(Received for publication, May 30, 1996, and in revised form, October 4, 1996)
From the Division of Cancer Pharmacology, Dana-Farber Cancer
Institute, Harvard Medical School, Boston, Massachusetts 02115, BASF Bioresearch Corp., Worcester, Massachusetts 01605, the § Department of Pharmacology, Kimmel Cancer Institute,
Philadelphia, Pennsylvania 19107, and the ¶ Department of
Radiation and Cellular Oncology, University of Chicago,
Chicago, Illinois 60637
The response of eukaryotic cells to ionizing
radiation (IR) includes induction of apoptosis. However, the signals
that regulate this response are unknown. The present studies
demonstrate that IR treatment of U-937 cells is associated with: (i)
internucleosomal DNA fragmentation; (ii) cleavage of poly(ADP-ribose)
polymerase; (iii) cleavage of protein kinase C ; and (iv) induction
of an Ac-DEVD-p-nitroanilide cleaving activity.
Overexpression of the cowpox protein CrmA blocked tumor necrosis factor
(TNF)-induced apoptosis but had no effect on IR-induced DNA
fragmentation or cleavage of poly(ADP-ribose) polymerase and protein
kinase C
. By contrast, overexpression of the baculovirus p35
protein blocked both IR- and TNF-induced apoptosis. The results further
demonstrate that the IR-induced proteolytic activity is directly
inhibited by the addition of purified recombinant p35, but not by CrmA. We show that the CPP32 protease is sensitive to p35 and not CrmA. We
also show that IR induces activation of CPP32 and that this event, like
induction of apoptosis, is sensitive to overexpression of p35 and not
CrmA. These findings indicate that IR-induced apoptosis involves
activation of CPP32 and that this CrmA-insensitive apoptotic pathway is
distinct from those induced by TNF and certain other stimuli.
The response of eukaryotic cells to ionizing radiation
(IR)1 includes cell cycle arrest and
activation of DNA repair. The available evidence suggests that IR
induces these effects by direct interaction with DNA or through the
formation of reactive oxygen intermediates that damage DNA and cell
membranes (1). In the event of irreparable damage, IR-treated cells
also undergo programmed cell death or apoptosis. Few insights, however,
are available regarding the signals that control induction of apoptosis
in the IR response. While p53 is required for optimal apoptosis induced
by IR (2, 3), the precise role of this tumor suppressor in regulating cell death is poorly understood. Other studies have shown that Bcl-2
and Bcl-xL inhibit IR-induced apoptosis (4-7). Several proteins, including poly(ADP-ribose) polymerase (PARP) (8), lamin B1
(9), DNA-dependent protein kinase (10), the 70-kDa protein
component of the U1 small nucleoprotein (11), and topoisomerase I (12),
have been shown to be cleaved during apoptosis. Recent studies have
also shown that protein kinase C (PKC) is proteolytically activated
during IR-induced apoptosis (13). Cleavage of PARP and PKC
is
blocked by overexpression of Bcl-2 and Bcl-xL (13, 14). The
mechanistic basis for the anti-apoptotic effects of Bcl-2 and
Bcl-xL is unclear.
Other work has supported the involvement of aspartate-specific cysteine
proteases in induction of apoptosis (15). The nematode death effector
Ced-3 is a cysteine protease (16) that has significant homology with
the interleukin-1 converting enzyme (ICE) (17). The finding that
overexpression of ICE or Ced-3 induces apoptosis has supported
involvement of the ICE/Ced-3 family of proteases in the cell death
pathway (18). Related ICE/Ced-3 homologs include Nedd2/Ich-1 (19, 20),
CPP32/YAMA/apopain (21-23), TX/Ich-2/ICErel-II (24-26),
ICErel-III (26), Mch2 (27), Mch3/ICE-LAP3/CMH-1 (28-30), Mch4, and Mch5 (31). CPP32, Mch3, and Ced-3, but not ICE, cleave PARP
after aspartate (22, 23, 28). More direct evidence for involvement of
an ICE-like protease in apoptosis comes from studies utilizing the
cowpox virus protein CrmA (32) and the baculovirus protein p35 (33),
which are direct inhibitors of at least certain members of this family.
Overexpression of CrmA inhibits the induction of apoptosis in diverse
models, including engagement of the Fas receptor and treatment with
tumor necrosis factor (TNF)
(34-36). Similarly, the p35
gene encodes an inhibitor that blocks apoptosis in insect and
mammalian cells (37-40).
The present studies demonstrate that, in contrast to TNF, IR induces apoptosis by a CrmA-insensitive, p35-sensitive mechanism. The results support a distinct signaling cascade responsible for IR-induced cell death.
Human U-937 myeloid leukemia
cells (American Type Culture Collection, Rockville, MD) were grown in
RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine
serum, 100 units/ml penicillin, 100 µg/ml streptomycin, and 2 mM L-glutamine. CrmA cDNA was cloned into
the pEF1 vector (provided by Dr. Ping Li, BASF, Worcester, MA), which
was constructed by ligation of human elongation factor 1 promoter
from pEF321-CAT (41) into a vector modified from pBluescript. p35 was
cloned into the pEF2 vector, which was made by substituting the
cytomegalovirus promoter of pcDNA3 with elongation factor 1
promoter (33, 41). To generate a CrmA- or p35-overexpressing line,
U-937 cells were cotransfected by electroporation (Gene Pulsar,
Bio-Rad, 0.25 V, 960 microfarads) with pcDNA3 and pEF1-CrmA or
pEF2-p35 (33, 41). Transfectants were selected in the presence of 400 µg/ml geneticin sulfate. Irradiation was performed with a
-ray
source (cesium 137, Gamma Cell 1000, Atomic Energy of Canada, Ltd.,
Ontario) at a fixed dose rate of 13 gray/min. Cells were also treated
with TNF (42).
Total cellular RNA was isolated as described (43). RNA (20 µg/lane) was separated in agarose/formaldehyde gels, transferred to nitrocellulose filters, and hybridized to the 32P-labeled DNA fragments corresponding to the entire CrmA or p35 open reading frames. Hybridizations were performed as described (7).
Analysis of DNA FragmentationDNA was prepared as described (13) and separated in 2% agarose gels. The DNA was visualized by UV illumination after ethidium bromide staining.
Immunoblot AnalysisCell lysates were prepared as described
(13). Proteins were subjected to electrophoresis in an SDS-10%
polyacrylamide gel and then transferred to nitrocellulose membranes.
The membranes were blocked with 5% dried milk, 0.1% Tween 20, and
phosphate-buffered saline and were incubated with anti-CrmA polyclonal
antibody (raised against full-length CrmA protein), anti-PKC (Santa
Cruz Biotechnology, Santa Cruz, CA), anti-CPP32 polyclonal antibody
(raised against the large subunit of CPP32; amino acids 1-175), or
anti-Ich-1L (Transduction Laboratories, Lexington, KY).
Preparation of lysates and immunoblotting for PARP were carried out as
described using the C-2-10 anti-PARP monoclonal antibody (44). After
washing with phosphate-buffered saline/Tween, the membranes were
incubated with horseradish peroxidase-conjugated anti-mouse IgG
(Amersham) for anti-PARP and anti-Ich-1L or anti-rabbit IgG
(Amersham) for anti-CrmA, anti-PKC
, and anti-CPP32.
CrmA contained an
N-terminal polyhistidine linker and was expressed in Escherichia
coli MM294 cells under the control of a pL promoter as
described (25). Cell lysate supernatant was diluted 1:1 with buffer A
(50 mM HEPES, pH 7.5, 10% glycerol, 0.2 M
NaCl) and applied to a 5-ml NiSO4-charged Hi-Trap column (Pharmacia Biotech Inc.). The column was washed with 2% buffer B
(buffer A plus 200 mM imidazole), and the protein was
eluted with buffer B. Fractions containing CrmA as judged by SDS-PAGE were pooled and dialyzed at 4 °C against 20 mM Tris, pH
7.5. The sample was applied to an 8-ml Mono Q column (Pharmacia),
equilibrated with buffer C (20 mM Tris, pH 7.5, 5 mM dithiothreitol), and eluted with a gradient to buffer D
(buffer C plus 300 mM NaCl). p35 was expressed in E. coli MM294 cells and purified as described (33). ICE, Ich-1,
Ich-2, and CPP32 contained N-terminal polyhistidine linkers and were
expressed and purified as described (25).
Cell lysates were centrifuged at 900 × g for 10 min at 4 °C. Protease assays included 178 µl of reaction buffer (100 mM HEPES, pH 7.5, 20% v/v glycerol, 5 mM dithiothreitol, and 0.5 mM EDTA), 2 µl of 10 mM acetyl-Asp-Glu-Val-Asp-p-nitroanilide (Ac-DEVD-pNA) in Me2SO (100 µM final concentration; California Peptide Research, Inc., Napa, CA), and 20 µl of cell lysate. Samples were incubated at 30 °C, and enzyme-catalyzed release of p-nitroanilide was monitored at 405 nm for 30 min in a microtiter plate reader (Molecular Devices Inc., Sunnyvale, CA). In certain experiments, the cell lysate diluted into assay buffer was first incubated with varying amounts of inhibitors at room temperature for 30 min. Inhibition of purified ICE homologs by CrmA was performed as described (33) using Ac-DEVD-pNA (500 µM) as a substrate for Ich-1, Ich-2, and CPP32 and Ac-YVAD-pNA (125 µM) as a substrate for ICE. Values for nanomoles of pNA released were calculated from those observed in A405 values using a standard curve.
Previous studies have demonstrated that treatment of U-937 cells
with IR is associated with the induction of apoptosis (6, 7, 45). To
determine whether this event is sensitive to inhibitors of ICE-like
proteases, we prepared U-937 cell clones that overexpress CrmA and p35
(Fig. 1A). IR treatment of both U-937 and
U-937/CrmA cells resulted in a pattern of internucleosomal DNA
fragmentation characteristic of apoptosis (Fig. 1B). By
contrast, there was little if any DNA fragmentation in irradiated
U-937/p35 cells (Fig. 1B). Immunoblot analysis of CrmA in
U-937/CrmA cell lysates and comparison of signals that were obtained
with recombinant CrmA indicated a concentration of approximately 1 ng/20 µg of lysates (data not shown). Since CrmA blocks TNF-induced
apoptosis in other cell types (34), we asked whether the U-937/CrmA
cells were sensitive to this agent. The results demonstrate that both CrmA and p35 block apoptosis induced by TNF treatment (Fig.
1C). These findings indicated that IR induces apoptosis by a
distinct CrmA-insensitive mechanism.
To determine the involvement of CrmA-insensitive proteases in
IR-induced apoptosis, we assessed cleavage of the 116-kDa PARP protein
to an 85-kDa fragment (8). As expected, IR treatment of U-937 cells
resulted in PARP cleavage (Fig. 2). Similar findings were obtained in irradiated U-937/CrmA cells, but there was no detectable cleavage of PARP in IR-treated U-937/p35 cells (Fig. 2). IR
also induces cleavage of PKC (13). PKC
cleavage to a 40-kDa
fragment was unaffected in IR-treated U-937/CrmA cells, and this event
was sensitive to overexpression of p35 (Fig. 2). Taken together, these
results demonstrate that IR induces cleavage of PARP and PKC
by one
or more p35-sensitive, CrmA-insensitive protease(s).
To assay for protease activity directly, we incubated cell lysates with
Ac-DEVD-pNA and monitored release of
p-nitroanilide. Lysates from U-937 cells exhibited an
increase in peptide cleavage activity that was detectable at 4 h
and maximal at 6 h after IR exposure (Fig.
3A). The kinetics of Ac-DEVD-pNA
cleavage corresponded temporally with IR-induced internucleosomal DNA
fragmentation in U-937 cells (13). Lysates from irradiated U-937/CrmA
cells exhibited a kinetically similar but less pronounced induction of
Ac-DEVD-pNA cleavage activity (Fig. 3A). By
contrast, there was little if any induction of such activity in lysates
from irradiated U-937/p35 cells (Fig. 3A). To further define
the effects of CrmA and p35 on Ac-DEVD-pNA cleavage, we
incubated lysates from irradiated U-937 cells with the purified
recombinant anti-apoptotic proteins. While CrmA had little apparent
effect, addition of p35 was associated with complete inhibition (Fig.
3B). CPP32 has been shown to be responsible for proteolytic
cleavage of Ac-DEVD-pNA (23). To address the potential role
of a CPP32-like activity in IR-induced apoptosis, we compared the
effects of the peptidic CPP32 inhibitor Ac-DEVD-cho (23) on IR-induced
Ac-DEVD-pNA cleaving activity with those observed after
adding purified CPP32 to lysates of unirradiated cells (Fig.
3C). The finding that the IR-induced protease activity in
cell lysates is approximately 3-fold more sensitive to Ac-DEVD-cho
inhibition than CPP32 suggests that the induced activity is due to
CPP32-like enzymes, perhaps including CPP32 itself.
ICE, Ich-1, Ich-2, and CPP32 are all potently inhibited by p35 (33).
CrmA, in contrast, displays about 104-fold selectivity for
ICE compared with CPP32 (23), although inhibition of CPP32 by CrmA has
also been reported (22). To resolve this discrepancy and to address the
differential effects of CrmA and p35 observed on IR-induced apoptosis,
we measured CrmA inhibition of purified ICE, Ich-1, Ich-2, and CPP32
(Fig. 4A). CrmA inhibited ICE and Ich-2
approximately 50% at a 1:1 molar ratio (approximately 1 µM each) but was without effect on Ich-1 and CPP32 at up
to a 10:1 molar ratio (Fig. 4A). Since the IR-induced Ac-DEVD-pNA cleaving activity is p35-sensitive and
CrmA-insensitive, these results further support involvement of a
CPP32-like protease in IR-induced apoptosis. To directly assess
activation of CPP32, we assayed lysates from irradiated cells for
cleavage of the proenzyme to its active subunits (21, 23, 31).
IR-treated U-937 and U-937/CrmA cells exhibited CPP32 activation, while
there was little if any cleavage of the proenzyme in irradiated
U-937/p35 cells (Fig. 4B). By contrast, there was no
detectable effect of IR treatment on Ich-1L levels (Fig.
4B). Taken together, the findings indicate that IR-induced
apoptosis is associated with activation of CPP32.
Eukaryotic cells respond to lethal doses of IR with induction of
apoptosis (8, 45, 46). Bcl-2 blocks IR-induced apoptosis (4-6).
Similar findings have been obtained in cells that overexpress Bcl-xL (7). Other studies have demonstrated that Bcl-2 and Bcl-xL inhibit IR-induced proteolytic cleavage of PKC
(13). These results have suggested that Bcl-2 and Bcl-xL
function upstream to IR-induced activation of certain members of the
ICE/Ced-3 family of cysteine proteases. More recent work has shown that
Bcl-2 and Bcl-xL block staurosporine-induced, but not
Fas-induced PARP cleavage and CPP32 activation (14). By contrast, the
finding that Bcl-2 and Bcl-xL block Fas-induced cleavage of
PKC
(13) suggests that these anti-apoptotic proteins may function
upstream only to certain cysteine proteases. The present studies were
performed to further define the apoptotic pathway in irradiated cells.
The CrmA and p35 proteins were overexpressed in U-937 cells because of
their reported functions in inhibiting members of the ICE/Ced-3 family.
The results suggest that IR activates a CrmA-insensitive, p35-sensitive
pathway.
CrmA is a member of the serpin family that inhibits ICE by forming an
active site-directed complex (32, 47). CrmA expression blocks apoptosis
induced by: (i) nerve growth factor withdrawal of chicken neurons (48);
(ii) cytotoxic T cell-mediated cytolysis (49); (iii) detachment of
cells from an extracellular matrix (50); (iv) serum withdrawal of Rat1
fibroblasts (20); and (v) activation of Fas or TNF receptors (34).
However, the present studies demonstrate that IR-induced apoptosis is
not blocked by overexpression of CrmA. This finding is not explainable
by insufficient CrmA expression since the U-937/CrmA cells were
insensitive to TNF-induced apoptosis. Moreover, a recent report has
shown that CrmA has no effect on IR-induced apoptosis of mouse lymphoma
cells (51). We also demonstrate that IR-induced PARP and PKC
cleavage is not affected by CrmA. Further, IR-induced
Ac-DEVD-pNA cleaving activity was detectable in U-937/CrmA
cells and was not inhibited by >10 µM CrmA in the assay
mixture (Fig. 3C) CPP32 was also not inhibited by CrmA at
concentrations of up to 10 µM (Fig. 4). Collectively, these data suggest that, in contrast to TNF, IR induces apoptosis via
one or more cysteine proteases (such as CPP32) that are
CrmA-insensitive.
The baculovirus p35 protein inhibits apoptosis in cells from insects,
nematodes, and mammals (33, 39, 52, 53). Other studies have shown that
p35 inhibits the proteolytic activity of Ced-3, ICE, CPP32, Ich-1, and
Ich-2, but not granzyme B (33, 54). The present studies demonstrate
that p35 blocks IR-induced Ac-DEVD-pNA cleaving activity. We
also found that p35 blocks IR-induced apoptosis and cleavage of PARP
and PKC. These findings support a role for ICE/Ced-3-like proteases
in IR-induced apoptosis. However, the sensitivity of TNF-induced, but
not IR-induced, apoptosis to CrmA supports the involvement of distinct
proteases in the two processes. Sphingomyelin hydrolysis and ceramide
production have been identified in TNF-treated cells (55, 56). This
pathway is also activated in cells exposed to IR (57). Since ceramide has been shown to mediate apoptosis (58), sphingomyelin hydrolysis may
contribute to induction of apoptosis by both IR and TNF. While there
may be central signals for inducing apoptosis by diverse agents, the
insensitivity of IR-induced apoptosis to CrmA distinguishes the
protease(s) activated in irradiated cells from those involved in
TNF-treated cells.
Recent work has demonstrated that Fas-mediated apoptosis is associated
with sequential activation of ICE and then CPP32-like activity (59). In
the present studies, there was little if any effect of IR treatment on
YVAD-cleaving activity (data not shown). These findings and the
demonstration that IR-induced apoptosis is insensitive to
overexpression of CrmA support the lack of ICE involvement. Thus,
ICE-like proteases may be involved in Fas- (59) and TNF-induced
apoptosis but not in IR-treated cells. Collectively, the present
results indicate that IR induces activation of CPP32. While PARP
cleavage can be mediated by CPP32 and/or Mch3 (23, 28), PKC is
cleaved by CPP32 and not ICE, Ich-1, Ich-2, Mch2, Mch3, or
ICErel-III (60). Thus, IR-induced cleavage of PARP and PKC
is consistent with activation of CPP32. Studies of
IR-induced Ac-DEVD-pNA cleavage and sensitivity of this
activity to Ac-DEVD-cho and p35, but not CrmA, also support involvement of CPP32. Moreover, IR-induced cleavage of the CPP32 proenzyme to its
subunits is in concert with activation of this protease. Thus, our
findings indicate that IR-induced apoptosis is associated with CPP32
activation and that the apoptotic signals induced by irradiation differ
from those associated with TNF- or Fas-mediated cell death.