Nature Cell Biology
5, 1051 - 1061 (2003)
Published online: 9 November 2003; | doi:10.1038/ncb1063
There is an Erratum (January 2004) associated with this Article.
Cytochrome c binds to inositol (1,4,5) trisphosphate receptors, amplifying calcium-dependent apoptosisDarren Boehning1, 5, Randen L. Patterson1, 5, Leela Sedaghat1, Natalia O. Glebova1, Tomohiro Kurosaki4
& Solomon H. Snyder1, 2, 31 Department of Neuroscience, Johns Hopkins University School of Medicine, 725 N. Wolfe Street, Baltimore, Maryland 21205, USA. 2 Departments of Pharmacology and Molecular Sciences, Johns Hopkins University School of Medicine, 725 N. Wolfe Street, Baltimore, Maryland 21205, USA. 3 Departments of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, 725 N. Wolfe Street, Baltimore, Maryland 21205, USA. 4 Department of Molecular Genetics, Institute for Liver Research, Kansai Medical University, and Laboratory for Lymphocyte Differentiation, RIKEN Research Center for Allergy and Immunology, Moriguchi 570-8506, Japan. 5 These authors contributed equally to this work.
Correspondence should be addressed to Solomon H. Snyder ssnyder@jhmi.eduMitochondrial cytochrome c release and inositol (1,4,5) trisphosphate receptor (InsP3R)-mediated calcium release from the endoplasmic reticulum mediate apoptosis in response to specific stimuli. Here we show that cytochrome c binds to the InsP3R during apoptosis. Addition of 1 nM cytochrome c blocks calcium-dependent inhibition of InsP3R function. Early in apoptosis, cytochrome c translocates to the endoplasmic reticulum where it selectively binds InsP3R, resulting in sustained, oscillatory cytosolic calcium increases. These calcium events are linked to the coordinate release of cytochrome c from all mitochondria. Our findings identify a feed-forward mechanism whereby early cytochrome c release increases InsP3R function, resulting in augmented cytochrome c release that amplifies the apoptotic signal.Calcium signalling mediates many cellular functions, including programmed cell death through influences on free radical production, organelle disruption and proteolytic enzyme activation1,
2. Increases of cytosolic calcium during cell death can arise from several sources, including damage to external membranes, hyper-activation of ligand gated ion channels or release from calcium stores in the endoplasmic reticulum3,
4. During cell death, mitochondrial calcium overload activates the permeability transition pore (PTP), resulting in transient mitochondrial depolarization and decreased ATP production. PTP gating may also cause release of mitochondrial proteins that activate apoptotic pathways5. One of these proteins, cytochrome c, is a small haem-containing enzyme of the electron transport chain. In many models of apoptosis, cytochrome c is released from the inter-membrane space of mitochondria into the cytosol5. Cytochrome c release can be rapid and coordinated between all mitochondria through an unknown mechanism6. Released cytochrome c then binds to apoptotic protease activating factor-1 (APAF-1), which recruits and activates caspase-9 to form the apoptosome5. Caspase-9 activation results in the cleavage and activation of caspases-3 and -7, initiating a proteolytic cascade that ultimately results in cell death.
Recent research reveals a privileged communication between mitochondria and endoplasmic reticulum calcium stores7,
8. Apoptotic stimuli such as staurosporine (STS) and ceramide elicit InsP3R-induced increases in mitochondrial calcium, resulting in PTP-mediated decrease in mitochondrial membrane potential and release of cytochrome c4.
Using a combination of yeast two-hybrid analysis and biochemical techniques, we now demonstrate the direct binding of cytochrome c to InsP3R during apoptosis. As little as 1 nM cytochrome c abolishes calcium-mediated inhibition of InsP3-mediated calcium release. In early stages of apoptosis, cytochrome c translocates from mitochondria to endoplasmic reticulum-containing fractions, binding specifically to InsP3R. Cytochrome c translocation to the endoplasmic reticulum is altered in DT40 cells that lack all three InsP3R isoforms. Furthermore, imaging studies reveal large, sustained, increases in cytosolic calcium during early apoptosis that are concomitant with InsP3R-cytochrome c interactions, as revealed by fluorescence resonance energy transfer (FRET). These cytochrome c−InsP3R interactions provide a molecular mechanism whereby InsP3R and mitochondria can work in concert to execute the apoptotic cell death programme.
Results
Cytochrome c binds InsP3R, altering function
To identify intracellular proteins that interact with InsP3R, we conducted a yeast two-hybrid screen of a rat brain cDNA library with the type-1 InsP3R as bait (Fig. 1a). Cytochrome c was identified as one of the interacting proteins. Transformation of yeast with varying truncations of InsP3R revealed that cytochrome c binds selectively to amino acids 2589−2749 (the carboxy-terminal tail) of the InsP3R. To confirm the binding of these two proteins, we used a glutathione S-transferase (GST) pulldown assay in which purified cytochrome c was incubated with GST−InsP3R fusion proteins (Fig. 1b). As in the yeast two-hybrid analysis, only the C-terminal tail bound to cytochrome c. Full-length InsP3R from crude cerebellum lysate and InsP3R purified to homogeneity bound to cytochrome c−agarose (Fig. 1c), indicating a direct association.
 | | |
InsP3R activity is biphasically regulated by cytosolic calcium levels. At resting calcium concentrations ( 100 nM), increases in calcium augment InsP3-induced calcium release9,
10. In contrast, concentrations of calcium greater than 5  M diminish the ability of InsP3 to release calcium, presumably representing a safety mechanism to prevent a feed-forward excessive release of calcium9,
10. To examine the influences of cytochrome c on InsP3R function, we measured calcium release from microsomal vesicles prepared from COS cells cotransfected with type-I InsP3R and the sarcoplasmic-endoplasmic reticulum calcium ATPase (SERCA; Fig. 1d)11. At the optimal calcium concentration for InsP3R activity (250 nM), concentration−response relationships for InsP3-mediated 45Ca2+ release were the same in the presence or absence of cytochrome c. At 1 nM cytochrome c, however, calcium-induced inhibition of InsP3R calcium release was abolished (Fig. 1e). Cytochrome c had no effect on SERCA activity in the absence of InsP3 using this assay (data not shown). High concentrations of cytochrome c (100 nM) were functionally indistinguishable from 1 nM cytochrome c, whereas 100 pM cytochrome c had no measurable effect on channel function (data not shown).
Cytochrome c binds to InsP3R during apoptosis
Most InsP3Rs in cells are associated with the endoplasmic reticulum, whereas cytochrome c is a classic mitochondrial constituent. We hypothesized that cytochrome c translocates to the endoplasmic reticulum during apoptosis to interact with InsP3R. To explore this possibility, we induced apoptosis in HeLa and PC12 cells with the broad-spectrum kinase inhibitor STS (STS) and monitored levels of cytochrome c by subcellular fractionation. We examined fractions containing mitochondria and heavy endoplasmic reticulum (P2), fractions containing microsomal (light) endoplasmic reticulum (P3) and the cytosolic (S3) fraction (Fig. 2a). At 4 and 6 h after STS treatment we observed a decrease of cytochrome c immunoreactivity in the P2 fraction of HeLa cells that was matched by corresponding increases in the P3 fraction, with little cytochrome c detected in the S3 fraction. Extended exposures revealed small but detectable increases in cytochrome c in the S3 fraction in response to STS (data not shown). The crude microsomal fraction P3 contains light endoplasmic reticulum, but can be also contaminated by other cellular constituents. To ensure that the P3 fraction was devoid of mitochondria, we monitored the mitochondrial enzyme cytochrome c oxidase (CytOx), which we found exclusively in the P2 fraction at all times. Haem oxygenase-2 (HO2) is a microsomal constituent12 that we detected in P2 and P3, but not S3, demonstrating that the S3 fraction is devoid of endoplasmic reticulum membranes. Lactate dehydrogenase (LDH), a cytoplasmic marker, is evident in S3, but not P2 and P3, with no time-dependent alterations. CytOx, HO2 and LDH immunoreactivity were also used as loading controls for each experiment.
| | Figure 2. Cytochrome c translocates to the endoplasmic-reticulum-binding InsP3R during apoptosis in HeLa cells. | | |  | (a) Cytochrome c translocates from the mitochondria (P2) to the endoplasmic reticulum (P3) fraction during apoptosis. Apoptosis was induced with 1 M STS for 0, 4 or 6 h. Cells were then harvested, fractionated and processed. CytOx (mitochondria), HO2 (endoplasmic reticulum) and LDH (cytosol) were used to check purity of the fractions and to ensure equal protein distribution in all lanes. (b) P3 fractions (100 g) from STS-treated HeLa cells were immunoprecipitated with anti-InsP3R antibody, pre-immune serum or anti-SERCA2 antibody. Immunoprecipitates were then immunoblotted with anti-cytochrome c, -InsP3R and -SERCA antibodies. Cytochrome c coprecipitates in an STS-dependent manner only in InsP3R immunoprecipitates. (c) The cytosol-enriched (S3) fraction was used to measure caspase-3 activity (see Methods). Panels a−c are from the same experimental treatment, and are representative of four separate determinations. (d) Inhibiting protein synthesis eliminates cytochrome c immunoreactivity in the P3 fraction at 0 h. This experiment was performed as in a, except that cells were pre-incubated with 10 g ml-1 cycloheximide (CHX) for 2 h before STS treatment. Cycloheximide treatment does not interrupt cytochrome c translocation to the endoplasmic reticulum. Cycloheximide accelerates the appearance of cytochrome c in the cytosol (S3) fraction and is concurrently associated with a large increase in caspase activity relative to no CHX treatment, as shown in f. (e) Experiment was performed as in b, except that the P3 fraction used was from subcellular fractionation in d. Cycloheximide does not interrupt cytochrome c−InsP3R binding. (f) As was performed as in c, except that the S3 fraction used was from the subcellular fractionation in d. Panels d−f are from the same experimental treatment, and are representative of three separate determinations.
 Full Figure and legend (64K) |
| |
The time-dependent decrease of cytochrome c in P2 and corresponding increase in P3 implies a translocation of cytochrome c from mitochondria to endoplasmic reticulum during apoptosis. To ascertain whether cytochrome c in the endoplasmic reticulum is associated with InsP3R, we performed immunoprecipitations with an antibody to the type-1 InsP3R (Fig. 2b). We detected co-immunoprecipitation of the two proteins from the P3 fraction 4 h after STS treatment, with increased co-immunoprecipitation at 6 h. We did not observe co-immunoprecipitation of cytochrome c with SERCA, or when pre-immune InsP3R serum was used, demonstrating the specificity of this interaction. Similar experiments were conducted with the type-3 InsP3R, providing essentially the same results (data not shown). As a result of the poor quality of available type-2 InsP3R antibodies and the low endogenous levels of type-2 InsP3R in most tissues13, we have been unable to conduct experiments to monitor its interactions with cytochrome c. We confirmed that STS treatment induces apoptosis in our preparations by monitoring caspase-3-like catalytic activity in the S3 fraction at 4 and 6 h after STS treatment (Fig. 2c). DNA fragmentation was also observed at later times (12−24 h; data not shown). We performed similar experiments in HeLa cells using the physiological ligand C2-ceramide to induce apoptosis. Ceramide elicits the same effects as STS, with cytochrome c selectively partitioning to the P3 fraction at early times (see Supplementary Information, Fig. S1).
At the zero time point before STS treatment, we detected cytochrome c in the P3 fraction (Fig. 2a). As the P3 fraction is devoid of mitochondrial contamination, we wondered whether the cytochrome c in the P3 fraction at the zero time point might represent newly synthesized protein associated with ribosomal membranes. To explore this possibility, we treated HeLa cells with the protein synthesis inhibitor cycloheximide (Fig. 2d). Cycloheximide treatment abolished cytochrome c in the P3 fraction at the zero time point. This treatment also resulted in the appearance of cytochrome c at 4 and 6 h in the cytosolic S3 fraction, which is devoid of cytochrome c in the absence of cycloheximide treatment, suggesting that cycloheximide increases the susceptibility to STS-induced apoptosis in HeLa cells. This hypothesis is supported by the fourfold increase in cytosolic caspase activity in cycloheximide-treated preparations, compared with those treated only with STS (Fig. 2c, f). In cycloheximide- and STS-treated cells, InsP3R and cytochrome c also co-immunoprecipitated in an STS-dependent manner (Fig. 2e).
To ensure that the translocation of cytochrome c from mitochondria to the endoplasmic reticulum is a general phenomenon, we examined STS-induced apoptosis in PC12 cells (Fig. 3a). As in HeLa cells, PC12 cells manifest increased cytochrome c associated with the P3 fraction in response to STS, with a concomitant decrease in the P2 fraction. Similarly to HeLa cells, we did not detect cytochrome c release into the S3 fraction early in STS-induced apoptosis, and the distribution among subcellular fractions of InsP3R, HO2 and LDH was the same in PC12 cells as in HeLa cells. CytOx immunoreactivity was not present in the S3 or P3 fractions, and endoplasmic reticulum contamination was not observed in the S3 fraction. InsP3R and cytochrome c also co-immunoprecipitated from PC12 cell lysates in a STS-dependent manner (Fig.3b). PC12 cells demonstrated a time-dependent increase in caspase activity in cytosolic fractions (Fig. 2c), similar to HeLa cells.
| | |
The cytochrome c−InsP3R interaction suggests that during initial phases of apoptosis, cytochrome c is released from mitochondria to bind to InsP3R in the closely adjacent endoplasmic reticulum. This would diminish calcium-induced inhibition of InsP3-mediated release of calcium, presumably resulting in mitochondrial and cytosolic calcium overload. As increases in caspase activity occur within the time course of cytochrome c translocation to the P3 fraction, we examined whether the broad-spectrum caspase inhibitor carbobenzoxy-valyl-alanyl-aspartyl-[O-methyl]-fluoromethylketone (Z-VAD-FMK) affects cytochrome c translocation. Z-VAD-FMK completely inhibited caspase activity induced by STS in HeLa cells (Fig. 4b). However, Z-VAD-FMK has no effect on cytochrome c translocation to the P3 fraction (Fig. 4a, b), demonstrating that translocation is independent of caspase activation.
| | |
InsP3R can be cleaved by caspases and the calcium-sensitive proteolytic enzyme, calpain, and such cleavage may participate in apoptotic cell death14,
15. To ascertain whether cleavage of InsP3R by these enzymes occurs during the early time phase of apoptosis that we examined, we monitored InsP3R protein levels in HeLa cells treated with Z-VAD-FMK and the calpain inhibitor MDL-28170 (Fig. 4c). We detected only a small amount of the reported cleavage fragment of InsP3R15 in response to STS, and this was not blocked by calpain inhibition (Fig. 4c; indicated by an asterisk). Z-VAD-FMK blocked this limited degradation, consistent with caspase cleavage. Caspase-3 proteolytic activity was confirmed by monitoring cleavage of the known caspase substrate poly(ADP ribose) polymerase (PARP). We also examined the influence of caspase and calpain inhibitors in PC12 cells, exploring later as well as earlier times after STS treatment. Similarly to HeLa cells, InsP3R levels remained constant 4 and 8 h after STS treatment, whereas PARP cleavage was observed (Fig. 4d). At later times, we observed a progressive decrease in InsP3R, with almost complete absence at 24 and 48 h. Z-VAD-FMK partially reduced this decline, but we could not determine whether this effect reflects inhibition of enzymatic activity cleaving InsP3R or is secondary to prevention of apoptosis by Z-VAD-FMK. The calpain inhibitor failed to alter the decline of InsP3R at later times. As little to no InsP3R degradation is observed at the early times of STS-induced apoptosis, we speculate that InsP3R function is maintained during the early stages of apoptosis and that cytochrome c may be influencing calcium homeostasis in response to apoptotic stimuli.
The translocation of cytochrome c to the P3 fraction and its binding to InsP3R imply that endoplasmic reticulum-associated InsP3R is responsible for the translocation. To examine this hypothesis directly, we used DT40 B lymphocytes, in which all three isoforms of the InsP3R are deleted16 (Fig. 5a, b). Although these cells have been demonstrated to be fairly resistant to apoptosis16, they still provide an excellent model for this study as we are examining the redistribution of cytochrome c within the population of cells. Using a physiological stimulus, we induced apoptosis by activating the B-cell receptor with anti-immunoglobulin M (IgM) for 24 h at concentrations of 2 or 10 g ml-1. As in HeLa and PC12 cells, anti-IgM treatment of wild-type DT40 cells elicited a decrease of cytochrome c in the P2 fraction and an increase in P3 fraction (Fig. 5a). Because the CytOx antibody used in this study does not recognize the chicken isoform, we used antibodies against the avian mitochondrial protein prohibitin as a marker for mitochondria17. Prohibitin was abundant in the P2 fraction, but not in the P3 or S3 fractions. SERCA and InsP3R were abundant in the P2 fraction, with lower levels in the P3 fractions, whereas LDH was restricted to the S3 fraction. In DT40 cells lacking InsP3R, anti-IgM depleted cytochrome c from the P2 fraction and resulted in its appearance in S3 but not P3. As in HeLa and PC12 cells, cytochrome c was found in the P3 fraction of untreated triple-InsP3R knockout cells (Fig.5b). We could not ascertain whether this was the result of new protein synthesis because translational inhibitors also blocked anti-IgM-induced apoptosis (data not shown). Thus, in cells lacking InsP3R, cytochrome c fails to translocate to the endoplasmic reticulum but instead appears in the cytoplasm, establishing that InsP3R mediates binding of cytochrome c to the endoplasmic reticulum.
| | |
Cytochrome c−InsP3R binding in intact cells
To evaluate cytochrome c−InsP3R interactions in intact cells, we used PC12 cells stably transfected with enhanced yellow fluorescent protein (EYFP)-labelled cytochrome c, then transiently transfected the cells with enhanced cyan fluorescent protein (ECFP) fused to the amino terminus of the InsP3R. If cytochrome c and the InsP3R come within sufficient proximity, ECFP (donor) emission at 480 nm will be quenched and EYFP (acceptor) emission at 535 nm will increase. Increases in the 535:480-nm ratio reflect a close association of the two proteins (less than 100 Å)18.
We observed a gradual increase of the emission ratio over time in response to STS, beginning at 1 h and extending through a 5-h time course (Fig. 6a−c). Surface plot display of the 535:480-nm ratio in two PC12 cells demonstrates the spatial increases in InsP3R-cytochrome c interactions (Fig. 6a). We observed these increases at very early times in extranuclear compartments of the cell periphery, which is consistent with mitochondrial localization in PC12 cells (Figs 6a and 7a). This increase in the 535:480-nm ratio varied among individual cells (Fig.6b), demonstrating that STS induction of apoptosis is asynchronous among a population of cells. The increase in the emission ratio, averaging approximately 0.3, is comparable with values obtained for numerous other interacting proteins19,
20. No increase was observed if STS was omitted (Fig. 6c), or in cells where cytosolic CFP (unattached to the InsP3R) was transfected (Fig. 6d). These experiments establish that cytochrome c and InsP3R are bound to each other during STS-induced apoptosis.
| | |
Cytochrome c−InsP3R binding and cytosolic calcium
To ascertain whether cytochrome c−InsP3R binding mediates apoptosis-associated intracellular calcium increases, we conducted simultaneous imaging of intracellular calcium and cytochrome c disposition (Fig. 7). In HeLa cells stably expressing YFP−cytochrome c and treated with STS, we observed gradual increases of calcium levels with an oscillation frequency of 20 min (Fig. 7b). At 270 min, calcium levels reached a plateau and no further oscillations were evident. This sustained increase in cytosolic calcium continued until 480 min without further oscillations, and no releasable calcium stores were observed after treatment with 1 M ionomycin at the 480-min timepoint (data not shown). To ensure that the observed oscillations were not caused by the destruction of the endoplasmic reticulum calcium stores by STS, we used thapsigargin at various time-points after treatment with STS, and do not see a marked decrease in thapsigargin-releasable endoplasmic reticulum calcium until 300 min, a time-point after which the oscillations are no longer evident (see Supplementary Information, Fig S3). Furthermore, these oscillations are not dependent on caspase activity, as HeLa cells pre-treated with caspase inhibitors displayed identical calcium oscillations in response to STS treatment (see Supplementary Information, Fig. S2).
Cytochrome c fluorescence was initially punctate and reflects localizations in mitochondria, as established by colocalization with the mitochondrial stain MitoTracker Red (Fig. 7a). Single-cell imaging revealed a spike of calcium immediately preceding a coordinated release of cytochrome c throughout the cell (Fig. 7c). After this spike, calcium levels remained at the pre-spike level indefinitely. Averaged traces from large numbers of PC12 cells stably expressing YFP−cytochrome c revealed a similar increase in intracellular calcium, as in HeLa cells, but with a much longer timescale (11.5 h) and less prominent oscillations (data not shown). Similarly to HeLa cells, imaging of individual PC12 cells revealed calcium spikes immediately preceding the coordinate cytochrome c release (Fig. 7d). Importantly, buffering of cytosolic calcium with BAPTA-AM in both HeLa and PC12 cells markedly reduced the release of cytochrome c from mitochondria (see Supplementary Information, Fig. S4), providing further evidence that increases in cytosolic calcium coordinate the release of cytochrome c from mitochondria.
To demonstrate that the interaction between cytochrome c and InsP3R is important in vivo for calcium release and apoptosis, we tested whether exogenously expressed cytochrome c-binding domain of InsP3R (amino acids 2589−2749) would function as a dominant-negative inhibitor of STS-induced calcium oscillations and cytochrome c release in HeLa cells. As a negative control, we also examined the effects of exogenously expressed amino acids 923−1581 of the InsP3R. We verified that the InsP3R peptide fragments were expressed by western blotting (Fig. 8a). HeLa cells overexpressing either of the peptides exhibited no change in calcium release in response to 10 M ATP when compared with control (Fig. 8b), demonstrating that endogenous InsP3R function is not affected. YFP−cytochrome c HeLa cells were transfected with each peptide before testing for changes in calcium responses and cytochrome c release in response to 1 M STS (Fig. 8c). Transfected cells were visualized by cotransfecting nuclear-targeted DsRed2. Expression of the C-terminal InsP3R peptide (amino acids 2589−2749) markedly reduced the frequency and magnitude of the calcium transients induced by STS. Calcium transients were not observed for 90 min in cells overexpressing the C terminus, whereas calcium transients began in control cells within minutes of STS exposure. Significant cytochrome c release was observed as early as 270 min in cells overexpressing the control peptide of the InsP3R. Cytochrome c release was not detected in cells overexpressing the C-terminal peptide earlier than 12 h (data not shown), demonstrating the C-terminal peptide markedly influences the kinetics of cytochrome c release in HeLa cells in response to STS treatment. The observed effects of the dominant-negative peptide may result from blocking cytochrome c-mediated activation of APAF-1. However, the peptide did not block cytochrome c-mediated caspase activation in cytosolic extracts (see Supplementary Information, Fig. S2a). Furthermore, broad-spectrum caspase inhibitors did not block translocation of cytochrome c to the InsP3R (Fig. 4a) or STS-induced calcium oscillations (see Supplementary Information, Fig. S2b). These results demonstrate that cytochrome c translocation to the endoplasmic reticulum and altered calcium homeostasis resulting in cell-wide cytochrome c release occur upstream of cytochrome c-mediated caspase activation.
| | |
Discussion
A major finding of our study is that cytochrome c binding to InsP3R prevents calcium-dependent inhibition of InsP3R function. The ability of low-micromolar concentrations of calcium to inhibit InsP3R receptor activity has been well-characterized in multiple systems, including intact cells, and is thought to reflect a feedback mechanism to prevent excess release of calcium9,
10. At a concentration of 1 nM, cytochrome c abolishes this action, an extremely high potency that supports physiologic relevance. By preventing calcium feedback inhibition of InsP3R activity, cytochrome c would substantially amplify InsP3R-mediated calcium release.
Our data support the hypothesis that altered calcium release by InsP3R during apoptosis reflects direct binding of cytochrome c to the InsP3R channel. Fluorescent imaging studies reveal that calcium spiking and cytochrome c release are temporally linked. Peptide competition of cytochrome c binding to the InsP3R demonstrates that the in vivo coordinate release of cytochrome c and altered calcium homeostasis is dependent on cytochrome c binding to the InsP3R and initiating calcium release.
The coordinate release of cytochrome c from mitochondria throughout the cell has not been previously explained at the molecular level. We suggest that a feed-forward cycle of calcium release from InsP3R and cytochrome c release from mitochondria provides a molecular mechanism. Cytochrome c binding to InsP3R adjacent to mitochondria would sensitize InsP3R to increased calcium release, causing mitochondrial and cytosolic calcium overload and further cytochrome c release. Therefore, small amounts of cytochrome c released early in apoptosis would first function at endoplasmic reticulum membranes to alter calcium handling. Subsequent global cytochrome c release from all mitochondria would supply the cytosolic cytochrome c necessary for caspase activation and completion of the apoptotic cascade.
Our findings reconcile previous reports suggesting a gradual apoptotic release of cytochrome c into the cytosol21,
22 with reports suggesting coordinate release6. The presumed cytosolic fractions in some earlier studies21,
22 were 10,000g supernatants, which contain endoplasmic reticulum. Here we show this fraction to be the principal target of early cytochrome c release. Imaging studies have revealed rapid and coordinate release of cytochrome c into the cytoplasm6. As cytochrome c binds to InsP3R in close apposition to mitochondria, cytochrome c released and bound to endoplasmic reticulum membranes would not be resolved by confocal or epifluorescence microscopy. We hypothesize that a slow, graded, release of cytochrome c primes InsP3R for increased calcium release, eventually triggering the massive cell-wide cytochrome c release coordinated through a calcium transient.
The importance of InsP3R in apoptosis has been emphasized by numerous studies using genetic manipulations to reduce InsP3R levels. Type-1-InsP3R-deficient lymphocytes are resistant to apoptotic stimuli23, and treatment of cells with type-3 InsP3R antisense RNA retards apoptosis24,
25. DT40 B cells lacking all forms of InsP3R display reduced apoptotic cell death16. Although these cells are resistant to apoptosis, almost a third do progress through the programmed cell death pathway, perhaps through increased cytosolic cytochrome c activating APAF-1 in the absence of increased cytosolic calcium. Alternatively, calcium release through InsP3R may function as an apoptotic signal amplifier, as suggested for cytochrome c26,
27. If true, apoptosis could still occur in the absence of InsP3R, albeit much less efficiently.
Methods
Materials.
STS, MDL-28170 and Z-VAD-FMK were purchased from BIOMOL Research Laboratories, Inc. (Plymouth Meeting, PA). Complete protease inhibitor cocktail without EGTA was from Roche Applied Science (Indianapolis, IN). Cytochrome c−agarose, BSA−agarose, protein A−Sepharose and all other chemicals were from Sigma (St. Louis, MO).
Yeast two-hybrid.
The Matchmaker 3 yeast two-hybrid system from Clontech (Palo Alto, CA) was used. AH109  -gal yeast was used. InsP3R fragments were cloned into pGADT7 (-gal acceptor domain) vector and were screened against a rat brain and human foetal kidney library (Clontech) in accordance with the manufacturer's specifications. Expression of these fragments was determined by western blotting using antibodies from Clontech, corresponding to the expression vector. Positive clones grew on minimal SD agar (Clontech) lacking adenine, histidine, leucine and tryptophan, and had -gal activity.
GST pulldown assay.
GST−InsP3R proteins (5 g) were incubated with 5 g cytochrome c in 500 l binding buffer (150 mM NaCl, 50 mM Tris at pH 7.8, 1% Triton X-100 and 1 mM EDTA) for 1 h at 4 °C with rotation. 50 l of 50% GST−Sepharose (Amersham Biosciences, Piscataway, NJ) was added and the incubation was continued for 30 min at 4 °C. The beads were washed four times with 1 ml binding buffer and separated on an SDS−PAGE gel. One half of the gel containing the GST fusion proteins was stained with Coomassie Blue, and the remainder of the gel containing cytochrome c was transferred to nitrocellulose and blotted.
Calcium release measurements.
Calcium release through recombinant type-I (SII+) InsP3R was measured exactly as described previously11. Calcium concentrations in all solutions were calibrated with a calcium-selective mini-electrode, and confirmed with fluorescent dyes. When the effects of cytochrome c were examined, it was added simultaneously with InsP3. In the absence of InsP3, SERCA activity was not affected by the addition of cytochrome c. At the concentrations used in this study, cytochrome c did not change the calcium concentration of the assay solution, as determined with a calcium mini-electrode.
Cell lines.
Human HeLa cells were obtained from the ATCC and cultured in DMEM supplemented with 10% foetal bovine serum (FBS), 2 mM L-glutamine and 1% penicillin-streptomycin. PC12 cells were cultured in DMEM supplemented with 10% horse serum, 5% FBS, 2 mM L-glutamine and 1% penicillin-streptomycin. DT40-WT and DT40-KO cells16 were cultured in RPMI 1640 supplemented with 10% FBS, 50 M 2-mercaptoethanol, 0.5 mg ml-1 G418, 1 mM glutamine and 1% penicillin-streptomycin. PC12 and HeLa cells stably expressing YFP−cytochrome c were generated by transfection with PvuI-linearized plasmid and selection in G418, exactly as suggested by the plasmid manufacturer (pcDNA3.1; Invitrogen). HeLa and PC12 cells were selected and maintained in 1 mg ml-1 and 1.5 mg ml-1 G418, respectively. YFP−cytochrome c expression was confirmed in isolated clones by confocal microscopy and western blotting. Mitochondrial localization was confirmed by colocalization with MitoTracker Red (Molecular Probes, Eugene OR).
Expression constructs.
InsP3R and SERCA-2b: The type-I InsP3R SII+ splice variant and SERCA-2b in pcDNA3.1 have been described elsewhere11 and were kindly provided by S.K. Joseph (Thomas Jefferson University, Philadelphia, PA).
CFP−InsP3R: Type-I InsP3R was digested with XbaI and sequentially partially digested with KpnI through use of an internal KpnI site in the InsP3R sequence. The 9300bp insert was then sub-cloned into the KpnI−XbaI sites of pECFP-C1 (Clontech). The pECFP-C1 plasmid was propagated in the dam- E. Coli strain SCS110 (Stratagene, La Jolla, CA) because of overlapping Dam methylation of the XbaI site. This construct expresses CFP as an N-terminal fusion with the InsP3R. This fusion was chosen because its N terminus is physically associated with the C-terminal tail of the InsP3R28. This fusion protein was localized to endoplasmic-reticulum-like membranes, as determined by confocal microscopy (data not shown).
YFP−cytochrome c: EYFP was fused to the N terminus of cytochrome c using single-overlap extension (SOE). This required PCR of EYFP and cytochrome c separately, then combining the two templates into a single PCR reaction using a 5' primer at the beginning of the EYFP sequence and a 3' primer on the cytochrome c sequence. The overlap region was engineered onto the reverse EYFP primer and forward cytochrome c primer, and contains BglII, XhoI and HindIII restriction sites necessary for facilitating removal and insertion of other cDNAs. PCR primers for EYFP were 5'-AATGCTAGCGCCACCATGGTGAGCAAGGGCGAG-3' (EYFP-F) and 5'-CGAAGCTTGAGCTCGAGATCTGAGCTTGTACAGCTCGTCCAT-3' (EYFP-R). PCR primers for cytochrome c were 5'-CTCAGATCTCGAGCTCAAGCTTCGATGGGTGATGTTGAAAAAG-3' (CytC-F) and 5'-AATTCTAGATTATTCATTAGTAGCCTTTTT-3' (CytC-R). EYFP-F contains an NheI site and a Kozak sequence, and CytC-R contains a stop codon and XbaI site. The cDNA for YFP was from Clontech, and cytochrome c was isolated from a rat brain cDNA library (Clontech) using primers CytC-F and CytC-R, and was sequenced to confirm that no mutations were generated by PCR. The final PCR product was sub-cloned into the NheI and XbaI sites of pcDNA3.1 (Invitrogen).
C-terminal-tail: amino acids 2589−2749 of the rat type-I InsP3R were amplified using PCR, incorporating a 5' EcoR1 and 3' Not1 restriction sites and cloned in frame to the EcoR1-Not1 sites of pCMV-HA (Clontech). PCR primers used were 5'-TTAAGAATTCGGATCGACACCTTTGCTGACCTG-3' and 5'-TTTTGCGGCCGCTTAGGCTGGCTGCTGTGGGTT-3'.
InsP3R-1 923−1581: amino acids 923 to 1581 of the rat type I InsP3R were amplified using PCR incorporating a 5' EcoR1 and 3' Xho1 restriction sites and cloned in-frame to the EcoR1−Xho1 sites of pCMV-Myc (Clontech). PCR primers used were 5'- GCA GAA TTC GGT CTA TCC ATG GAG TTG GG-3' and 5'- GCA CTC GAG GCG GGC TGA TAA CCG CCA GTT -3'.
Antibodies.
Rabbit polyclonal antiserum against the type-I InsP3R was generated by injecting New Zealand White male rabbits with the peptide CRIGLLGHPPHMNVNPQQPA coupled to keyhole limpet hemocyanin (Pierce Biotechnology, Rockford, IL). Initial injection was with Complete Freund's Adjuvant, and boost injections at days 14, 21 and 49 were with Incomplete Freund's Adjuvant. Rabbit injections, bleeds, and housing were performed by Cocalico Biologicals (Reamstown, PA). Antisera generated against this peptide have been reported previously29 and are specific for the type-I InsP3R. The production of rabbit anti-HO2 antisera has been described previously30. Mouse anti-Chicken IgM clone M-4 was from Southern Biotech (Birmingham, AL). Mouse anti-cytochrome c was purchased from Zymed Laboratories (San Francisco, CA). Mouse anti-cytochrome c oxidase subunit IV (CytOx) was from Molecular Probes. Rabbit anti-Prohibitin was from Research Diagnostics (Flanders, NJ). Goat anti-lactate dehydrogenase (LDH) was purchased from Chemicon (Temecula, CA). Mouse anti-PARP was from Alexis Biochemicals (San Diego, CA). Mouse anti-sarco-endoplasmic reticulum ATPase-2 (SERCA2) was from Affinity BioReagents (Golden, CO).
Subcellular fractionation.
After inducing apoptosis with STS or anti-IgM treatment, cells were harvested by gently scraping plates with a cell scraper and washing once with cold PBS. The washed pellet was subjected to one freeze−thaw cycle in liquid nitrogen. Pellets were resuspended in 1 ml buffer A (250 mM sucrose, 10 mM Tris-HCl at pH 7.5, 1 mM EGTA, 1mM phenyl methylsulphonyl fluoride (PMSF) and one protease inhibitor pill). Cells were then homogenized on ice using a 1-ml glass Dounce homogenizer with a tight-fitting pestle until 95% of cells were disrupted, as judged by Trypan blue staining. Crude lysates were centrifuged at 1,000g for 15 min at 4 °C to remove nuclei and unbroken cells. The supernatant was collected and the pellet (P1) discarded. The low-speed supernatant was then subjected to a 10,000g centrifugation for 15 min, yielding the 10,000g pellet (P2). The supernatant from the P2 pellet was centrifuged at 15,000g to completely rid the supernatant of any remaining mitochondria. Finally, the 15,000g supernatant was separated into cytosol (S3) and light membrane (P3) fractions by centrifugation at 100,000g for 1 h. The 100,000g supernatant was collected as the S3 fraction and the pellet was resuspended in 40−70 l of buffer A. The P2 fraction was washed twice by resuspending cells in 100 l buffer A and pelleting (10,000g for 15 min). After the final wash, the P2 fraction was resuspended in 50−100 l buffer A. All fractionations were repeated a minimum of four times with essentially identical results.
Co-immunoprecipitation.
Cell lysis buffer (150 mM NaCl, 50 mM Tris-HCl at pH 7.8, 1% Triton X-100 and 1 mM EDTA) was added to 100 g of P3 fraction to bring samples to a total volume of 500 l. Samples were cleared of insoluble debris by centrifugation at 10,000g. Anti-InsP3R-1, pre-immune sera, or anti-SERCA2 antibodies and protein A−Sepharose beads were added and incubated on a rotator for 30 min at 4 °C. The protein A−Sepharose beads were washed three times with lysis buffer and quenched with 20 l of SDS sample buffer. Co-immunoprecipitates were resolved by SDS−PAGE and analysed by western blotting.
Cytochrome c−agarose binding.
Cerebellar microsome or COS-7 cell lysates were incubated with 20 l cytochrome c−agarose (approximately 200 g total cytochrome c) in a final volume of 0.5 ml for 2 h at 4 °C in the dark with rotation. Beads were washed four times with 0.5 ml cell lysis buffer and separated on SDS−PAGE gels. BSA coupled to the same agarose matrix as cytochrome c was used as a negative control. When used, InsP3Rs were purified to homogeneity exactly as described previously31.
Caspase activity.
Caspase activity was performed on S3 fractions using the EnzChek caspase-3 assay kit #1, essentially after the manufacturers protocol (Molecular Probes). This kit measures the increase in fluorescence generated by the cleavage of the aminomethylcoumarin (AMC)-labelled caspase-3 substrate Z-DEVD-AMC. The production of fluorescent substrate was monitored continuously every 5 min for approximately 2 h in a 96-well dish using an automated fluorescent plate reader with data acquisition software (LS55 luminescence spectrometer, PerkinElmer Instruments, Shelton, CT). The slope of the linear regression drawn through each time point was used to determine change in fluorescence over time for each sample. A standard curve using known amounts of AMC was used to convert fluorescent values to specific catalytic activity.
FRET.
PC12 cells stably expressing YFP−cytochrome c were cultivated in glass bottom 35-mm dishes (MatTek corporation, Ashland, MA). CFP−InsP3R and ECFP cDNAs were transfected with Lipofectamine 2000 (Invitrogen), and were used in experiments 48 h after transfection. Cells were imaged on a Zeiss Axiovert 135 inverted microscope with a 40 Plan Neofluar lens (Carl Zeiss MicroImaging, Thornwood, NY) equipped with a Photometrics CoolSNAP HQ cooled charge-coupled device camera (Roper Scientific, Tucson, AZ). Cells were imaged in phenol-free Leibovitz's L-15 medium supplemented with 10% horse serum, 5% fetal bovine serum and 2 mM L-glutamine overlaid with mineral oil at 37 °C with a heated stage. Dual-emission ratiometric imaging was controlled with IPLab 3.5.5 software (Scanalytics, Inc., Fairfax, VA) using excitation filter D436/10, beam splitter 86002bs, and emission filters D470/30 and HQ535/30 (Chroma Technology Corp., Brattleboro, VT). Emission filters were alternated using a computer-controlled filter wheel (LUDL electronics, Hawthorne, NY). Cells were excited at 436 nm and 535:470-nm ratiometric images were acquired every 5 min for 350 min. Data in Fig. 6c and d were pooled from at least three separate experiments for each trace (n values are indicated).
Calcium imaging.
Calcium measurements were performed as previously described32 using the Intracellular Imaging calcium imaging system and InCyt2 imaging software (Intracellular Imaging, Cincinnati, OH). Fura-2 AM-loading (Molecular Probes) was performed for 1 h at 37 °C for PC12 cells stably expressing YFP−cytochrome c, and 15 min at 37 °C for HeLa cells stably expressing YFP−cytochrome c. Cells were treated with 1 M STS at time zero and incubated at 37 °C until imaging began. Resting calcium levels in cell lines were similar: that is, 100−200 nM. Images at 340 nm, 380 nm and 485 nm were collected every 10 s for 30 min in HeLa cell experiments and every 20 s for 1 h in PC12 cell experiments. A 10-min gap between each time point was necessary for data acquisition and compilation. All measurements shown are an average of 50 cells, unless otherwise stated, and are representative of a minimum of three, and in most cases, a larger number of independent experiments.
Note: Supplementary Information is available on the Nature Cell Biology website.
Received 10 September 2003; Accepted 20 October 2003; Published online: 9 November 2003.
REFERENCES
-
Berridge, M.J., Lipp, P. & Bootman, M.D. The versatility and universality of calcium signalling. Nature Rev. Mol. Cell Biol. 1, 11−21 (2000). | Article | PubMed | ISI | ChemPort |
-
Orrenius, S., Zhivotovsky, B. & Nicotera, P. Regulation of cell death: the calcium-apoptosis link. Nature Rev. Mol. Cell Biol. 4, 552−565 (2003). | Article | PubMed | ISI | ChemPort |
-
Ferrari, D. et al. Endoplasmic reticulum, Bcl-2 and Ca(2+) handling in apoptosis. Cell Calcium 32, 413−420 (2002). | Article | PubMed | ISI | ChemPort |
-
Hajnoczky, G., Csordas, G., Madesh, M. & Pacher, P. Control of apoptosis by IP(3) and ryanodine receptor driven calcium signals. Cell Calcium 28, 349−363 (2000). | Article | PubMed | ISI | ChemPort |
-
Strasser, A., O'Connor, L. & Dixit, V.M. Apoptosis signaling. Annu. Rev. Biochem. 69, 217−245 (2000). | Article | PubMed | ISI | ChemPort |
-
Goldstein, J.C., Waterhouse, N.J., Juin, P., Evan, G.I. & Green, D.R. The coordinate release of cytochrome c during apoptosis is rapid, complete and kinetically invariant. Nature Cell Biol. 2, 156−162 (2000). | Article | PubMed | ISI | ChemPort |
-
Rizzuto, R. et al. Close contacts with the endoplasmic reticulum as determinants of mitochondrial Ca2+ responses. Science 280, 1763−1766 (1998). | Article | PubMed | ISI | ChemPort |
-
Csordas, G., Thomas, A.P. & Hajnoczky, G. Quasi-synaptic calcium signal transmission between endoplasmic reticulum and mitochondria. EMBO J. 18, 96−108 (1999). | Article | PubMed | ISI | ChemPort |
-
Bezprozvanny, I., Watras, J. & Ehrlich, B.E. Bell-shaped calcium-response curves of Ins(1,4,5)P3- and calcium-gated channels from endoplasmic reticulum of cerebellum. Nature 351, 751−754 (1991). | Article | PubMed | ISI | ChemPort |
-
Finch, E.A., Turner, T.J. & Goldin, S.M. Calcium as a coagonist of inositol 1,4,5-trisphosphate-induced calcium release. Science 252, 443−446 (1991). | PubMed | ISI | ChemPort |
-
Boehning, D. & Joseph, S.K. Functional properties of recombinant type I and type III inositol 1, 4,5-trisphosphate receptor isoforms expressed in COS-7 cells. J. Biol. Chem. 275, 21492−21499 (2000). | Article |
|
