Calcium-dependent, interleukin 1-converting enzyme inhibitor-insensitive degradation of lamin B1 and DNA fragmentation in isolated thymocyte nuclei.

Recent work suggests that the proteolytic degradation of the nuclear lamins is a common event in apoptosis, although the nature of the proteases involved is still not clear. Our previous work showed that the degradation of lamin B1 in glucocorticoid-treated thymocytes occurs via a Ca2+-sensitive mechanism and that exogenous Ca2+ promotes lamin degradation in isolated thymocyte nuclei from untreated cells. Here we demonstrate that peptide-based inhibitors of the interleukin 1beta-converting enzyme family of cysteine proteases (Tyr-Val-Ala-Asp fluoromethyl ketone) and of the nuclear scaffold multicatalytic proteinase (Ala-Pro-Phe chloromethyl ketone) block the degradation of lamin B1 to a 21-kDa fragment in thymocytes treated with glucocorticoid, the Ca2+-mobilizing agent thapsigargin, or antibodies to the T cell receptor. However, among a panel of inhibitors specific for several different proteases implicated in apoptosis, only tosylphenylalanyl chloromethyl ketone and the nuclear scaffold protease inhibitor block lamin degradation, histone H1 cleavage, and DNA fragmentation in isolated thymocyte nuclei incubated with Ca2+. Overexpression of human BCL-2 in nuclei by stable transfection resulted in an inhibition of Ca2+-stimulated lamin degradation and DNA fragmentation, suggesting that endogenous nuclear BCL-2 regulates activation of the nuclear scaffold protease. The results demonstrate the existence of an alternative pathway of lamin degradation and DNA fragmentation mediated by a resident Ca2+-stimulated nuclear protease that is not directly dependent upon activation of the interleukin 1beta-converting enzyme family of cell death regulators.

Recent work suggests that the proteolytic degradation of the nuclear lamins is a common event in apoptosis, although the nature of the proteases involved is still not clear. Our previous work showed that the degradation of lamin B 1 in glucocorticoid-treated thymocytes occurs via a Ca 2؉ -sensitive mechanism and that exogenous Ca 2؉ promotes lamin degradation in isolated thymocyte nuclei from untreated cells. Here we demonstrate that peptide-based inhibitors of the interleukin 1␤-converting enzyme family of cysteine proteases (Tyr-Val-Ala-Asp fluoromethyl ketone) and of the nuclear scaffold multicatalytic proteinase (Ala-Pro-Phe chloromethyl ketone) block the degradation of lamin B 1 to a 21-kDa fragment in thymocytes treated with glucocorticoid, the Ca 2؉ -mobilizing agent thapsigargin, or antibodies to the T cell receptor. However, among a panel of inhibitors specific for several different proteases implicated in apoptosis, only tosylphenylalanyl chloromethyl ketone and the nuclear scaffold protease inhibitor block lamin degradation, histone H1 cleavage, and DNA fragmentation in isolated thymocyte nuclei incubated with Ca 2؉ . Overexpression of human BCL-2 in nuclei by stable transfection resulted in an inhibition of Ca 2؉ -stimulated lamin degradation and DNA fragmentation, suggesting that endogenous nuclear BCL-2 regulates activation of the nuclear scaffold protease. The results demonstrate the existence of an alternative pathway of lamin degradation and DNA fragmentation mediated by a resident Ca 2؉ -stimulated nuclear protease that is not directly dependent upon activation of the interleukin 1␤-converting enzyme family of cell death regulators.
Apoptosis is a process of physiological cell death characterized by a series of stereotyped morphological alterations and endogenous endonuclease activation leading to the generation of large (50 kb 1 ) and oligonucleosomal DNA fragments (1,2). Although inhibitors of macromolecular synthesis prevent apoptosis in many circumstances, it is currently thought that the essential enzymatic components of the effector pathway are constitutively expressed in all mammalian cells beyond the blastomere stage of embryonic development (3). The identities of these cell death mediators are currently not known, but a major recent breakthrough toward their identification has come from work on the genes that regulate programmed cell death in the nematode Caenorhabditis elegans (4). This work has revealed that the C. elegans death promoter ced-3 is structurally homologous to a growing family of proteins related to human interleukin 1␤-converting enzyme (ICE) (5), a cysteine protease. Subsequent research has shown that specific viral and peptide-based inhibitors of the ICE family block apoptotic cell death in many mammalian models (4 -9), indicating that ICE and its homologs play crucial roles in the effector phase of apoptosis.
Recent work has shown that the nuclear lamins are cleaved by one or more members of the ICE family during apoptosis. Early work by Kaufmann (10) demonstrated that the oligonucleosomal DNA fragmentation induced by cancer chemotherapeutic agents in a human myelogenous leukemia line (HL-60) was associated with the degradation of lamin B 1 (and several other nuclear polypeptides). Subsequent work has shown that lamin degradation is associated with DNA fragmentation and cell death in several other models of apoptosis (11)(12)(13)(14) and with chromatin condensation and endonuclease activation in isolated nuclei incubated with apoptosis-promoting cell extracts (14,15) or Ca 2ϩ (11). The identities of the lamin proteases are not known, although preliminary evidence suggests that an ICE family member may be directly involved (15), and a Ca 2ϩdependent serine protease exhibiting "chymotrypsin-like" activity that is associated with the nuclear scaffold (NS) has been implicated in another (16). A direct role for lamin degradation in chromatin condensation and DNA fragmentation is attractive, as lamin solubilization is associated with chromatin condensation in both apoptosis and mitosis, the spacing of lamin binding sites in chromatin (periodicity of approximately 20 -50 kb) mirrors the sizes of large DNA fragments produced at the commitment phase of apoptosis (17)(18)(19)(20)(21), and the kinetics of lamin degradation and 50-kb DNA fragment formation overlap (11).
Much of our previous work has been concerned with identifying the biochemical mechanisms that regulate endogenous endonuclease activation and cell death in immature thymocytes undergoing apoptosis in response to a variety of stimuli. We have shown that early, sustained increases in the cytosolic Ca 2ϩ concentration are commonly involved, and Ca 2ϩ buffering agents can interfere with endonuclease activation and cell death (22). It is possible that these Ca 2ϩ alterations exert direct effects on components of the apoptotic effector machinery resident in the nucleus, as a number of investigators have demonstrated that typical morphological alterations and chromatin cleavage can be induced in isolated nuclei from untreated cells upon incubation in the presence of exogenous Ca 2ϩ (23)(24)(25)(26)(27). However, to date the proteases responsible for Ca 2ϩ -dependent lamin cleavage in apoptotic cells and isolated nuclei have not been characterized, nor has the possible interrelationship between protease activation and DNA fragmentation been defined. To address these issues was the aim of the present study.

EXPERIMENTAL PROCEDURES
Cells and Reagents-Thymocyte suspensions prepared from 4 -8week-old BALB/c mice were maintained in RPMI 1640 medium supplemented with 10% fetal calf serum, antibiotics, and 50 M 2-mercaptoethanol. The T-T hybridoma 2B4.11 was maintained in the RPMI 1640 medium described. Transfectants were generated by electroporation at 100 V and 800 microfarads using 1 ϫ 10 7 2B4.11 cells and 10 g of a linearized plasmid containing human bcl-2 under the control of the SV40 promoter/enhancer (Bil2-Neo; generously provided by Dr. Timothy McDonnell, The University of Texas M. D. Anderson Cancer Center). Cells were then selected for 2 weeks in the presence of 0.4 mg/ml G418 (Geneticin, Life Technologies, Inc.), and BCL-2 expression in the resistant cells was verified by immunoblotting with the anti-human BCL-2 monoclonal antibody 6C8. The metastatic murine melanoma line K1735 clone M2X21 was maintained in modified Eagle's medium supplemented with 10% fetal calf serum, vitamins, and sodium pyruvate.
Transfectants were generated by calcium phosphate precipitation using the Bil2-Neo plasmid as described previously (28). Cells were then selected for 3 weeks in the presence of 0.4 mg/ml G418, and BCL-2 expression in resistant colonies was determined by immunoblotting. Viability was assessed by trypan blue exclusion.
Peptide inhibitors of ICE (YVADfmk, zVADfmk, based on a cleavage site in interleukin-1␤) and CPP32 (DEVDfmk, based on a cleavage site in PARP) were purchased from Enzyme Systems Products, Inc. (Dublin, CA). Another peptide inhibitor of ICE (YVADcmk) and an inhibitor of the NS protease (APFcmk) were from Bachem (King of Prussia, PA). All other inhibitors were purchased from Sigma.
Isolation of Nuclei-Cells were lysed in 0.1% Nonidet P-40 and nuclei isolated by centrifugation at 200 ϫ g through a glycerol cushion as described previously (11). Nuclei were washed twice in incubation buffer (150 mM KCl, 5 mM MgCl 2 , 25 mM Tris, pH 7.5) and resuspended in the same buffer for incubation. The purity of representative preparations from each cell type was verified by electron microscopy.
DNA Fragmentation Analysis-Quantification of DNA fragmentation was accomplished using the diphenylamine reagent as described previously (29). Briefly, cells were harvested by centrifugation, lysed in 0.5 ml of a buffer containing 0.5% Triton X-100, 25  mM EGTA, and 10 mM EDTA for 15 min on ice, and samples were centrifuged for 20 min at 12,000 ϫ g to separate DNA fragments (supernatants) from intact chromatin (pellets). The DNA content in each fraction was determined using the diphenylamine reagent (30); results are expressed as the percentage of total DNA in each sample which resisted sedimentation at 12,000 ϫ g. Alternatively, the DNA present in the centrifuged supernatants was precipitated overnight at Ϫ20°C following the addition of 2 volumes of isopropyl alcohol and NaCl (to 0.5 M final concentration). After centrifugation at 12,000 ϫ g for 10 min, precipitates were harvested, dried, and incubated for 1 h in TE buffer (10 mM Tris, pH 8.0, 1 mM EDTA) containing 0.2 mg/ml proteinase K and 1 mg/ml RNase A. The DNA fragments were resolved by electrophoresis for 1 h at 75 V on 1.5% agarose gels preimpregnated with ethidium bromide. The DNA fragments were then photographed.
Field-inversion gel electrophoresis (FIGE) (FIGE mapper, Bio-Rad) was used to detect large DNA fragments. Aliquots of nuclei were mixed with equal volumes of a melted plug solution containing 150 mM NaCl, 10 mM Tris, pH 7.5, 10 mM EDTA, and 1.2% low melting point agarose, transferred to wells of a FIGE plug mold, and incubated for 15 min at 4°C. Plugs were then ejected and incubated for 16 h at 37°C in a lysis buffer containing 50 mM Tris, pH 8.0, 50 mM EDTA, 1% sarcosyl, and 0.2 mg/ml proteinase K. Plugs were then washed three times for 1 h at 4°C in a buffer containing 10 mM Tris and 1 mM EDTA, and the plugs were inserted into the wells of a 1% agarose gel. DNA fragments were then resolved by FIGE at a forward voltage of 180 V and a reverse voltage of 120 V for 16 h using a switch time ramp of 0.1-3.5 s (program 4 provided by the manufacturer). Gels were then stained with ethidium bromide and photographed.
Immunoblotting-Cells were lysed for 5 min at 37°C and rotated for 1 h at 4°C in a stringent lysis buffer containing 150 mM NaCl, 1% Triton X-100, 0.5% sodium dodecyl sulfate (SDS), 0.5% sodium deoxycholate, 1 mM phenylmethylsulfonyl fluoride, and 25 mM Tris, pH 7.5 (lamin B 1 and histone H1 experiments) or for 1 h at 4°C in the same buffer lacking SDS and sodium deoxycholate (BCL-2 experiments). Debris was pelletted by centrifugation for 5 min at 12,000 ϫ g, and pellets and/or supernatants were solubilized for 5 min at 100°C in Laemmli's SDS-polyacrylamide gel electrophoresis sample buffer containing 100 mM dithiothreitol. Polypeptides were resolved at 100 V on 12% gels and electrophoretically transferred to nitrocellulose (0. instructions. To detect histone H1, membranes were probed overnight with a mouse monoclonal antibody to histone H1 (Biogenex Inc., 1/1000 in block) and 1 h with sheep anti-mouse Ig (1/2000 in TBS-T) and developed by ECL. To detect human BCL-2, membranes were blocked overnight, probed for 1 h with 6C8 (1 g/ml in TBS-T), 1 h with rabbit anti-hamster Ig (1/2000 in TBS-T), and 1 h with donkey anti-rabbit Ig (1/2000 in TBS-T) and developed by ECL.

RESULTS
Previous work using specific peptide-based protease inhibitors has implicated two distinct proteases (ICE family members and the Ca 2ϩ -dependent NS protease) in the cleavage of lamin B 1 in apoptotic cells and in a cell-free system. We therefore tested the effects of these peptide inhibitors on lamin B 1 cleavage and DNA fragmentation in thymocytes induced to undergo apoptosis upon exposure to glucocorticoid, the Ca 2ϩmobilizing agent thapsigargin, and antibodies to the T cell receptor complex. The peptide inhibitor of ICE, YVADfmk (7-9, 31), effectively inhibited lamin B 1 cleavage in response to all three stimuli (Fig. 1A) and partially prevented both spontaneous and induced DNA fragmentation in the cells (Fig. 1, B and C). However, a peptide inhibitor of the NS protease, APFcmk (16), was even more potent, completely blocking lamin cleavage (Fig. 1A) and DNA fragmentation (Fig. 1, B and C) in all circumstances. Both inhibitors also exerted corresponding effects on the decrease in cell size normally associated with apoptosis in thymocytes (measured by forward scatter by fluorescence-activated cell sorter, not shown) and cell death (Fig.  1D). Together, these results confirm other recent findings indicating that peptide inhibitors of the ICE family (31,32) or the NS protease (32) interfere with all of the biochemical events of apoptosis in whole thymocytes that are exposed to certain apoptotic stimuli.
Recently, several groups have used different cell-free systems to study the regulation of various components of the apoptotic effector machinery (7,14,33,34). However, we and others have shown that glucocorticoids, thapsigargin, and T cell receptor triggering all induce apoptosis in thymocytes via mechanisms that involve alterations in intracellular Ca 2ϩ homeostasis (22) and that exogenous Ca 2ϩ can directly promote the nuclear changes of apoptosis in isolated nuclei in the absence of cytoplasm (11,18,24,27,35). Moreover, it has been suggested that Ca 2ϩ -mediated DNA fragmentation involves unidentified serine proteases constitutively present in isolated rat liver nuclei (36). We therefore investigated the involvement of various proteases in DNA fragmentation observed in nuclei incubated with Ca 2ϩ . Preliminary work demonstrated that exogenous Ca 2ϩ promoted DNA fragmentation in isolated thymocyte nuclei in a dose-dependent fashion (Fig. 2A). The response was characterized by the formation of both large 30 -50-kb (Fig.  2B) and oligonucleosomal (not shown) DNA fragments. Similarly, Ca 2ϩ promoted the cleavage of lamin B 1 to a distinct fragment of roughly 21 kDa (Fig. 2C) and the solubilization and cleavage of histone H1 (Fig. 2D)  cleavage of both lamin B 1 and histone H1 suggested that activation of endogenous nuclear proteases was associated with Ca 2ϩ -stimulated DNA fragmentation in isolated thymocyte nuclei. We therefore employed a panel of protease inhibitors to characterize further the activities involved. Two different calpain inhibitors, the serine protease inhibitor TLCK and YVADfmk both failed to affect DNA fragmentation (Fig. 3, A and B), lamin B 1 cleavage (Fig. 3C), or histone H1 solubilization and cleavage (Fig. 3D). Similarly, another peptide-based ICE family antagonist, DEVDfmk (37), had no effect on Ca 2ϩ -mediated proteolysis or DNA fragmentation in isolated thymocyte nuclei over a wide dose range (1-200 M, not shown). The only two inhibitors that were effective in this system were the serine protease antagonist TPCK and the NS protease antagonist APFcmk, which blocked all of the Ca 2ϩ -induced biochemical events (Fig. 3) in a dose-dependent fashion (Fig. 4). Thus, a protease sensitive to TPCK and APFcmk, but insensitive to ICE family and calpain antagonists, is required for Ca 2ϩ -mediated lamin B 1 and histone H1 solubilization and cleavage and for endonuclease activation in isolated thymocyte nuclei.
Previous work has shown that the nuclease inhibitors aurintricarboxylic acid (ATA) and Zn 2ϩ , are capable of inhibiting DNA fragmentation and cell death in thymocytes undergoing apoptosis (38) and can interfere with Ca 2ϩ -induced DNA fragmentation in isolated nuclei (18,26,36). However, whether these agents directly inhibit endonuclease activation or whether they interfere with the protease activation required for endonuclease activation has not been established. Incubation with a concentration of Zn 2ϩ which blocks apoptosis in whole thymocytes (26) did not prevent the cleavage of lamin B 1 , whereas ATA partially prevented the response (Fig. 3C). The nuclease inhibitor ATA also partially inhibited large DNA fragment formation (Fig. 5A) and largely inhibited oligonucleosomal DNA fragmentation (Fig. 5B), consistent with previous observations. Interestingly, submillimolar concentrations of Zn 2ϩ totally abrogated formation of both the large and oligonucleosomal DNA fragments (Fig. 5, A and B), whereas at no dose did Zn 2ϩ inhibit lamin B 1 cleavage (Fig. 5C). Together, these observations suggest that the effects of ATA on large fragment formation may be caused in part by inhibition of the Ca 2ϩ -stimulated NS protease, but the effects of Zn 2ϩ can be dissociated from NS protease inhibition and in this system may be linked solely to effects on nuclease activation.
Previous work has shown that calmodulin antagonists are capable of inhibiting DNA fragmentation in thymocytes exposed to apoptotic triggers (29,39), and it has been suggested that the NS protease may be regulated by calmodulin (40). We therefore tested the effects of the calmodulin antagonist calmidizolium on DNA fragmentation and lamin B 1 cleavage in isolated nuclei. Calmidizolium blocked Ca 2ϩ -induced oligonucleosomal DNA fragmentation in a dose-dependent fashion (Fig. 6A); however, it failed to prevent cleavage of lamin B 1 (Fig. 6B) or the accumulation of 30 -50-kb DNA fragments (Fig. 6C). Members of the BCL-2 family play evolutionarily conserved roles in regulating apoptotic cell death, although the points at which they act within the pathway remain unclear, as effects of BCL-2 upstream and downstream (33,34) of ICE family protease activation have been suggested. Recent efforts from our laboratory and others have demonstrated that a significant fraction of total cellular BCL-2 protein is localized to the nuclear envelope (28,41,42), and we have demonstrated that overexpression of BCL-2 can block Ca 2ϩ -stimulated oligonucleosomal DNA fragmentation in isolated nuclei from a Rastransformed murine fibroblast line (28). To investigate further the effects of BCL-2 on the nuclear events of apoptosis, we transfected the T cell hybridoma 2B4.11, an extensively characterized cell line that responds to apoptotic triggers in fashion similar to thymocytes (43), with human BCL-2. Our choice of this line was reinforced by a recent report confirming that isolated nuclei from the parental line contain up to three polypeptides (49,47, and 45 kDa) termed inducible Ca 2ϩ /Mg 2ϩ lymphocyte endonuclease, which exhibit Ca 2ϩ -dependent nuclease activity inhibitable by Zn 2ϩ (44). Expression of BCL-2 in whole cells was confirmed by immunoblotting (not shown), and preliminary experiments confirmed that the cells were resistant to induction of apoptosis in response to glucocorticoid or thapsigargin (Fig. 7, A and B). Consistent with previous studies, isolated nuclei from these cells contained substantial levels of human BCL-2 (Fig. 7C). Strikingly, incubation of nuclei from the BCL-2 transfectants in the presence of Ca 2ϩ failed to stimulate lamin cleavage (Fig. 7D) or DNA fragmentation (Fig. 7E) for up to 2 h, although both responses were observed after longer times of incubation (not shown). To test whether the effects of BCL-2 were cell type-specific, we conducted analogous experiments with isolated nuclei from pooled BCL-2 transfectants of a metastatic murine melanoma line (K1735 clone M2X21) (45). Again, Ca 2ϩ -mediated cleavage of lamin B 1 and DNA fragmentation were dramatically reduced in nuclei from the BCL-2 transfectants compared with the levels observed in nuclei from the parental cells (Fig. 7, D and E). Thus, suppression of Ca 2ϩ -dependent protease activation appears to represent one important aspect of BCL-2 function in the nucleus that may explain, at least in part, its effects on the Ca 2ϩ -sensitive pathway of endonuclease activation. DISCUSSION Previous work has shown that lamin degradation is associated with apoptosis in several different model systems. In this study, we investigated the nature of the proteases responsible for this event in whole thymocytes exposed to several different apoptotic stimuli and in isolated nuclei incubated with exogenous Ca 2ϩ . The results indicate that specific peptide-based inhibitors of the ICE family (YVADfmk) and of the NS protease (APFcmk) blocked lamin cleavage, DNA fragmentation, and cell death in whole cells following T cell receptor triggering or treatment with glucocorticoid or the intracellular Ca 2ϩ -mobilizing agent thapsigargin. Because the NS protease inhibitor was significantly more potent than the ICE antagonist in blocking both lamin cleavage and DNA fragmentation, it is possible that APFcmk affects a process downstream of the ICE family in the pathway of apoptosis in thymocytes. Consistent with this notion, of a panel of inhibitors specific for several different proteases which were implicated previously in apoptosis, only fragmentation was quantified using the diphenylamine reagent. Mean Ϯ S.D., n ϭ 3. Panel C, dose-dependent effects of Zn 2ϩ on lamin cleavage. Isolated nuclei were incubated for 1 h with or without 5 mM Ca 2ϩ in the absence or presence of the indicated concentrations of Zn 2ϩ , and proteolysis of lamin B 1 was measured by immunoblotting. the serine protease inhibitor TPCK and the NS protease-specific peptide APFcmk blocked lamin degradation and DNA fragmentation in isolated thymocyte nuclei that were incubated with Ca 2ϩ . Because both compounds possess a phenylalanine residue proximal to the chloromethyl ketone derivative (the position that determines inhibitor selectivity), it appears likely that both compounds affect the same protease in nuclei. Identifying this protease will be the subject of our future investigation.
Our results confirm other work indicating that a TPCKsensitive protease is required for Ca 2ϩ -mediated endonuclease activation in whole cells and isolated nuclei (36,46). Although the nature of this interaction between proteases and endonucleases in the apoptotic pathway is still not known, the present results support a hypothetical scenario (18) in which Ca 2ϩ -dependent activation of the NS protease promotes the degradation of the chromatin-associated polypeptides lamin B 1 and histone H1, resulting in the sequential exposure of the lamin binding sites and internucleosomal regions to a Ca 2ϩ -dependent endonuclease. Because the lamin binding sites are spaced every 20 -50 kb and the histone H1 attachment sites every 180 bases in chromatin, this model would account for the stepwise pattern of DNA cleavage observed in apoptotic thymocytes and other cell types. From these data it might be postulated that chromatin cleavage is mediated by a constitutively active endonuclease whose action is restricted solely by DNA accessibility. However, this idea does not appear to be correct, as Zn 2ϩ preferentially blocks the production of both large and oligonucleosomal DNA fragments without interfering detectably with proteolysis of lamin B 1 . Conversely, Lazebnik and colleagues (15) have shown that it is possible to inhibit lamin degradation without blocking oligonucleosomal DNA fragmentation with ICE antagonists in nuclei incubated with cytoplasmic extracts from cells arrested at the S and M phases of the cell cycle (S/M extracts). Moreover, our results support the idea suggested by Sun and Cohen (20) that independent endonuclease activities are responsible for the 50-kb and oligonucleosomal DNA fragments, as we have found that formation of the latter but not the former is blocked by calmodulin antagonists in isolated nuclei. In addition to removing chromatinbound proteins, it is possible that proteolytic processing directly affects endonuclease activation, an idea that is supported by preliminary work suggesting that two candidate Ca 2ϩ -dependent thymocyte endonucleases appear to exist as high molecular weight precursors in untreated cells (47,48), and it may account for the appearance of the 47-and 45-kDa inducible Ca 2ϩ /Mg 2ϩ lymphocyte endonuclease polypeptides in isolated nuclei from 2B4.11 cells exposed to glucocorticoid or anti-T cell receptor antibodies prior to the onset of apoptosis (44).
The nuclear localization, Ca 2ϩ -dependence, and peptide inhibitor sensitivity of the protease implicated in lamin cleavage and DNA fragmentation in this study are all specific characteristics of the NS protease. Although most of the molecular characteristics of this protease are still not known, preliminary work indicates that it is structurally related to the multicatalytic proteinase complex, otherwise known as the proteosome (16). The observation that elevations in the cytosolic Ca 2ϩ concentration trigger chromatin condensation and nuclear envelope breakdown during mitosis (49,50) coupled with the ability of cytosolic Ca 2ϩ increases to activate the NS protease directly (16) raises the possibility that the NS protease is not only involved in certain examples of apoptosis but also in some aspects of mitosis. Circumstantial support for this hypothesis is provided by the observation that disregulated proliferation in hyperplastic nodules and hepatocellular carcinomas is associated with elevated NS protease activity (51,52), and a pep- tide inhibitor of the NS protease blocks chemical transformation in a murine fibroblast line (53). Coordinate regulation of catabolic enzymes in apoptosis and mitosis is an attractive possibility given the observed similarities in cellular morphology as well as the known capacity of oncogenes (most notably Myc) (54) and cyclin-dependent kinases (55) to promote both responses. It is also possible that the proteosome itself may play some role in the cytoplasmic events of apoptosis, as increases in ubiquitin expression (56) and in enhanced ubiquitination of cellular proteins (57) are linked to programmed cell death in the moth Manduca sexta and in irradiated human peripheral blood lymphocytes, respectively.
Recently, a number of investigators have used cell-free systems in an attempt to identify the biochemical machinery necessary for the nuclear features of apoptosis. These studies were initiated by Lazebnik and co-workers (14), who demonstrated that S/M extracts were capable of inducing characteristic chromatin condensation, nuclear lamina disruption, and DNA fragmentation in isolated nuclei from control cells. Importantly, lamin degradation and endonuclease activation in nuclei incubated with S/M extracts (and other cell-free systems) occurred in the absence of exogenous Ca 2ϩ (14), and lamin degradation and DNA fragmentation were insensitive to TPCK and the peptide inhibitor of the NS protease (15). Together with the results reported here, these data indicate that there are at least two pathways to lamin degradation and DNA fragmentation in isolated nuclei, one mediated by activated ICE family members and the other by Ca 2ϩ -induced activation of the NS protease. It is still formally possible that the ICE pathway activates one of the Ca 2ϩ -independent enzymatic activities of the NS protease, an issue that requires further investigation.
Despite intensive investigation, the mechanisms underlying the inhibition of apoptosis by BCL-2 are still not clear. In this study, we demonstrated that the amount of BCL-2 constitutively present in isolated nuclei from two divergent cell lines transfected with BCL-2 is sufficient to block Ca 2ϩ -mediated lamin degradation and DNA fragmentation. Precisely how it does so is not known, although we have recently obtained FIG. 7. Effects of overexpression of BCL-2 in two different transfectants. Panel A, effects of glucocorticoid and thapsigargin on the viability of 2B4.11 (white bars) and 2B4/BCL2.1 (black bars) cells. Cells were incubated for 16 h in the absence or presence of 100 nM thapsigargin or 10 M methylprednisolone (MPS), and cell viability was measured by trypan blue exclusion. Mean Ϯ S.D., n ϭ 3. Panel B, inhibition of DNA fragmentation by BCL-2 in whole cells. Cells were incubated in the absence or presence of 100 nM thapsigargin for 8 h, and DNA fragmentation was analyzed by agarose gel electrophoresis. Results are characteristic of three independent experiments. Panel C, expression of BCL-2 in isolated nuclei detected by immunoblotting. Panel D, inhibition of Ca 2ϩ -stimulated lamin cleavage by BCL-2. Isolated nuclei were incubated with or without 5 mM Ca 2ϩ for 1 h, and the p21 fragment of lamin B 1 was detected by immunoblotting. Panel E, inhibition of Ca 2ϩ -stimulated DNA fragmentation by BCL-2. Isolated nuclei from 2B4.11 (white bars) or 2B4/BCL2.1 cells (black bars) were incubated for 2 h in the absence or presence of 5 mM Ca 2ϩ , and DNA fragmentation was quantified using the diphenylamine reagent. Mean Ϯ S.D., n ϭ 3. evidence that BCL-2 may exert its effects, at least in part, via its action on an energy-dependent Ca 2ϩ uptake system localized to the nuclear envelope (42). Because the NS protease is apparently localized to the nuclear envelope as well, and we presented evidence previously that nuclear Ca 2ϩ uptake promotes endonuclease activation (27), this activity of BCL-2 may account for the inhibition of protease and nuclease activation observed. Alternatively, because recent work demonstrated that the nuclear Ca 2ϩ pool regulates transport of large (Ͼ10 kDa) polypeptides through nuclear pores (58), it is also possible that BCL-2 exerts its effects on nuclei by blocking translocation of certain cytosolic proteins (i.e. p53, Myc, ICE homologs, DNase I) that may cooperate with the NS protease to promote lamin/histone cleavage and DNA fragmentation. These possibilities are not mutually exclusive, and both are under investigation at present.
An important question raised by this study concerns why specific inhibitors of either the ICE family or the NS protease are capable of inhibiting lamin B 1 degradation and DNA fragmentation in whole thymocytes, whereas only the latter prevent these changes in isolated nuclei supplemented with Ca 2ϩ . It is possible that the ICE family is involved in promoting the Ca 2ϩ alterations that trigger nuclear NS protease activation or that the pathways of apoptosis triggered in thymocytes by glucocorticoids, thapsigargin, and anti-T cell receptor antibodies in thymocytes are actually initiated within the nucleus. The latter would have important implications for current thinking concerning the location of the apoptotic effector machinery in general, as the results obtained with cytoplasmic apoptosispromoting extracts and enucleated cells (59,60) imply that all of the important biochemical alterations tied to commitment to cell death occur in the cytoplasm. Both of these explanations for the likely interrelationship between the ICE family and the NS protease are testable hypotheses that are under investigation at present.