Dysregulated Ryanodine Receptors Mediate Cellular Toxicity

THE JOURNAL OF BIOLOGICAL CHEMISTRY © 2003 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 278, No. 31, Issue of August 1, pp. 28856 –28864, 2003 Printed in U.S.A.

Dysregulated Ryanodine Receptors Mediate Cellular Toxicity RESTORATION OF NORMAL PHENOTYPE BY FKBP12.6* Received for publication, December 6, 2002, and in revised form, April 29, 2003 Published, JBC Papers in Press, May 16, 2003, DOI 10.1074/jbc.M212440200

Christopher H. George‡§, Gemma V. Higgs‡, John J. Mackrill¶, and F. Anthony Lai‡ From the ‡Department of Cardiology, Wales Heart Research Institute, University of Wales College of Medicine, Heath Park, Cardiff CF14 4XN, United Kingdom and the ¶Department of Biochemistry, University College Cork, Cork, Ireland

Ca2ⴙ homeostasis is a vital cellular control mechanism in which Ca2ⴙ release from intracellular stores plays a central role. Ryanodine receptor (RyR)-mediated Ca2ⴙ release is a key modulator of Ca2ⴙ homeostasis, and the defective regulation of RyR is pathogenic. However, the molecular events underlying RyR-mediated pathology remain undefined. Cells stably expressing recombinant human RyR2 (Chinese hamster ovary cells, CHOhRyR2) had similar resting cytoplasmic Ca2ⴙ levels ([Ca2ⴙ]c) to wild-type CHO cells (CHOWT) but exhibited increased cytoplasmic Ca2ⴙ flux associated with decreased cell viability and proliferation. Intracellular Ca2ⴙ flux increased with human RyR2 (hRyR2) expression levels and determined the extent of phenotypic modulation. Co-expression of FKBP12.6, but not FKBP12, or incubation of cells with ryanodine suppressed intracellular Ca2ⴙ flux and restored normal cell viability and proliferation. Restoration of normal phenotype was independent of the status of resting [Ca2ⴙ]c or ER Ca2ⴙ load. Heparin inhibition of endogenous inositol trisphosphate receptors (IP3R) had little effect on intracellular Ca2ⴙ handling or viability. However, purinergic stimulation of endogenous IP3R resulted in apoptotic cell death mediated by hRyR2 suggesting functional interaction occurred between IP3R and hRyR2 Ca2ⴙ release channels. These data demonstrate that defective regulation of RyR causes altered cellular phenotype via profound perturbations in intracellular Ca2ⴙ signaling and highlight a key modulatory role of FKBP12.6 in hRyR2 Ca2ⴙ channel function.

Basic cellular homeostatic processes such as gene transcription, protein synthesis, and division are dependent on the precise spatio-temporal coordination of intracellular Ca2⫹ release into the cytoplasm and its subsequent sequestration into Ca2⫹ storage organelles or its extrusion from the cell (1). Defective Ca2⫹ handling has been implicated in many pathological conditions, including cardiac abnormalities, neurodegenerative disease, polycystic kidney disease, and cancer (2– 6). Much attention has focused on the encryption of complex Ca2⫹ signals underpinning homeostasis in healthy cells and how these are altered in pathogenic states to invoke cell death pathways (7, 8). Cell survival is thought to involve the “silencing” of cell death machinery, for example, pro-apoptotic elements are re* This work was supported by British Heart Foundation Research Fellowship FS2000020 (to C. H. G.) and by British Heart Foundation Grant PG99087 (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 “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § To whom correspondence should be addressed. Tel.: 44-292-0744431; Fax: 44-292-074-3500; E-mail: [email protected].

fractory to Ca2⫹-driven signals in healthy cells, and alterations of the frequency/amplitude encryption of Ca2⫹ signals invoke specific cell death programs (9). In Ca2⫹-driven cell death, the Ca2⫹ storage status of the endoplasmic reticulum (ER)1 and the levels of cytoplasmic free Ca2⫹ ([Ca2⫹]c) are proposed as the main determinants of cell fate, but the relative contribution of these in triggering cell death pathways is widely debated. Augmented ER Ca2⫹ load in COS cells caused cell death (10), whereas depletion of ER Ca2⫹ was toxic in SH-SY5Y neurons (11). Similarly, depletion of ER Ca2⫹ appeared to have a protective function for HeLa cells (12), in contrast to pancreatic B-cells where increased ER Ca2⫹ storage protected from nitric oxide-induced apoptosis (13). Studies in hippocampal neurons showed that increased [Ca2⫹]c was sufficient to mediate cell death (14) and in astrocytes similar elevation in [Ca2⫹]c is achieved solely from influx of extracellular Ca2⫹ with negligible involvement of intracellular Ca2⫹ stores (15). Conversely, it has been shown that a moderate and sustained elevation of cytoplasmic Ca2⫹ in neurons is protective (16). Other cellular elements contribute to modulation of the intracellular Ca2⫹ signal. The close apposition of mitochondria with the ER is implicated in shaping cellular Ca2⫹ phenomena (17) and mitochondrial function appears crucial in both cell survival (18) and destruction (9). The spatial organization of Ca2⫹ release channels, ryanodine receptors (RyR), or inositol 1,4,5-trisphosphate receptors (IP3R) cells appears to be of key importance in determining cellular fate. Increased density of IP3R and RyR triggers cell death (19 –21), and the finding that both IP3R and RyR can directly sense cytoplasmic and luminal Ca2⫹ environments (22–24) suggests that Ca2⫹ itself is a key transducer in the activation of programmed cell death. We have investigated the specific role of RyR in mediating altered cellular phenotype. RyRs are large homotetrameric channels (molecular mass, ⬃2.3 MDa) that mediate Ca2⫹ release from sarcoplasmic reticulum (SR) in skeletal and cardiac muscle. Recent evidence predicts that in situ RyRs do not exist as discrete entities but act as scaffolding complexes that integrate a multitude of regulatory signals from accessory proteins, including calmodulin, FK506-binding protein (FKBP), protein kinase A, and protein phosphatases (25, 26). It has been demonstrated that altered stoichiometry between RyR and inter1 The abbreviations used are: ER, endoplasmic reticulum; [Ca2⫹]c, cytoplasmic Ca2⫹; RyR, ryanodine receptor; hRyR, human RyR; IP3, inositol 1,4,5-trisphosphate; IP3R, IP3 receptor; SR, sarcoplasmic reticulum; FKBP, FK506-binding protein; CHO, Chinese hamster ovary cells; BAPTA-AM, 1,2-bis(2-aminophenoxy)ethane-N,N,N⬘,N⬘-tetraacetic acid tetrakis (acetoxymethyl ester); WT, wild-type; GST, glutathione S-transferase; PBS, phosphate-buffered saline; PI, propidium iodide; KRH, Krebs-Ringer-Hepes medium; 4-CMC, 4-chloro-m-cresol; Emmax, wavelength for maximum emission; Exmax, wavelength for maximum excitation; DiOC6, 3,3⬘-dihexyloxacarbocyanine iodide.

28856

This paper is available on line at http://www.jbc.org

hRyR2 Interaction with FKBP12.6 Prevents Cell Death acting accessory proteins results in pathogenic Ca2⫹ signaling. For example, dissociation of the RyR2䡠FKBP12.6 complex induces an aberrant leak of Ca2⫹ from the SR, which is implicated in the pathogenesis of heart failure (27, 28). We have generated CHO cells stably expressing discrete levels of recombinant human cardiac RyR (CHOhRyR2) and showed that a selective interaction occurred between hRyR2 and co-expressed FKBP12.6, which critically modulated the Ca2⫹ release activity of hRyR2 in situ (29). This study investigates the empirical observation that CHOhRyR2 exhibit increased cellular toxicity and decreased proliferation when compared with wild-type CHO cells. We demonstrate that these phenotypic changes are underscored by perturbations in cellular Ca2⫹ handling specifically mediated by hRyR2. Normal cellular phenotype was restored by modulating hRyR2-mediated Ca2⫹ release via co-expression of FKBP12.6, which highlighted the role of FKBP12.6 in homeostatic Ca2⫹ signaling and cell function. Our data indicate that expression of hRyR2 in an FKBP12.6-deficient context leads to increased cytoplasmic Ca2⫹ flux, without an accompanying change in mean cytoplasmic Ca2⫹ levels, which is deleterious to cell survival. EXPERIMENTAL PROCEDURES

All cell culture materials, plastic ware, and LipofectAMINE 2000 were obtained from Invitrogen (Paisley, UK). Rapamycin, thapsigargin, and ryanodine were from Calbiochem (Nottingham, UK). AlamarBlue was from BIOSOURCE International (Nivelles, Belgium). Fluo3-AM, BAPTA-AM, DiOC6 (3), and Alexa-conjugated secondary antibodies were from Molecular Probes (Leiden, Netherlands). All other chemicals were from Sigma (Poole, UK). Cell Culture and Analysis of Recombinant Protein Expression—Chinese hamster ovary (CHOhRyR2) cells stably expressing discrete levels of hRyR2 (29) were routinely cultured in supplemented nutrient Ham’s F-12 medium (containing fetal calf serum (10% (v/v), amphotericin B (2.5 ␮g/ml), and G418 sulfate (500 ␮g/ml)). CHOWT cells were cultured in supplemented nutrient Ham’s F-12 medium lacking G418 sulfate. In some experiments, cells were transfected with plasmid pCR3 (Invitrogen, UK) encoding human FKBP12 or FKBP12.6 using LipofectAMINE2000 in a ratio of 0.8 ␮g:1 ␮g (DNA:lipid) according to the manufacturer’s protocol, and transfection efficiencies of ⬎85% were routinely achieved. FKBP-hRyR2 interaction was disrupted following incubation of cells with rapamycin (5 ␮M). Immunoblotting analysis of recombinant hRyR2 in post-nuclear supernatants obtained from wild-type CHO (CHOWT) and CHOhRyR2 was performed as described using an RyR2-specific antibody, pAb129 (30). The cellular expression of endogenous and recombinant FKBP12 and FKBP12.6 isoforms was determined using a rabbit polyclonal antiFKBP12 antibody (see below) in immunoblotting analysis of post-nuclear supernatants (100 ␮g) obtained from CHOWT and CHOhRyR2 following separation of proteins on 20% (v/v) acrylamide SDS-PAGE (29). Immunofluorescent localization of endogenous IP3R and recombinant hRyR2 in CHOhRyR2 cells was performed using pAb-40 (anti-IP3R, 1:250 dilution) (31) and pAbN-19 (anti-RyR, 1:500 dilution) (Santa Cruz Biotechnology), respectively, followed by subsequent detection using donkey secondary antibodies conjugated to Alexa488 (IP3R) and Alexa546 (RyR) as previously described (29). The extent of co-localization between IP3R and RyR signals was quantified using Scanware software (Leica). DiOC6 (3) (0.5 ␮M) was used to visualize the ER in RyR-deficient CHOWT cells. Production of Anti-FKBP12 Polyclonal Antiserum—Recombinant human FKBP12-glutathione S-transferase (hFKBP12-GST) fusion protein was generated as previously described (32). hFKBP12-GST (250 ␮g), complexed with glutathione-Sepharose beads (Amersham Biosciences) and resuspended in PBS (250 ␮l) after extensive washing in PBS, was emulsified with an equal volume of Freund’s Complete adjuvant (first immunization) or Freund’s Incomplete adjuvant (second and third immunizations) prior to subcutaneous injection into adult New Zealand White female rabbits (University College Cork Biological Services Unit). Purified hFKBP12 (50 ␮g) (Sigma) emulsified in Freund’s Incomplete adjuvant as above was used as the final immunogen. Antiserum was isolated from whole blood following clotting and was stored in sodium azide (0.02% (w/v)). Determination of Cell Death Modality and Cell Proliferation—Cell viability was determined 36 h post-seeding using Trypan blue (0.4%

28857

(w/v) in PBS (NaCl, 140 mM; KCl, 2.7 mM; Na2HPO4, 10 mM; NaH2PO4, 2 mM; pH 7.4) and propidium iodide (PI; 1 ␮g/ml in PBS) permeability assays (33). The plasma membranes of viable cells are impermeable to Trypan blue and PI, and only cells that were stained positively by both Trypan blue and PI were confirmed to be non-viable. Cells were visualized using phase (Trypan blue) and fluorescent microscopy (PI; Emmax ⬎ 620 nm), respectively. All assays were performed on adherent cells, and there was negligible detachment of cells from the culture surface throughout these experiments as determined by hemocytometric measurement of aspirated culture medium (data not shown). The extent of apoptotic cell death was quantified using a fluorometric TdT-mediated dUTP nick-end labeling (TUNEL) assay (DeadEndTM, Promega, UK). Briefly, resting cells or cells preincubated with modulators of intracellular Ca2⫹ handling (see below) were fixed in paraformaldehyde (3% (v/v) in PBS), permeabilized in Triton X-100 (0.2% (v/v) in PBS), and incubated in buffer containing terminal deoxynucleotidyl transferase enzyme (25 units) for 1 h at 37 °C to incorporate fluorescein12-dUTP into caspase-fragmented DNA, a phenotypic marker of apoptosis. Cells were washed in 2⫻ SSC (300 mM NaCl, 30 mM sodium citrate; pH 7.0), and apoptotic cells were identified by the presence of highly fluorescent nuclei (Emmax, 511 nm). Determination of Cellular Proliferation—The proliferative capacity of sub-confluent CHOWT and CHOhRyR2 cells was determined on the same cell populations at 24-h intervals for 96 h. Cells were incubated with medium containing AlamarBlue (10% (v/v)) for 5 h prior to fluorometric analysis of aspirated medium (LS50B, PerkinElmer Life Sciences). Cellular metabolic activity reduces AlamarBlue from a nonfluorescent (blue) form into a fluorescent (red) form (Exmax, 560 nm; Emmax, 590 nm). The oxidized and reduced forms of AlamarBlue are freely cell permeable and do not contribute to cellular toxicity (34). The seeding density of cells was adjusted to ensure that the cells were sub-confluent throughout these procedures. The data obtained using AlamarBlue cell viability assays was corroborated by hemocytometric measurement of cell number performed in parallel experiments. Measurement of [Ca2⫹]c, ER Ca2⫹ Load, and Intracellular Ca2⫹ flux—The Ca2⫹-dependent fluorescence of fluo3 was calibrated in situ in streptolysin O (200 units/ml)-permeabilized cells (35) loaded with fluo3-AM (15 ␮M), which were incubated with extracellular solutions containing known [Ca2⫹] (17 nM to 39 ␮M) to clamp [Ca2⫹]c (Molecular Probes). The apparent Kd (Kd,app) of 580 nM was generated using GraphPad Prism software. [Ca2⫹]c was calculated from fluorescent data using: [Ca2⫹] ⫽ Kd,app (F ⫺ Fmin)/(Fmax ⫺ F) (36). Fmax and Fmin were determined following the addition of ionomycin (1 ␮M) and EGTA (5 mM), respectively, at the end of experiments. Cells grown on coverslips were loaded with fluo3-AM (15 ␮M) for 1 h at 23 °C in Krebs-Ringer-Hepes medium (KRH; 120 mM NaCl, 25 mM Hepes, 4.8 mM KCl, 1.2 mM KH2PO4, 1.2 mM MgCl2, 5.5 mM glucose, 1.3 mM CaCl2; pH 7.4), and immediately prior to experiments, cells were transferred to KRH containing nominal free Ca2⫹ (KRH-Ca2⫹, where EGTA (1 mM) replaced CaCl2). Resting and agonist-stimulated [Ca2⫹]c was measured in cells using a confocal microscope (SP2, Leica, Heidelberg, Germany) in bidirectional scan mode (512- ⫻ 64-pixel resolution) controlled with Leica software. To specifically trigger Ca2⫹ release through hRyR2, 4-chloro-m-cresol (4-CMC; 0.5 mM) was added to cells, and Ca2⫹ release through endogenously expressed IP3R (37) was triggered by the intracellular generation of IP3 from PIP2 via the activation of purinergic (P2Y) receptors using ATP (1 mM) (38). To investigate the effects of blocking RyR-mediated Ca2⫹ release, cells were exposed to ryanodine (1 mM) for 15 h or for 2 h prior to agonist-stimulated Ca2⫹ release. IP3R-mediated Ca2⫹ release was inhibited by cytoplasmic loading of heparin using a hypo-osmotic pinocytic lysis procedure (39). Briefly, cells were incubated in Leibowitz L-15 medium containing sucrose (0.5 M), polyethylene glycol 1000 (10%, v/v), and heparin (5 mg/ml) for 5 min to form heparin-containing pinocytic vesicles. Cells were exposed to hypo-osmotic media (6 parts L-15 medium to 4 parts water) for 2 min to promote pinocytic lysis and the distribution of heparin throughout the cytoplasm, prior to rinsing and further culturing in normal nutrient F-12 media. This procedure, which resulted in the loading of ⬃95% cells as determined by cytoplasmic localization of fluorescent fluorescein isothiocyanate-dextran, did not contribute to cell toxicity (data not shown). In some experiments, cells were loaded with BAPTA-AM (0.1 ␮M), a cytoplasmic Ca2⫹ chelator, for 15 h prior to experiments. The ER Ca2⫹ load status was estimated from the measurement of peak Ca2⫹ release following addition of thapsigargin (5 ␮M) to CHOWT and CHOhRyR2 cells maintained in KRH-Ca2⫹ (29). Thapsigargin-induced Ca2⫹ release in CHOWT cells was designated as 100%. The variability in the intracellular Ca2⫹ flux in resting (non-stimu-

28858

hRyR2 Interaction with FKBP12.6 Prevents Cell Death lated) CHOhRyR2, when compared with CHOWT, was estimated following calculation of the F ratio (40). The F ratio, defined as S.D.a2/S.D.b2 (where S.D.a is the standard deviation of [Ca2⫹]c values determined in CHOhRyR2 cells, and S.D.b is the standard deviation of [Ca2⫹]c values calculated in CHOWT cells for each experimental condition), numerically describes the amplitude variability in resting cytoplasmic Ca2⫹ flux. All experiments were performed in medium containing nominal free Ca2⫹ to negate the contribution of extracellular Ca2⫹ influx. RESULTS AND DISCUSSION

FIG. 1. CHOhRyR2 cells exhibit altered cellular phenotype. A, post-nuclear cell lysates (250 ␮g) prepared from CHOWT, CHOL, and CHOH were separated using 4% (v/v) acrylamide SDS-PAGE, and recombinant hRyR2 was detected following immunoblotting analysis using an anti-RyR2 antibody (pAb129). Rabbit cardiac microsomes (CM, 25 ␮g) prepared as described previously (56) were used as a source of native RyR2. Densitometric analysis determined that the expression of hRyR2 in CHOH was 42 ⫾ 9% greater than in CHOL, whereas CHOWT did not express detectable levels of RyR2. In B: panel a, the morphology of a typical population of CHOWT cells (stained with DiOC6 (3) to visualize the ER) is shown. The arrow depicts a cell exhibiting altered morphology. Bar represents 20 ␮m. Panels b and d, CHOH (b) or CHOL (d), stained with pAb129, were characterized by a significant proportion (⬃40% and ⬃20%, respectively) of small, rounded cells exhibiting altered morphology (arrowed). The bar represents 20 ␮m. Panels c and e, highlighted cells in b and d (boxed) were scanned at higher magnification to reveal extensive membrane blebbing. The bar represents 5 ␮m. Panels are representative of at least four separate experiments. In C: panel a, the extent of apoptosis (black squares) or necrosis (white squares) in CHOWT and CHOhRyR2 cells in the presence or absence of

Overexpression of hRyR2 Induces Cell Necrosis—Chinese hamster ovary (CHOhRyR2) cells stably expressing discrete levels of hRyR2 (CHOH (high hRyR2 expression) and CHOL (low hRyR2 expression), originally designated CHO#3.2 and CHO#3.20, respectively) have been generated previously (29). Immunoblotting analysis of post-nuclear supernatants obtained from CHOH, CHOL, and CHOWT confirmed the rank order of recombinant hRyR2 expression (molecular mass, ⬃560 kDa) was CHOH ⬎ CHOL ⬎⬎ CHOWT (RyR-deficient) (Fig. 1A). The normal morphology of CHOWT cells is shown with typically 2– 4% of cells exhibiting altered morphology (Fig. 1B, panel a, arrowed). In contrast, CHOH, a cell line stably expressing high levels of hRyR2, were characterized by a dramatically increased prevalence of small, morphologically altered cells (Fig. 1B, panel b, arrowed), a small proportion of which exhibited significant cell surface membrane blebbing (Fig. 1B, panel c), a phenotypic marker of apoptotic cell death. CHOL cells, which express lower levels of hRyR2 (Fig. 1A), also exhibited pronounced cellular toxicity (Fig. 1B, panel d, arrowed) with evidence of apoptosis (Fig. 1B, panel e). We determined a marked reduction in the viability of resting CHOhRyR2 when compared with CHOWT, which was proportional to the expression levels of hRyR2, i.e. viability of CHOH ⬍ CHOL ⬍⬍ CHOWT (Fig. 1C, panel b). The normal modulation of RyR2 requires a multitude of accessory proteins, some of which are not endogenously expressed in CHO cells (21, 29) and thus our CHOhRyR2 cell model provides a powerful platform to determine the impact of defective hRyR2 Ca2⫹ signaling on cell function. Inhibition of hRyR2 Ca2⫹ release by incubation of cells with ryanodine (1 mM) fully restored the viability of CHOhRyR2 cells to levels measured in CHOWT (Fig. 1C, panel b), indicating that the increased cell death measured in CHOhRyR2 was mediated by the expression of functional hRyR2. Analysis of the mode of cell death in resting CHOhRyR2 identified significantly elevated levels of necrotic cell death, which was prevented following ryanodine-mediated inhibition of hRyR2 (Fig. 1C, panel a). These data are in broad agreement with findings that heterologous expression of RyR produces marked phenotypic changes when compared with wild-type host cell lines (21, 41, 42), but our results extend these findings by demonstrating that cell death in resting (non-stimulated) cells expressing functional hRyR2 predominantly occurs via necrosis. However, the precise mechanistic basis of necrosis is unknown in these cells, and thus throughout this study the term “necrosis” refers to a currently undefined mechanism of cell death that is distinct from apoptosis. Increased Cytoplasmic Ca2⫹ Flux via hRyR2 Mediates Cell Necrosis—Because the resting [Ca2⫹]c in CHOhRyR2 was effectively constant (Fig. 2A), the decreased viability of CHOhRyR2 did not arise from sustained elevations in cytoplasmic Ca2⫹. ryanodine (1 mM) was quantified as described under “Experimental Procedures.” *, p ⬍ 0.005 when compared with CHOWT. Panel b, the viability of untreated cells (white bars) or cells exposed to ryanodine (1 mM; black bars) was quantified using Trypan Blue/propidium iodide cell exclusion assays. Data are expressed as mean values ⫾ S.E. (n ⫽ 6) and are derived from at least 700 cells in each experiment. *, p ⬍ 0.005 when compared with CHOWT and CHOL; **, p ⬍ 0.01 when compared with CHOWT and CHOH.

hRyR2 Interaction with FKBP12.6 Prevents Cell Death

FIG. 2. FKBP12.6 suppresses cytoplasmic Ca2ⴙ flux and restores cell viability. A, resting [Ca2⫹]c were measured in CHOWT and CHOhRyR2 cells with or without co-expression of FKBP12.6 or FKBP12. Data were acquired every 30 ms from cells in nominal free extracellular Ca2⫹. Traces shown are representative of at least six separate experiments. To disrupt FKBP12.6-RyR2 interaction, rapamycin (5 ␮M) was added 15 h prior to measurements. In B: panel a, resting [Ca2⫹]c in cells was calculated as in A. Panel b, the F ratio was calculated from the fluorescent data obtained from CHOhRyR2 and CHOWT cells and numerically describes the variability in the cytoplasmic Ca2⫹ signal (see “Experimental Procedures”). Panel c, the peak [Ca2⫹] release following the addition of thapsigargin (5 ␮M) was used to estimate ER Ca2⫹ load. Data are given as the relative ER Ca2⫹ load when compared with that measured in untreated CHOWT cells (100%). Where indicated, rapamycin was added 15 h prior to measurement. Data are given as means ⫾ S.E. (n ⫽ 4). *, p ⬍ 0.005 when compared with CHOWT and CHOL cells; **, p ⬍ 0.01 when compared with CHOWT and CHOH; and #, p ⬍ 0.005 when compared with untreated CHOWT and CHOhRyR2 cells. In C: immunoblotting analysis of post-nuclear supernatants (100 ␮g) obtained from control cells (left panel) or cells expressing recombinant FKBP12.6 (middle panel) or FKBP12 (right panel) using a rabbit polyclonal anti-FKBP12 antiserum (1:1000 dilution). The antibody crossreacts with FKBP12 and FKBP12.6. Rabbit cardiac homogenates (CH; 25 ␮g) were used as a source of FKBP isoforms. In D: the mode of cell death (a) and the overall cell viability (b) was determined in control cells (⫺) or following the expression of FKBP12 (12) or FKBP12.6 (12.6) in the absence or presence of rapamycin (white and black bars, respectively). Data are given as means ⫾ S.E. (n ⫽ 6, ⬎700 cells per experiment). In a: *, p ⬍ 0.01 when compared with CHOWT and in b: *, p ⬍ 0.005 when compared with CHOWT and CHOL; **, p ⬍ 0.01 when compared with CHOWT and CHOH cells.

Despite similar resting [Ca2⫹]c in CHOWT and CHOhRyR2 (Fig. 2B, panel a), we observed that increased expression of hRyR2 was associated with a concordant increase in “noise” (flux) in resting Ca2⫹ signals recorded in CHOhRyR2 cells (Fig. 2A, “control” traces). We hypothesized that the increased amplitude of

28859

FIG. 3. The effects of cytoplasmic Ca2ⴙ modulators on Ca2ⴙ handling in CHOhRyR2 cells. A, resting [Ca2⫹]c was determined in cells incubated in ryanodine (1 mM) and BAPTA-AM (0.1 ␮M) or loaded with heparin (5 mg/ml) via hypo-osmotic lysis of pinosomes (see “Experimental Procedures”). Data were acquired every 30 ms and are representative of at least six separate experiments. In B: the resting [Ca2⫹]c (a), F ratio (b), and ER Ca2⫹ load (c) were determined as described in the legend accompanying Fig. 2. Data are given as means ⫾ S.E. (n ⫽ 5); *, p ⬍ 0.01 when compared with CHOWT and CHOL cells; **, p ⬍ 0.01 when compared with CHOWT and CHOH cells; and #, p ⬍ 0.001 when compared with untreated cells. C, the extent of apoptosis (black squares) or necrosis (white squares) (a) and overall cell viability (b) were determined in cells incubated in ryanodine, BAPTA, or heparin (all for 15 h). Data are given as means ⫾ S.E. (n ⫽ 6) and were obtained from ⬎800 cells in each experiment. In a: *, p ⬍ 0.005 when compared with CHOWT; in b: *, p ⬍ 0.005 when compared with CHOWT and CHOL cells and **, p ⬍ 0.005 when compared with CHOWT and CHOH cells.

Ca2⫹ flux measured in CHOhRyR2, which is clearly apparent in Fig. 2A, arose as a consequence of aberrant Ca2⫹ efflux from the ER via hRyR2 and that steady-state resting [Ca2⫹]c was achieved by a concordant increase in Ca2⫹ sequestration via hRyR2-dependent up-regulation of the SR/ER Ca2⫹-ATPase in CHOhRyR2 cells (29). We estimated the variability in the intracellular Ca2⫹ flux in resting (non-stimulated) CHOhRyR2 by calculating the F ratio (40). The F ratio (as defined under “Experimental Procedures”) numerically described the amplitude variability in resting cytoplasmic Ca2⫹ flux. Quantifying this amplitude variability of [Ca2⫹]c (Fig. 2B, panel b (40)) identified CHOWT and CHOhRyR2 to have very distinct F ratios, despite similar mean [Ca2⫹]c (Fig. 2B, panel a). Thus, similar resting [Ca2⫹]c masked the degree of variability in intracellular Ca2⫹ flux CHOhRyR2 cells. Furthermore, the magnitude of cytoplasmic Ca2⫹ flux correlated with the expression levels of hRyR2 and the rank order of variation in intracellular Ca2⫹ flux was CHOH ⬎ CHOL ⬎⬎ CHOWT (Fig. 2B, panel b). To investigate more precisely a causative role of hRyR2 in mediating the increased cytoplasmic Ca2⫹ flux in CHOhRyR2

28860


FIG. 4. Ca2ⴙ release through hRyR2 mediates cellular toxicity following agonist-stimulation. A and C, peak [Ca2⫹]c release was determined following specific activation of recombinant hRyR2 (4-CMC, 0.5 mM) (A) or endogenous IP3R (ATP, 1 mM) (C) in untreated cells (white bars) or those that had been preincubated in ryanodine (1 mM) (black bars) or heparin (5 mg/ml) (gray bars). Cells were transferred to KRH-Ca2⫹ immediately prior to Ca2⫹ measurement. Data represents mean [Ca2⫹]c ⫾ S.E. (n ⫽ 6) where *, p ⬍ 0.01 when compared with CHOWT and CHOL cells; **, p ⬍ 0.01 when compared with CHOWT and CHOH cells; and #, p ⬍ 0.005 within grouping. B and D, the extent of apoptosis (white squares) and necrosis (black squares) (a) and the overall cell viability (b) was determined 15 h after the addition of 4-CMC (B) or ATP (D) to untreated cells (white bars) or those that had been preincubated in ryanodine (1 mM) (black bars) or heparin (5 mg/ml) (gray bars). This time period (15 h) permitted the full development of apoptotic and necrotic phenotype following agonist-stimulation of cells (21). In a: *, p ⬍ 0.01 when compared with CHOWT and in b; *, p ⬍ 0.005 when compared with CHOWT and CHOL cells and **, p ⬍ 0.005 when compared with CHOWT and CHOH cells. Data (means ⫾ S.E.) was obtained from ⬎700 cells in each experiment (n ⫽ 6).

cells, we co-expressed FKBP12.6, a potent modulator of RyR2 function (27, 29) and determined its effects on Ca2⫹ handling and cellular phenotype in CHOhRyR2 cells. Polyclonal antiserum raised against FKBP12, a homologue of FKBP12.6 that does not functionally interact with RyR2 (29, 43, 44), crossreacted with FKBP12.6 owing to a high degree of sequence homology (85%) (Fig. 2C). FKBP12 exhibits an extensive tissue distribution (43) and was endogenously expressed in CHOWT and CHOhRyR2 cells. In contrast, FKBP12.6 is expressed at far lower levels in tissues (45) and was not detected in post-nuclear fractions obtained from these cells (Fig. 2C, left panel). High level expression of recombinant FKBP12.6 in CHOhRyR2 cells (Fig. 2C, middle panel) was associated with significantly decreased intracellular Ca2⫹ flux (Fig. 2, A and B, panel b) and a concordant restoration of CHOhRyR2 viability (Fig. 2D). In contrast, co-expression of FKBP12 (Fig. 2C, right panel) had little effect on the magnitude of cytoplasmic Ca2⫹ flux (Fig. 2, A and B, panel b) and did not affect cell viability (Fig. 2C) consistent with the absence of FKBP12-hRyR2 interaction. However, in complete agreement with ryanodine inhibition of hRyR2 (Fig. 1), FKBP12.6 expression was associated with a reduced necrosis in the few non-viable cells (⬃5% of the total population),

FIG. 5. Intracellular Ca2ⴙ release channels extensively co-localize within CHOhRyR2. The subcellular distribution of recombinant hRyR2 (red pixels) and endogenous IP3R (green pixels) were visualized in CHOH and CHOL cells as described under “Experimental Procedures.” Coincident pixels that could be directly superimposed appear yellow (merge). The bar represents 15 ␮m.

which persisted (Fig. 2D, panel a). Disruption of specific protein-protein interaction between FKBP12.6-hRyR2 using rapamycin (46) reversed the inhibitory effect of FKBP12.6 on


28861

FIG. 6. FKBP12.6 inhibition of hRyR2 protects CHOhRyR2 from agonist-induced cell death. A and C, peak [Ca2⫹]c release was determined following specific activation of recombinant RyR2 (4-CMC, 0.5 mM) (A) or endogenous IP3R (ATP, 1 mM) (C) in control cells (⫺) or following expression of FKBP12.6 (⫹) in the absence or presence of rapamycin (5 ␮M) (white and black bars, respectively). Rapamycin was added 15 h prior to measuring [Ca2⫹]. Cells were transferred to KRH-Ca2⫹ immediately prior to Ca2⫹ measurement. Data illustrate mean [Ca2⫹]c ⫾ S.E. (n ⫽ 6) where *, p ⬍ 0.005 when compared with CHOWT and CHOL cells and **, p ⬍ 0.01 when compared with CHOWT and CHOH cells. B and D, the extent of apoptosis (black squares) and necrosis (white squares) (a) and the overall cell viability (b) in CHOWT and CHOhRyR2 cells with (⫹) or without (⫺) expression of FKBP12.6 was determined 15 h after the addition of 4-CMC (B) or ATP (D) to trigger Ca2⫹ release through RyR2 or IP3R, respectively. Data (mean ⫾ S.E.) was obtained from at least 700 cells in each experiment (n ⫽ 6). In a: *, p ⬍ 0.01 when compared with CHOWT and in b: *, p ⬍ 0.005 when compared with CHOWT and CHOL cells and **, p ⬍ 0.01 when compared with CHOWT and CHOH cells.

hRyR2-mediated cytoplasmic Ca2⫹ flux (Fig. 2, A and B, panel b) and abolished the pro-survival effects of FKBP12.6 in CHOhRyR2 (Fig. 2D). Interestingly, the restoration of cell viability following co-expression of FKBP12.6 in CHOhRyR2 was associated with a marked potentiation of the ER Ca2⫹ store (Fig. 2B, panel c) suggesting that super-filling ER Ca2⫹ stores did not directly contribute to CHOhRyR2 cell death. However, there was close correlation between the magnitude of cytoplasmic Ca2⫹ flux and the extent of cell toxicity (R2 ⫽ 0.937), strongly indicating that increased cytoplasmic Ca2⫹ flux, independent of resting [Ca2⫹]c or ER Ca2⫹ load status, is a key determinant of CHOhRyR2 cell viability. Furthermore, the finding that FKBP12.6 expression had a marked pro-survival effect on CHOhRyR2 by inhibiting the increased cytoplasmic Ca2⫹ flux through hRyR2 may go some way to providing a mechanistic basis for the use of FKBP12.6-mediated strategies as a candidate therapy in preventing the abnormal SR Ca2⫹ efflux underpinning heart failure (47). Although it is widely accepted and clearly demonstrated in the present work that hRyR2 Ca2⫹ release is modulated by interaction with FKBP12.6, controversy exists as to the precise nature of RyR2-FKBP12 interaction (48, 49). In the present context of our CHOhRyR2 cell system, we show that FKBP12 does not interact functionally with hRyR2. We next investigated the effect of biochemical modulation of hRyR2-mediated cytoplasmic Ca2⫹ flux on cell phenotype. In agreement with the data obtained following co-expression of FKBP12.6 in CHOhRyR2 cells, ryanodine-mediated inhibition of recombinant hRyR2 did not affect [Ca2⫹]c (Fig. 3B, panel a), yet significantly reduced cytoplasmic Ca2⫹ flux (Fig. 3, A and B, panel b) and restored cellular viability (Fig. 3C, panel b) by prevention of necrotic cell death (Fig. 3C, panel a). This prosurvival effect of ryanodine in CHOhRyR2 was associated with augmentation of the ER Ca2⫹ capacity (Fig. 3B, panel c). We used BAPTA to directly investigate the phenotypic effects of suppressing cytoplasmic Ca2⫹ flux in CHOhRyR2 and CHOWT

cells. However, because normal cell processes are dependent on coordinated fluctuations in [Ca2⫹]c, BAPTA treatment (⬎0.5 ␮M) was associated with a dose-dependent toxicity (data not shown), and thus lower concentrations were used (0.1 ␮M), because this adequately suppressed cytoplasmic Ca2⫹ fluctuations (Fig. 3A) yet did not adversely affect the cell viability during these experiments. Incubation of cells with BAPTA-AM (0.1 ␮M, 15 h), a potent inhibitor of apoptosis (50) (Fig. 3C, panel a), decreased resting [Ca2⫹]c and ER Ca2⫹ load (Fig. 3B, panels a and c, respectively), markedly suppressed intracellular Ca2⫹ flux (Fig. 3, A and B, panel b) and restored CHOhRyR2 cell viability (Fig. 3C, panel b). Inhibition of endogenously expressed IP3R following cytoplasmic loading of heparin had little effect on resting [Ca2⫹]c, cytoplasmic Ca2⫹ flux, and ER Ca2⫹ load (Fig. 3B, panels a– c, respectively), and accordingly, CHOhRyR2 viability remained low (Fig. 3C). Taken together, our findings do not predict a clear role for mean resting [Ca2⫹]c or altered ER Ca2⫹ load in the decreased cell viability of CHOhRyR2 but strongly indicate that CHOhRyR2 toxicity is underscored by increased hRyR2-mediated cytoplasmic Ca2⫹ flux. Cells are tolerant of moderate cytoplasmic Ca2⫹ fluxes, which encrypt homeostatic signaling events (7, 8), and the dramatic loss of cell viability in CHOhRyR2 highlights the extent to which expression of dysregulated hRyR2 perturbs the normal mode of Ca2⫹ signaling. Apoptosis Is Triggered following Agonist Activation of hRyR2 or via Functional Interaction with Endogenous IP3R—Our results point to a pivotal role of increased hRyR2 Ca2⫹ release in initiating necrotic cell death in resting CHOhRyR2, and hence we investigated the phenotype resulting from activation of hRyR2 in agonist-stimulated cells. Addition of 4-CMC (0.5 mM), a potent and specific activator of hRyR2 Ca2⫹ release (51), increased [Ca2⫹]c in CHOhRyR2, and the magnitude of Ca2⫹ release was proportional to the levels of hRyR2 expressed in each cell type (CHOH ⬎ CHOL) (Fig. 4A). No Ca2⫹ was released in CHOWT following addition of 4-CMC (Fig. 4A), confirming

28862


that CHOWT does not express functional levels of RyR, and CHOWT cell viability remained high (Fig. 4C). Ryanodine treatment abolished 4-CMC-induced Ca2⫹ release in CHOhRyR2 populations, whereas inhibition of endogenous IP3R using heparin had no effect on Ca2⫹ release triggered by 4-CMC (Fig. 4A). This moderate increase in [Ca2⫹]c upon 4-CMC addition to CHOH and CHOL cells (0.32 ⫾ 0.04 and 0.17 ⫾ 0.03 ␮M, respectively) was associated with a profound decrease in cellular viability via apoptosis (Fig. 4B). Ryanodine inhibition of hRyR2 prevented apoptosis, whereas heparin had no effect on 4-CMC-induced cellular toxicity (Fig. 4B, gray bars) highlighting a role for hRyR2 Ca2⫹ release in determining the extent and mode of cell death in CHOhRyR2. The initiation of apoptotic pathways following the activation of skeletal muscle RyR (RyR1) in heterologous expression systems has been reported (21), but our data highlight a fundamental switch in cell death mechanisms upon agonist-induced Ca2⫹ mobilization. Although apoptosis and necrosis are proposed to share common initiating programs (50, 52), when taken together our data provide strong evidence that different modes of Ca2⫹ release in resting (increased cytoplasmic Ca2⫹ flux) and activated cells (“global” Ca2⫹ mobilization) initiate distinct cell death pathways (necrosis and apoptosis, respectively) in CHOhRyR2 cells. It was necessary to determine whether the profound loss of CHOhRyR2 viability following 4-CMC addition was due to (a) the mode of Ca2⫹ mobilization or (b) the magnitude of Ca2⫹ release from intracellular stores. To address this issue, ATP (1 mM) was used to activate IP3R-mediated Ca2⫹ release via the purinergic (P2Y) receptor generation of IP3 (38). Unlike 4-CMC, which only elicited Ca2⫹ release in CHOhRyR2 cells, ATP elicited a large [Ca2⫹]c increase in CHOWT (0.71 ⫾ 0.08 ␮M), CHOH (0.92 ⫾ 0.11 ␮M), and CHOL (0.88 ⫾ 0.09 ␮M) (Fig. 4C). Note that the ATP-induced Ca2⫹ release was augmented in CHOhRyR2 cells when compared with CHOWT cells, in agreement with the finding that the expression of hRyR2 augments ER Ca2⫹ load (Fig. 2B, panel c). Importantly, in CHOWT cells maintained in nominal free extracellular Ca2⫹, ATP addition did not decrease cell viability, whereas a similar magnitude of Ca2⫹ release in CHOhRyR2 produced significant cell death via apoptosis, the extent of which was proportional to the expression levels of hRyR2 (Fig. 4D). Cytoplasmic loading of heparin abolished IP3R-dependent Ca2⫹ release in all cells (Fig. 4C) and prevented cell death in CHOhRyR2 (Fig. 4D) indicating that the ATP-induced toxicity in CHOhRyR2 cells required the activation of endogenous IP3R. Importantly, following incubation of CHOhRyR2 cells with ryanodine, the magnitude of ATP-induced Ca2⫹ release was decreased (Fig. 4C) and resulted in a remarkable preservation of cell viability (Fig. 4D). These results raise the intriguing possibility that the pronounced loss of viability in CHOhRyR2 following IP3R-activation is specifically mediated by Ca2⫹ release from ryanodine-sensitive stores via hRyR2. In accord with this hypothesis, this loss did not occur in RyR-deficient CHOWT cells where viability remained intact following 4-CMC or ATP addition (Fig. 4, A–D). There was extensive intracellular co-localization between recombinant hRyR2 and endogenous IP3R in CHOH and CHOL cells (90 ⫾ 7% and 94 ⫾ 5% coincidence (yellow), respectively (n ⫽ 6)) with both Ca2⫹ release channels displaying a perinuclear latticelike distribution characteristic of the ER network (Fig. 5). It has been shown that spatially coincident IP3R and RyR Ca2⫹ release channels participate in functional “cross-talk” (53, 54), and indeed, our results suggest that activation of IP3R (heparin-sensitive) triggered Ca2⫹ release via recombinant hRyR2 (ryanodine-sensitive), which critically modulated the CHOhRyR2 phenotype. These data also provided compelling evidence that the loss of cell viability in agonist-stimulated

FIG. 7. Decreased cellular proliferation in CHOhRyR2 populations can be restored by hRyR2 inhibition. A, the proliferation rates of CHOWT (open circle), CHOH (closed circle), and CHOL (square) were measured using AlamarBlue fluorescence at 590 nm (F590 nm; see “Experimental Procedures”). a, untreated cells; b, cells cultured in medium supplemented with ryanodine (1 mM); c, cells co-expressing FKBP12; d, cells co-expressing FKBP12.6. Data are plotted as means ⫾ S.E. (n ⫽ 6); *, p ⬍ 0.01 when compared with CHOWT and CHOL cells and **, p ⬍ 0.01 when compared with CHOWT and CHOH cells. B, AlamarBlue fluorescence (F590) in untreated cells (a), cells incubated with ryanodine (1 mM) (b), or cells co-expressing FKBP12 (c) or FKBP12.6 (d) was determined in parallel with hemocytometric measurement of cell number. CHOWT, CHOH, and CHOL are represented by open circles, closed circles, and squares, respectively. The ratios of F590:cell number were normalized in each instance to data obtained from CHOWT cells on day 1 (assigned 1) and are plotted as means ⫾ S.E. (n ⫽ 4).

CHOhRyR2 was underpinned by the mode of Ca2⫹ release (i.e. via hRyR2) rather than the magnitude of Ca2⫹ release. We currently do not know the precise mechanism by which activation of IP3R triggers hRyR2 Ca2⫹ release in these cells, or conversely, how activation of recombinant hRyR2 fails to trigger Ca2⫹ release via IP3R, but it is likely that cellular machinery required to permit functional interaction between these Ca2⫹ release channels (25, 53, 54) are present in CHOhRyR2. Clearly, a more refined analysis of sub-cellular Ca2⫹ mobilization is required to more fully elucidate the molecular basis of this phenomenon. We demonstrated a profound effect of FKBP12.6, but not FKBP12, in restoring the viability of resting CHOhRyR2 cells via

hRyR2 Interaction with FKBP12.6 Prevents Cell Death the specific modulation of hRyR2 Ca2⫹ release (Fig. 2), and thus we investigated whether the beneficial effects of FKBP12.6 persisted in agonist-stimulated CHOhRyR2 cells. In CHOWT, co-expression of FKBP12.6 did not affect Ca2⫹ release triggered by 4-CMC or ATP (Fig. 6, A and C, respectively), confirming that FKBP12.6 did not functionally interact with endogenous IP3R and had no impact on viability or mode of cell death (Fig. 6B, panels b and a, respectively). In contrast, FKBP12.6 inhibited 4-CMC or ATP-triggered Ca2⫹ release in CHOhRyR2 cells (Fig. 6, A and C, respectively) and preserved CHOhRyR2 cell viability by inhibiting apoptosis (Fig. 6, B and D). Rapamycin antagonized the beneficial effects of FKBP12.6 on intracellular Ca2⫹ release (Fig. 6, A and C) and CHOhRyR2 viability (Fig. 6, B and D) further emphasizing that the effects of FKBP12.6 on cell phenotype were mediated by specific protein-protein interactions with hRyR2. Abnormal Intracellular Ca2⫹ Handling Mediates Decreased CHOhRyR2 Proliferation—Finally, we investigated the effects of the altered Ca2⫹ handling in CHOhRyR2 on cellular function by measuring their proliferative capacity. Increased expression levels of hRyR2 were associated with decreased cellular proliferation (CHOH, 37 ⫾ 14%; CHOL, 59 ⫾ 10% when compared with CHOWT (100%) during the first 72 h) (Fig. 7A). However, although cells did not reach confluency during the experiment, the rate of proliferation slowed as the cell density increased (days 1– 4) (Fig. 7A, panels a– d), suggesting that the proliferation of CHOWT and CHOhRyR2 was sensitive to the ambient cellular environment. Modulation of hRyR2 by ryanodine (Fig. 7A, panel b) and FKBP12.6 (Fig. 7A, panel d) restored CHOhRyR2 cellular proliferation rates to those measured in CHOWT cells (Fig. 7B). Expression of FKBP12 had no effect on the cell growth (Fig. 7A, panel c). Incubation of cells with heparin or BAPTA (even at 0.1 ␮M) for the duration of the experiments resulted in total cell death (data not shown), presumably due to the sustained inhibition of normal homeostatic signaling events, which are dependent on endogenous IP3R Ca2⫹ release (1). These results indicate that biochemical (ryanodine) or physiological (FKBP12.6) modulation of hRyR2, which inhibits pathological fluxes in cytoplasmic Ca2⫹ (Figs. 2–3), restored normal cell proliferation in populations of CHOhRyR2. Thus, an important demonstration in the present study is that the co-expression of recombinant FKBP12.6 was necessary and sufficient to regulate RyR2, whereas FKBP12 did not interact with RyR2 and had no discernible effect on cellular phenotype in our model. Although we cannot exclude the possibility that the decreased proliferation of CHOhRyR2 reflected decreased ATP availability due to the energy demands of continuously pumping Ca2⫹ from the cytoplasm back into the ER, we propose that ATP depletion is not causative of abnormal phenotype. The ratio of AlamarBlue fluorescence, a robust indicator of cellular metabolism to actual cell number was similar in all cells at all time points directly indicating that the expression of hRyR2 (Fig. 7B, panel a) or following its modulation of hRyR2 using ryanodine (Fig. 7B, panel b) or co-expression of FKBP12 or FKBP12.6 (Fig. 7B, panels c and d, respectively) did not compromise the metabolic activities of CHOH and CHOL cells, which were indistinguishable from those of CHOWT (Fig. 7B, panels a– d). Second, expression-profiling analysis of CHOH demonstrated that increased Ca2⫹ flux mediates precise Ca2⫹dependent (ryanodine-sensitive) transcriptional alterations in intracellular signaling networks (55). In view of our data, we propose that increased Ca2⫹ flux results in the specific modulation of signaling pathways, which are directly causative of the CHOhRyR2 phenotype. The finding that the expression levels of hRyR2 determined

28863

the extent of intracellular Ca2⫹ dysregulation and the degree of phenotypic modulation is an important conclusion of the present work. The CHOhRyR2 cell model represents a powerful platform to study the role of RyR2 on cellular function in an FKBP12.6-deficient context. Furthermore, our demonstration that reconstitution of the RyR2䡠FKBP12.6 complex in intact cells inhibited aberrant ER-cytoplasmic Ca2⫹ flux and resulted in complete restoration of normal cellular phenotype represents a significant advance in the characterization of RyRmediated signaling in cell function. It would be interesting to determine whether similar phenotypes are established as a consequence of defective functionality of other Ca2⫹ regulatory molecules or whether those elucidated in this study represent RyR-specific alterations in cellular phenotype. However, in the present context, our data suggest that increased Ca2⫹ flux in resting CHOhRyR2 precisely modulates signaling programs involved in cell viability and proliferation, and our cell model could enable the determination of the impact of defective RyRdriven Ca2⫹ signaling on other diverse cellular processes. Acknowledgments—We thank S. Zissimopoulos for the FKBP12 and FKBP12.6 plasmids and M. B. Hallett for helpful discussions. REFERENCES 1. Berridge, M. J., Lipp, P., and Bootman, M. D. (2000) Nat. Rev. Mol. Cell. Biol. 1, 11–21 2. Anger, M., Lompre, A. M., Vallot, O., Marotte, F., Rappaport, L., and Samuel, J. L. (1998) Circulation 98, 2477–2486 3. Marks, A. R. (2000) Circ. Res. 87, 8 –11 4. Chan, S. L., Mayne, M., Holden, C. P., Geiger, J. D., and Mattson, M. P. (2000) J. Biol. Chem. 275, 18195–18200 5. Somlo, S., and Ehrlich, B. (2001) Curr. Biol. 11, 356 –360 6. Vanden Abeele, F., Skryma, R., Shuba, Y., Van Coppenolle, F., Slomianny, C., Roudbaraki, M., Mauroy, B., Wuytack, F., and Prevarskaya, N. (2002) Cancer Cell 1, 169 –179 7. Dolmetsch, R. E., Lewis, R. S., Goodnow, C. C., and Healy, J. I. (1997) Nature 386, 855– 858 8. Dolmetsch, R. E., Pajvani, U., Fife, K., Spotts, J. M., and Greenberg, M. E. (2001) Science 294, 333–339 9. Hajnoczky, G., Csordas, G., Madesh, M., and Pacher, P. (2000) Cell Calcium 28, 349 –363 10. Ma, T. S., Mann, D. L., Lee, J. H., and Gallinghouse, G. J. (1999) Cell Calcium 26, 25–36 11. Nguyen, H. N., Wang, C., and Perry, D. C. (2002) Brain Res. 924, 159 –166 12. Pinton, P., Ferrari, D., Rapizzi, E., Virgilio, F. D., Pozzan, T., and Rizzuto, R. (2001) EMBO J. 20, 2690 –2701 13. Oyadomari, S., Takeda, K., Takiguchi, M., Gotoh, T., Matsumoto, M., Wada, I., Akira, S., Araki, E., and Mori, M. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 10845–10850 14. Kitao, Y., Ozawa, K., Miyazaki, M., Tamatani, M., Kobayashi, T., Yanagi, H., Okabe, M., Ikawa, M., Yamashima, T., Stern, D. M., Hori, O., and Ogawa, S. (2001) J. Clin. Invest. 108, 1439 –1450 15. Fern, R. (1998) J. Neurosci. 15, 7232–7243 16. Franklin, J. L., and Johnson, E. M. (1992) Trends Neurosci. 12, 501–508 17. Rizzuto, R., Pinton, P., Carrington, W., Fay, F. S., Fogarty, K. E., Lifshitz, L. M., Tuft, R. A., and Pozzan, T. (1998) Science 280, 1763–1766 18. Tinel, H., Cancela, J. M., Mogami, H., Gerasimenko, J. V., Gerasimenko, O. V., Tepikin, A. V., and Petersen, O. H. (1999) EMBO J. 18, 4999 –5008 19. Jayaraman, T., and Marks, A. R. (1997) Mol. Cell. Biol. 17, 3005–3012 20. Blackshaw, S., Sawa, A., Sharp, A. H., Ross, C. A., Snyder, S. H., and Khan, A. A. (2000) FASEB J. 14, 1375–1379 21. Pan, Z., Damron, D., Nieminen, A. L., Bhat, M. B., and Ma, J. (2000) J. Biol. Chem. 275, 19978 –19984 22. Xu, L., and Meissner, G. (1998) Biophys. J. 75, 2302–2312 23. Koizumi, S., Lipp, P., Berridge, M. J., and Bootman, M. D. (1999) J. Biol. Chem. 274, 33327–33333 24. Miyakawa, T., Mizushima, A., Hirose, K., Yamazawa, T., Bezprozvanny, I., Kurosaki, T., and Iino, M. (2001) EMBO J. 20, 1674 –1680 25. Mackrill, J. J. (1999) Biochem. J. 337, 345–361 26. Marx, S. O., Reiken, S., Hisamatsu, Y., Gaburjakova, M., Gaburjakova, J., Yang, Y. M., Rosemblit, N., and Marks, A. R. (2001) J. Cell Biol. 154, 699 –708 27. Marx, S. O., Reiken, S., Hisamatsu, Y., Jayaraman, T., Burkhoff, D., Rosemblit, N., and Marks, A. R. (2000) Cell 101, 365–376 28. Yano, M., Ono, K., Ohkusa, T., Suetsugu, M., Kohno, M., Hisaoka, T., Kobayashi, S., Hisamatsu, Y., Yamamoto, T., Kohno, M., Noguchi, N., Takasawa, S., Okamata, H., and Matsuzaki, M. (2000) Circulation 102, 2131–2136 29. George, C. H., Sorathia, R., Bertrand, B. M. A., and Lai, F. A. (2003) Biochem. J. 370, 579 –589 30. Tunwell, R. E. A., Wickenden, C., Bertrand, B. M. A., Shevchenko, V. I., Walsh, M. B., Allen, P. D., and Lai, F. A. (1996) Biochem. J. 318, 477– 487 31. Mackrill, J. J., Challiss, R. A. J., O’Connell, D. A., Lai, F. A., and Nahorski, S. R. (1997) Biochem. J. 327, 251–258 32. Mackrill, J. J., O’Driscoll, S., Lai, F. A., and McCarthy, T. V. (2001) Biochem. Biophys. Res. Commun. 285, 52–57

28864


33. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., and Smith, J. A. (1994) Current Protocols in Molecular Biology, Wiley and Sons, New York 34. Matute-Bello, G., Liles, W. C., Steinberg, K. P., Kiener, P. A., Mongovin, S., Chi, E. Y., Jonas, M., and Martin, T. R. (1999) J. Immunol. 15, 2217–2225 35. Li, Y., Kranias, E. G., Mignery, G. A., and Bers, D. M. (2002) Circ. Res. 90, 309 –316 36. Hallett, M. B., Davies, E. V., and Pettit, E. J. (1996) Comp. Methods Enzymol. 9, 591– 606 37. Monkawa, T., Miyawaki, A., Sugiyama, T., Yoneshima, H., Yamamoto-Hino, M., Furuichi, T., Saruta, T., Hasegawa, M., and Mikoshiba, K. (1995) J. Biol. Chem. 270, 14700 –14704 38. Rhee, S. G. (2001) Annu. Rev. Biochem. 70, 281–312 39. Okada, C. Y., and Rechsteiner, M. (1982) Cell 29, 33– 41 40. Armitage, P., and Berry, G. (1994) Statistical Methods in Medical Research, 3rd Ed., Blackwell, Oxford 41. Bhat, M. B., Hayek, S. M., Zhao, J., Zang, W., Takeshima, H., Weir, W. G., and Ma, J. (1999) Biophys. J. 77, 808 – 816 42. Tong, J., McCarthy, T. V., and MacLennan, D. H. (1999) J. Biol. Chem. 274, 693–702 43. Timerman, A. P., Onoue, H., Xin, H. B., Barg, S., Copello, J., Wiederrecht, G., and Fleischer, S. (1996) J. Biol. Chem. 271, 20385–20391 44. Carmody, M., Mackrill, J. J., Sorrentino, V., and O’Neill, C. (2001) FEBS Lett. 505, 97–102

45. Xin, H. B., Senbonmatsu, T., Cheng, D. S., Wang, Y. X., Copello, J. A., Ji, G. J., Collier, M. L., Deng, K. Y., Jeyakumar, L. H., Magnuson, M. A., Inagami, T., Kotlikoff, M. I., and Fleischer, S. (2002) Nature 416, 334 –337 46. Kaftan, E., Marks, A. R., and Ehrlich, B. E. (1996) Circ. Res. 78, 990 –997 47. Yano, M., Kobayashi, S., Kohno, M., Doi, M., Tokuhisa, T., Okuda, S., Suetsugu, M., Hisoaka, T., Obayashi, M., Ohkusa, T., Kohno, M., and Matsuzaki, M. (2003) Circulation 107, 477– 484 48. Jeyakumar, L. H., Ballester, L., Cheng, D. S., McIntyre, J. O., Chang, P., Olivey, H. E., Rollins-Smith, L., Barnett, J. V., Murray, K., Xin, H. B., and Fleischer, S. (2001) Biochem. Biophys. Res. Commun. 281, 979 –986 49. Shou, W., Aghdasi, B., Armstrong, D. L., Guo, Q., Bao, S., Charng, M. J., Mathews, L. M., Scheider, M. D., Hamilton, S. L., and Matzuk, M. M. (1998) Nature 391, 489 – 492 50. Thibaut, S., Bourre, L., Hernot, D., Rousset, N., Lajat, Y., and Patrice, T. (2002) Apoptosis 7, 99 –106 51. Zorzato, F., Scutari, E., Tegazzin, V., Clementi, E., and Treves, S. (1993) Mol. Pharmacol. 44, 1192–1201 52. Xu, K., Tavernarakis, N., and Driscoll, M. (2001) Neuron 31, 957–971 53. Haak, L. L., Song, L. S., Molinski, T. F., Pessah, I. N., Cheng, H., and Russell, J. T. (2001) J. Neurosci. 21, 3860 –3870 54. White, C., and McGeown, J. G. (2002) J. Physiol. 542, 725–733 55. George, C. H., and Lai, F. A. (2003) Biophys. J. 84, 390a 56. Currie, S., and Smith, G. (1999) Cardiovasc. Res. 41, 135–146