(Received for publication, January 22, 1997, and in revised form, March 11, 1997)
From the Department of Molecular Biology, Cell
Biology, and Biochemistry, Brown University, Providence, Rhode
Island 02912 and the ¶ Department of Biology, Howard University,
Washington, D. C. 20059
The apoptotic cysteine protease, caspase-3, is
expressed in cells as an inactive 32-kDa precursor from which 17 kDa
(p17) and 12 kDa (p12) subunits of the mature caspase-3 are
proteolytically generated during apoptosis. Two amino acid sequences,
ESMDS (amino acids 25-29) and IETD
S (amino acids 172-176), in
the precursor have been defined as the cleavage sites for the
production of the p17 and p12 subunits. Using a cell-free assay system,
we demonstrate that the caspase-3 precursor appears to be cleaved first
at the IETD
S site, producing the p12 subunit and a 20-kDa (p20)
peptide. Subsequently, the p20 is cleaved at the ESMD
S site,
generating the mature p17 subunit. The cleavage at the IETD
S site
required a protease activity that was selectively inhibited by the
peptide, Ac-IETD-CHO (acetyl-IETD-aldehyde), and other protease
inhibitors, such as the cowpox viral serine protease inhibitor, CrmA,
and N-
-tosyl-L-phenylalanine chloromethyl
ketone. The protease that catalyzed the cleavage at the ESMD/S site
was selectively inhibited by another peptide, Ac-ESMD-CHO
(acetyl-ESMD-aldehyde). More interestingly, the caspase-3 inhibitor,
Ac-DEVD-CHO, but not the caspase-1 inhibitor, Ac-YVAD-CHO, also
selectively inhibited the protease activity that cleaves at the
ESMD
S site. This indicated that the cleavage at the ESMD
S site
was either autocatalytic or that it required a caspase-3-like activity.
In summary, we demonstrate that production of the p17:p12 form of
caspase-3 is a sequential two-step process and appears to require two
distinct enzymatic activities.
In recent years, evidence has accumulated that the degradation of certain proteins by members of the caspase family is a general biochemical event taking place in cells undergoing apoptosis (for reviews, see Refs. 1-3). One caspase family member, caspase-3 (4-6), has been studied extensively. In particular, several of the cellular protein targets of caspase-3 have been identifed. These include the DNA repair enzyme poly(ADP-ribose) polymerase (PARP)11component of the U1 small nuclear ribonucleoprotein (8), and the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs) (9-11). Interestingly, inhibition of caspase-3 or caspase-3-like proteases in various cells has been shown to block apoptosis (6, 12-17). In addition, the functional inactivation of the caspase-3 gene in knock-out mice results in the profound absence of apoptosis in certain tissues and lethality shortly after birth (18). Therefore, caspase-3 appears to be an extremely biologically relevant apoptotic protease.
All caspase family members are initially synthesized as inactive
precursors and require proteolytic processing themselves to generate
the two subunits that form the active protease. This suggests that the
apoptotic machinery may be regulated, in part, by a proteolytic cascade
(for review, see Ref. 19). In the case of caspase-3, the mature enzyme
is formed from 17 kDa (p17) and 12 kDa (p12) subunits, which are
produced from a 32-kDa precursor (4-6). There are two amino acid
sequences, ESMDS (amino acids 25-29) and IETD
S (amino acids
172-176), in the 32-kDa caspase-3 precursor protein that have been
defined as the cleavage sites for the production of the p17 and p12
subunits (5) (see Fig. 1). At both sites, the cleavage should occur
between the aspartic (Asp) and serine (Ser) residues (5). However, the
process of how the caspase-3 precursor is converted into the mature
caspase-3 remains uncharacterized, and the cellular proteases that
cleave these sites have not been identified.
We have recently shown that when the human promyelocytic leukemia HL-60 cell line was induced to undergo apoptosis by exposure to staurosporine (STS), one of the early biochemical events was the conversion of the 32-kDa caspase-3 precursor into the mature protease that was responsible for the subsequent proteolysis of DNA-PKcs (9). In addition, we have also demonstrated that apoptosis in HL-60 cells triggered by STS was independent of the synthesis of new proteins (20), indicating that HL-60 cells constitutively express all of the factors necessary for apoptosis including the caspase-3 precursor proteases. Therefore, cell extracts from HL-60 cells should provide an ideal system for the identification of the proteases that can catalyze the conversion of the caspase-3 precursor protein into an active apoptotic protease.
A recent study by Liu et al. (21) using a cell-free system has identified cytochrome c and dATP as two essential factors required for the induction of the proteolytic activation of caspase-3 in vitro. In our study, we have modified this in vitro assay system and investigated the process of proteolytic processing of caspase-3. Our results show that this process appears to involve two proteolytic steps and to be catalyzed by two different proteases. Lastly, our study provides insight into why certain protease inhibitors, including the caspase-3 selective inhibitor Ac-DEVD-CHO, can block apoptosis.
Fetal bovine serum, STS, EDTA, dithiothreitol,
phenylmethylsulfonyl fluoride,
N--tosyl-L-phenylalanine chloromethyl ketone (TPCK), aprotinin, antipain, leupeptin, peroxidase-conjugated goat
anti-mouse IgG antibodies, and rat heart cytochrome c were purchased from Sigma. The tetrapeptide inhibitor of caspase-3, acetyl-DEVD-aldehyde (Ac-DEVD-CHO) (5, 7), was purchased from Bachem
Bioscience, Inc. (Torrance, CA). The tetrapeptide inhibitor of
caspase-1, acetyl-YVAD-aldehyde (Ac-YVAD-CHO) (22), was purchased from
Oncogene Sciences, Inc. (Uniondale, NY). Acetyl-ESMD-aldehyde (Ac-ESMD-CHO) and acetyl-IETD-aldehyde (Ac-IETD-CHO) were synthesized by Quality Controlled Biochemicals, Inc. (Hopkinton, MA). The purity of
these peptides was approximately 98%. The cowpox viral serine protease
inhibitor, CrmA, was purchased from Kamiya Biomedical Co. (Seattle,
WA). Monoclonal antibodies against human DNA-PKcs have been
described (9). Mouse monoclonal anti-human caspase-3 antibodies were
purchased from Transduction Laboratories, Inc. (Lexington, KY). The ECL
Western blot analysis reagents were purchased from Amersham Life
Science, Inc. (Arlington Heights, IL).
HL-60 cells were originally obtained from the American Type Culture Collection. The cells were cultured in RPMI 1640 medium supplemented with 20% fetal bovine serum, 100 units/ml penicillin and 50 units/ml streptomycin.
Preparation of Cell ExtractsFive liters of HL-60 cells
were pelleted and washed in 200 ml of phosphate-buffered saline three
times. After the final wash, the cell pellet was resuspended in an
equal volume of hypotonic buffer (10 mM HEPES, pH 7.5, 5 mM MgCl2, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl, 50 µg/ml each of leupeptin,
aprotinin and antipain). The cell suspension was incubated on ice for
15 min. The cells were then broken by passing them 5 times through a
syringe with a 25-gauge needle. This resulted in the lysis of over 99%
of the cells. The lysate was first centrifuged at 12,500 × g for 10 min at 4 °C. Then, the supernatant was
centrifuged at 100,000 × g for 1 h at 4 °C.
The 100,000 × g supernatant (S-100) was
transferred to another tube, and protein concentration was determined
and diluted to a final concentration of 10 mg/ml with the hypotonic
buffer. The final preparation also contained 5% glycerol. This
preparation was kept at 85 °C.
Conditions described by Liu et al. (21) were modified for this assay. Briefly, 20 µl of the S-100 preparation was mixed with 0.4 µg of cytochrome c and 7.5 mM EDTA in a final volume of 25 µl. The mixture was incubated at 37 °C for 30-60 min. The reaction was terminated by addition of 5 µl of 6 × SDS sample loading buffer and heated at 100 °C for 5 min. Alternatively, 3 µl (~500 ng) of a purified DNA-PK solution (9) was added to the mixture, and the incubation was continued for another 60 min before it was terminated. When protease inhibitors were studied, the extract was first incubated with an inhibitor for 10 min at 20 °C and then mixed with cytochrome c and incubated at 37 °C. Proteins were then subjected to SDS-polyacrylamide gel electrophoresis (12% for caspase-3 analysis and 5% for DNA-PK proteolysis analysis). The proteins were transferred onto nitrocellulose paper by electrophoresis in a Trans-Blot chamber, and proteins were identified by Western blot analysis using antibodies and ECL Western reagents. The Western blot analysis conditions for DNA-PKcs have been described (9). For detection of caspase-3, a new protocol was developed, which ensured detection of the p17 protein. Briefly, the blot was incubated in a TBST-albumin solution (10 mM Tris, pH 7.5, 150 mM NaCl, 0.025% Tween 20, 0.1% bovine serum albumin) containing 10% low fat dry milk powder for 1 h at room temperature. The blot was rinsed in TBST-albumin solution once, incubated with the mouse anti-human caspase-3 antibody in the TBST-albumin solution (1 µg/ml) for 20 min at room-temperature, and then washed in TBST-albumin solution for 10 min. The blot was subsequently incubated with peroxidase-conjugated goat anti-mouse IgG antibodies in TBST-albumin solution (1:1000 dilution) for 20 min at room temperature and washed in TBST-albumin solution for 30 min, and then the caspase-3 protein was detected using ECL Western detection reagents.
Liu et al. (21)
recently showed that in the presence of dATP, cytochrome c,
and in a cytosolic extract from Hela cells, caspase-3 precursors,
which were translated in vitro, were
converted into active caspase-3. Using a similar approach, we studied
the effect of cytochrome c on the status of the endogenous
32-kDa caspase-3 precursor present in cytosolic extracts derived from HL-60 cells. Incubation of the HL-60 S-100 extract with 0.4 µg of
cytochrome c at 37 °C for 60 min resulted in the complete
loss of the 32-kDa caspase-3 precursor and was accompanied by the
appearence of the p17 subunit (which is the only subunit recognized by
the antibody), indicating that the 32-kDa precursor was proteolytically processed (Fig. 2A, lane 3). This
effect appeared to be specifically induced by cytochrome c
since incubation of the S-100 extract alone at 37 °C for either 0 or
60 min did not induce the processing of the 32-kDa caspase-3 precursor
(Fig. 2A, lanes 1 and 2,
respectively). To confirm these results, the presence of caspase-3
activity was monitored using purified DNA-PK as a substrate (9-11).
Thus, purified DNA-PK was incubated with either untreated or cytochrome
c-treated HL-60 S-100 extract and subsequently subjected to
Western blot analyses. Only the cytochrome c-treated S-100
extract contained an activity that cleaved the DNA-PKcs
into a 150-kDa fragment (Fig. 2B, lane 3), which
was identical to the size observed in vivo in apoptotic
cells (9). Therefore, cytochrome c treatment of HL-60 S-100
extract induced the appearance of an active caspase-3.
The kinetics of cytochrome c-induced proteolytic processing of caspase-3 were determined. Cytochrome c induced the appearance of the p17 subunit within 10 min and complete conversion occurred within 30 min (Fig. 2C). It was interesting to note that during the initial 10-20 min, a small amount of a 20-kDa (p20) peptide was also detected (Fig. 2C, lanes 2 and 3). However, the level of this p20 peptide was far below that of the p17 peptide. According to the reported work on the molecular cloning of the caspase-3 gene (4, 5), the p17 polypeptide is contained within p20 (see also Fig. 1). The lack of significant accumulation of p20 in the assay suggested that p20 was rapidly converted into the p17 subunit under the assay conditions (see also below). Thus, efficient and complete conversion of endogenous precursor caspase-3 into mature p17 and p12 subunits was achieved by addition of cytochrome c to HL-60 cytoplasmic extracts.
Selective Inhibition of Cleavage at the ESMDThe predicted cleavage
sites in the 32-kDa caspase-3 precursor are defined by the amino acid
sequences ESMDS (amino acids 25-29) and IETD
S (amino acids
172-176), with the cleavage presumed to occur between the Asp and Ser
residues at both sites (5). Aldehyde peptide analogs of cleavage site
amino acid sequences have been shown to function as highly selective
inhibitors of caspase-1 (22) and caspase-3 (5, 7, 9). Therefore, we
synthesized acetyl-ESMD-aldehyde (Ac-ESMD-CHO) and acetyl-IETD-aldehyde (Ac-IETD-CHO) peptide analogs and used them as potential site-selective inhibitors in our caspase-3 processing assay. The Ac-ESMD-CHO peptide
analog, in a range of 2.5-5 µM, blocked the formation of
the p17 subunit and concomitantly induced the accumulation of the p20
peptide (Fig. 3A). In striking contrast, the
presence of the Ac-IETD-CHO peptide analog, in a concentration range of 0.5 µM, blocked the formation of the p17 subunit and
concomitantly induced the accumulation of the 32-kDa precursor (Fig.
3B). In particular, the Ac-IETD-CHO peptide did not induce
accumulation of the p20 product. Thus, the peptide analogs Ac-EMSD-CHO
and Ac-IETD-CHO both inhibit caspase-3 processing but they result in
the accumulation of different reaction products.
A Caspase-3-selective Inhibitor Preferentially Blocks Cleavage at the ESMD
To extend the above results,
we used two additional well characterized aldehyde peptides,
Ac-DEVD-CHO, which is a highly selective inhibitor of caspase-3 (5, 7,
9), and Ac-YVAD-CHO, which is a highly selective inhibitor of caspase-1
(22). Addition of the Ac-DEVD-CHO inhibitor to our assay, in a
concentration range of 50-100 nM, completely blocked the
production of the p17 subunit and resulted in the concomitant
accumulation of p20 (Fig. 3C), indicating that Ac-DEVD-CHO
blocked processing of caspase-3 at the ESMDS site in the p20
peptide. In contrast, the caspase-1 inhibitor, Ac-YVAD-CHO, did not
have any effect on the proteolytic processing of caspase-3 (Fig.
3D), consistent with our earlier studies (9). These results
suggested that (i) a caspase-3-like proteolytic activity was required
for cleavage at the ESMD
S site in the p20 peptide and (ii) that the
pattern of inhibition induced by Ac-DEVD-CHO was similar to that
observed with Ac-EMSD-CHO.
Two additional inhibitors were found to have
a profound inhibitory effect on proteolysis of the caspase-3 precursor.
TPCK, which can inhibit serine proteases (30), and, in some studies, apoptosis (25-27) prevented the 32-kDa precursor from being processed at all in a dose-dependent fashion (Fig.
4A). CrmA, which is a cowpox viral protein that has been
shown to inhibit the activity of caspase-1 and caspase-3 and block
apoptosis (6, 12, 23, 24), also prevented the 32-kDa caspase-3
precursor from being processed at all (Fig. 4B). These
findings indicated that (i) both TPCK and CrmA can inhibit
caspase-3 activation and (ii) that their pattern of
inhibition was similar to that observed with Ac-IETD-CHO.
In this study, we have demonstrated that in a cell-free assay system cytochrome c induces the proteolytic conversion of the 32-kDa caspase-3 precursor into the mature caspase-3. Furthermore, our results showed that this process involved two sequential cleavage steps and that each step was catalyzed by a unique proteolytic activity. Finally, our results provide additional insight into the mechanisms by which protease inhibitors block apoptosis.
Proteolytic Processing of Caspase-3Molecular cloning of the
caspase-3 gene and biochemical analysis of the purified, active
caspase-3 showed that caspase-3 was initially synthesized as an
inactive 32-kDa precursor (4-6), which was subsequently processed by
proteolytic cleavage at ESMDS (amino acids 25-29) and IETD
S
(amino acids 172-176) to produce the p12 and p17 subunits that form
the mature caspase-3 (5) (see Fig. 1). Our results indicate that this
process appears to require two sequential cleavage steps
(Fig. 5). Cleavage at the IETD/S site appeared to occur
first, yielding a mature p12 subunit and a p20 intermediate product.
However, it should be pointed out that since good antibodies that can
recognize the p12 subunit are not available now, the exact status of
this subunit during the assay period remains speculative. Subsequently,
a cleavage occurred at the ESMD
S site in the p20 intermediate,
yielding the mature p17 subunit. Together, the p12 and p17 subunits
then form the mature caspase-3 (Fig. 5). Thus, blocking the cleavage at
IETD
S with Ac-IETD-CHO, CrmA, or TPCK completely abrogated the
appearance of any processed products of caspase-3, whereas blocking the
cleavage at ESMD
S with Ac-EMSD-CHO or Ac-DEVD-CHO allowed the
primary cleavage at IETD
S to occur but not the secondary cleavage at
EMSD
S (Figs. 3 and 4). Thus, our data provide strong evidence for a
sequential two-step processing mechanism.
The Nature of the Protease That Cleaves the IETD
Our
results suggest that the protease activity required for the cleavage
step at the IETDS site in the 32-kDa caspase-3 precursor has some
novel properties. First, it is sensitive to the sequence-specific peptide inhibitor, Ac-IETD-CHO (Fig. 3B), and it is also
sensitive to CrmA and TPCK (Fig. 4). CrmA is a cowpox viral protein
that can inhibit several members of the caspase family including
caspase-1 (28, 29), caspase-3 (6), caspase-6 (34), and caspase-8 (32).
However, it is unlikely that, in our system, CrmA is inhibiting the
caspase-1 or caspase-3 activities. First, Northern blot analysis of
HL-60 cells found no evidence for the expression of the caspase-1 gene
(data not shown). Second, the presence of the highly selective caspase-1 inhibitor, Ac-YVAD-CHO (22), had no effect on the process of
proteolytic activation of caspase-3 (Fig. 3D). Third, the
cleavage at the IETD
S site proceeded normally in the presence of the
caspase-3 inhibitor (Fig. 3C). Therefore, it is unlikely that caspase-1 or caspase-3-like proteases are involved with this event. It remains to be seen if other caspases, such as caspase-6 and
caspase 8, might be responsible for cleaving the caspase-3 IETD
S
site. This is potentially likely since both activated caspase-6 (33,
34) and caspase-8 (32) have been shown to initiate the proteolytic
activation process of caspase-3 in vitro, and the activity
of both caspases is inhibitable by CrmA (32, 34).
TPCK is a general inhibitor of chymotrypsin-like serine proteases (30),
and while it has not been shown to have an inhibitory effect on
caspases, it has been observed, in some studies, to have an inhibitory
effect on apoptosis (25-27). Thus, it is possible that TPCK, at the
concentrations used in our assay, directly inhibited the protease that
cleaves at the IETDS site, thus inhibiting the processing of the
caspase-3 precursor (Figs. 4A and 5). If this is true, then
both TPCK and CrmA may be inhibiting the same protease. Alternatively,
it is also possible that the TPCK-sensitive protease is different from
that of CrmA. This would suggest that the TPCK-sensitive protease acts
upstream of the CrmA-sensitive protease. Experiments to test this
possibility are underway.
One surprising finding in this study is
the observation that the highly selective caspase-3 inhibitor,
Ac-DEVD-CHO (5, 7, 9, 14), blocked only the second cleavage step at the ESMDS site in the p20 peptide (Fig. 3C). This prevented
the conversion of the p20 intermediate peptide into the mature p17
subunit. Apparently, the inhibitory effect of Ac-DEVD-CHO was selective
upon this cleavage step, since it had no observable effect on the first
cleavage step at the IETD
S site in the 32-kDa precursor (Fig.
3C). This suggests that, unlike the protease activity that
was required for cleaving the IETD
S site, either caspase-3 itself
or, more likely, a caspase-3-like proteolytic activity is specifically required for cleaving the ESMD
S site. It could be argued that the
HL-60 cell S-100 extract contained a trace amount of endogenous, active
caspase-3, which was responsible for generating the p17 subunit once
the p20 product was available. However, this appears to be unlikely for
the following reasons. First, we have not been able to detect the
presence of p17 in the untreated S-100 extract even in Western blots
with overloaded protein samples (data not shown). Second, the effective
concentration of Ac-DEVD-CHO (50-100 nM; see Fig.
3C) that was required for the inhibition of this protease
activity was relatively high. In assays that have been described for
assessing the inhibitor potency of the Ac-DEVD-CHO peptide on purified
caspase-3, the Ki values have been estimated to be
less than 1 nM (5). The fact that Ac-DEVD-CHO inhibited the
protease activity that cleaves the ESMD
S site in a range of 50-100
nM suggests that this activity could not be attributed to
the presence of a trace amount of the active caspase-3 in the extract.
Rather, it is likely that this activity is simply caspase-3-like.
Aternatively, it could be speculated that p20 and p12 may form a
partially active intermediate protease that undergoes autocatalysis at
the ESMDS site, producing the fully active caspase-3 consisting of
p17 and p12 (Fig. 5). Unlike the fully active caspase-3, which is
extremely sensitive to the peptide inhibitor Ac-DEVD-CHO (5), the
partially active p20/p12 intermediate protease may have a much lower
affinity for the Ac-DEVD-CHO peptide and thus be less sensitive to it
than the fully active p17/p12 protease. Experimentally, it should be
possible to prove or disapprove this hypothesis. Introduction of a
point mutation into the ESMD
S site that makes it uncleavable should
give rise to the production of only p12 and p20 peptides. If the
p20/p12 complex has any proteolytic activity, then it may be able to
quantify this using caspase-3 substrates such as PARP (5, 7) and
DNA-PKcs (9-11).
It is worthy of note that CrmA, TPCK, and Ac-DEVD-CHO blocked conversion of the inactive 32-kDa caspase-3 precursor into the mature caspase-3 (Figs. 3C and 4). All three inhibitors have been shown, by various investigators, to block apoptosis in a variety of cells under many different treatments (6, 12, 14, 23-27, 31). The reason CrmA and Ac-DEVD-CHO block apoptosis was thought to be due to their direct inhibitory effect on the active apoptotic caspase-3 (5, 6). The results of this study have revealed that both inhibitors are also capable of preventing the conversion of the inactive caspase-3 precursor into the active protease.
Finally, although TPCK has been found to inhibit apoptosis in several studies (25-27), the mechanism by which it mediates this effect was unclear. In particular, it has been well characterized that TPCK has no direct inhibitory effect on caspase-3 (5, 9). Therefore, we suggest that at least one mechanism by which TPCK prevents apoptosis is that it can inhibit a novel protease that participates in the process of the conversion of the caspase-3 precursor to the active apoptotic protease.