(Received for publication, May 5, 1997)
From the Department of Molecular & Experimental Medicine, The Scripps Research Institute, La Jolla, California 92037
Fas-driven apoptosis in Jurkat cells results in the inactivation of cytochrome c with cessation of oxygen consumption. Overexpression of Bcl-2 was found to protect against acidification and apoptosis mediated by Fas ligation in these cells. Bcl-2 is present in the outer mitochondrial membrane, but the molecular mechanism by which it protects cells is unknown. Because Bcl-2 projects into the mitochondrial intermembrane space and cytochrome c is located in the intermembrane space, we considered the possibility that Bcl-2 might protect cytochrome c from inactivation during Fas-mediated apoptosis. The present study shows that 1) in Jurkat cells, cytochrome c inactivation during Fas-driven apoptosis requires the permeabilization of the outer mitochondrial membrane; and 2) the post-mitochondrial fraction from CEM cells that overexpress Bcl-2 both prevents and reverses cytochrome c inactivation.
Apoptosis refers to a metabolic program that when activated causes a cell to commit suicide according to a stereotyped sequence of events that culminate in the fragmentation of the cell and the ingestion of its remains by mononuclear phagocytes. The net effect of this program is to annihilate the cell without releasing any of its contents into the surrounding medium. Among the characteristic events that occur during apoptosis are the transfer of phosphatidylserine to the outer leaflet of the cell membrane; the cleavage of certain proteins such as poly(ADP-ribose) polymerase by the caspases, a set of aspartate-specific cysteine proteases that are activated during apoptosis; intracellular acidification; extensive cross-linking of intracellular proteins; and the destruction of the genome with the release of nucleosome-sized fragments of DNA (1).
Several earlier studies have suggested that cytochrome c is an important participant in the apoptosis program. We found that the cytochrome was inactivated during Fas-mediated apoptosis in Jurkat cells though it remained associated with the mitochondria (2). Inactivation was mediated by a heat-labile cytosolic factor that we now call cytochrome c-inactivating factor of apoptosis (CIFA).1 Other groups, however, reported that cytochrome c was lost from mitochondria during apoptosis and that cytochrome c once released into the cytosol promoted the processing of caspase-3 to its active form, leading to the completion of apoptosis (3-5).
To try to clarify some of these questions and to learn more about the mechanism by which CIFA inactivates cytochrome c, we examined the effect of CIFA-containing post-mitochondrial (PM) fraction on electron transport in isolated mitochondria. In the present study, we show that the action of CIFA depends in part on the permeabilization of the mitochondrial outer membrane, and demonstrate that its action is opposed by the anti-apoptotic protein Bcl-2.
Jurkat, CEM/Neo, and CEM/Bcl-2 cell lines were maintained in RPMI 1640 with 5% fetal calf serum and 2 mM L-glutamine. Apoptosis was induced by incubation with 50 ng/ml anti-Fas antibody (clone CH-11, Kamiya Biomedical Co., Thousand Oaks, CA) at a density of 107 cells/ml for 5 to 60 min in RPMI 1640 supplemented with 1% fetal calf serum. Controls were incubated in parallel. For incubations extending to 90 min, 5% fetal calf serum was included. Extent of apoptosis was evaluated by acridine orange staining and scoring of condensed or fragmented nuclei.
Preparation of Cell FractionsOne hundred million cells
were suspended in 1 ml of cavitation buffer (100 mM
sucrose, 20 mM MOPS (pH 7.4, except where noted), 1 mM EGTA, 1 mg/ml bovine serum albumin, 10 µM
triethanolamine, 5% Percoll, 1 mM phenylmethylsulfonyl
fluoride, 10 µM each aprotinin, pepstatin A, and
leupeptin) and disrupted by nitrogen cavitation. Except where noted,
digitonin (0.01%) was added to the cavitation buffer. After
cavitation, a portion of the cell lysate was stored at 70 °C until
used for spectrophotometry. The remainder was centrifuged twice at
2500 × g for 5 min at 4 °C to remove unbroken cells
and nuclei, which were resuspended in 1 ml of cavitation buffer and
stored at
70 °C as the nuclear fraction until used for
spectrophotometry. Electron microscopy of the nuclear pellet demonstrated the presence of nuclei and some mitochondria, as well as
occasional intact cells. Centrifugation of the 2500 × g supernatant at 10,000 × g for 15 min at
4 °C provided a pellet representing the mitochondrial fraction,
which was resuspended either in respiration buffer (0.25 M
sucrose, 0.1% bovine serum albumin, 10 mM
MgCl2, 10 mM K/Hepes, 5 mM
KH2PO4, pH 7.4) or in 10,000 × g supernatant (the PM fraction) as noted in figure legends, and used immediately for oxygen uptake measurements. Alternatively, the
mitochondrial fraction was resuspended in mitochondrial buffer (300 mM sucrose, 1 mM EGTA, 20 mM MOPS
(pH 7.4), 1 mg/ml bovine serum albumin, and 10 µM each
aprotinin, pepstatin, and leupeptin) and stored at
70 ° for
spectroscopy. The PM fraction was also stored at
70 °C. Although
this PM fraction contained some mitochondria, freeze-thawing eliminated
the contribution to cytochrome c activity as measured by the
oxygen electrode.
Cytochrome c activity was measured in an oxygen electrode essentially as described previously (2), reducing the cytochrome with ascorbate plus tetramethylphenylenediamine (TMPD). All results presented are representative of two or three similar experiments and are corrected for azide-independent oxygen uptake.
Measurement of Cytochrome ConcentrationCytochromes were
measured spectrophotometrically. Difference spectra were recorded using
a Perkin-Elmer Lambda 18 spectrophotometer over the wavelength range of
650-400 nm. Mitochondrial fractions (derived from 2 × 108 to 1 × 109 cells) were suspended in
1.4 ml of 50 mM potassium phosphate buffer, pH 7.0, by
aspiration. Under these conditions, all the cytochromes present were
oxidized as determined by pilot experiments in which a small quantity
of potassium ferricyanide was added (as oxidant) to the reference
cuvette. Reduction of cytochrome c was achieved by the
addition of 2 mM ascorbate/200 µM TMPD (final concentrations) to the sample cuvette. The sample was scanned at
intervals from 30 s to 10 min after the addition of the reductant. To ensure that full reduction of cytochrome c had occurred,
5 µM potassium cyanide was added (to inhibit cytochrome
oxidase), and the spectrum was rerecorded. Finally, the sample was
fully reduced by the addition of a few grains of sodium dithionite
before recording the dithionite-reduced minus air-oxidized spectrum. The latter spectrum was used to estimate the content of cytochrome oxidase (cytochrome a/a3). Cell cavitates,
nuclear fractions, and PM fractions (derived from 2 × 108 to 1 × 109 cells in 1.4 ml) were
found to be partially reduced. The cytochrome c content was
determined by recording ascorbate/TMPD reduced minus ferricyanide-oxidized difference spectra. The total cytochrome content
was estimated after the addition of dithionite to the sample cuvette.
The following extinction coefficients were used: cytochrome
c (550 nm) = 18.5 mM1
cm
1 (6) and cytochrome oxidase (cytochrome
a/a3) (605 nm) = 16 mM
1 cm
1 (7).
We previously showed that cytosol from Jurkat cells that had been
driven into apoptosis with anti-Fas IgM (Fas cytosol) contained CIFA, a
heat-labile factor that inactivated electron transport through
cytochrome c in digitonin-permeabilized cells (2). In the
present studies, we examined the effect of CIFA on isolated mitochondria. In accord with earlier findings in whole cells, we found
that CIFA blocked cytochrome c-mediated electron transport in mitochondria, as measured by N3-sensitive
ascorbate-dependent oxygen uptake. Our experiments further
showed that CIFA was generated as early as 5 min after Fas ligation
(Fig. 1) and that its effect on
cytochrome c in isolated mitochondria was blocked by the
general caspase inhibitor ZVAD-fluoromethylketone (ZVAD·fmk),
suggesting that CIFA may be generated by a caspase or may itself be a
death protease (Fig. 1).
For the foregoing studies, the cells had been treated with digitonin, which permeabilizes the outer membranes of the mitochondria. To determine whether the outer membrane influences the action of CIFA, we carried out experiments with mitochondria prepared from control and anti-Fas-treated cells in the absence of digitonin (designated digitonin-free mitochondria). Outer membranes of mitochondria isolated in this way retain their natural impermeability. In digitonin-free mitochondria, cytochrome c-dependent ascorbate oxidation was resistant to brief incubation with Fas PM fraction (141.4 ± 19.5 and 136.5 ± 34.1 ng at oxygen/min/1.6 × 108 cell eq of mitochondria, respectively, for mitochondria incubated for 5 min with PM fraction from normal and Fas-treated cells). It thus appears that the outer membrane protects cytochrome c from CIFA. Nevertheless, cytochrome c inactivation appeared to occur during apoptosis in intact cells because exposure of intact cells to anti-Fas IgM for 90 min led to a fall in oxygen consumption (using TMPD/ascorbate as electron donor), reflecting an inhibition by 31 and 42% in two independent experiments (apoptosis assessed by nuclear morphology (chromatin condensation/nuclear fragmentation) amounted to 48 and 44%, respectively). This finding implies that during apoptosis, the outer mitochondrial membrane is rendered permeable to CIFA, allowing the factor access to the cytochrome c in the intermembrane space.
We have shown that cytoplasmic acidification to pH 6.8 and below is a feature of Fas-mediated apoptosis (8, 9). To determine whether this fall in pH might be responsible for the permeabilization of the mitochondrial outer membrane within the apoptotic cells, we examined the effect of lowering the cytosolic pH on the inactivation of cytochrome c in digitonin-free mitochondria (i.e. mitochondria with an intact outer membrane). For this purpose, we prepared digitonin-free mitochondria from Fas-treated cells and incubated them with Fas PM fraction prepared by disrupting cells in cavitation buffer adjusted at 37 °C to pH 6.8 or 7.2. We found that cytochrome c-dependent oxygen consumption was reduced by almost 25% when mitochondria were treated with Fas PM fraction at pH 6.8 compared with incubation at pH 7.2 or incubation with control PM fraction at either pH (Table I). Thus, the mitochondrial outer membrane becomes permeable to CIFA at a pH value achieved by cells undergoing apoptosis. Interestingly, digitonin-free mitochondria prepared from control cells were not susceptible to CIFA at either pH, suggesting that Fas ligation modifies the outer mitochondrial membrane as well as generating CIFA.
|
Because Bcl-2 is located in the outer mitochondrial membrane (10, 11),
it was of interest to examine the effect of this anti-apoptotic protein
on cytochrome c inactivation. For this purpose, we used the
T-lymphoblast cell line CEM stably transfected with Bcl-2 (CEM/Bcl-2)
or the empty expression vector
(CEM/Neo)2 (9). Expression of
Fas is similar in the two lines (9). Treatment of these cells with
anti-Fas IgM as described above led within 30 min to morphologic
features of apoptosis in 52.5 ± 2.5% of the CEM/Neo cells, but
in <10% of the CEM/Bcl-2 cells. In addition, digitonin-treated
mitochondria from Bcl-2-expressing cells were completely resistant to
inactivation by CIFA, while the mitochondria from CEM/Neo cells were
susceptible to CIFA (Fig. 2).
Although PM fraction from Fas-treated CEM/Neo cells could inactivate
cytochrome c, PM fraction from Fas-treated CEM/Bcl-2 cells
could not (not shown). It appeared that overexpression of Bcl-2 either
prevented the formation or nullified the action of CIFA in the same PM
fraction. When we incubated digitonin-treated CEM/Neo mitochondria with
PM fraction from control CEM/Bcl-2 cells for 5 min, washed them, and
then incubated them with CIFA (i.e. PM fraction from CEM/Neo
Fas-treated cells), we found that they were resistant to cytochrome
c inactivation (Fig. 3). In
contrast, pretreatment of mitochondria with PM fraction from control
CEM/Neo cells did not confer resistance. Thus a protective factor from CEM/Bcl-2 cells was able to interact with the mitochondria in such a
way as to confer resistance to CIFA.
More surprising was the finding that the inactivation of cytochrome c by CIFA could be reversed in part by PM fraction from the CEM/Bcl-2 cells. This was demonstrated by treating normal mitochondria with apoptotic PM fraction from Jurkat cells, then reisolating them and incubating them either with CEM/Neo PM fraction or CEM/Bcl-2 PM fraction, and finally reisolating them again and measuring rates of cytochrome c-dependent oxygen uptake. Rates of oxygen uptake by control and apoptotic mitochondria incubated with CEM/Neo PM fraction were comparable to rates seen in experiments in which the mitochondria had not been incubated with CEM/Neo PM fraction (not shown). Incubation of the apoptotic mitochondria with CEM/Bcl-2 PM fraction, however, resulted in a substantial restoration of oxygen uptake (Table II). Incubation of control mitochondria with the CEM/Bcl-2 PM fraction had no effect.
|
Several recent reports have suggested that during apoptosis, most of the cytochrome c is released from the mitochondria, subsequently promoting caspase activation in the cytosol (3, 12, 13). Our observation that cytochrome c-dependent oxygen uptake by apoptotic mitochondria can be partially or completely reversed by CEM/Bcl-2 PM fraction, however, is not fully consistent with these reports. We therefore investigated the localization of cytochrome c in subcellular fractions from normal and apoptotic cells, comparing the quantities of cytochrome c with the amounts of cytochrome oxidase in these fractions to distinguish cytochrome c release from mitochondrial contamination. The results (Table III) show that the distribution of cytochrome c among nuclear, mitochondrial, and post-mitochondrial fractions was the same in normal and apoptotic cells. In the nuclear, mitochondrial, and PM fractions, the ratios of cytochrome c to cytochrome oxidase were roughly equal, suggesting that by far the majority of the cytochrome c in the nuclear and post-mitochondrial fractions can be accounted for by mitochondrial contamination. Electron microscopy of the nuclear fraction confirmed a similar amount of mitochondrial contamination in both control and Fas-treated samples (not shown). Finally, the amount of cytochrome c (and other mitochondrial cytochromes) in the PM fraction did not vary with the length of incubation in anti-Fas IgM or the extent of apoptosis (measured by nuclear condensation), suggesting that if mitochondrial cytochrome c release occurs in this system, it was not apparent even when more than 50% of the cells have undergone chromatin condensation and nuclear fragmentation.
|
Mitochondria have been implicated in several of the events of apoptosis (14-17), including caspase activation and chromatin cleavage (3, 12, 13). Earlier studies from our laboratory showed that within 5 min after Fas ligation, PM fraction from Jurkat cells acquired a ZVAD-sensitive factor (CIFA) capable of inactivating electron transport through cytochrome c. This observation raised the question of access because cytochrome c is normally separated from the CIFA-containing cytosol by the outer mitochondrial membrane, whereas the earlier investigations had been conducted with mitochondria whose outer membranes had been permeabilized with digitonin. It was, therefore, important to examine the effect of CIFA on mitochondria with intact outer membranes. The studies described above showed that CIFA could not penetrate the outer membranes of normal mitochondria, but the outer membranes of apoptotic mitochondria were permeable to CIFA, provided the pH of the buffer in which the mitochondria were suspended was on the acid side of neutrality. Our results suggest that a potential mechanism for the penetration of the outer mitochondrial membrane by CIFA is a Fas-dependent alteration in the membrane that causes its permeability to increase as the cytoplasm acidifies.
Bcl-2, a protein located inter alia in the outer mitochondrial membrane, was found to protect mitochondria against the effects of CIFA. Bcl-xL, a member of the Bcl-2 family, (18) is structurally related to ion channels (19, 20), suggesting that the effect of Bcl-2 might be mediated through an effect on ion transport across this membrane. Bcl-2 must have protected the mitochondria against CIFA through at least one mechanism unrelated to transport, however, since its protective effect was seen in mitochondria treated with digitonin, a procedure that renders the outer membrane permeable to molecules at least as large as trypsin (molecular mass 24 kDa) (21). Our observations with Bcl-2 suggest the possibility of a direct interaction between Bcl-2 and cytochrome c or CIFA. Regardless of mechanism, however, we propose that the protection conferred on cytochrome c by Bcl-2 is able to prevent or delay mitochondrial dysfunction and the other apoptotic events in which the cytochrome is thought to participate (3, 12-17).
Several recent reports have suggested that during apoptosis, cytochrome c is released from the mitochondria and supports the activation of cytosolic caspases (3-5). Our finding that Bcl-2 is able to partially reverse the CIFA-mediated inactivation of cytochrome c led us to study cytochrome c localization in our system. The ability of Bcl-2 to reverse the inactivation of electron transport through cytochrome c indicates that the cytochrome was still present within the mitochondria and argues against its loss from the mitochondria or its irreversible modification by CIFA. The results of spectroscopic measurements also strongly suggest that cytochrome c remained associated with the mitochondria as the cells executed their apoptosis program. It therefore appears that the egress of cytochrome c from the mitochondria is not always required for caspase activation and apoptosis and that, in this system at least, the subcellular location of cytochrome c did not change during apoptosis.
We gratefully acknowlege the technical assistance of Mr. Grant Meisenholder.