| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Ralph H. Johnson Veterans Administration and the Departments of Medicine and
Biochemistry and Molecular Biology, Medical University of South Carolina, South Carolina 29425, USA
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: targeting • ceramide • mitochondria • apoptosis • Bcl-2 • cytochrome c
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. Mitochondria manifest signs of outer membrane and/or inner membrane permeabilization when exposed to a variety of proapoptotic second messengers. Thus, cytochrome c, which is normally confined in the mitochondrial intermembrane space, is found in the cytosol of cells undergoing apoptosis (3
, 4)
. Cytosolic cytochrome c forms a complex with Apaf-1 and procaspase-9, resulting in activation of caspase-9, which then processes and activates other caspases, such as caspase-3, to orchestrate the biochemical execution of programmed cell death (5)
. Proapoptotic Bcl-2 family proteins, including Bax, Bak, and Bid, induce mitochondrial membrane permeabilization and cytochrome c release (6
7
8)
. In contrast, the antiapoptotic Bcl-2 family proteins are capable of preventing cytochrome c release while also significantly inhibiting cell death (9)
.
The sphingolipid ceramide is now well established as an endogenous regulator of apoptosis in many cancer cell systems, and in response to many cytotoxic and chemotherapeutic agents. A close association between the production of ceramide and the onset of programmed cell death has been well established (10)
. In fact, a number of apoptosis-promoting agents, including tumor necrosis factor 
(TNF-) (11)
, chemotherapeutic drugs (12
, 13)
, ischemia/reperfusion (14)
, Fas antigen activation (15)
, irradiation (16)
, and corticosteroids (17)
, can generate ceramide by the induction of sphingomyelin (SM) hydrolysis through the action of sphingomyelinases (SMases) and/or through the de novo pathway (18
, 19)
. Changes in endogenous levels of ceramide in response to these agents occur after the onset of the first biochemical signals of apoptosis such as activation of the initiator caspases, caspase-8, -9, and -10, and usually prior to the activation of the executioner caspases such as caspase-3, -6, and -7 (20
21
22)
. In addition, studies with inhibitors of ceramide metabolism, such as PDMP (23)
and D-MAPP (24)
, show that these agents also induce apoptosis, most probably as a consequence of elevating intracellular ceramide levels. In many apoptotic programs, inhibition of de novo generation of ceramide with fumonisin B1 prevents, at least in part, the development of apoptosis (18
, 19)
. Also, hepaptocytes from acid SMase knockout mice are resistant to apoptosis (25)
.
Previous studies have shown that the major cellular pool of SM present on the outer leaflet of the plasma membrane is not involved in the ceramide pathway of apoptosis (26
, 27)
. However, we had shown that endogenous expression of SMase was able to induce apoptosis, implying that certain intracellular pools of ceramide may be the biologically relevant pools (28)
. Because of this apparent compartmental difference in cellular ceramide, in this study, the Bacillus cereus sphingomyelinase (bSMase) protein was targeted to different intracellular compartments in order to generate ceramide in these compartments and determine in which of them ceramide generation can induce apoptosis.
Our results demonstrate that only when bSMase was targeted to mitochondria did cells undergo apoptosis; its targeting to the inner plasma membrane, the cytoplasm, the endoplasmic reticulum (ER), the Golgi apparatus, or the nucleus was ineffective. This occurred despite the increase of ceramide levels in all these compartments. Further, the results show that cell death induced by mitochondrial targeting of bSMase requires the catalytic activity of SMase, induces the release of cytochrome c, and is prevented by the overexpression of Bcl-2 protein, indicating that bSMase targeted to mitochondria is a specific inducer of apoptosis.
These results indicate for the first time that the generation of ceramide in mitochondria specifically induces cell death. These data have important implications for apoptosis, as most apoptotic processes converge on mitochondria. They also provide a novel approach for examining the subcellular topology of lipid-mediated cell regulation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
32P]ATP was from Amersham Pharmacia Biotech (Arlington Heights, IL). Exogenous bSMase from Staphylococcus aureus was from Sigma (St. Louis, MO). TNF- was from PeproTech (Rocky Hill, NJ). All DNA sequences were obtained at the protein and DNA sequencing facility of the Medical University of South Carolina.
Generation of expression constructs
bSMase was cloned into pBS vector (Stratagene, La Jolla, CA) by polymerase chain reaction (PCR) as described previously(28)
. This construct was then used as a template to perform the PCR reactions in this study.
Construction of bSMase-green fluorescent protein (GFP) fusion protein
The bSMase gene was first amplified by using a 5'-CGGGGTACCATGGAAGCATCTACAAATCAAAATG primer containing a KpnI site and a 3'-TCGGGGCCCTCTTCATAGAAATAGTCGCCTCTACTGGA primer containing an ApaI site. The PCR fragment was then subcloned into the TOPO TA cloning vector and amplified. It was next digested from the TOPO vector by using KpnI and ApaI restriction enymes and subcloned into the same sites in pEGFP-N1 (Clontech) to generate pEGFP-N1/bSMase construct, with GFP being at the carboxyl terminus of bSMase.
Construction of bSMase-GFP with targeting signals
The pEGFP-N1/bSMase construct was used for PCR of the bSMase-GFP fragment with a 5'-AAACTGCAGGAAGCATCTACAAATCAAAATGATACATTAAAA bSMase primer containing PstI site and a 3'- CATTCTCGAGCTTGTACAGCTCGTCCATGCCGAGAGTGAT GFP primer containing XhoI site. The amplified bSMase-GFP fragment was then subcloned into the TOPO TA cloning vector and amplified. To prepare constructs with targeting sequences for the nucleus, ER, and mitochondria, the bSMase-GFP fragment was digested from the TOPO vector by using PstI and XhoI restriction enzymes and then subcloned into the same sites in pCMV/myc vectors (Invitrogen) containing different intracellular targeting sequences, which also contained an myc epitope tag. For cytoplasmic targeting, the pCMV/myc vector did not contain any targeting sequence. These constructs were named pCMV/bSMase-GFP/Nuc, pCMV/bSMase-GFP/ER, pCMV/bSMase-GFP/Mito, and pCMV/bSMase-GFP/Cyto, representing the bSMase-GFP targeted to the nucleus, ER, mitochondria, and cytoplasm, respectively. For the Golgi construct, the targeting sequence was obtained by PCR using a 5'-CCATGGCTAGGCTTCGGGAGCCGCTCCTGAGC primer containing the NcoI site and a 3'-CTCGAGGGCCCCTCCGGTCCGGAGGTCCCCGCA primer containing the XhoI site and the vector pECFP-golgi (Clontech) as the template. In parallel, a bSMase-GFP fragment having XhoI site at the 5'-CTCGAGGAAGCATCTACAAATCAAAATGATACATTA primer and NotI site at the 3'-GCGGCCGCACTTGTACAGCTCGTCCATGCCGAGAGTGAT primer was obtained by PCR using the pEGFP-N1/bSMase as a template. The amplified fragments from the two PCRs were then sucloned by double cloning into the TOPO TA cloning vector. The whole fragment Golgi targeting sequence-bSMase-GFP was then digested from the TOPO vector by using NcoI and NotI and subcloned into the pCMV/myc vector. The generated construct was then named pCMV/bSMase-GFP/Golgi.
For the generation of the inner plasma membrane construct, the farnesylation targeting sequence fused to GFP was obtained by PCR with a 5'-GGGCCCATGTGAGCAAGGGCGAGGAGCTGTTCA GFP primer containing the ApaI site and a 3'-farnesyl targeting sequence primer containing the XhoI site using the pEGFP-F vector (Clontech) as a template. After PCR, the obtained fragment was subcloned into the TOPO TA cloning vector. In parallel, the GFP fragment was removed by digestion by using ApaI and XhoI restriction enzymes from the pCMV/bSMase-GFP/cyto construct. The farnesyl-GFP fragment was then digested from the TOPO vector and subcloned into the digested pCMV/bSMase-GFP/cyto vector. The construct generated was named pCMV/bSMase-GFP/PM.
Mutation of aspartic acid 295 into glycine in bSMase was performed by PCR with a 5'- CGGGGTACCATGGAAGCATCTACAAATCAAAATG primer containing a KpnI site and a 3'- TCGGGGCCCTCTTCATAGAAATAGTCGCCTCTACTGGA-TAATCACCAGAGT primer containing the ApaI site using pEGFP-N1/bSMase as a template. The PCR product bSMase-D295G was subcloned into the pEGFP-N1 vector within the same restriction sites. The bSMase-D295G-GFP fragment was then digested and subcloned into the pCMV/Mito vector to generate the pCMV/bSMase-D295G-GFP/Mito construct.
All the control vectors used (pCMV/GFP/Nuc, pCMV/GFP/ER, pCMV/GFP/Mito, pCMV/GFP/Cyto, pEGFP/GFP/PM, pCMV/GFP/Golgi) were empty vectors containing the GFP targeted to the different compartments.
As shown in Fig. 1
, the farnesylation signal from c-Ha-ras was fused to the carboxyl terminus of the bSMase-GFP fusion protein (29)
. This farnesylation signal directs the fusion protein to the inner face of the plasma membrane. The mitochondrial targeting sequence isolated from subunit VIII of human cytochrome c oxidase would presumably cause the bSMase-GFP to be translocated to the inner mitochondrial membrane/mitochondrial matrix (30)
. The ER-targeting sequence from a mouse Vh chain consisted of two signals, the first signal targets the fusion protein into the ER compartment and the second SEKDEL signal allows the retention of the protein in the ER (31)
. For the generation of a protein that targets the Golgi apparatus, the amino-terminal 81 amino acids of human ß1,4-galactosyltransferase (GT) was fused to the amino-terminal bSMase-GFP fusion protein. This region of human GT contains the membrane-anchoring signal peptide that targets the fusion protein to the trans-medial region of the Golgi apparatus (32)
. The nuclear targeting sequence from the SV40 large T antigen has been triplicated and was fused to the carboxyl-terminal bSMase-GFP fusion protein (33)
. All the vectors were sequenced to confirm that they had the correct sequence.
|
Cell culture and transient transfection
MCF7 breast cancer cells were maintained in RPMI 1640 medium (Life Technologies, Gaithersburg, MD) supplemented with 10% FCS in 5% CO2 at 37°C in a humidified incubator. MCF7 cells overexpressing Bcl-2 (MCF7/Bcl-2) were grown in RPMI 1640 supplemented with 10% FCS and 100 mg/ml Geneticin. For transfection, cells were seeded at 105 cells per 60-mm dish. Twenty-four hours after plating, they were transfected transiently with control vectors containing GFP (pCMV/GFP) or vectors containing bSMase-GFP (pCMV/bSMase-GFP) using Superfect and 3.5 µg of each plasmid per dish. After 3–4 h of incubation with the mixture, the cells were washed with PBS and fresh medium was added. For cell viability and cytochrome c release assays, MCF7 cells were cotransfected with pCMV vectors plus pEGFP-N1 empty vector in a ratio of 10:1 to enhance fluorescence.
Cell viability assay
After transfection, cells were washed twice with PBS then incubated with 1 µg/ml annexin V conjugated to Alexa Fluor 594 (Molecular Probes) in annexin-binding buffer (10 mM HEPES pH 7.4, 140 mM NaCl, and 2.5 mM CaCl2) at 37°C. After 30 min of incubation, cells were viewed under a fluorescence microscope, and GFP-positive cells were counted for annexin V binding. Different fields (at least 300 cells) were counted for each experiment.
Localization of the bSMase-GFP fusion products
Cells in 35-mm dishes were first transfected with 1 µg of pCMV/GFP vectors or pCMV/bSMase-GFP vectors as described above. They were then fixed and stained with 1 µg/ml DAPI (Molecular Probes). In some experiments when the localization of the mitochondrial fusion protein was investigated, cells transfected with the mitochondria-targeted bSMase-GFP fusion protein were first loaded with 25 nM of Mitotracker red CMXRos for 20 min prior to fixation. The GFP images were collected by excitation at 395 nm and emission at 507 nm. To avoid fluorescent cross-talk, green GFP and Mitotracker red CMXRos fluorescence measurements were taken sequentially.
Immunostaining
Cytochrome c release was followed by immunostaining after 24 or 48 h in transfected cells or after 18 h in cells treated with TNF-. At the end of the treatment time, cells were washed three times with PBS followed by fixation in freshly prepared 3.7% formaldehyde for 10 min. The fixed cells were washed three times with PBS for 15 min each, followed by permeabilization in 0.15% Triton X-100 in PBS for 15 min. The cells were then blocked for 60 min in blocking buffer (2% bovine serum albumin in PBS) followed by a 4-h incubation with a mouse monoclonal antibody against cytochrome c (1:200) (PharMingen). The cells were washed three times for 10 min each in blocking buffer followed by 1-h incubation with a rhodamine-labeled goat anti-mouse IgG (1:200) (Jackson Immunology). The cells were washed three times for 10 min in PBS, stained with 1 µg/ml DAPI, and then examined under a fluorescence microscope.
Ceramide measurements
Total endogenous ceramide levels were measured by using the diacylglycerol kinase (DGK) method as described previously (34)
.
SMase activity assay
Transfected cells were lysed in 25 mM Tris buffer pH 7.4 containing 0.1% Triton X-100, 5 mM EDTA, 1 mM phenylmethylsulfonylfluoride, and 2 mg/ml each of chymostatin, leupeptin, antipain, and pepstatin A. Aliquots of 5–10 µl (20 µg of protein) of the total homogenate were added to 100 µl of reaction mixture containing 50 mM Tris (pH 7.4), 5 mM MgCl2, 5 mM dithiothreitol, 10 nmol phosphatidylserine, and 10 nmol [choline-methyl-14C]sphingomyelin (
100,000 cpm). After 1 h of incubation at 37°C, the reaction was stopped by the addition of 1.5 ml of chloroform/methanol (2:1), and the phases were separated by the addition of 200 µl of water. After centrifugation, 400 µl of the upper phase was counted by liquid scintillation.
Other procedures
Protein concentration was determined by using the Bradford assay.
Statistical analysis
For statistical analysis, the t test for paired sample means was used. In case of the results expressed as a percentage of control, the statistical analysis was performed on the actual values.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A, treatment of MCF7 cells with exogenous bSMase for 0.5, 1, and 3 h, caused a significant increase of ceramide levels (25-fold) in these cells, as a result of hydrolysis of SM in the outer leaflet of the plasma membrane. Evaluation of apoptosis by the trypan blue exclusion assay showed that there was no induction of cell death (Fig. 2B
). These results indicate that ceramide generation at the outer leaflet of the plasma membrane in MCF7 cells does not induce cell death. These data are in accord with previous observations (28
, 35)
.
|
Targeting of bSMase-GFP fusion protein to different intracellular compartments
The results from exogenous treatment of MCF7 cells with bSMase suggested that perhaps intracellular pools of SM and ceramide are involved in cell death. Thus, our goal was to define the intracellular pools where ceramide generation induces cell death. To this end, bSMase was first fused with GFP, and then the fusion gene was subcloned into mammalian vectors containing sequences that target the fusion protein to the following intracellular compartments: inner plasma membrane, cytoplasm, mitochondria, ER, Golgi apparatus, and nucleus (see Materials and Methods and Fig. 1
). The control vectors contained the GFP with targeting sequences to the respective compartments. These vectors were then used to transiently transfect MCF7 cells. Initially, the transfection efficiency of all constructs was determined and was found to be within the same range of 30–35% (data not shown). Next, the localization of each fusion protein was investigated. As shown in Fig. 3
A, the patterns of fluorescence indicated that the tagged proteins were indeed expressed in the individual cellular compartment to which they were targeted: the plasma membrane protein localized predominately to the periphery of cells, the cytoplasmic protein showed a typical diffuse pattern throughout the cytoplasm, the nuclear protein signal colocalized with a specific nuclear probe DAPI, the ER protein displayed a fine reticular network all over the cytoplasm and around the nuclear envelope, the Golgi protein showed punctate structures in the perinuclear region, and the mitochondrial protein displayed a punctate distribution characteristic of mitochondria. As an example, and to ascertain the localization of the fusion protein targeted to mitochondria, transfected MCF7 cells were incubated with Mitotracker, a specific mitochondrial marker. Fig. 3B
shows that the Mitotracker red signal coincided very strongly with the bSMase-GFP signal, and the yellow color in the overlay image indicates that the two signals colocalize.
|
Next, SMase activity was measured in MCF7 cells transfected with all constructs. As shown in Fig. 4
, transfection of MCF7 cells with bSMase constructs resulted in an increase of neutral SMase activity as compared with control cells transfected with vectors containing the GFP targeted to the specific compartments. The time course of the peak of activity for all constructs was within 24–48 h, but the magnitude of this increase varied depending on each construct. Thus, the activity increased about five- to sevenfold when cells were transfected with the cytoplasmic, Golgi, and nuclear constructs. However SMase activity increased only twofold with the mitochondrial and plasma membrane constructs. Interestingly, SMase activity increased 60-fold when cells were transfected with the ER construct. These observed differences of activity were most probably caused by differences in protein expression and/or stability as revealed by Western blot analysis using an antibody directed against the GFP or against the myc epitope (data not shown). Although the SMase activity increased to a different extent, these results indicate that all constructs expressed an active protein.
|
When ceramide levels were measured, a concomitant increase in total cellular ceramide levels was observed (Fig. 5
). Different degrees of the intracellular increase in ceramide levels occurred in the different transfectants. The ER construct showed the highest increase (about twofold), whereas all other constructs showed an increase of 10–60% as compared with controls (cytoplasm 10%, inner plasma membrane 15%, mitochondria 25%, nucleus 30%, and Golgi apparatus 60%). This variable increase in ceramide levels could be a result of either the different levels of expression of the fusion proteins or, more likely, the variations in SM levels within each compartment. It should be noted that the ceramide measurements were performed on whole cells. Thus, the small increases seen may be quite significant when localized to the respective compartments.
|
Mitochondrial targeting of bSMase induced cell death
The effect of the overexpression of the different bSMase targeted constructs on cell death was next examined by using annexin V binding of the GFP-positive cells at 24, 48, and 72 h. As shown in Fig. 6
A, at 48 h, targeting the bSMase to the inner plasma membrane, cytoplasm, ER, Golgi apparatus, or nucleus was without any effect on cell death. By contrast, and very interestingly, the targeting of bSMase into mitochondria induced about 25% cell death (Fig. 6A
). Results similar to those at 48 h were obtained for all the constructs at 24 and 72 h (data not shown). Thus, when cell death was monitored as a function of time for the pCMV/bSMase-GFP/Mito construct, approximately 20% cell death was observed at 24 h after transfection but increased somewhat by 72 h (Fig. 6B
). Cell death for vector-transfected cells was about 5% for the whole period studied. These results imply that generation of ceramide in plasma membrane, Golgi apparatus, ER, and nucleus does not appear to be involved in apoptosis and that ceramide needs to be localized to mitochondria to induce cell death. Moreover, the results with the soluble or cytosolic bSMase also suggest that not much SM is accessible in cytosolic-facing membranes and that these pools are not involved in cell death.
|
An active enzyme was required to induce cell death
To demonstrate that mitochondrial ceramide generation induces cell death in a specific manner, the ability of an inactive enzyme mutant to induce cell death was evaluated. It was previously shown that a point mutation of aspartate 295 to glycine of bSMase completely abolishes the activity of this enzyme (36)
. Thus, the mitochondrial vector (pCMV/bSMase-GFP/Mito) was changed at position D295 into G295 of bSMase to produce a catalytically inactive enzyme. The mutant mitochondria- targeted bSMase indeed localized to mitochondria (data not shown). As expected, SMase activity and ceramide levels measured in MCF7 cells transfected with this mutant were not increased (Fig. 7
A, B). Furthermore, when cell death was monitored by using annexin V binding, this mutant failed to induce cell death as compared with the mitochondria-targeted native bSMase (Fig. 7C
). These results indicate that the cell death induced by mitochondrial targeting of bSMase requires SMase catalytic activity.
|
Cell death induced by mitochondrial ceramide generation was prevented by Bcl-2
Previous studies from our laboratory showed that the overexpression of Bcl-2 protein in MCF7 cells protected these cells from apoptosis induced by ceramide (21)
. Also, Bcl-2 overexpression has been shown to prevent ceramide-induced death in multiple models (12
, 37)
. Thus, it was important to investigate the effect of Bcl-2 overexpression on cell death induced by the mitochondrial targeting of bSMase. As shown in Fig. 8
A, transfection of MCF7 cells with the pCMV/bSMase-GFP/Mito construct resulted in approximately 20% cell death. In contrast, cells overexpressing Bcl-2 protein were resistant to mitochondrial targeting of bSMase and showed no cell death (Fig. 8A
). Similar observations were obtained when cells were treated with TNF-, which was used as a positive control in these experiments. Figure 8B
shows that ceramide levels in MCF7/Bcl-2 cells were still increased after pCMV/bSMase-GFP/Mito transfection, indicating that Bcl-2 did not protect cells by interfering directly with SMase activity and ceramide production.
|
These results indicate that cell death induced by mitochondrial ceramide is blocked by Bcl-2 and suggest that the effect of Bcl-2 on the ceramide pathway is downstream of ceramide production and at the level of the mitochondria.
Mitochondrial ceramide generation induced cytochrome c release
To determine the mechanism by which mitochondrial ceramide activates the apoptotic program, cytochrome c release from mitochondria into the cytosol was investigated, because this appears to be a major mechanism mediating the mitochondrial pathway of apoptosis. As shown in Fig. 9
A, TNF- treatment induced a change in the cytochrome c staining pattern from a punctate mitochondrial profile in untreated cells to a diffuse cytosolic distribution. When cells were transfected with the pCMV/bSMase-GFP/Mito construct, a similar redistribution pattern of cytochrome c from the mitochondria to the cytoplasm was observed (Fig. 9A
). In contrast, the overexpression of the bSMase-D295G-GFP/Mito mutant did not affect the distribution of cytochrome c, as demonstrated by a persistent punctate staining. Quantitation of cytochrome c release for the different treatments is shown in Fig. 9B
. TNF- induced a marked release of cytochrome c, with 60% of cells releasing cytochrome c. In Mito-bSMase-transfected cells, cytochrome c was released in about 25% of transfected cells, whereas the mutated D295G-SMase did not affect the release of cytochrome c. These results indicate that mitochondrially generated ceramide induces a specific apoptotic program.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To address these issues, we aimed at elevating ceramide levels selectively in individual cellular compartments and studying the effects of this increase on cell death. Constructs that target the bSMase protein to the inner leaflet of the plasma membrane, the cytoplasm, the mitochondria, the ER, the Golgi apparatus, or the nucleus were generated. The results showed that targeting of bSMase to all the cellular compartments studied resulted in an increase in total ceramide levels.
When cell death was investigated, the results demonstrated that, of all the compartments studied, generation of only mitochondrial ceramide caused cell death as measured by the externalization of phosphatidylserine, an early event in apoptosis. Further, investigation of the cell death induced by mitochondrial ceramide generation revealed that this event is specific because: 1) a mutant inactive enzyme targeted to mitochondria failed to increase ceramide and to induce cell death, and 2) cell death was blocked by Bcl-2 overexpression. These results led us to investigate the possibility of a direct link between ceramide and mitochondrial cell death pathways.
The involvement of mitochondria in apoptosis and the existence of a mitochondrial pathway of cell death are now well established. According to recent understanding, mitochondrial pathways of apoptosis cause the release of the mitochondrial protein cytochrome c into the cytosol, resulting in the processing of caspase-9 and the downstream caspases such as caspase-3, leading ultimately to cell death. Another pathway involves the release of apoptosis-inducing factor, which may function independently of caspases (38)
.
Studies aimed at elucidating the mechanisms by which mitochondrial ceramide induces cell death showed that ceramide acts at least in part by inducing the release of cytochrome c from mitochondria and causing the activation of the downstream cascade of cell death. A mutant enzyme was unable to release cytochrome c, and Bcl-2 blocked cytochrome c release (data not shown), demonstrating again the specificity of the events induced by mitochondrial ceramide generation.
These results have several important implications. One immediate implication of this study is that mitochondrial SM and ceramide pools exist and are involved in apoptosis. The involvement of SM and ceramide in intracellular signaling raised important questions as to their cellular localization. In particular, the recent identification of a mitochondrial ceramidase implies the existence of a mitochondrial pool of ceramide and possibly other metabolic precursors (e.g., SM) or metabolites (e.g., sphingosine) of ceramide. The most widely held belief was that SM is located almost exclusively within the outer leaflet of the plasma membrane (39)
. However, other studies of the topology of SM suggested that it may also be present in other compartments. A number of studies of erythrocytes (40
, 41)
and nucleated mammalian cells (42
, 43)
suggested that SM may exist within the inner leaflet of the plasma membrane. In addition, studies of the biosynthesis of SM showed that it is synthesized in the cis and medial Golgi apparatus, suggesting that yet another intracellular compartment contains SM (44)
. Unpublished data from our laboratory with highly purified rat liver mitochondria treated with bSMase show a significant increase in ceramide levels, further suggesting the presence of a pool of SM in mitochondria. In addition, we observed several enzyme activities of sphingolipid metabolism, such as SMase and sphingomyelin synthase, to be present in mitochondria-associated membranes (J. Usta, Y. A. Hannun, and L. M. Obeid, unpublished observations).
Although the increase in total ceramide levels was moderate, this increase can become very relevant when considering the localization of the generated ceramide in only one compartment. For example, if mitochondrial membranes account for about 10% of total membranes, then a 20% increase in total ceramide represents approximately a threefold elevation in mitochondrial ceramide. Thus, a substantial increase in ceramide in mitochondrial membranes would appear as a small increment in total ceramide. This moderate increase could be the result of the presence of small amounts of total SM, and this is supported by previous studies showing that the pool of SM hydrolyzed in response to TNF- and other apoptosis-inducing agents is usually 10–30% of total SM. In addition, the presence of ceramide-metabolizing enzymes in these compartments such as mitochondrial ceramidase could also attenuate ceramide levels. Thus, the ability of SMase to induce ceramide accumulation in mitochondria establishes the existence of a mitonchondrial pool of SM. Together with the discovery of a mitochondrial ceramidase (45)
, these observations indicate the presence of a metabolic pathway of sphingolipids in mitochondria.
The localization of these pools within the mitochondrion is not known. The mitochondrial-targeting sequence directs the bSMase-GFP fusion protein to the mitochondrial matrix. Thus, the bSMase would probably act on a pool of SM present on the inner mitochondrial membrane to generate ceramide on that side of the mitochondrial membrane. However, bSMase could act on an outer mitochondrial membrane pool of SM during its transport into mitochondria. In either case, and given the nonpolar nature of ceramide, the ceramide generated in either mitochondrial membrane can flip-flop from one side of the membrane to the other to find its target.
These data are also consistent with some in vitro studies showing a direct interaction between ceramide and mitochondria. For example, a direct inhibition of complex III of the mitochondrial respiratory chain by ceramide was demonstrated (46)
. Ceramide was also shown to induce production of reactive oxygen species in intact mitochondria (47)
and in cells (48)
. It is therefore highly likely that a mitochondrial pool of ceramide is involved in these processes.
Further, results from this study also suggest the hypothesis that distinct pools of SM are involved in distinct functional responses, and which pool of SM is involved in the SM cycle is a question of debate. Several studies (26
, 27
, 49
, 50)
support the conclusion that SMase acts at the cytosolic side of the plasma membrane or in some other intracellular membranes. However, Tepper et al. (51)
showed recently that ceramide generated during the execution phase of apoptosis is derived from SM initially located on the outer leaflet of the plasma membrane, which gains access to a cytosolic SMase by flipping to the inner leaflet in a process of lipid scrambling, paralleling phosphatidylserine externalization. In contrast to these results, we show here that a mitochondrial pool of SM is involved in cell death, whereas the plasma membrane SM pools seem not to be involved in the initial launching of cell death. It is conceivable that a later phase of ceramide derives from hydrolysis of plasma membrane of SM. The role of this ceramide has not been defined.
SM has also been shown to be associated with specific lipid domains called rafts (52)
, and it was suggested that the pool of SM present in these domains is coupled to ceramide generation through the action of acid SMase (53
, 54)
. Recently, a neutral SMase activity has been shown to be associated with caveolae (55)
. In addition, the authors suggested that this caveolar neutral SMase activity is involved, at least partially, in TNF- signaling. It is conceivable that exogenous bSMase application or targeting the bSMase with a farnesyl sequence results in a protein that fails to access these SM pools.
Although some of the mechanisms and key players of mitochondrial cell death have been deciphered, several biochemical events remain to be ordered in the mitochondrial death pathway.
We previously demonstrated that ceramide acts upstream of Bcl-2 (12)
. Bcl-2 phosphorylation has been shown to be important for its antiapoptotic function. A possible mechanism identifying a role for ceramide in mitochondria and apoptosis involves activation of a protein phosphatase as a target of ceramide. Ceramide has been found to specifically activate a mitochondrial protein phosphatase 2A (PP2A), which rapidly and completely dephosphorylates Bcl-2, leading to cell death (56)
. Conversely, Bcl-2 phosphorylation by protein kinase C (PKC) has been suggested (57)
, and PKC has been shown to be inactivated by ceramide in cells and in vitro (58)
, most probably through the action of ceramide-activated protein phosphatase. Whether mitochondrial ceramide is involved in activation of a mitochondrial or extramitochondrial protein phosphatase remains to be determined.
Another potential function of mitochondrial ceramide is enhancement of the translocation to the mitochondria of the proapoptotic members of the Bcl-2 family such as Bax, Bad, and Bid. Preliminary results from our laboratory show that the generation of mitochondrial ceramide induces the translocation of Bax from the cytosol to the mitochondria (unpublished observations). In addition, ceramide was shown to potentiate the induction of mitochondrial permeability transition by Bax (59)
.
In this study, we showed that the generation of ceramide at the level of mitochondria induces cell death. These results are important at three levels. First, mitochondrial ceramide appears to be specifically associated with cell death, as the generation of ceramide in other cellular compartments was without any effect on cell death. Second, other studies showed that the generation of ceramide in other cellular compartments induces distinct cellular responses. For example, although treatment with exogenous bSMase did not affect cell death, it has been shown that this generation of ceramide at the outer leaflet of the plasma membrane is associated with inhibition of PKC-induced nuclear factor-
B activation (60)
, and with inhibition of platelet-derived growth factor-induced phosphatidyl 3-kinase activity (54)
. Thus, ceramide action appears to be compartmentalized. Third, the studies undertaken here for determination of compartment-specific function of bioactive lipids provide a novel and potentially generalized approach for probing specific responses to lipids generated in specific compartments.
An important conclusion from these observations is the presence of topologically restricted pools of ceramide that are associated with specific physiological responses. In a broader scope, these observations support the hypothesis that cells contain distinct and unique pools of lipids that may be biochemically, metabolically, and functionally distinct.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication June 29, 2001. Accepted for publication August 22, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B in normal and SV40-transformed human fibroblasts. J. Biol. Chem. 275,14760-14766This article has been cited by other articles:
|
|
 |
|
  V. Michel and M. Bakovic The solute carrier 44A1 is a mitochondrial protein and mediates choline transport FASEB J, August 1, 2009; 23(8): 2749 - 2758. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Yabu, A. Shimuzu, and M. Yamashita A Novel Mitochondrial Sphingomyelinase in Zebrafish Cells J. Biol. Chem., July 24, 2009; 284(30): 20349 - 20363. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. E. Modrak, E. Leon, D. M. Goldenberg, and D. V. Gold Ceramide Regulates Gemcitabine-Induced Senescence and Apoptosis in Human Pancreatic Cancer Cell Lines Mol. Cancer Res., June 1, 2009; 7(6): 890 - 896. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Novgorodov, T. I. Gudz, and L. M. Obeid Long-chain Ceramide Is a Potent Inhibitor of the Mitochondrial Permeability Transition Pore J. Biol. Chem., September 5, 2008; 283(36): 24707 - 24717. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. L. Holland and S. A. Summers Sphingolipids, Insulin Resistance, and Metabolic Disease: New Insights from in Vivo Manipulation of Sphingolipid Metabolism Endocr. Rev., June 1, 2008; 29(4): 381 - 402. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. J. Siskind, L. Feinstein, T. Yu, J. S. Davis, D. Jones, J. Choi, J. E. Zuckerman, W. Tan, R. B. Hill, J. M. Hardwick, et al. Anti-apoptotic Bcl-2 Family Proteins Disassemble Ceramide Channels J. Biol. Chem., March 14, 2008; 283(11): 6622 - 6630. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Stiban, L. Caputo, and M. Colombini Ceramide synthesis in the endoplasmic reticulum can permeabilize mitochondria to proapoptotic proteins J. Lipid Res., March 1, 2008; 49(3): 625 - 634. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Yu, S. A. Novgorodov, D. Chudakova, H. Zhu, A. Bielawska, J. Bielawski, L. M. Obeid, M. S. Kindy, and T. I. Gudz JNK3 Signaling Pathway Activates Ceramide Synthase Leading to Mitochondrial Dysfunction J. Biol. Chem., August 31, 2007; 282(35): 25940 - 25949. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Kitatani, J. Idkowiak-Baldys, J. Bielawski, T. A. Taha, R. W. Jenkins, C. E. Senkal, B. Ogretmen, L. M. Obeid, and Y. A. Hannun Protein Kinase C-induced Activation of a Ceramide/Protein Phosphatase 1 Pathway Leading to Dephosphorylation of p38 MAPK J. Biol. Chem., December 1, 2006; 281(48): 36793 - 36802. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. E. Modrak, D. V. Gold, and D. M. Goldenberg Sphingolipid targets in cancer therapy. Mol. Cancer Ther., February 1, 2006; 5(2): 200 - 208. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Novgorodov, Z. M. Szulc, C. Luberto, J. A. Jones, J. Bielawski, A. Bielawska, Y. A. Hannun, and L. M. Obeid Positively Charged Ceramide Is a Potent Inducer of Mitochondrial Permeabilization J. Biol. Chem., April 22, 2005; 280(16): 16096 - 16105. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. E. Modrak, T. M. Cardillo, G. A. Newsome, D. M. Goldenberg, and D. V. Gold Synergistic Interaction between Sphingomyelin and Gemcitabine Potentiates Ceramide-Mediated Apoptosis in Pancreatic Cancer Cancer Res., November 15, 2004; 64(22): 8405 - 8410. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Sakurai, M. Nagahama, and M. Oda Clostridium perfringens Alpha-Toxin: Characterization and Mode of Action J. Biochem., November 1, 2004; 136(5): 569 - 574. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Marchesini, W. Osta, J. Bielawski, C. Luberto, L. M. Obeid, and Y. A. Hannun Role for Mammalian Neutral Sphingomyelinase 2 in Confluence-induced Growth Arrest of MCF7 Cells J. Biol. Chem., June 11, 2004; 279(24): 25101 - 25111. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Ochi, M. Oda, H. Matsuda, S. Ikari, and J. Sakurai Clostridium perfringens {alpha}-Toxin Activates the Sphingomyelin Metabolism System in Sheep Erythrocytes J. Biol. Chem., March 26, 2004; 279(13): 12181 - 12189. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. V. de Avalos, Y. Okamoto, and Y. A. Hannun Activation and Localization of Inositol Phosphosphingolipid Phospholipase C, Isc1p, to the Mitochondria during Growth of Saccharomyces cerevisiae J. Biol. Chem., March 19, 2004; 279(12): 11537 - 11545. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Perry and N. D. Ridgway The role of de novo ceramide synthesis in the mechanism of action of the tricyclic xanthate D609 J. Lipid Res., January 1, 2004; 45(1): 164 - 173. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B.-J. Kroesen, S. Jacobs, B. J. Pettus, H. Sietsma, J. W. Kok, Y. A. Hannun, and L. F. M. H. de Leij BcR-induced Apoptosis Involves Differential Regulation of C16 and C24-Ceramide Formation and Sphingolipid-dependent Activation of the Proteasome J. Biol. Chem., April 18, 2003; 278(17): 14723 - 14731. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Marchesini, C. Luberto, and Y. A. Hannun Biochemical Properties of Mammalian Neutral Sphingomyelinase2 and Its Role in Sphingolipid Metabolism J. Biol. Chem., April 11, 2003; 278(16): 13775 - 13783. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. A. Cowart, Z. Szulc, A. Bielawska, and Y. A. Hannun Structural determinants of sphingolipid recognition by commercially available anti-ceramide antibodies J. Lipid Res., December 1, 2002; 43(12): 2042 - 2048. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Luberto, D. F. Hassler, P. Signorelli, Y. Okamoto, H. Sawai, E. Boros, D. J. Hazen-Martin, L. M. Obeid, Y. A. Hannun, and G. K. Smith Inhibition of Tumor Necrosis Factor-induced Cell Death in MCF7 by a Novel Inhibitor of Neutral Sphingomyelinase J. Biol. Chem., October 18, 2002; 277(43): 41128 - 41139. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Garcia-Ruiz, A. Colell, A. Morales, M. Calvo, C. Enrich, and J. C. Fernandez-Checa Trafficking of Ganglioside GD3 to Mitochondria by Tumor Necrosis Factor-alpha J. Biol. Chem., September 20, 2002; 277(39): 36443 - 36448. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Le Stunff, I. Galve-Roperh, C. Peterson, S. Milstien, and S. Spiegel Sphingosine-1-phosphate phosphohydrolase in regulation of sphingolipid metabolism and apoptosis J. Cell Biol., September 16, 2002; 158(6): 1039 - 1049. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. A. Hannun and L. M. Obeid The Ceramide-centric Universe of Lipid-mediated Cell Regulation: Stress Encounters of the Lipid Kind J. Biol. Chem., July 12, 2002; 277(29): 25847 - 25850. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
