards shown are myosin (205 kDa), @-galactosidase (116 kDa), phosphorylase (97.4 kDa), bovine serum albumin (66 kDa), ovalbumin (45 kDa) and carbonic anhydrase (29 kDa). B, immunoblot of samples prepared and run in an identical way to that described above (50 pg of protein/ lane). Efficiency of electrophoretic transfer of the M, -400,000 protein was greater than 70% and that of the fragments greater than 80%, as determined by comparison of Coomassie-stained gels of control (not blotted) and blotted gels.
Treatment at a SR protein:trypsin ratio of 10,OOO:l for 1 min caused cleavage of a portion of the Ca2+ pump protein (Fig.  lA, lane 3 ) , as indicated by the appearance of a major polypeptide band of M, -55,000 (29, 30); however, a longer incubation time of 5 min at a 10 times higher trypsin concentration (SR protein:trypsin ratio of 1OOO:l) was required to effect disappearance of the 110-kDa band.
The proteolytic fragments of the Ca2+ release channel which retained immunoreactivity following degradation were identified by immunoblot staining (Fig. lB), using a rabbit antiserum raised against the purified rat skeletal muscle -400-kDa Ca2+ release channel protein. Comparison of the minus trypsin control lanes (lane 1 of Fig. 1, A and B ) shows that the antibody specifically recognized the -400-kDa polypeptide among the major protein bands. The antibody also visualized an additional minor band which migrated just below the -400-kDa protein and whose intensity varied with different membrane preparations. Most likely its appearance is due to partial degradation of the -400-kDa polypeptide (9). The plus trypsin control lane (lane 2 of Fig. 1, A and B ) indicates minimal degradation of the -400-kDa band when soybean trypsin inhibitor protein and DIFP were added to the vesicles before the addition of trypsin. This suggested rapid and effective inactivation of trypsin in the vesicle suspension under our experimental conditions.
Immunoblots of membranes digested with trypsin are shown in the last five lanes of Fig. 1B. Mild proteolysis (at a SR protein:trypsin ratio of 10,OOO:l for 1 min) resulted in the loss of the -400-kDa protein and a concomitant appearance of six major immunoreactive polypeptides of M , -250,000, -230,000, -220,000, 205,000, 160,000, and 65,000 and four minor, with respect to quantity and/or immunoreactivity, bands of M, 190,000, 125,000, 55,000, and 50,000 (Fig. lB,  lane 3 ) . More severe proteolytic conditions (SR protein:trypsin ratio of 1OOO:l) caused the immediate appearance of three major polypeptides of M, 160,000,65,000, and 27,000, and minor bands of M , -220,000, 190,000, 125,000, 55,000, and 50,000 after 1 min of trypsin exposure (Fig. lB, lune 4 ) . Faint bands of M, 110,000 and 90,000 also present after 1 min were more pronounced after a 5-min exposure (Fig. lB,lune 5 ) as the higher molecular weight bands were substantially decreased. No major immunoreactive bands were observed after 60 min of trypsin exposure (Fig. lB, lane 7).
45Cu2+ Release- Fig. 2 compares the 45Ca2+ release behavior of undigested and trypsin-digested heavy SR vesicles. Vesicles were passively loaded with 1 mM 45Ca2+ and then diluted into a medium which either inhibited or activated the SR Ca2+ release channel. 45Ca2+ release was slow in the Ca2+ release channel inhibiting medium which contained 10 mM M g + and 10 PM ruthenium red. This allowed determination of the amounts of 45Ca2' (23 nmol/ml protein) trapped by the vesicles. About 75% of the trapped 45Ca2+ was released in less than 30 s on dilution of the vesicles into the Ca2+ release channel activating medium which contained 10 PM free Ca2+. Some of the 45Ca2+ (-5 nmol/mg protein) remained with the vesicles for longer times because not all of them contained the Ca2+ release channel (21). channel inhibiting medium. At longer time intervals, 60 min, a partial breakdown of the 45Ca2' permeability barrier was observed for vesicles diluted into the medium containing 10 mM M e and 10 pM ruthenium red. Inability of the vesicles to trap 45Ca2' did not appear to be due to the presence of a Caz+ release channel which could not be closed by Mg2+ and ruthenium red, based on the following two observations. First, prolonged trypsin digestion similarly decreased the amounts of 45Ca2+ trapped by light SR vesicles lacking the channel (not shown). Second, control experiments with the channel-impermeable molecule ['4C]sucrose showed that trypsin digestion decreased the amounts of [ 1 4 C ]~~~r~~e trapped by the vesicles to an extent similarly as observed for 45Ca2+ in the channel inhibiting medium in Fig. 2.
There occurred several noticeable changes in the initial rate of Ca2+-induced Ca2+ release after trypsin digestion of the vesicles. Fig. 2 shows that trypsin digestion for 20 min caused a decrease in the rate of 45Ca2' efflux without appreciably affecting the total amount of 45Ca2+ that was released by the vesicles in 150 s after dilution into the 10 pM Ca2' release medium. The effects of more limited trypsin digestion on the initial rates of 45Ca2+ efflux were determined with the use of a rapid mixing and quench apparatus (Fig. 3). Treatment for 1 min at a SR protein:trypsin ratio of 1OOO:l resulted in a small decrease of the rate of Ca2'-induced 45Ca2+ release. More extensive treatment for 1 and 15 min at a SR protein:trypsin ratio of 1OO:l had an opposite effect in that the vesicles lost their 45Ca2+ stores with a faster rate (1.5-and 3.5-fold) than the undigested Vesicles. Taken together, the data of Figs. 2 and 3 suggest that mild trypsin digestion is without a marked effect on Ca2+-induced Ca2+ release channel activity. At an intermediate stage of trypsin digestion the channel is activated, whereas extensive proteolysis appeared to result in channel inactivation. This latter stage was difficult to quantitate due to a general breakdown of the permeability barrier of the vesicles. The Ca2+ release channel of skeletal SR contains regulatory binding sites for Ca2+, M%+, and adenine nucleotides (25,31). The effects of trypsin digestion on these sites was tested by determining the Ca" release behavior of the vesicles in nanomolar and micromolar Ca2+ media without or with added M e and adenine nucleotide (Fig. 4). Digestion of the vesicles for 5 and 30 min at a SR protein:trypsin ratio of 1OO:l

Effect of trypsin digestion on Effect of trypsin digestion on
increased and subsequently decreased, respectively, the rate of 45Ca2+ efflux in the nanomolar Ca2+ medium. In the three vesicle preparations, 45Ca2+ release was accelerated by the addition of 5 mM AMP to the nanomolar Caz+ medium or by increasing the free Ca2+ concentration from <lo-' M to 6 X M. In the undigested vesicles and after digestion of 5 min, a majority of the vesicles released their 45Ca2+ stores in less than 30 s. Digestion for 30 min resulted in a small decrease of the amounts of 45Ca2+ released after 30 s in the micromolar Ca2+ and nucleotide containing media. Addition of 1 mM M e to the micromolar Ca" release medium reduced the rate of suggest that trypsin digestion modifies SR 45Ca2+ release activity without causing a loss of regulation of the channel by Ca", M F , and adenine nucleotide.
Single Channel Measurements-Another more direct way of measuring Ca2+ release channel activity involves the incorporation of single channels into planar lipid bilayers (7, 8). Due to the low frequency of incorporation of only one or two Ca2+ release channels into the bilayer, the effects of trypsin were observed in only five separate recordings, two of which are shown in Fig. 5. The current fluctuations of a single skeletal Ca" release channel which was recorded with 50 mM Ca2+ trans as the current carrier are shown in Fig. 5A. In the upper truce, the channel was partially activated by 2.5 pM free Ca2+ in the cis chamber (the SR cytoplasmic side of the channel). In the second trace of Fig. 5A, an increase in channel activity was observed after trypsin (10 pg/ml) was added to the cis chamber. Channel open time (Po) increased from 0.07 (A, upper trace) to 0.60 (A, second truce) 6 min after the addition of trypsin without a change of unit conductance (100 pS in 50 mM Ca2+). The last trace of Fig. 5A shows that after 9 min of trypsin exposure channel activity ceased. This abrupt inactivation was preceded by a short period of reduced channel activity. Channel inactivation was considered to be irreversible since an increase in cis Ca2+ or the addition of ATP were ineffective in reactivating the channel (not shown). This behavior contrasted with that observed during the trypsininduced activation phase where the channel remained sensitive to further activation by ATP and inhibition by M g + (not shown). In two recordings, trypsin did not appreciably activate the SR Ca2+ release channel when added to the trans chamber (the SR lumenal side of the channel) for 30 min. Fig. 5B shows the effects of trypsin on another single Ca2+ release channel which was activated (Po = 0.69) by 2.5 p~ Ca2+ and 1 mM ATP cis. In this situation, trypsin did not significantly alter channel activity until about 10 min after the addition of trypsin, when the channel abruptly, and irreversibly, ceased to conduct Ca2+.
f'H1Ryanodine Binding-Ryanodine is a neutral plant alkaloid which binds with high affinity and specificity to the SR Ca2+ release channel (9-11). The drug has a dual effect in that nanomolar concentrations open the channel, whereas at concentrations above 10 pM, ryanodine completely closes the skeletal channel (32, 33).
[3H]Ryan~dine binding was measured in the presence of micromolar free Ca2+ and 2.5 mM AMP-PCP, a nonhydrolyzable ATP analog (Fig. 6). The two ligands nearly optimally open the skeletal SR Ca2+ release channel (8) and thereby favor use-dependent activation and inactivation of the channel by ryanodine (32). Figure 6A shows mM HEPES, pH 7.4, in the trans chamber before (upper truce), and 6 and 9 min (middle and lower traces), after the addition of 10 pg of trypsin/ml to the cis chamber. Recordings were filtered at 300 Hz and sampled at 1 kHz. Holding potential was 0 mV. B, a separate channel was recorded as in A except that the cis chamber contained in addition 1 mM ATP. Lower trace was recorded 10.5 min after the addition of trypsin to the cis chamber. It should be noted that the single channel experiments were carried out at a nominally extremely low SR protein:trypsin ratio due to perfusion of the cis (and trans) chambers after fusion of a single Ca2+ release channel (8).
vesicles (inset, Fig. 6A). In trypsin-treated vesicles (at a SR protein:trypsin ratio of 400:l) the number of high affinity binding sites was appreciably reduced after a digestion time of 3 min. No high affinity binding was detected in vesicles treated with trypsin for 30 min. The data of Fig. 6A also indicate the presence of additional low affinity [3H]ryanodine binding sites which appeared to be less affected by trypsin digestion. In undigested vesicles, the total number of high and low affinity binding sites was 50 k 15 pmol/mg protein, suggesting that for each high affinity binding site there are approximately three low affinity binding sites. Rearrangement of the binding data in the form of a Hill plot yielded a non-linear curve which at free ryanodine concentrations in the range of 0.3-3 and 50-1000 nM could be approx- [3H]Ryanodine binding to both control (minus trypsin) and trypsindigested vesicles was determined as described under "Experimental Procedures." In parallel experiments, it was established that in each condition, equilibrium of [3H]ryanodine binding was reached. Specific binding of [3H]ryanodine is shown in A and in the form of a Scatchard plot (inset of A ) and Hill plots ( B ) , and was obtained by assuming that nonspecific binding is linear with ryanodine concentration. Nonspecific binding was defined as the difference between total and specific binding and amounted to 8 nmol/mg protein at 1 mM [3H] ryanodine. Hill plots were obtained assuming that specific binding of [3H]ryanodine amounted to 54 pmol/mg protein (one high and three low affinity binding sites in control vesicles). Data shown are from one representative set of results chosen from three separate experiments performed with vesicles digested to slightly different extents by using varying SR protein:trypsin ratios and times. imated by two straight lines (Fig. 6 B ) . From the slopes, n values of 0.95 and 0.35 were obtained, respectively. Hill plots of binding data of vesicles treated with trypsin at a protein ratio of 400:l for 3 min showed a less non-linear behavior, whereas binding to more extensively digested vesicles (for 30 min) could be reasonably well fitted by a straight line with an n value of 0.75. These results suggested that in trypsindigested vesicles there occurred a significant change in the affinity and cooperativity of [3H]ryanodine binding.

Effects of
Sedimentation Coefficients-The apparent sedimentation coefficient of the Chaps-solubilized [3H]ryanodine receptor-Ca2+ release channel complex was determined by sucrose density gradient centrifugation. In the control condition in the absence of trypsin digestion, a single peak of bound radioactivity was observed in the lower half of the gradient (Fig. 7). Bound radioactivity comigrated with a minor protein peak with an apparent sedimentation coefficient of 30 S and composed of polypeptides of M, -400,000 (9). Trypsin treatment of membranes for 5 min at a protein:trypsin ratio of 1OOO:l followed by Chaps solubilization resulted only in a small shift of the S value (Fig. 7). However, a more dramatic decrease of the S value from 30 S to -10 S was observed after digestion for 120 min. This reduction in sedimentation rate suggested that there had occurred a significant change in size and/or protein conformation of the trypsin-digested, Chapssolubilized receptor complex.  (Fig. 8). In control experiments, an essentially identical fragmentation pattern of the -400-kDa band was observed on immunoblots when vesicles were not labeled (Fig. 1B) or were prelabeled with a saturating concentration of ryanodine (20 PM) prior to trypsin digestion, suggesting a similar pattern of degradation for the unlabeled and labeled receptor. Tryptic digestion at a SR protein:trypsin ratio of 500:l for 1 min resulted in a 1.5-fold decrease of the 45Ca2+ efflux rate, measured by diluting the vesicles into a nanomolar free Ca2+ release medium, without an appreciable effect on [3H]ryanodine binding or the sedimentation behavior of the channel complex. Continued exposure to trypsin resulted in an increase of the 45Ca2+ release rate, a decrease of the S value of the solubilized complex from 30 S to about 15 S and a loss of high affinity [3H]ryanodine binding when assayed after trypsinization.  Vesicles were also first labeled with 2 nM [3H]ryanodine before being treated with trypsin. In this case, membranes digested at a protein:trypsin ratio of 500:l for 30 min retained greater than 80% of the bound [ 3 H ]~a n~d i n e (Fig. 8). By comparison, high affinity [3H]ryanodine binding was reduced to less than 1% when vesicles were labeled after digestion at the same SR protein:trypsin ratio for 30 min. Fig. 8 further shows that digestion at a lower SR protein:trypsin ratio, 100:1, for 30 min decreased the S value of the solubilized complex to -10 S and the 45Ca2+ efflux rate to a value close to that of undigested vesicles. The effects of more extensive digestion were not compared because of a loss of the general permeability barrier of the vesicles (see Fig. 2) and a large decrease in the amount of bound [3H]ryanodine on solubilization with Chaps (not shown).

DISCUSSION
Heavy junctional-derived SR vesicles contain a ligandgated Ca2+ release channel which is composed of polypeptides of M , -400,000. Earlier studies showing the rapid degradation of high molecular mass SR proteins by Ca2+-activated protease (16) have been corroborated by this report and other recent observations that the -400-kDa channel protein is highly susceptible to proteolysis (9,18,20). The appearance of smaller peptides of -300 and 160 kDa upon isolation of the ryanodine receptor, in addition to the -400-kDa protein, initially suggested them to be distinct subunits of a heteromeric complex (34, 35). However, more recent evidence has revealed that, when purified in the presence of protease in-hibitors, the -400-kDa protein is the major constituent of the ryanodine receptor complex (9-11), indicating that the -300and 160-kDa proteins were derived from proteolytic fragmentation caused by endogenous proteases (9). In this regard, other recent studies suggesting the involvement/association of proteins smaller than the -400-kDa channel protein with SR Ca2+ release (36-38), may conceivably be related to its acute lability. These developments in defining the subunit composition of the SR Ca2+ release channel are the converse of that which occurred for the muscle nicotinic acetylcholine receptor channel complex, where early purification studies identified a single ( a ) subunit (-40 kDa), which turned out to have derived from proteolysis of a heteropentameric complex ( a2. p . y . 6 ) comprising subunits of >40 kDa (39,40).
Mild trypsin digestion of SR membranes resulted in the transient appearance of protein bands of 115-170 kDa which underwent further rapid proteolysis (18). In the present study, the use of a specific anti-30 S complex antiserum has allowed a more definitive determination of the tryptic fragmentation pattern of the Ca2+ release channel proteins and has enabled its correlation with the [3H]ryanodine binding, sedimentation behavior, and functional characteristics of the Ca2+ release channel. Our studies indicate that short term trypsin digestion of ryanodine-labeled and unlabeled vesicles at a high SR protein:trypsin ratio results in the complete disappearance of the -400-kDa channel band on SDS gels and the corresponding appearance of a series of lower molecular mass bands between -250 and 27 kDa upon immunoblot analysis (Fig.  lB, lunes 3 and 4 ) without an appreciable effect on [3H] ryanodine binding or the sedimentation behavior of the Chaps-solubilized channel complex. In addition, the rate of 45Ca2+ release from the vesicles decreased maximally 1.5-fold under conditions of partial activation of the SR Ca2+ release channel, i.e. when vesicles were diluted into media containing nanomolar (Fig. 8) or micromolar (Fig. 3) free Ca2+.
More extensive trypsin exposure produced several more marked changes in the structure and function of the channel complex. Disappearance of polypeptides with M , > 70,000 on immunoblots of ryanodine-labeled (not shown) and unlabeled membranes (Fig. lB, lunes 6 and 7) correlated with a decrease of the apparent S value of the solubilized receptor complex from 30 S to 10-15 S. Although [3H]ryanodine bound to the receptor prior to digestion remained bound, significant differences in the affinity and interaction of the binding sites were noticed when [3H]ryanodine binding was studied after trypsin digestion. Since a similar number of total binding sites (high and low affinity) were measured in the postlabeling experiments, trypsin digestion appeared to induce protein conformational changes in the complex which specifically altered the high affinity sites without causing a measurable loss of low affinity [3H]ryanodine binding sites. This surprising finding may help to elucidate the subunit composition and stoichiometry of the Ca2+ channel oligomeric complex. An additional observation was that more extensive trypsin digestion resulted in a transient increase of the &Ca2+ efflux rate (Figs. 3 and 8). That this increase was specifically due to an activation of the high conductance SR Ca2+ release channel was strongly suggested by the following two observations. First, trypsin digestion did not appreciably increase the Ca2+ permeability of light SR vesicles which lacked the Ca2+ release channel. During trypsin digestion the SR Ca2+ pump in heavy and light SR vesicles can be cleaved into two peptides of M , -55,000 and 50,000 (29, 30) which appeared as one major broad band in our gel system (Fig. 1A ). Although further slow cleavage of the Ca2+ pump protein has been reported to render SR vesicles "leaky" to Ca2+ by uncoupling Ca2+ transport from Release Release ryanodine receptor channel protein channel protein channel protein Ryanodine ATP hydrolysis (41,42) and this may have occurred under our experimental conditions, we did not observe a significant increase in the number of leaky vesicles. In this regard, it should also be noted that  have recently found that loss of Ca2+ accumulation by trypsindigested vesicles is not due to the cleavage of the Ca2+ pump protein, but rather to that of some other component of the SR membrane. A second observation supporting a direct action of trypsin on the Ca2+ release channel was that in planar bilayers trypsin increased the fraction of open time of the Ca2+-activated channel. Subsequent to channel activation, a brief period of reduced activity was noticed which was followed by a complete and irreversible loss of channel activity. Trypsin digestion did not induce sub-levels of conductance, a phenomenon previously observed when the purified skeletal 30 S channel complex was reconstituted into the bilayers (9), and was ineffective in changing channel activity when added to the lumenal side of the channel. Our measurements further revealed that the trypsin-digested channel remained sensitive to activation by Ca2+ and adenine nucleotide and inhibition by M F and ruthenium red. As a consequence, the extent of channel activation depended on the composition of the releasehilayer media ( Figs. 4 and 5).
In a recent study, under conditions of extensive digestion of the -400-kDa protein, the junctional processes of heavy vesicles were visualized by electron microscopy without discernible alterations in ultrastructure (18). This observation suggested that the proteolytic fragments were stabilized by multiple noncovalent interactions which only can be dissociated by strong detergents. In the present study, a decrease in the apparent sedimentation coefficient of the trypsindigested, Chaps-solubilized receptor complex from 30 S to 10 S was observed with retention of prelabeled [3H]ryanodine binding (Fig. 8). Since ultrastructural studies with the purified channel complex have revealed the presence of four peripheral protein "loops" which project from an electron-dense core region of the complex (9), a full or partial removal of the protein loops, during detergent solubilization of the membranes would appear to be the most plausible explanation for the decrease in S value. Retention of [3H]ryanodine to the -10 S complex would therefore suggest that the central portion of the complex comprises the high affinity ryanodine binding site and is held together by noncovalent bonds which are not broken by the mild detergent Chaps.
In conclusion, we have presented evidence that the cytoplasmic transverse tubule-SR junction spanning portion of the SR Ca2+ release channel is highly sensitive to proteolysis. Our results indicate that the channel retains its ability to conduct Ca2+ after cleavage of the -400-kDa channel subunits to fragments identifiable by immunoblotting of less than 70,000 daltons. In addition, although measurement of channel activity and [3H]ryanodine binding affinity after proteolysis suggested the occurrence of protein structural changes, the channel remained sensitive to activation by Ca2+ and adenine nucleotide and inhibition by M e and ruthenium red.