| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 278, Issue 36, 34042-34050, September 5, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Program in Apoptosis and Cell Death Research, The Burnham Institute, La Jolla, California 92122
Received for publication, May 15, 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
For the executioner caspases, proteolysis between the large and small subunit is thought to be the fundamental activating event (4–8). All caspases possess N-terminal extensions, and in the case of the in initiators these are required for recruitment to the respective activation complexes. In contrast, a consensus role for the short N-terminal extensions of the executioners (Fig. 1A) has yet to be established. Initial findings suggest that the N-terminal peptides of caspase-3 and -7 have no effect on activity or the ability to be activated in vitro. However, in vivo both caspases seem to require the removal of the N-peptides for efficient activation. Indeed, in some cells caspase-3 removes the N-peptide of caspase-7 before the latter is activated by an initiator granzyme B, a serine protease that activates caspases during T-cell-mediated killing (9).
|
Other roles attributed to the N-peptide are the silencing of caspase-3 (10) and caspase-6 (11) and prevention of nuclear import of Xenopus caspase-7 (12). The later hypothesis is appealing, but reports are conflicting regarding the subcellular localization of caspase-7. Although it is generally agreed that the zymogen of caspase-7 is cytosolic, subcellular fractionation experiments have suggested that active caspase-7 relocalizes to the nucleus (12), microsomes (13), or mitochondria (14) during apoptosis. Some of those discrepancies may be artifacts of the experimental procedure given the fact that the pI of caspase-7 lacking the N-peptide is very different than the one of the zymogen (pI = 8.0 versus 5.5) and may cause the protein to precipitate upon N-peptide removal. Secondly and most importantly, apoptosis is associated with pH and ionic strength changes as well as nuclear-cytoplasmic barrier disruption (15) and may result in mixing and/or relocalization of compartment specific markers used in subcellular fractionation experiments. Finally, usage of EGFP fusion protein as a tracker for caspase localization (12, 16) may also be a source of experimental variation because of potential dimerization of green fluorescent proteins, natural tendency to localize to the nucleus and increase in molecular weight of the caspase complex. In the present study we have tested a number of these hypotheses for the function of the enigmatic N-peptide of caspase-7.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Mammalian Expression Constructs—The cDNA for human caspase-7
(GenBankTM acc. no. NM_001227
[GenBank]
) was use as a template for all constructs.
All pcDNA3 constructs were subcloned into KpnI and XhoI
sites by PCR with oligonucleotides adding the appropriate restriction sites
flanking the initiating methionine codon at the 5'-end and the stop
codon at the 3'-end. The FLAG epitope was added to the C terminus
(Fig. 1B) of each
construct by PCR with the following reverse oligonucleotide
5'-ctcgagctacttgtcatcgtcgtccttgtagtcttgactgaagtagagttcc (FLAG
coding sequence underlined). The 
N deletion mutant was made by PCR with
oligonucleotide 5'-cccaagcttggtaccgccatgaagccagaccggtcctcgttt
that replaces residues 1–24 by a Kozak sequence (underlined). All
mutations were made using Quick-Change XL mutagenesis protocol (Stratagene, La
Jolla, CA) with the following oligonucleotides and the corresponding reverse
sequence: C186A (catalytic cysteine residue also known as 285 in the caspase-1
numbering system), 5'-cttcattcaggctgcccgagggacc; NC (D23A),
5'-gcaaatgaagattcagtggctgctaagccagaccgg; D198A,
5'-ggcatccaggccgctagcgggcccatcaatg; D206A,
5'-ccatcaatgacacggccgctaatcctc. All constructs were sequenced
to ensure DNA sequence integrity.
Recombinant Protein Expression and Purification—Recombinant caspases were expressed in Escherichia coli using as C-terminal His-tagged fusion proteins using the pET expression system (Novagen, Madison, WI). All constructs were subcloned into NdeI and XhoI sites of pET-23b(+) using the same strategy as for pcDNA3 constructs. The initiating methionine was contained within the NdeI site used in the subcloning. Proteins were expressed in BL21(DE3) E. coli stain (Novagen) and purified by Ni2+-affinity chromatography as previously described (17). The NC and C186A mutants were generated as described above and the M45A mutant was obtained with the oligonucleotide 5'-gaagaaaaatgtcaccgcgcgatccatcaag and its reverse sequence.
Immunoblot Analysis—Cell extracts were prepared in mRIPA
buffer (50 mM Tris, pH 7.4, 100 mM NaCl, 1% Nonidet
P-40, 0.5% deoxycholic acid, 0.1% SDS) with general protease inhibitors (1
mM EDTA, 1 mM 1,10-phenanthroline, 50 µM
3,4-dichloroisocoumarin, 10 µM E-64, and 10 µM
leupeptin). Polycaspase inhibitors were added when required (100
µM
z-VAD-FMK1 and 100
µM Ac-DEVD-CHO). Lysates were centrifuged at 18,000 x
g for 15 min at 4 °C to remove cellular debris. Nuclear extracts
were prepared from the insoluble material of the mRIPA extracts by boiling
samples in 4 M urea and 1% SDS for 10 min. DNA was sheered by 10
passages through a 27-gauge needle. Samples from 2 x 105
cells were resolved on 8–18% acrylamide gradient gels, transfer to
polyvinylidene difluoride membrane in 10 mM CAPS, pH 11, and 10%
methanol at constant current (0.4 A for 40–60 min) as previously
described (18). Ensuing blots
were processed to immunoblotting with the various antibodies and corresponding
HRP-conjugated secondary antibodies (1:3000, APBiotech, Piscataway, NJ) and
SuperSignal detection reagents (Pierce Chem. Co., Rockford, IL). Cytosolic
extracts for cytochrome c were obtained using an hypertonic buffer
(250 mM sucrose, 70 mM KCl, 137 mM NaCl, 4.3
mM Na2HPO4, 1.4 mM
KH2PO4, pH 7.2, 200 µg/ml digitonin, and protease
inhibitors) as previously described
(19). The following antibodies
were used: caspase-7 (monoclonal, 1:1000), cytochrome c (monoclonal,
1:1000) and PARP (monoclonal, 1:1000, BD Pharmingen, San Diego, CA); hsp90
(monoclonal, 1:2500) and protein kinase C 
, (polyclonal, 1:2000, Santa
Cruz, CA); lamin B1, (monoclonal, 1:1000, Zymed Laboratories Inc.,
San Francisco, CA); anti-FLAG (M2, monoclonal, 5 µg/ml, Sigma).
Enzymatic Assays—Transfected 293A cells grown in 60-mm
dishes were prepared in 50 µl of mRIPA buffer, left on ice for 10 min and
centrifugation at 18,000 x g for 20 min. Protein concentration
was determined using DC protein assay (Bio-Rad, Hercules, CA) using bovine
serum albumin as a standard. Similar amount of protein were assayed for
caspase activity using 100 µM of fluorogenic Ac-DEVD-AFC
substrate (Bachem Biosciences Inc., King of Prussia, PA) in caspase buffer (10
mM Pipes, pH 7.2, 100 mM NaCl, 10% sucrose, 0.1% Chaps,
1 mM EDTA, and 10 mM dithiothreitol)
(17). Amidolytic activity was
measured on a f-max Molecular Device spectrofluorometer at 37 °C
(EX
= 405 nm, EM = 510 nm). Recombinant
enzymes were titrated to determine the exact active site concentration as
previously described (17) and
assays were carried in caspase buffer using either Ac-DEVD-AFC or the
equivalent chromogenic substrate Ac-DEVD-pNA (Abs. 405 nm). Enzymatic assays
using recombinant or 35S-labeled in vitro-translated
proteins were performed in caspase buffer at 37 °C in 20–50 µl
reaction volume. Laemmli loading buffer was added, and samples were resolved
by SDS-PAGE, stained with GelCode Blue stain reagent (Pierce) and dried.
In vitro translation was done using the TNT in
vitro translation kit (Promega Corp., Madison, WI); an equivalent amount
of radioactive substrate was used for each assay.
Affinity Capture Assay—Cells were transfected in 60-mm plates with the indicated constructs 24–36 h prior labeling. Media was replaced with warm media containing 1 µM biotinyl-VAD(OMe)-FMK (ICN Pharmaceutical inc., Irvine, CA) and incubated for 5 h at 37 °C. Cells were harvested, washed and lysed in 0.5 ml of mRIPA buffer with the protease inhibitor mixture described above. Samples were immunoprecipitated with anti-FLAG (M2, 10 µg/ml), anti-caspase-3 (1:250) or anti-caspase-6 (10 µl of crude serum, generous gift of Dr. S. Krajewski) antibodies at 4 °C for 16 h. Protein-A/G agarose beads (30 µl slurry, Santa Cruz Biotechnology) were added and incubated for 1 h at 4 °C. Immune complex was recovered by centrifugation and washed three times with mRIPA buffer. Samples were resolved on SDS-PAGE gels and blotted to polyvinylidene difluoride membrane as described above. Biotinylated proteins were revealed using streptavidin-HRP (0.2 µg/ml, Sigma) in i-Block buffer (Tropix Inc., Bedford, MA) and SuperSignal reagents.
DNA Fragmentation Assays—293A cells grown in 90-mm dishes were transfected, harvested, washed, and lysed in 0.2 ml of lysis buffer (20 mM Tris, pH 8.0, 10 mM EDTA, and 0.2% Triton X-100) and left on ice for 15 min. Lysate was clarified by centrifugation at 18,000 x g for 10 min and DNase-free RNase was added to 50 µg/ml and incubated at 37 °C for 1 h. SDS was added to 0.1% and proteinase K to 0.1 mg/ml and further incubated at 50 °C for 16 h. Samples were extracted twice with phenol:chloroform and ethanol-precipitated with 0.3 M sodium acetate. DNA was resuspended in water and analyzed on a 1.5% agarose gel containing ethidium bromide (20).
Fluorescence Microscopy—Cells were grown in slide chambers and transfected with the appropriate plasmid. Cells were washed twice with warm PBS and fixed with methanol for 2 min. Nonspecific binding sites were blocked with 5% nonfat dry milk in phosphate-buffered saline for 1 h at 37 °C, and samples were processed using anti-FLAG monoclonal antibody (M2, 10 µg/ml) and a FITC-conjugated secondary antibody (Molecular Probes, Eugene, OR). Samples were stained with DAPI (250 nM) and mounted with VectaShield (Burlingame, CA).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
N-caspase-7), and one contains a single amino
acid mutation (D23A) making the N-peptide non-cleavable (NC-caspase-7,
Fig. 1B). These were
compared with wild-type caspase-7 for their ability to initiate death in
recipient human 293A cells. Expression of FLAG-tagged caspase-7 in 293A cells
is sufficient to cause some cell death but deletion of the N-peptide greatly
increased the effect whereas prevention of N-peptide removal abrogates it
(Fig. 2A). This
differential effect is overcome at very high expression levels (3 µg
plasmid/60-mm dishes) upon which all caspase-7 constructs were able to produce
massive cell death (not shown). In all transfection experiments the onset of
cell death, measured by detachment from the dish, was sooner with
N-caspase-7 than wild type or NC-caspase-7. This is not due to
disparity in protein levels since immunoblot analysis of FLAG-tagged caspase-7
constructs from cell extracts revealed equivalent expression levels
(Fig. 2B,
top). Furthermore, the amount of N-caspase-7 present seems to
be less as detected by a caspase-7 antibody
(Fig. 2B,
middle). It is noteworthy that substantial processing of the
N-peptide of WT caspase-7 occurs during the expression period. The catalytic
mutant, C186A, had no effect on cell viability and is not processed.
|
To quantify caspase activation induced by the constructs we performed a
time-course analysis using different amounts of transfected DNA. Transfected
cells were harvested at various time points and the caspase activity in cell
extracts was measured using Ac-DEVD-AFC substrate
(Fig. 2, C and
D). In our hands saturation occurs when more than 2.5
µg/60-mm dish are used. We thus used DNA amounts up to 2.0 µg to detect
changes in the activation profile and data for 0.5 and 2.0 µg are shown.
Caspase activity was minimal at 12 h post-transfection for both conditions but
increases afterward to reach a maximum at 38 h before declining. Most
interestingly, at lower expression levels N-caspase-7 activity
increased dramatically at 18 h as compare with WT/NC-caspase-7 suggesting a
faster activation rate (Fig.
2C); similar results were obtained using 0.25 µg of
transfected DNA (data not shown). This difference was not apparent when higher
amounts of DNA were used (Fig.
2D and data not shown). Even in the latter condition
widespread cell death was delayed by 
6 h for wild-type and NC-caspase-7
as compared with N-caspase-7 samples. This is despite the fact that no
more DEVDase activity was observed, and suggests that removal of the N-peptide
is important for the caspase-7 killing effect. As a measure of the amount of
caspase (DEVDase) activity that is generated following caspase-7 transfection
we compared the results with cell death induced by ectopic expression of the
intrinsic pathway activator Bax. Transfection of caspase-7 results in a
10-fold increase of DEV-Dase activity in cell extracts as compared with the
maximum activity obtained in Bax-transfected cells suggesting that the
measured activity is mainly attributed to caspase-7 (data not shown).
Cleavage within the linker region of caspase-7 is strictly required for its activation (4, 6, 21, 22). This region contains 2 cleavage sites that could each potentially result in activation of the zymogen (Fig. 1B). Therefore we generated Asp/Ala mutations at each site for expression in 293A cells (Fig. 3). Analysis using anti-caspase-7 and anti-FLAG antibodies shows that mutation of D198A abrogates cleavage of the linker region whereas mutation of D206A did not impede cleavage (Fig. 3, A and B). It is noteworthy that removal of the N-peptide was observed for all constructs probably owing the to ability of caspase-7 overexpression to activate endogenous caspase-7 (data not shown). Albeit weak, a cleavage fragment was observed with the D198A mutant that prompted us to test whether or not this species was active or simply an aberrant cleavage product. To do this we employed an affinity-labeling strategy to capture active caspases (22). Transfected cells were incubated with the biotinylated cell permeable irreversible polycaspase inhibitor, biotinyl-VAD(OMe)-FMK. Caspase-7 was then immunoprecipitated with anti-FLAG antibody and biotinylated proteins were detected with streptavidin-HRP (Fig. 3C). It is only when Asp198 was available for cleavage that a robust labeling of caspase-7 was observed confirming that caspase-7 activation as a consequence of overexpression is a sequential event requiring processing of Asp198 without prior to cleavage at Asp206.
|
Caspase-7 Is the Most Downstream Caspase—Although caspase-7 is generally accepted to be a downstream (executioner) caspase (9, 23) its overexpression might lead to the activation of other caspases, in which case cell death would be a combination of multiple executioner (caspase-3, -6, and -7) activity and not solely that of caspase-7. Furthermore most caspases, and particularly caspase-3, exert overlapping substrate specificity with caspase-7 (24, 25) making the simple readout of DEVDase activity unreliable in diagnosis. Immunoblot analysis failed to reveal any processed caspase-3 or -6 in cell extracts of transfected cells (data not shown). However, because active caspase amounts undetectable by standard Western analysis may still be significant in cell death (26), we used affinity capture of caspases as a more sensitive approach. Using this method all three catalytically competent caspase-7 constructs were readily labeled (Fig. 4A). However, when caspase-3 (Fig. 4B) and caspase-6 (Fig. 4C) were selectively immunoprecipitated from transfected cell lysates no labeling was detected. The intensity of the labeled caspase-3 and -6 in a positive control extract, programmed to undergo caspase activation by Bax-transfection, suggests that none of those caspases were activated by the expressed caspase-7. Importantly, the lack of caspase-3 and -6 activation following overexpression of caspase-7 indicates that none of the apical caspases (caspase-8, -9, or -10) able to activate executioner zymogens had become activated.
|
Cell Death by Caspase-7—Having established that caspase-7
overexpression does not lead to the activation of other caspases, we looked
for processing of indicator death substrates known to be caspase-3 or -7
substrates. We choose PARP
(27) as a nuclear substrate
and PKC (28) as a
cytosolic substrate. Expression of caspase-7 results in cleavage of PARP to
the diagnostic 86-kDa fragment, and a 36-kDa fragment characteristic of
apoptosis is generated from full-length PKC
(Fig. 5). In both cases,
N-caspase-7 was more effective in cleaving the substrates. It is
noteworthy that Bax overexpression results in more PARP cleavage than simple
ectopic expression of N-caspase-7. This suggests that whereas caspase-7
can cleave PARP in vivo, a collaborating caspase facilitates access
to PARP, possibly by enhancing nuclear entry. We examined lamin B1,
a known substrate of caspase-6
(29,
30), and found no cleavage
whereas almost the entire pool of lamin B1 is cleaved in
Bax-transfected cells. This demonstrates that caspase-7 overexpression does
not generate excessive nonspecific proteolysis.
|
We further examined DNA fragmentation
(Fig. 5B), a hallmark
of apoptosis (31). The extent
of DNA fragmentation was comparable to Bax-transfected cells with wild-type
and N-caspase-7, slightly reduced with NC-caspase-7, and absent from
cells transfected with an empty plasmid or caspase-7 catalytic mutant. To
determine whether apoptosis induced by caspase-7 implicates the mitochondrial
pathway we also checked for release of pro-apoptotic mediators from
mitochondria in transfected cells. No increase in cytosolic cytochrome
c (Fig. 5C)
or SMAC/DIABLO (data not shown) was detected as compared with positive control
cells transfected with Bax, ruling out participation of mitochondria in cell
death mediated by caspase-7.
Human Caspase-7 Is Not a Nuclear Caspase—Shortly after the
N-peptide of caspase-7 is a stretch of basic residues (K38KKK,
Fig. 1B) reminiscent
of a nuclear localization signal. Removal of the N-peptide during apoptosis
could expose this signal and cause relocation of caspase-7 as suggested for
the Xenopus ortholog
(12). We thus examined the
subcellular localization of FLAG-tagged caspase-7. However, because cell
death, for example by caspase-7 overexpression, is accompanied by important
morphologic changes we examined the subcellular localization of the respective
catalytic mutants in COS-7 cells (Fig.
6A). No difference in staining pattern was found for any
of the full-length, N, or NC-caspase-7 constructs; all were mainly
excluded from the nucleus. This is consistent with subcellular fractionation
experiments showing that both N and NC-caspase-7 catalytic mutant were
fully cytosolic (data not shown). Our results do not exclude that a small
portion of caspase-7 may find its way to apoptotic nuclei, but they rule out
the possibility that removal of the N-peptide allows an active transport or
accumulation of human caspase-7 in the nuclei.
|
Removal of the N-Peptide Alters Self-activation of
Caspase-7—The enhanced apoptosis attributable to N-peptide removal
may reflect an increase in the inherent enzymatic activity of caspase-7. To
test this we determined the kinetic parameters of recombinant proteins
expressed in E. coli. Proteins were expressed as C-terminal His tag
fusions and purified to near homogeneity by
Ni2+-affinity chromatography as described
(17). Using active
site-titrated enzymes, we established that N-caspase-7 is enzymatically
similar to wild type or NC-caspase-7. Indeed, Km and
kcat values for N-caspase-7 and NC-caspase-7 on the
fluorogenic substrate Ac-DEVD-AFC and the chromogenic substrate Ac-DEVD-pNA
were identical within experimental error
(Table I). It is thus clear
that any phenotypes observed in transfection experiments using the
corresponding proteins are unlikely to be due to differences in catalytic
activity but rather to some property of the N-peptide itself.
|
We noticed that during purification of the various recombinant protein from
E. coli, NC-caspase-7 was mainly present as a zymogen after 5 h of
expression whereas N-caspase-7 was mainly processed suggesting that the
N-peptide may delay auto-catalytic activation of the zymogen. Time-course
analysis of caspase-7 activation in E. coli allowed us to monitor the
conversion of the zymogen that is induced by autolytic cleavage
(Fig. 7A). Activation
of N-caspase-7 occurred concomitantly with cleavage of the inter-chain
linker region at Asp198 (Fig.
1B), as determined by Edman degradation, as early as 3-h
post-induction to produce large and a small subunits
(Fig. 7A,
bottom). In comparison, autolytic conversion of NC-caspase-7 was
delayed by about 2 h. In both cases a second cleavage occurs resulting in the
trimming of the small subunit N terminus at Asp206. These data
indicate that the N-peptide plays a role in stabilizing the zymogen, at least
in vitro under the abnormally high concentrations obtained during
expression in E. coli. However, it is possible that differences in
expression level or onset of expression influenced activation, so we designed
an experiment in which the processing of pro-caspase-7 by added caspase-7
could be more carefully controlled. We used the catalytic mutant (C186A) forms
of N and NC-caspase-7 to abrogate any complications from additional
activity of the processed zymogens. These were in vitro translated in
the presence of [35S]methionine and used to assay the ability of
recombinant caspase-7 to process the zymogens. The presence of a secondary
translation product initiated at Met45 was observed (marked by the
asterisk); its mutation to alanine removed the secondary translation product
but did not change the outcome of the results in the experiments described
below.
|
We observed that processing of the labeled N-caspase-7 zymogen was
more efficient than for the NC-caspase-7 zymogen for the same catalytic amount
of added active recombinant caspase-7 (Fig.
7B). This suggests that the N-peptide interferes with
conversion of the zymogen and could partially explain why the onset of cell
death is delayed in 293A cells for WT/NC-caspase-7
(Fig. 2). We conclude that
removal of the N-peptide does not affect the intrinsic enzymatic activity of
the enzyme, but increases the rate of zymogen processing. Consequently, it is
pro-caspase-7 acting as a substrate, not caspase-7 acting as an enzyme, that
is influenced by the N-peptide.
The N-peptide of Caspase-7 Prevents Activation in Vivo but Not in
Vitro—Self-processing is not thought to be the endogenous pathway
to caspase-7 activation used in mammalian cells induced to undergo apoptosis.
Indeed this is thought to be the role of apical caspase-8 and -9 (reviewed in
Refs.
1–3).
We thus investigated the role of the N-peptide in the activation of
procaspase-7 by caspase-8 in vitro
(Fig. 8A). Serial
dilutions of recombinant caspase-8 were used to process recombinant catalytic
mutant versions of N-caspase-7 and NC-caspase-7. We introduced the M45A
mutation to eliminate a secondary translation product frequently observed when
the proteins are expressed in E. coli. This mutation is not expected
to alter any properties of caspase-7 since it is not conserved in mouse or in
the hamster orthologs (both have an alanine at this position). Cleavage of
caspase-7 zymogen by caspase-8 generates two large and two small subunit
derivatives corresponding to proteolysis of the linker region at the two
activation sites (see above and Fig.
1B). Significantly, no difference was observed in the
rates of cleavage of the two forms of caspase-7
(Fig. 8A). A similar
conclusion was drawn using caspase-9 as an activator although 10-fold less
enzyme was required to process the same amount of zymogen (data not shown).
Thus the N-peptide does not alter the ability of apical caspases to activate
pro-caspase-7 in vitro.
|
To test whether the N-peptide influences zymogen activation in vivo we expressed catalytic mutant caspase-7 constructs in 293A cells, where they act as a reporter of processing without themselves participating in the process. To analyze cleavage products, FLAG-tagged proteins were immunoprecipitated and revealed with anti-FLAG antibody (which results in clearer detection of the analyzed protein). Small subunit related bands produced from the pro-caspase-7 (Fig. 8C) are similar to the ones observed in the in vitro assay and should represent caspase-8-mediated processing. This processing was greatly reduced for NC-caspase-7. NC-caspase-7 is processed, and therefore activated, far more slowly than WT-caspase-7. We conclude that the N-peptide regulates the activation of caspase-7 by the extrinsic pathway (caspase-8) in vivo, but not in vitro. The reasons for this are discussed below.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
The size and nature of the N-terminal peptides defines different caspase
groups. Initiators such as caspase-8, -9, and -10 contain relatively long
peptides that encompass distinct recognized folding domains (death effector
domains or caspase recruitment domains) required for their recruitment to
activator complexes. In these complexes the apical caspases attain catalytic
activity, becoming able to activate executioner caspase-3 and -7 by direct
limited proteolysis. The zymogens of the executioners contain relatively short
N-peptides without a recognizable folding pattern, and the function of these
N-peptides is elusive. The caspase-3 and -7 N-peptides are removed during
apoptosis and they are highly conserved from frogs to man. For example,
although human and Xenopus pro-caspase-7 share limited amino acid
identity within their respective N-terminal domain, both contain the
DSVD
A cleavage site for the N-peptide and the KKKK sequence described in
Fig. 1. This likely reflects a
function preserved in all vertebrate orthologs of caspase-7.
One proposal for the N-peptide of caspase-7 is that it covers a partial nuclear localization signal comprising the KKKK segment (12). While this may be the case with Xenopus caspase-7, with which the forgoing study was conducted, it does not seem to be the case for human caspase-7, since we were unable to observe any nuclear accumulation of full-length or N-peptide deleted caspase-7 in COS-7 cells. Massive overexpression of caspase-7 can result in some nuclear translocation, but this is likely to be an artifact of forced overexpression. Consequently, the postulated function of the N-peptide in regulating nuclear translocation has not been conserved in vertebrates.
Another role that has been proposed for the N-peptide is the silencing of the zymogen form of caspase-3 (10). The general conclusion from this paper was that the caspase-3 N-peptide retains the protease in an inactive state. At first glance it seems that a similar function could be ascribed to the caspase-7 N-peptide, since its deletion renders recipient mammalian cells more sensitive to death by caspase-7 transfection. However, we are not aware of instances where apoptosis is triggered at the effector stage, where caspase-7 resides, without participation of upstream initiator caspases and our results on death mediated by ectopic expression of caspase-7 do not signify that this is a route in vivo. They simply reveal that the N-peptide of caspase-7 alters the properties of pro-caspase-7 as a substrate of caspase-7, but importantly not as a substrate of the initiator caspase-8 or -9.
Our data suggest that differences in the rate of activation of wt, N
and NC pro-caspase-7 are due to the zymogens acting as substrates for
processing mediated by caspase-7, but significantly not the physiologic
activators caspase-8 and -9 in vitro. This suggests that the presence
of the N-peptide either hinders access of caspase-7 to the zymogen linker, or
acts as a weak inhibitor of caspase-7 itself. The latter possibility is ruled
out by our kinetic data that show essentially identical catalytic parameters
for each of the active forms of caspase-7. The former possibility suggests
that the N-peptide alters the conformation of the zymogen in a subtle manner.
However, the atomic resolution structures available for full-length caspase-7
or pro-caspase-7 do not reveal density for residues prior to Ser47,
either because there is no defined structure or because of mobility with
respect to the fixed bulk of the catalytic domains in the crystals
(4,
6,
39). In addition, there are no
contacts visible between the N-peptide and the catalytic domain. This
observation weakens the hypothesis that the N-peptide can directly modulate
caspase-7 structure, activity or activatability, and so another mechanism must
be considered for the function of the N-peptide.
There is a major difference in the processing pattern of caspase-3 and -7 that sheds light on the differential function of their respective N-peptides. Caspase-3 activation, both in recombinant and natural settings, occurs by inter-domain cleavage followed by N-peptide removal (40). In stark contrast, in the same settings the activation of caspase-7 occurs first by N-peptide removal, and it has been shown that caspase-3 removes the N-peptide of caspase-7 before it is activated by the cascade initiator granzyme B in MCF-7 cells (9). Here we show that it is also true for triggering by caspase-8, since prevention of N-peptide removal diminishes caspase-7 interdomain processing in vivo. More importantly, we also show that initial removal of the N-peptide is essential for the efficient ultimate activation of caspase-7 in cell transfection experiments, and therefore presumably also in vivo.
There is still no clear evidence of the reason for the requirement of N-peptide removal in caspase-7 activation. However, the data are consistent with a physical sequestration of the zymogen from its apical activators. It seems that this sequestration is overcome when the N-peptide is removed from procaspase-7, after which it can be activated by the normal route of interdomain cleavage. Part of the mechanism may reside in the inherent properties of the N-peptide; it is a highly negatively charged region. Indeed, the isoelectric point of the full-length zymogen is much lower than the one of the protein lacking the N-peptide. Classical protein sequestration involves binding to a structural protein or organelle. The highly charged N-peptide and the following region could serve such a purpose. However, most subcellular fractionation and cell-free studies report endogenous caspase-7 as primarily cytosolic in healthy cells (13, 14), so it is not clear where the sequestering partner resides. But the results of our study underscore the importance of subcellular localization in the apoptotic process. In vitro, including in hypotonic cytosolic extracts (9) pro-caspase-7 is activated by caspase-8 and granzyme B in an N-peptide independent manner, but in whole mammalian cells these cascade initiators are hindered in their access until the N-peptide is removed, presumably by caspase-3.
The study presented here shows that caspase-7 is not simply a redundant executioner. It appears to require a two-step activation mechanism in vivo, though not in vitro (9). It is fully capable of conducting most of the apoptotic program, but is highly dependent on the activity of caspase-3 to remove the block due to the N-peptide before the canonical activation process begins. Clearly further studies are required to identify the molecular mechanism(s) by which the N-peptide prevents self-activation and sequestration of caspase-7. Undoubtedly, determining the structure of the N-peptide will shed light on both issues and provide insights on the role of caspase-7 in the apoptotic program.
|
FOOTNOTES |
|---|
To whom correspondence should be addressed: The Burnham Institute, 10901 N.
Torrey Pines Rd. La Jolla, CA 92037. Tel.: 858-646-3114; Fax: 858-713-6274;
E-mail:
gsalvesen{at}burnham.org.
1 The abbreviations used are: z-VAD-FMK,
benzyloxycarbonyl-VAD-fluoromethylketone; AFC, 7-amido-4-fluoromethylcoumarin;
CHO, aldehyde; Chx, cycloheximide; Ac-, acetyl-; DAPI,
4',6-diamidino-2-phenylindole; GrB, granzyme B; PARP, poly(ADP-ribose)
polymerase; pNA, paranitroanilide; CAPS, 3-(cyclohexylamino)propanesulfonic
acid; HRP, horseradish peroxidase; Pipes, 1,4-piperazinediethanesulfonic acid;
CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid; FITC,
fluorescein isothiocyanate.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  F. L. Scott, G. J. Fuchs, S. E. Boyd, J.-B. Denault, C. J. Hawkins, F. Dequiedt, and G. S. Salvesen Caspase-8 Cleaves Histone Deacetylase 7 and Abolishes Its Transcription Repressor Function J. Biol. Chem., July 11, 2008; 283(28): 19499 - 19510. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Twiddy, G. M. Cohen, M. MacFarlane, and K. Cain Caspase-7 Is Directly Activated by the ~700-kDa Apoptosome Complex and Is Released as a Stable XIAP-Caspase-7 ~200-kDa Complex J. Biol. Chem., February 17, 2006; 281(7): 3876 - 3888. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Feeney and A. C. Clark Reassembly of Active Caspase-3 Is Facilitated by the Propeptide J. Biol. Chem., December 2, 2005; 280(48): 39772 - 39785. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J. Zandy, S. Lakhani, T. Zheng, R. A. Flavell, and S. Bassnett Role of the Executioner Caspases during Lens Development J. Biol. Chem., August 26, 2005; 280(34): 30263 - 30272. [Abstract] [Full Text] [PDF]
|
|
), 
) and harvested at the
indicated time; the 0-h time point was when cells were transfected. Cell
lysates was prepared in mRIPA buffer and assayed for caspase activity using
Ac-DEVD-AFC substrate as described under "Experimental
Procedures." Background activity (empty plasmid) never reached more than
0.2 µM/min/mg and is omitted for clarity.
(10 ng/ml) and Chx (10 µg/ml) and
cell extracts were prepared in mRIPA containing protease inhibitors,
immunoprecipitated with anti-FLAG (M2) and visualized by immunoblotting to
reveal the processed small subunit. The asterisk marks a secondary
translation product initiated at Met45. NC-caspase-7 C186A tends to
reveal aberrant cleavage surrounding the N-peptide when the NC mutation is
introduced.