| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Unité de Recherche en Immunologie Cellulaire et Moléculaire, Inserm U364, Faculté de Médecine, 06107 Nice Cedex 2, France; and
Laboratoire d'Anatomie Pathologique, Hôpital Pasteur, 06002 Nice Cedex 1, France
1Correspondence: Unité de Recherche en Immunologie Cellulaire et Moléculaire, Inserm U 364, Faculté de Medecine, Avenue de Valombrose, 06107-Nice Cedex 2, France. E-mail : lanteri{at}unice.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: CAD • DFF • hyperosmotic shock • ICAD
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The
apoptotic process is driven by a genetically determined cell suicide
program, and alterations of this process can contribute to diseases
that include cancer or degenerative disorders (2)
.
Cells committed to apoptotic death show typical structural changes such
as volume reduction, membrane blebbing, chromatin condensation, and DNA
degradation in high molecular weight (50–300 kb) as well as
oligonucleosomal fragments (3
, 4)
. The execution phase of
apoptosis is characterized by a strictly coordinated and stereotyped
process, independent of the apoptogenic agent. Indeed, various
initiators of PCD converge in inducing mitochondrial alterations, which
play a central role in apoptotic cell death (5)
. These
alterations include formation of a megaspore at contact sites between
the mitochondrial membranes, matrix swelling, loss of the inner
membrane electrochemical gradient (
m), and
cytochrome c release, which accompanies PCD in every
circumstances studied up to now (6)
. Once released in the
cytosol, it promotes the formation of a multiprotein complex, which
leads to the activation of caspase-9 and of the downstream caspase-3
(7
, 8)
.
Caspases are a family of proteases endowed with a specificity toward
aspartate residues, which are proteolytically processed from their
inactive zymogenic form to active heterodimeric enzymes
(9)
. In the last few years, the central role played by
these proteases in the execution phase of apoptosis has been firmly
established. Among them, caspase-3 is directly involved in the cleavage
of many protein substrates (9)
. One of these is the
inhibitory subunit of the heterodimeric DNA fragmentation factor
(DFF-45) (10)
or its murine homologue ICAD (inhibitor of
caspase-activated DNase; 11
). These proteins sequester in
an inactive form the endonuclease that causes the internucleosomal DNA
degradation, termed CAD (caspase-activated DNase) (12
, 13)
, DFF-40 (14)
, or CPAN (caspase-activated
nuclease; 15
). During the apoptotic process, ICAD/DFF-45
cleavage by caspase-3 releases CAD/DFF-40, thus activating it
(11)
. Indeed, neither caspase-3-deficient cells nor mice
lacking ICAD/DFF-45 display the typical DNA laddering caused by
nucleosomal DNA fragmentation when apoptosis is triggered (16
, 17)
. CAD/DFF-40 activity has also been proposed to be
responsible for chromatin condensation (14
, 17
, 18)
and to
generate the high molecular weight fragments of DNA that appear before
its oligonucleosomal degradation (19)
.
In the execution phase of apoptosis, the observed cytosolic shrinkage
is very important in that it prevents cells from swelling up to
lysis and releasing their potentially toxic cytoplasmic content into
the surrounding tissues (1)
. Cells passively shrink when
exposed to a hyperosmotic environment, and a hyperosmotic shock can be
used to induce apoptosis in some cell types (20
21
22
23)
.
Thus, cell shrinkage seems to have both a general function during the
PCD program and to be a proapoptotic signal by itself. Cells can also
reduce their volume by a loss of K+, Cl-, and
organic osmolytes (as betaine, taurine, myoinositol, sorbitol). Organic
osmolytes and Cl- permeate through outwardly rectifying
Cl- channels (ORCCs; reviewed in ref 24
). Therefore, these
channels could play a role in the apoptotic cytosolic shrinkage.
In this study, we provide evidence that exposure of hematopoietic cells to a hyperosmotic shock activates apoptosis. Regardless of the nature of the apoptotic agent, our results indicate that DFF-45 cleavage is not sufficient per se to activate the DFF endonuclease, whose activity could be controlled also by a Cl- efflux-dependent process. Moreover, we could observe a dissociation between oligonucleosomal DNA fragmentation and other nuclear catabolic events that accompany apoptosis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cells and culture media
Human T leukemia Jurkat cells, human lymphoblastic leukemia CEM
cells, and human monocytic THP-1 cells were grown in RPMI 1640 culture
medium supplemented with 5% fetal bovine serum (Gibco BRL,
Gaithersburg, Md.) and 2 mM L-glutamine. The final osmolarity of the
medium was 290 mosM. Hyperosmotic conditions (500 mosM) were obtained
by directly adding 210 mM mannitol, sorbitol, myoinositol, betaine, or
taurine or 105 mM NaCl to the culture medium. Hypoosmotic conditions
(200 mosM) were obtained by adding deionized water to the culture
medium.
Flow cytometric analysis
Cytometric recordings of m and cell
surface exposure of PS were performed simultaneously, as described
(25)
. Briefly, after induction of apoptosis,
106 Jurkat cells were resuspended in HEPES buffer
(10 mM HEPES, 150 mM NaCl, 5 mM CaCl2). Cells
were then incubated for 15 min at 37°C in FITC-conjugated Annexin-V
(Boehringer Mannheim, Indianapolis, Ind.) and chloromethyl Xarosamine
(CMXRos, 50 nM; Molecular Probes, Eugene, Oreg.). After placing the
samples in HEPES buffer containing 1 µg/ml of propidium iodide
(Sigma), they were analyzed on a FACScan flow cytometer (Becton
Dickinson, Mountain View, Calif.). Data acquisition and analysis were
performed using CellQuest software. We used forward and side scatters
to eliminate debris; Annexin-V FITC (FL1) and CMXRos (FL2) fluorescent
signals were then analyzed on a density plot. In control experiments,
cells were incubated in the presence of the mitochondrial uncoupling
agent carbonyl cyanide m-chlorophenol-hydrazone.
DNA fragmentation assays
Oligonucleosomal DNA fragmentation was analyzed by
electrophoresis on agarose gel and by using the cytofluorometric
TdT-mediated dUTP-X nick end labeling (TUNEL) technique. After
induction of apoptosis 3 x 106 cells were
lysed in a buffer containing 10 mM Tris, 1 mM EDTA, and 0.2% Triton
X-100, pH 8.0. Samples were incubated in 100 µg/ml RNase A (30 min,
37°C) and 100 µg/ml proteinase K (10 min, 56°C), and DNAs were
precipitated in 0.5 M NaCl-isopropanol, washed in 70% ethanol, and
loaded on a 1.5% agarose gel. Internucleosomal DNA fragmentation was
studied in vitro on isolated nuclei or high molecular weight
calf thymus DNA (Boehringer Mannheim). Nuclei were prepared by lysing
5 x 106 control Jurkat cells in 10 mM NaCl,
10 mM Tris/HCl, pH 7.5, 3 mM MgCl2, 20% glycerol
in the presence of phosphatase and protease inhibitors (1 mM vanadate,
1 µg/ml leupeptin, 1 µM pepstatin, 1 mM PMSF) and at 4°C. Cells
were further disrupted with 15 strokes in a tissue grind pestle
(Kontes, Vineland, N.J.); thereafter, 0.5% Nonidet P-40 was
added and cytosols were eliminated by centrifugation. Either nuclei or
2.5 µg high molecular weight DNA were incubated for 2 h at
37°C with 450 µg of cytosolic extracts in a final volume of 300
µl. DNA was then extracted following the above protocol, but the
lysis step was suppressed when using high molecular weight DNA.
Cytosolic extracts were prepared by lysing Jurkat cells in 10 mM NaCl,
20 mM Tris/HCl, pH 7.5, 1 mM CaCl2, 1% Nonidet P-40 with
phosphatase and protease inhibitors (see above). When indicated, NaCl
concentration was augmented after lysis to 135 mM or 250 mM or lysis
was performed at a different pH (6 or 9). The TUNEL technique was
applied on 106 cells by the use of the in
situ cell death detection kit (Boehringer Mannheim) following
manufacturer instructions.
High molecular weight DNA fragmentation was assessed by pulsed-field gel electrophoresis (PFGE). Briefly, cells were washed and resuspended in phosphate-buffered saline and mixed with a double volume of 1% low melting point agarose. The mixture was allowed to solidify and blocks were treated overnight at 50°C with a solution containing: 250 mM EDTA, pH 8.0, 1% sarkosyl, and 100 µg/ml proteinase K. Samples were then washed in 10 mM Tris, 1 mM EDTA, pH 8.0. PFGE was performed at 12°C using a handmade CHEF-type gel box, with 20 s pulses for 20 h at 6 V/cm. The gel was then stained with ethidium bromide (0.5 µg/ml) and photographed.
Western immunoblot analysis
Cytosolic extracts were prepared as described above, but with
135 mM NaCl in the lysis buffer, and loaded on sodium dodecyl
sulfate-polyacrylamide gels. Proteins were then blotted to Immobilon-P
membranes (Millipore, Bedford, Mass.) and transferred 3 h at 500
mA, 4°C in transfer buffer (20 mM Tris-base, 150 mM glycine, 20%
methanol). Nonspecific binding was blocked by 1 h incubation in a
buffer composed of 3% bovine serum albumin, 0.5% gelatin, 0.1% Tween
20, 1 mM EDTA, 150 mM NaCl, 10 mM Tris/HCl, pH 7.4. Antibodies were
incubated at 4°C overnight and horseradish peroxidase-conjugated
secondary antibodies were added for 1 h at room temperature.
Proteins were visualized by enhanced chemiluminescence (Amersham,
Little Chalfont, U.K.). To detect activated caspases, cytosolic
fractions were incubated for 10 min with 20 µM Ac-YVK(bio)D-CHO,
resolved on a 15% sodium dodecyl sulfate-polyacrylamide gel, and
transferred as described above. Labeled caspases were visualized with
horseradish peroxidase-conjugated streptavidin and enhanced
chemiluminescence.
Electron microscopy analysis
To perform electron microscopy (EM) analysis,
107 cells were fixed in 2.5% glutaraldehyde in
phosphate-buffered saline for 1 h at 4°C and then postfixed in
1% osmium tetroxide. After dehydration in a series of graded ethanol
baths (30% to 100%) and then in propylene oxide, cells were embedded
in Epon (Taab, France). Cell sections (80–200 nm) were obtained using
a Reichert Ultracut E microtome and stained with uranyl acetate. Grids
were examined with a Jeol 1200 EXII electron microscope.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and the breakdown of
the inner mitochondrial membrane electrochemical gradient
(m). Cytofluorometric analysis was
performed only on cells that displayed plasma membrane integrity as
measured by propidium iodide exclusion. The lower right quarters of
diagrams depicted in Fig. 1
A correspond to cells exhibiting both a loss of
m and plasma membrane outward flipping of
PS. Positive cells for both types of apoptotic markers accumulated
independent of the substance used to increase extracellular osmolarity,
indicating that by no means were counterbalancing mechanisms efficient
in Jurkat cells. The low number of apoptotic cells in NaCl
hyperosmotic conditions was due to experimental variability and was
significantly higher than in controls (values ± SD
were 13.31 ± 6.5% in NaCl vs.1.43 ± 0.47% in basal,
n=7; P<0.001 with a Student's
t test analysis). All the compounds strongly induced
oligonucleosomal DNA fragmentation, with the remarkable exception of
NaCl (Fig. 1
B). This result was confirmed by TUNEL
cytofluorometric analysis, as shown in Fig. 1
C, where
the efficiency of NaCl and mannitol to induce double-strand DNA nicking
was compared. This observation was reproduced on the lymphoblastic
leukemia CEM cell line and the monocytic THP-1 cell line. As
exemplified in Fig. 1
D, mannitol but not NaCl caused the
characteristic DNA laddering pattern, whereas both effectors induced
the loss of m and PS plasma membrane flipping (not
shown).
|
Dose-response analysis on Jurkat cells confirmed that after 4 h in
500 mosM and 600 mosM media, nucleosomal DNA fragmentation was
detectable when cells were treated with mannitol but not with NaCl
(Fig. 2
A). Instead, loss of m and surface exposure
of PS were observed with both effectors (Fig. 2
C). A weaker hyperosmotic shock (400 mosM) could
not trigger any apoptosis hallmark independent of the compound added to
the medium (Fig. 2
A, C). In 600 mosM media, the
strong apoptosis induction caused an upward shift of the CMXRos-stained
cell population (see also Fig. 5
B), probably because
under these conditions, apoptotic cells underwent a strong cell volume
reduction that artifactually increased the CMXRos intracellular
concentration. We verified by using the mitochondrial uncoupling agent
carbonyl cyanide m-chlorophenol-hydrazone that in the presence of 600
mosM NaCl, the cells situated in the right part of the figure did
present a collapsed m.
|
|
By exposing cells to a 500 mosM medium, the m
collapse was measurable as soon as 1 h, whereas flipping of PS on
the cell surface appeared only after 2 h both with mannitol and
NaCl. The mannitol-induced oligonucleosomal DNA cleavage occurred
concomitantly to PS flipping (Fig. 2
B, D). Cell exposure
for up to 8 h to a 500 mosM NaCl hyperosmotic medium was not
sufficient to induce any DNA laddering (not shown).
Both mannitol and NaCl hyperosmotic shocks activate caspases
Cytosolic lysates of Jurkat cells exposed to mannitol or
NaCl hyperosmotic shocks were incubated with the biotinylated
tetrapeptide Ac-YVK(bio)D-CHO, which labels with low selectivity the
cleaved (i.e., activated) forms of caspases. The two apoptotic
conditions displayed the same pattern of Ac-YVK(bio)D-CHO
affinity-labeled caspases (Fig. 3
A). Pretreatment with Ac-DEVD-CHO, a blocker of
caspase-3-like enzymatic activities, strongly reduced the
Ac-YVK(bio)D-CHO labeling, likely because of a competition between the
two compounds for the same enzymatic site. Moreover, both mannitol and
NaCl treatment induced generation of the activated 17 kDa form of
caspase-3, the enzyme directly responsible for DFF activation (Fig. 3B
). Ac-DEVD-CHO partially inhibited the formation of this
fragment and induced the accumulation of a 20 kDa band. This probably
corresponds to a cleaved form of caspase-3 which has retained its
amino-terminal domain, suggesting the existence of a two-step cleavage
of the enzyme blocked by Ac-DEVD-CHO after processing of the
carboxyl-terminal fragment. Furthermore, poly (ADP-ribose) polymerase
(PARP), a nuclear target of caspase-3-like proteases (27)
,
was cleaved into the expected 85 kDa fragment irrespective of the
effector used to increase extracellular osmolarity (Fig. 3C
). Ac-DEVD-CHO abolished both PARP cleavage (Fig. 3C
) and mannitol-induced oligonucleosomal DNA fragmentation
(Fig. 3D
). A time course analysis was performed to rule out
an artifactual caspase activation due to cell extract preparation.
However, neither caspase-3 nor PARP cleavage was measurable up to
1 h after the hyperosmotic shock (data not shown).
|
In addition, in both hyperosmotic conditions, Ac-DEVD-CHO induced
the accumulation of a cell population that exhibited a low
mitochondrial CMXRos uptake, but was not labeled by FITC-annexin V
(lower left quarters of diagrams in Fig. 3E
). Therefore,
caspase-3-like protease activation was required for cell surface PS
exposure but not for m dissipation.
Electron microscopy analysis of Jurkat cells after mannitol and
NaCl hyperosmotic shock
We compared by EM inspection Jurkat cells exposed to mannitol or
NaCl. The two hyperosmotic conditions caused the same nuclear
morphological alterations (wide arrows in Fig. 4
C, E), i.e., perinuclear chromatin condensation, disruption
of the nuclear envelope with karyorrhexis, and formation of apoptotic
bodies containing condensed nuclear material. Thus, both hyperosmotic
shocks led to complete nuclear degradation, the formation of cytosolic
vesicles (thin arrows in Fig. 4C, E
), and marked
mitochondrial swelling, with loss of internal membrane cristae (Fig. 4D, F
). Ruptures of the external mitochondrial
membrane were also visible (arrowhead in the inset of Fig. 4D
). Some cells displaying mitochondrial alterations did not
yet exhibit karyorrhexis (arrows in Fig. 4D
indicate the
nuclear envelope), thus confirming that mitochondria alterations are an
early event in the apoptotic cascade.
|
Alterations of Cl- efflux preclude internucleosomal
DNA degradation
As increasing extracellular NaCl creates an unfavorable
electrochemical gradient for Cl- efflux, a high
NaCl medium could alter the cytosolic ionic composition in such a way
as to inhibit nucleosomal DNA cleavage during the apoptotic process. A
loss of intracellular Cl- is normally associated
to cell volume reduction, which is a late apoptotic event common to
most apoptotic pathways. Therefore, we investigated whether
hyperosmotic NaCl could prevent the nucleosomal DNA fragmentation
induced by other proapoptotic stimuli, i.e., cross-linking of Fas by
using a specific agonist antibody or treatment with staurosporine
(STS). The DNA laddering elicited in Jurkat cells by these proapoptotic
agents was abolished when they were added in a medium made hyperosmotic
by increasing NaCl (Fig. 5
A). Under these conditions, various other apoptotic hallmarks
were nevertheless present such as loss of
m, cell surface exposure of PS (Fig. 5B
), as well as cleavage of caspase-3 and PARP (Fig. 5C, D
).
In addition to the inhibition of Cl- conductance
by using two unrelated ORCC blockers, 1,9-dideoxyforskolin (ddFSK) and
NPPB, abolished oligonucleosomal DNA fragmentation induced by exposure
to hyperosmotic mannitol (Fig. 6
A). Nevertheless, the persistence of PARP cleavage indicated
that they did not abrogate apoptosis (Fig. 6B
). Furthermore,
nucleosomal DNA cleavage induced by STS was also inhibited in the
presence of ddFSK or NPPB (Fig. 6A
). If the hyperosmotic
shock was provided with Na-gluconate, thus adding to the medium an
anion larger than Cl- and poorly permeable
through ORCCs, DNA laddering was again detectable, even if at a lower
extent, and DNA fragmentation induced by STS was not inhibited (Fig. 6C
). To investigate whether a Cl-
efflux through ORCCs could be sufficient to elicit nucleosomal DNA
fragmentation, we exposed cells to a 200 mosM hypoosmotic medium. ORCCs
are strongly activated under these conditions, but after 4 h no
DNA cleavage appeared (Fig. 6C
) and no other
apoptotic feature was measurable (not shown). To better delineate at
which phase of the PCD program a Cl- efflux
could intervene in the control of nucleosomal DNA fragmentation, we
observed that NPPB addition up to 2 h after apoptosis triggering
with mannitol could still inhibit the DNA laddering formation (Fig. 6D
).
|
Cl- efflux does not intervene in ICAD/DFF-45 cleavage
or high molecular weight DNA degradation
We asked whether activation of the heterodimeric endonuclease DFF
correlated with oligonucleosomal DNA fragmentation. Jurkat cells were
exposed both to STS and hyperosmotic mannitol, which caused nucleosomal
DNA degradation, and to conditions in which this event was not
detectable and Cl- efflux was impaired (namely,
STS or mannitol) in the presence of the two ORCC blockers or
hyperosmotic NaCl with or without STS. Remarkably, the cleavage of
ICAD/DFF-45, the inhibitory subunit of DFF, was always observed even if
it occurred at a variable extent, depending on the conditions
(Fig. 7
A).
|
To further study the CAD/DFF-40 endonuclease activity, in
vitro experiments were performed by incubating isolated nuclei of
control Jurkat cells with cytosolic lysates of the same cells after
apoptosis triggering. Lysates from cells exposed to mannitol or STS
could induce DNA fragmentation, but lysates from cells exposed to NaCl
(with or without STS) or mannitol in the presence of NPPB failed to
trigger any internucleosomal DNA cleavage (Fig. 7B
). This
experiment was performed in a buffer containing a low
Cl- concentration (10 mM); however, raising the
Cl- concentration to 250 mM did not alter the
DNA fragmentation pattern observed under any condition (not shown).
Cytosolic lysates were also incubated with high molecular weight calf
thymus DNA, which retains its nucleosomal proteins. The effects of STS
in a hyperosmotic NaCl medium (i.e., the condition with the strongest
cleavage of ICAD/DFF-45) and of mannitol were assayed at different
Cl- concentrations (Fig. 7C
):
the DNA laddering pattern occurred only in the presence of mannitol.
Intracellular pH also does not seem to be directly involved in
CAD/DFF-40 regulation, since the same pattern of DNA laddering was
reproduced in buffers with pH ranging from 6 to 9 (not shown).
Finally, the PFGE analysis reported in Fig. 7D
revealed the
presence of DNA fragments of ~50–150 kDa in all the cases in which
an internucleosomal DNA cleavage was not apparent (NaCl, NaCl with STS,
mannitol with NPPB). Instead, DNA fragments of lower molecular weight
were evidenced in the conditions enabling DNA laddering (e.g., mannitol
and STS).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, its substitution for Cl- creates more
favorable conditions for Cl- efflux. However, cytosolic
Cl- loss through ORCCs was not sufficient per
se to activate DNA cleavage since in hypoosmotic conditions,
where these channels are strongly activated (24)
, we were
unable to measure any apoptotic hallmark.
In any case, side effects of the ORCC blockers on targets other
than chloride channels cannot be excluded. Indeed, ddFSK has been
reported to inhibit P-glycoprotein activity (29)
,
depletion-induced Ca2+ influx in T cells (30)
,
and a K+ channel in epithelial cells (31)
,
with possible implications in the commitment to apoptosis. Moreover, in
our hands, apoptosis was always enhanced by NPPB (see Fig. 6
B and Fig. 7
A), possibly as a result of
a certain proapoptotic activity of this compound per se,
which was never observed with ddFSK (not shown). Accordingly, a
mitochondrial uncoupling activity of NPPB had been described
(32)
. Nevertheless, the proapoptotic effects that NPPB
might exert by acting on secondary targets cannot account for the
blocking effect it displayed on DNA laddering.
A decrease in cytosolic K+ concentration has been described
after apoptosis induction (33
34
35
36)
. This event has been
proposed to be necessary for caspase-3 and apoptotic endonuclease
activation (37)
or, alternatively, to be a prerequisite
for nuclear and DNA degradation downstream of caspase activation
(36)
. In most cells, a Cl- efflux is coupled
to the activation of K+ channels to maintain
electroneutrality, particularly during cytosolic volume decrease
(38)
. Theoretically, cells could use similar mechanisms to
shrink in late apoptotic phases, since blockade of a class of
K+ channels has been reported to inhibit
apoptosis-associated cell shrinkage (35)
. Therefore,
activation of ORCCs and K+ conductances might be concerted
during the apoptotic process, and the activity of CAD/DFF-40 could
require a reduction of the cytosolic ionic strength secondary to a
coordinated efflux of K+ and Cl-. Since with
our conditions the ORCC blocker NPPB could inhibit DNA laddering if
supplied up to 3 h after mannitol, the cytosolic Cl-
loss involved in the activation of DFF should occur in a late phase of
apoptosis, after the starting of irreversible death commitment steps.
In fact, loss of m was already detectable as soon as
1 h and caspase-dependent processes after 2 h of hyperosmotic
shock.
Recently, activation of ORCCs has been described in Jurkat cells after
Fas receptor triggering and their inhibition would correlate with a
partial block of apoptosis (39)
. On the contrary, we have
observed that the various maneuvers used to inhibit Cl-
efflux do not block PCD execution. Instead, in our hands,
Cl- efflux inhibition seems to specifically affect the DFF
activity rather than the whole apoptotic process. In our model, the
endonuclease activity per se or an upstream activator of
this enzyme could be under the regulation of a Cl- efflux.
The DFF heterodimer is activated by caspase-3 via the proteolytical
processing of the ICAD/DFF-45 inhibitor (11)
. However, we
could observe cleavage of caspase-3 to its active fragments as well as
formation of apoptotic bodies, chromatin condensation, and PS
externalization, which selectively depend on caspase-3 activity
(40)
, in the absence of DNA fragmentation. More
remarkably, proteolysis of ICAD/DFF-45 was measurable in all apoptotic
conditions tested, independent of the occurrence of DNA laddering. Even
if the degree of cleavage was greatly variable (Fig. 7
A), a complete proteolysis of ICAD/DFF-45 was observed
in certain conditions where nucleosomal fragmentation was not
detectable. To our knowledge this is the first report of an ICAD/DFF-45
cleavage without any oligonucleosomal DNA fragmentation. Together,
these experiments indicate that Cl- efflux affects DFF
activity somewhere downstream caspase-3 activation and ICAD/DFF-45
proteolysis.
In vitro experiments performed by incubating
apoptotic cytosolic fractions with either intact nuclei or high
molecular weight DNA did not reveal any oligonucleosomal DNA cutting
when apoptosis had been induced in conditions of altered
Cl- efflux. As lymphocyte cytosolic Cl-
concentration is around 60 mM (41)
and various apoptosis
agonists cause an intracellular acidification (42)
, we
tested lysates from cells exposed to various apoptotic conditions in
Cl- concentrations ranging from 10 mM to 250 mM and pH
ranging from 6 to 9. However, no modification of the nucleosomal DNA
cleavage was ever observed when Cl- efflux had been
altered (Fig. 7
B, C and data not shown). These data
argue against the hypothesis that cytosolic Cl- loss could
directly control CAD/DFF-40 enzymatic activity or that it could reduce
the access of the nuclease to the internucleosomal domains of DNA.
Alternatively, when Cl- efflux is prevented, CAD/DFF-40
could be partitioned into a cellular compartment from which it cannot
reach DNA. However, the simplest explanation of these findings is that
the impairment of Cl- efflux during apoptosis induction
inhibits in a persistent manner the activation of the endonuclease in
such a way that it cannot be restored in a further in
vitro assay by changing Cl- concentration.
CAD/DFF-40 could require a second covalent trigger in addition to its
release from ICAD/DFF-45 for a complete activation or, alternatively,
the removal of another inhibitory constraint. In both cases, the
process would be under the control of a Cl- efflux and the
dilution of cellular material in the in vitro
experiments would not permit recovery of the second hit necessary for
CAD/DFF-40 activation.
Moreover, we observed chromatin perinuclear condensation, loss of
nuclear integrity, and formation of apoptotic bodies even without
internucleosomal DNA fragmentation, implying that CAD/DFF-40 activity
is not necessary for these events to occur. This result is in contrast
with several reports indicating that activation of this enzyme is
mandatory for chromatin condensation (14
, 17
, 18)
and with
the hypothesis that oligonucleosomal degradation would coincide with
complete nuclear destruction (4)
. Nevertheless, in accord
with our results, a CAD-DFF-40-independent chromatin condensation has
recently been described in addition to the CAD-DFF-40-dependent one
(18)
, and the mitochondrial release of the
apoptosis-inducing factor (AIF) has been shown to induce chromatin
condensation on purified nuclei (43)
. Furthermore, we have
observed high molecular weight (50–150 kb) DNA fragments when DNA
laddering was absent. These large fragments are probably a transient
event that precedes smaller size DNA ruptures, as they were further
degraded under conditions that enabled DNA laddering. These data lend
support to a two-step model in which distinct endonuclease activities
are required for DNA processing in high molecular weight fragments and
subsequent internucleosomal degradation. Accordingly, cells displaying
only high molecular weight DNA fragmentation after apoptosis induction
have been described (44)
; this apoptotic change is also
triggered by the mitochondrial release of the AIF (43)
.
We have observed a mitochondrial swelling after apoptosis induction by
hyperosmotic shock, with loss of matrix cristae and ruptures in the
external mitochondrial membrane. Similar mitochondrial
changesduring PCD engagement have recently been proposed to be
responsible for cytochrome c release (45)
.
Intact mitochondria maintain a higher internal osmolarity than cytosol
(46)
; instead, after de-energization their volume is
determined by the cytosolic colloid-osmotic pressure. In our
experimental conditions, cells faced a hyperosmotic environment, but
the observed matrix swelling suggests that cytosolic colloid-osmotic
pressure was maintained at a lower level than in the mitochondrial
matrix. Furthermore, we have shown that the collapse of
m is independent from caspase-3-like activation, in
agreement with several reports showing that a m
breakdown is an early apoptotic event (47
48
49)
.
In summary, our results indicate that the degradative pathway leading to DNA destruction in apoptosis needs a complex regulatory mechanism, involving a control by a Cl- efflux downstream of the dissociation of CAD/DFF-40 from its inhibitory subunit. In turn, the effect of Cl- on the endonuclease is probably linked to the cytosolic shrinkage that accompanies late apoptotic phases. A better understanding of this phenomenon is needed in order to characterize the channel(s) involved and elucidate the mechanism through which Cl- efflux can be related to the activation of the apoptotic endonuclease.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication February 19, 1999. Revised for publication April 23, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
HindIII molecular weight
marker, BioLabs) and C) with the cytofluorometric
TUNEL technique after exposure to mannitol or NaCl. The peak on the
left represents cells with intact DNA; that on the right represents
apoptotic cells. D) DNA fragmentation due to mannitol or
NaCl in CEM and THP-1 cells is compared with that observed in Jurkat
cells.
