The calcium release channel of sarcoplasmic reticulum is modulated by FK-506-binding protein. Dissociation and reconstitution of FKBP-12 to the calcium release channel of skeletal muscle sarcoplasmic reticulum.

The ryanodine receptor/calcium release channel (CRC) of rabbit skeletal muscle terminal cisternae (TC) of sarcoplasmic reticulum (SR) has been found to be tightly associated with FK-506 binding protein (FKBP-12), the cytosolic receptor (immunophilin) for the immunosuppressant drug FK-506 (Jayaraman, T., Brillantes, A. M., Timerman, A. P., Fleischer, S., Erdjument-Bromage, H., Tempst, P., and Marks, A. (1992) J. Biol. Chem. 267, 9474-9477). In this study, a procedure is described to dissociate FKBP from TC and reconstitute human recombinant FKBP-12 back to the ryanodine receptor so that the role of the immunophilin on CRC activity can be assessed. Titration of TC vesicles with FK-506 dissociates FKBP from the ryanodine receptor. Sedimentation of FK-506-treated vesicles effectively separates the TC from the soluble FKBP-FK506 complex which remains in the supernatant. The FKBP-deficient TC vesicles have altered functional characteristics: 1) the ATP-stimulated calcium uptake rate of TC vesicles is reduced 2-fold; and 2) the threshold concentration of caffeine required to induce calcium release from TC vesicles is decreased. These changes appear to reflect modification of the calcium release channel since: 1) severalfold higher concentrations of FK-506 do not alter the calcium uptake rate of either longitudinal tubules of SR, or TC vesicles in the presence of ruthenium red; 2) human recombinant FKBP reassociates with FKBP-deficient TC but not with control TC or longitudinal tubules of SR; and 3) the reduced Ca2+ uptake rate in FKBP-deficient TC is restored to control values in the FKBP-reconstituted TC. These studies demonstrate that FKBP-12 modulates the CRC of rabbit skeletal muscle sarcoplasmic reticulum.

The ryanodine receptor/calcium release channel (CRC) of rabbit skeletal muscle terminal cisternae (TC) of sarcoplasmic reticulum (SR) has been found to be tightly associated with FK-506 binding protein (FKBP-12), the cytosolic receptor (immunophilin) for the immunosuppressant drug FK-506 (Jayaraman, T., Brillantes, A. M., Timerman, A. P., Fleischer, S., Erdjument-Bromage, H., Tempst, P., and Marks, A. (1992) J. Biol. Chem. 267,[9474][9475][9476][9477]. In this study, a procedure is described to dissociate FKBP from TC and reconstitute human recombinant FKBP-12 back to the ryanodine receptor so that the role of the immunophilin on CRC activity can be assessed. Titration of TC vesicles with FK-506 dissociates FKBP from the ryanodine receptor. Sedimentation of FK-506-treated vesicles effectively separates the TC from the soluble FKBP-FK506 complex which remains in the supernatant. The FKBP-deficient TC vesicles have altered functional characteristics: 1) the ATP-stimulated calcium uptake rate of TC vesicles is reduced 2-fold; and 2) the threshold concentration of caffeine required to induce calcium release from TC vesicles is decreased. These changes appear to reflect modification of the calcium release channel since: 1) severalfold higher concentrations of FK-506 do not alter the calcium uptake rate of either longitudinal tubules of SR, or TC vesicles in the presence of ruthenium red; 2) human recombinant FKBP reassociates with FKBP-deficient TC but not with control TC or longitudinal tubules of SR; and 3) the reduced Ca2+ uptake rate in FKBP-deficient TC is restored to control values in the FKBP-reconstituted TC. These studies demonstrate that FKBP-12 modulates the CRC of rabbit skeletal muscle sarcoplasmic reticulum.
Skeletal muscle contraction is triggered by calcium release * This work was supported in part by National Institutes of Health Grant HL32711 and the Muscular Dystrophy Association of America. 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.
$ Recipient of a National Research Service Award from the National Institutes of Health. from the sarcoplasmic reticulum (SR)' (1). Sarcoplasmic reticulum consists of two domains. The longitudinal tubules are composed largely of calcium pump membrane involved in calcium uptake enabling muscles to relax (2). The terminal cisternae (TC) of SR are junctionally associated with the transverse tubule by way of the foot structures to form the intracellular triad junction (3). It is across the triad junction that the stimulus of transverse tubule depolarization is transduced to Ca2+ release from the lumen of the TC which triggers contraction. The SR calcium release channel has been isolated from solubilized TC vesicles by virtue of its high affinity for the drug ryanodine (4)(5)(6)(7). The ryanodine receptor is morphologically identical to the foot structure (5). The receptor has been cloned and its protomer size, deduced from its predicted amino acid sequence, is 565,000 daltons (8)(9). The relative mass of the native receptor, as determined by scanning transmission electron microscopy, is 2.3 million (10). Therefore, the receptor is a homotetramer which displays distinct 4-fold symmetry as viewed by electron microscopy (11). The purified receptor reconstituted into planar lipid bilayers displays calcium ion channel activity which is modulated by drugs in a manner that is consistent with the modulation of calcium ion fluxes observed in isolated SR vesicles (6,12,13).
Cyclophilin and FK binding protein (FKBP-12) are the cytosolic receptors for the potent immunosuppressant drugs cyclosporin A and FK-506, respectively. These receptor proteins are referred to as immunophilins (14, 15). FK-506 and cyclosporin A are used to prevent graft rejection following organ transplantation. The drugs bind to their respective receptors and each immunophilin-drug complex blocks transcription of a set of early phase genes, including interleukin-2, required for activation of T-lymphocytes. Each immunophilin-drug complex is a potent inhibitor of calcineurin, a calmodulin-dependent/calcium-activated protein phosphatase (16). Inhibition of calcineurin appears to inhibit transcription of the interleukin-2 gene in T-lymphocytes, thereby inhibiting T-cell activation (17,18).
Immunophilins are highly conserved and broadly distributed in all cells. However, their physiological role in all cells, including T-lymphocytes, is unknown. Recently, the skeletal The abbreviations used are: SR, sarcoplasmic reticulum; BSA, bovine serum albumin; CHAPS, 3-[(3-cholamidopropyl)dimethylammoniol-1-propanesulfonate; CRC, calcium release channel; CsA, cyclosporin A DTT, dithiothreitol; FKBP-12 (or FKBP), FK-506 binding protein; IHM, imidazole homogenization medium; LT, longitudinal tubules; RR, ruthenium red; TBST, Tris-buffered saline-Tween; TC, terminal cisternae; PAGE, polyacrylamide gel electrophoresis. 22992 muscle ryanodine receptor has been found to be tightly associated with FKBP-12 (19,20). The finding that FKBP-12 is tightly associated with the calcium release channel of rabbit skeletal muscle sarcoplasmic reticulum provides a system to study the role of this immunophilin in skeletal muscle excitation-contraction coupling. In this study, we describe a procedure to dissociate FKBP-12 from TC and then rebind recombinant FKBP-12 to FKBP-deficient TC. As shown in this report, this procedure is useful to assess the role of FKBP-12 on ryanodine receptor function. General Methods-The protein concentration of SR and human recombinant FKBP-12 preparations were estimated by the folin reaction using bovine serum albumin as protein standard (21). SDS-PAGE was performed with a mini-slab gel apparatus (Hoeffer Scientific) using the buffer system described by Laemmli (22).

Isolation of Longitudinal and Terminal Cisternae Vesicles from Rabbit Skeletal Muscle
Membrane fractions referable to the terminal cisternae and longitudinal tubules of sarcoplasmic reticulum were isolated from New Zealand White rabbit skeletal muscle as described previously (5,23).

Drug Binding Assays
PHlRyanodine Binding-The concentration of the high affinity ryanodine-binding site in SR membranes was measured by Scatchard analysis of ryanodine-binding isotherms from 2 to 60 nM [3H]ryanodine. Ryanodine binding was measured essentially as described previously (5,24); briefly, SR vesicles (25 pg of TC or 100 pg of LT protein) were incubated for 1 h at room temperature in 0.125 ml of buffer containing 10 mM K-HEPES, pH 7.4, 1 M KCl, 25 p~ CaC12, and 2-60 nM [3H]ryanodine (8500 counts/min/pmol). Nonspecific binding was estimated in the presence of 8 p~ cold ryanodine. Free ligand was separated from bound by sedimenting the TC vesicles in a Beckman TL-100.1 rotor at 95,000 revolutions/min for 15 min at 2 "C. The pellets were rinsed once and resuspended in 0.2 ml of water then counted in 4 ml of Cytoscint (ICN, Cleveland, OH).
[3H]FK-816 was added directly to the binding mixture from a stock of the drug in ethanol (final ethanol concentration was 5%). Solubilization of the membranes in 0.5% CHAPS was required in order to separate bound from free ligand on LH-20 Sephadex columns. Additionally, 0.5% CHAPS prevented loss of the hydrophobic ligand frequently observed in aqueous medium either due to its insolubility or adsorption to the plasticware. Free [3H]FK-816 was separated from bound ligand on a 2-ml column of Sephadex LH-20 equilibrated in elution buffer (10 mM NaP04, 1 mM DTT, 0.25% CHAPS, and 0.01% NaN3). Typically 0.05 ml of binding mixture (containing 1.0 pg of TC, 5.0 pg of LT, or 1.0 ng of recombinant FKBP-12) was loaded onto a 2-ml LH-20 resin (packed in a 6-ml disposable screening column, Fisher Scientific). Bound ligand was eluted with 1.45 ml of elution buffer and collected in a 13 X 100mm test tube. 0.75 ml of the eluate was counted in 5.0 ml of Cytoscint. The free ligand absorbed to the LH-20 columns could be eluted with 12 ml of elution buffer allowing the columns to be reused several times.
Drug Binding Assays Utilizing a Sepharose 6B Spin Column-Binding of [3H]ryanodine and [3H]FK-816 to solubilized TC vesicles was also evaluated with a Sepharose 6B spin column to separate free from bound ligand. Spin columns composed of 2 ml of Sepharose 6B with 1 ml of Sephadex LH-20 carefully layered on top, were packed in 6-ml disposable screening columns (Fisher Scientific). Inclusion of 1 ml of LH-20 resin was found to reduce nonspecific ryanodine binding severalfold in comparison with a simple 2-ml Sepharose 6B column (Timerman and Fleischer 1992).' Both resins were preequilibrated in FK-816 binding buffer containing 1 M NaCl. The sample (50 or 100 pl) was loaded onto the LH-20 resin and washed into the column with 0.45 ml of FK-816 binding buffer containing 1 M NaCl. The column was centrifuged at 2 "C for 3 min at 800 revolutions/min in a Beckman TJ-6 tabletop centrifuge collecting the eluate (1.7 ml) in a 16 X 100-mm test tube. For comparison of [3H]ryanodine binding by the spin column method uersus control (TL 100.1 centrifugation method), TC vesicles (41 pg in 0.125 ml) were incubated for 1 h at room temperature as described above in the presence of 60 nM [3H] ryanodine (specific activity of 16,000 counts/min/pmol). After the incubation, samples (20 pg of TC protein) were either directly centrifuged in the TL-100.1 rotor or solubilized in 50 p1 of FK-816 binding buffer containing 1 M NaCl and 5 mg/ml BSA and then loaded onto the Sepharose 6B/LH-20 spin column. For comparison of [3H]FK-816 binding by the spin column method uersus control (2 ml LH-20 column method), TC vesicles (1.25 pg of protein in 0.125 ml) were incubated with 10 nM [3H]FK-816 in FK-816 binding buffer for 30 min at 37 "C. After incubation, samples (0.5 pg of protein) were applied to either a 2-ml LH-20 column or a 3-ml Sepharose 6B/LH-20 spin column.

Purification of Skeletal Muscle Ryanodine Receptor
The ryanodine receptor was purified from solubilized TC vesicles utilizing a modification of the sucrose gradient method described by Lai et al. (6). Briefly, TC vesicles were solubilized as described by Inui et al. (5) at 3 mg of protein/ml in 20 mM Tris-C1 pH 7.2, 1 M NaCl, 2 mM DTT, 2 pg/ml leupeptin, 0.1 mM phenylmethylsulfonyl fluoride, 4% CHAPS, and 2% soy bean lipid. The solubilized vesicles were incubated at room temperature for 1 h in the presence or absence of 10 p M FK-506. Solubilized vesicles (5 ml) were loaded on top of a step sucrose gradient containing 8 ml each 10, 15, 17.5, and 20% sucrose in solubilization buffer (lipid was omitted from the 15, 17.5, and 20% sucrose steps) and centrifuged at 2 "C for 2.5 h in a Beckman VTi-50 rotor at 45,000 revolutions/min. The tubes were fractionated from the bottom and purification of the ryanodine receptor was monitored by SDS-PAGE (6% resolving gel) stained with Coomassie Blue. Enriched fractions were combined, diluted with buffer (5 mM sodium-phosphate, pH 7.2, 2 mM DTT, 0.5 M KCl, 1% CHAPS) and concentrated by ultrafiltration (Centriprep 30) to about 0.2 mg/ml as determined by two-dimensional densitometry of the high molecular weight band (ryanodine receptor protomer) uersus bovine serum albumin.

The Ryanodine Receptor
Is Modulated by FKBP at a constant power of 2 watts/gel. One gel was visualized by silver staining while the other was transferred to an Immobilon-P membrane (Millipore, Bedford MA) for 30 min at 150 mA constant current in transfer buffer composed of 200 mM glycine, 25 mM Tris-Cl, pH 8.5, and 0.01% SDS. The blotted membrane was blocked in TBST solution (10 mM Tris-C1, pH 8.0, 0.55 M NaCl, and 0.05% Tween-20) containing 5% (w/v) dry milk protein for 1 h then probed for 90 min with anti-FKBP antiserum (19) at a dilution of 1:2000 in blocking buffer. The membrane was washed repetitively with TBST then probed for 1 h with secondary antibody (goat anti-rabbit IgG) conjugated to alkaline phosphatase at a 1:5000 dilution in blocking buffer. The membrane was again washed with TBST and developed with nitro blue tetrazolium and 5-bromo-4-chloro-3-indoyl phosphate substrates.

Calcium Loading and Release Assays
TC vesicles (2 mg/ml) were preincubated at room temperature for 1 h in IHM buffer containing 0-10 p~ FK-506. The vesicles were then stored on ice until the calcium loading and release assays were performed. The vesicles were often stored on ice for up to 7 h followed by quick freezing in liquid nitrogen and overnight storage at -80 "C. The loading rates of control vesicles was not compromised by such storage. The loading rates of FK-506-treated TC vesicles, however, was occasionally further reduced after prolonged storage. Calcium loading and caffeine-induced release assays in the presence of 100 mM phosphate were monitored spectrophotometrically with the calcium indicator antipyrylazo I11 essentially as described previously (4,27). Briefly, TC vesicles (25 pg of protein) were added to 1 ml (final volume) of loading medium (100 mM KP04, pH 7.0,5 mM antipyrylazo 111, 4 mM MgCl2, and 1 mM ATP), then aliquots of CaClz (7.5 or 15 nmol) were repetitively spiked into the cuvette, and the calcium uptake rate was monitored by dual wavelength spectrophotometry (710-790 nm) in a Hewlett Packard 8451A diode array spectrophotometer. After preloading the vesicles with 50-60 nmol of calcium (2.0-2.4 pmol of calcium/mg of TC), caffeine was added to induce calcium release. Following the release induced by caffeine, the total releasable calcium pool was measured by addition of the calcium ionophore A23187 (3 pglml).
To measure the enhancement of the calcium-loading rate by ruthenium red in the dissociation/reconstitution experiment, the basal calcium-loading rate in the medium described above was calculated from the average of two successive additions of 10 nmol of CaC12 in the absence of ruthenium red. Ruthenium red (3.5 p~) was then added to the medium, and the maximal uptake rate was calculated from a single addition of 20 nmol of CaC12. Calcium-loading rates of each sample were performed in duplicate.

Dissociation and Reconstitution of FKBP-12 with Terminal Cisternae Vesicles
Terminal cisternae vesicles were incubated in IHM at 5 mg/ml in the presence or absence of 6-10 p M FK-506 (these drug concentrations contain 10-16 times the number of FK-506 binding equivalents and no major differences were observed between 6 and 10 p M FK-506) for 90 min at room temperature. Samples of TC in 0.5 ml (2.5 mg of protein) were then loaded onto a 4-ml step gradient in a Beckman SW 56 rotor tube. The gradient was composed of sucrose solutions buffered with 5 mM imidazole-C1, pH 7.4, with steps of 12.5% (0.8 ml), 15% (1.25 ml), 17.5% (1.25 ml), and a 0.2-ml cushion of 50% sucrose. For reconstitution of FKBP-deficient TC vesicles, the 15% sucrose step contained recombinant FKBP-12 (200 pg, three to five times the molar equivalents of FK-506 loaded onto the gradient). TC vesicles were centrifuged in a Beckman SW 56 rotor at 2 "C for 1.5 h at 10,000 revolutions/min followed by an additional 1.5 h at 15,000 revolutions/min. The top layers were carefully aspirated before pipetting the vesicles from the interface of the 17.5-50% sucrose cushion in a volume of -0.2 ml. Typically 60-75% of the protein applied to the gradient was recovered in this procedure. Nonspecific association of FKBP-12 to the membrane was estimated by centrifugation of longitudinal tubules through an identical gradient containing FKBP-12 as well as control TC (not treated with FK-506). Sedimentation of longitudinal tubule fractions required additional centrifugation for 1 h at 20,000 revolutions/min. Reconstitution was monitored by Western blot analysis using antiserum to FKBP-12. Functional recovery was evaluated by measuring enhancement of the ATPstimulated calcium-loading rate by ruthenium red as described above. Competition binding studies (Fig. 2)  ' The ratio of FKBP/calcium release channel (homotetramer) was previously estimated to be l/foot structure. This was based on limited analysis by densitometry of the FKBP (14 kDa) band separated by SDS-PAGE of the purified ryanodine receptor (19) and not terminal cisternae as reported here. The underestimation appears due to a combination of experimental differences.  Fig. l . 4 ) and [3H] ryanodine binding isotherms from 2 to 60 nM [3H]ryanodine as described under "Experiment Procedures." Scatchard analysis of the ryanodine binding data gave only a straight line ( r = -0.976 f 0.016) compatible with a single high affinity ryanodine-binding site as described previously (24). The Kd for the high affinity ryanodinebinding site in the TC vesicles used in this study was 8.5 f 1.9 nM ( n = 5). The TC fraction binding data are expressed as the mean f S.D. for five preparations. The LT fraction binding data are expressed as the mean f the range for two preparations. With regard to the TCbinding data, the coefficient of variation (S.D./mean) of 18-20% for FK-816 and ryanodine binding indicates the absolute value of each receptor varied somewhat from preparation to preparation, yet the ratio of FKBP/foot structure was more consistent from preparation to preparation (the FKBP/foot ratio data have a coefficient of variation of only 6%).  Western blot analysis shown in Fig. 4B clearly indicates that all of the FKBP in untreated samples sediments with TC vesicles (Fig. 4B, sample 1 pellet ( P ) ) . Treatment with FK-506 (0.12-5.0 p~) results in a dose-dependent shift of about 80-90% of the FKBP into the supernatant fraction (Fig. 4B,  samples 2-5). The for dissociation of FKBP from TC vesicles is in the concentration range between 0.12 and 0.5 The Ryanodine Receptor Isolated from FK-506-treated Terminal Cisternae Lacks FKBP-Ryanodine receptor preparations were purified from solubilized T C vesicles by sucrose gradient centrifugation after incubation of the solubilized vesicles (in the presence or absence of 10 p~ FK-506). The ryanodine receptor prepared from control T C contains the 14 kDa band visualized by silver staining (lane 2 of Fig. 5A) which reacts positively to FKBP-12 antiserum (lane 2 of Fig.  5B). This band is greatly reduced or nearly absent from ryanodine receptor preparations isolated from solubilized T C vesicles which were preincubated for 1 h with 10 p~ FK-506 (lane 3 of Fig. 5, A and B).
The Calcium Loading Rate of TC Is Reduced by FK-506 Treatment-To evaluate the effect of FKBP-12 on the function of the calcium release channel, we compared the ATPstimulated calcium loading rate of control TC and TC treated with FK-506 (0-10 p~) .
The basis of this assay is that the calcium loading rate of TC vesicles is reduced by calcium leak from the vesicles via the calcium release channel (4). Therefore, drugs which close the calcium release channel (such as ruthenium red) enhance the calcium loading rate of T C vesicles. Ruthenium red has no effect on the loading rate of longitudinal tubules, which are devoid of ryanodine receptor. In contrast, drugs which activate the calcium release channel (such as ryanodine) reduce the calcium uptake rate of T C vesicles (4).
The calcium loading rate of T C vesicles is reduced after pretreatment with FK-506 in a concentration-dependent manner to about 35% of the control rate (Fig. 6). The ICs0 for FK-506 on the loading rate is approximately 0.2 p~. Therefore, the concentration of FK-506 required to alter TC func-pM FK-506.

22.
D-B 97- TC vesicles (2 mg/ml, equivalent to 0.25 FM FKBP-12) were incubated in IHM buffer containing 0-5.0 pM FK-506 for 1 h prior to centrifugation to yield pellets ( P ) and supernatants (S) as described under "Experimental Procedures." Equivalent volumes of P and S were separated by SDS-PAGE (15% gel) and stained with silver ( A ) or blotted and probed with FKBP antiserum ( B ) (19). In both A and B, 1-5 refers to FK-506 concentrations of 0, 0.12, 0.5, 1.2, and 5.0 pM. Lune FK refers to recombinant FKBP-12, and its position is indicated by the arrow on the right. The position of size markers (Bio-Rad), the resolving gel top (T), and dye front (D) are as indicated to the left.

FK
In B, the arrowhead to the right indicates IgG present in the vesicles which responds to secondary antibody alone. The data are representative of five different T C preparations. tion (Fig. 6) is in the same range as that required to dissociate FKBP-12 from the calcium release channel of TC vesicles (Fig. 4B). The loading rate of longitudinal tubules (which are devoid of ryanodine receptor) and TC vesicles in the presence of ruthenium red (which closes the calcium release channel) is not affected by 10 p~ FK-506 (Fig. 6). Dissociation of FKBP from the ryanodine receptor increases the Ca2' flux out of the TC likely due to enhanced activation of the calcium release channel as evidenced by the reduced TC calcium loading rate. We also find that rapamycin (10 p~) reduces the calcium loading rate of T C vesicles and dissociates FKBP from T C vesicles; however, CsA (10 p~) does not alter either the loading rate or association of FKBP with T C vesicles.
Treatment of TC with FK-506 Reduces the Threshold for Caffeine-induced Calcium Release-Pretreatment of TC with  FK-506 (0.5 p~) reduces the threshold concentration of caffeine required to induce calcium release from T C vesicles (Fig.  7). Control (untreated) vesicles (Fig. 7, A-C) more rapidly take up four to five aliquots of CaClz (7.5 or 15 nmol each) until they accumulate nearly 2 pmol of calcium/mg protein.

22-
The threshold of caffeine required to induce calcium release from untreated vesicles is 2.5 mM (Fig. 7C). This is to be contrasted with TC vesicles pretreated with 0.5 PM FK-506 ( Fig. 7, D and E ) . As expected from the results above (Fig. 6), the calcium loading rate is distinctly lower (e.g. compare A and D ) in treated vesicles. Additionally, the threshold of caffeine required for release is reduced to 1 mM caffeine (Fig.  7 0 ) . These results further indicate that dissociation of FKBP- 12 from the ryanodine receptor modulates the calcium release channel of rabbit skeletal muscle SR.

Rebinding of FKBP to FKBP-deficient TC
Vesicles and Reconstitution of Function-TC vesicles were preincubated for 90 min a t room temperature in the presence or absence of FK-506. Duplicate pairs of samples (0.5 ml each) were washed by centrifugation through identical sucrose gradients in which the 15% step of one gradient was supplemented with recombinant FKBP-12 (see "Experimental Procedures"). The vesicles were isolated from the gradient and analyzed for FKBP by Western blot analysis using antiserum against FKBP-12 (19). The results are summarized in Fig. 8A. Control TC vesicles (without FK-506 treatment) isolated from both control and FKBP-12-supplemented gradients contain similar amounts of FKBP-12 (Fig. 8A, lanes 1 and 2 ) while FK-506treated vesicles have a decreased content of FKBP (Fig. 8A,  lane 3 ) . Centrifugation of the FKBP-deficient vesicles through a sucrose gradient supplemented with recombinant FKBP in the 15% layer restores the level of FKBP-12 in the deficient vesicles to control levels (Fig. 8A, lane 4 ) . To evaluate the nonspecific interaction of FKBP-12 with the membrane, longitudinal tubule vesicles treated with FK-506 were centrifuged through similar gradients. Lanes 5 and 6 of Fig.  8A Fig. 8B; compare sample 3 versus sample 1 ). This is reflected in the 4.5-fold enhancement of the calcium loading rate with ruthenium red, as compared with the 2.4-fold ruthenium red enhancement in control TC (sample 1 ). The rebinding of FKBP to stripped vesicles (sample 4 ) restores Ca2+ loading to the control rate, and likewise the ruthenium red enhancement in the reconstituted vesicles is essentially the same as control TC. It should be emphasized that the calcium uptake rate in the presence of ruthenium red or 15 nmol/ 1 ml assay ( i e . calcium was added to concentrations of 7.5 or 15 p~) are indicated by the single arrowheads. The double arrowhead indicates the addition of the calcium ionophore A23187. It should be noted that for both control and FK-506-treatedTC, the extravesicular Ca2+ concentration after energized Ca2+ uptake is essentially the same for both (AOD = 0.090, which is equivalent to a free Ca2+ concentration of 1.4 p~) .
The data are typical of four T C preparations.
is similar in control, stripped, and reconstituted vesicles. Therefore, the increased ruthenium red enhancement observed in FKBP-deficient TC reflects the lower calcium loading rate resulting from the dissociation of FKBP-18 (sample 3 ) from the ryanodine receptor as previously described in Fig.  6. Rebinding of recombinant FKBP decreases the Ca2+ leak to control values (Fig. 8B, sample 4 ) . These data provide strong evidence that FKBP-12 modulates the gating properties of the calcium release channel of skeletal muscle sarcoplasmic reticulum. DISCUSSION We have previously reported that FKBP-12 is tightly associated with the ryanodine receptor/calcium release channel of rabbit skeletal muscle terminal cisternae (19). In this study, a procedure has been developed to dissociate FKBP-12 from the ryanodine receptor of TC and rebind FKBP to such FKBP-deficient TC. This dissociation and reconstitution procedure permits the evaluation of the effect of FKBP on the  [1][2][3][4] or LT vesicles (samples 5 and 6 ) at 5 mg/ml were incubated for 90 min in the presence (+) or absence (-) of FK-506. Duplicate samples were centrifuged through sucrose gradients supplemented with (+) or devoid of (-) 200 pg of recombinant FKBP-12 in the 15% sucrose step as described under "Experimental Procedures." Vesicles isolated from the gradient were monitored by Western blot analysis rateruthenium red) as indicated in the abscissa on the right. The ruthenium red enhancement ratio for the controls (sample 1 ) ranged from 2-to 3-fold. function of the ryanodine receptor. Most importantly, this study indicates that FKBP-12 specifically modulates the activity of the calcium release channel.

14-
A key breakthrough in this study was the development of a gentle procedure to dissociate FKBP-12 from the ryanodine receptor with FK-506. The clue came from studies which indicated slower binding of FK506 to TC as compared with recombinant FKBP although the binding affinities were similar (Fig. 1). An assay, using gel exclusion, designed to test FK-816 binding to the ryanodine receptor indicated that the smaller FKBP-FK506 complex is dissociated from the calcium release channel. The EC5o for FK-506 to dissociate FKBP from the ryanodine receptor was found to be in the range of 0.12 to 0.5 p~ (Fig. 4B). Since the concentration of FKBP in a 2 mg/ml suspension of TC is about 0.25 p~, these results show that titration with just a small excess of FK-506 equivalents leads to dissociation of FKBP from the ryanodine receptor.
We find that FKBP specifically modulates the activity of the calcium release channel. TC vesicles treated with FK-506 have a reduced calcium uptake rate due to enhanced leak of Ca2' from the CRC (Fig. 6). The concentration of FK-506 required to modulate TC function is similar to the concentration required to dissociate FKBP from the ryanodine receptor. That is, the IC,, to reduce the calcium uptake rate of TC vesicles and the EC5,, to dissociate FKBP-12 from the ryanodine receptor is in the same concentration range, between 0.2 to 0.3 p~ FK-506. The calcium uptake rate of longitudinal tubules (devoid of ryanodine receptor) and TC vesicles with ruthenium red (which closes the ryanodine receptor) is not reduced by 10 p~ FK-506. Thus, the effect of FK-506 appears to be specific for the ryanodine receptor/calcium release channel. A second index of altered ryanodine receptor function in FKBP-deficient TC vesicles is the 2-fold reduction in caffeine concentration required to stimulate caffeine-induced calcium release (Fig. 7).
The calcium loading rate of FKBP-deficient TC is restored by rebinding recombinant FKBP. The reassociation is specific to FKBP-deficient TC since FKBP does not bind to control TC or FK-506-treated longitudinal tubules. Our studies provide evidence that FKBP modulates the ryanodine receptor/ calcium release channel of skeletal muscle SR. In developing the dissociation and rebinding of FKBP methodology, the handling of TC vesicles had to be minimized to retain functionally competent vesicles. Washing the FK-506-treated TC and rebinding of recombinant FKBP was performed on a single sucrose gradient. FKBP was stripped from TC vesicles with excess FK-506. Recombinant FKBP was added in the 15% sucrose step at concentrations in excess of the total molar equivalents of FK-506 in the sample loaded onto the gradient. Under these conditions, FKBP readily reassociates with TC vesicles to comparable levels of control TC (Fig. 8A). Rebinding of FKBP restores the calcium loading rate of stripped vesicles to control values (Fig. 8B).
The stoichiometry of [3H]FK-816 to I3H]ryanodine binding in TC vesicles (Table I) indicates a molar ratio of approximately four FKBP/ryanodine receptor. The data suggest that each protomer of the homotetrameric calcium release channel is tightly associated with 1 mol of FKBP-12.
Although the action of FKBP in T-lymphocytes has been studied in detail, the physiological role of FKBP in this and all other cell types remains to be elucidated. FKBP is an I Is Modulated by FKBP 22999 ubiquitous and highly conserved protein which possesses peptidyl-prolyl cis-trum isomerase activity (29). The finding that FKBP is tightly associated with the calcium release channel of skeletal muscle opens the door to the study of the role of FKBP in muscle excitation-contraction coupling. The methodology developed in this study to dissociate FKBP from TC and rebind it is an important step in the study of the role of FKBP in calcium homeostasis in muscle.
In summary, in this study a procedure has been devised to dissociate the FKBP from the CRC in TC. We find that removal of FKBP-12 from the ryanodine receptor decreases Ca2+ loading which is explained by the increased tendency of the calcium release channel to open. Thus, FKBP appears to stabilize the closed conformation of the skeletal muscle ryanodine receptor and may thereby be important in modulating the gating kinetics of the calcium release channel during excitation-contraction coupling in skeletal muscle.