HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 41, 38185-38192, October 12, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/41/38185    most recent M100629200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Boulares, A. H. Articles by Smulson, M. E. Search for Related Content PubMed PubMed Citation Articles by Boulares, A. H. Articles by Smulson, M. E. Social Bookmarking                      What's this?

Roles of DNA Fragmentation Factor and Poly(ADP-ribose) Polymerase in an Amplification Phase of Tumor Necrosis Factor-induced Apoptosis^*

A. Hamid Boulares, Anna J. Zoltoski, Alexander Yakovlev^§, Ming Xu^¶, and Mark E. Smulson

From the  Department of Biochemistry and Molecular Biology and ^§ Department of Neuroscience, Georgetown University School of Medicine, Washington, D. C. 20007 and the ^¶ Department of Cell Biology, Neurobiology, and Anatomy, University of Cincinnati College of Medicine, Cincinnati, Ohio 45267

Received for publication, January 23, 2001, and in revised form, June 19, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

During apoptosis, endonucleases cleave DNA into 50-300-kb fragments and subsequently into internucleosomal fragments. DNA fragmentation factor (DFF) is implicated in apoptotic DNA cleavage; this factor comprises DFF45 and DFF40 subunits, the former of which acts as a chaperone and inhibitor of the catalytic subunit and whose cleavage by caspase-3 results in DFF activation. Disruption of the DFF45 gene blocks internucleosomal DNA fragmentation and confers resistance to apoptosis in primary thymocytes. The role of DFF-mediated DNA fragmentation in apoptosis was investigated in primary fibroblasts from DFF45^/ and control (DFF45^+/+) mice. DFF45 deficiency rendered fibroblasts resistant to apoptosis induced by tumor necrosis factor (TNF). TNF induced rapid cleavage of DNA into ~50-kb fragments in DFF45^+/+ fibroblasts but not in DFF45^/ cells, indicating that DFF mediates this initial step in DNA processing. The TNF-induced activation of poly(ADP-ribose) polymerase (PARP), which requires PARP binding to DNA strand breaks, and the consequent depletion of the PARP substrate NAD were markedly delayed in DFF45^/ cells, suggesting a role for DFF in PARP activation. The activation of caspase-3 and mitochondrial events important in apoptotic signaling, including the loss of mitochondrial membrane potential and the release of cytochrome c, induced by TNF were similarly delayed in DFF45^/ fibroblasts. DFF45^/ and DFF45^+/+ cells were equally sensitive to the DNA-damaging agent and PARP activator N-methyl-N'-nitro-N-nitrosoguanidine. Inhibition of PARP by 3-aminobenzamide partially protected DFF45^+/+ cells against TNF-induced death and inhibited the associated release of cytochrome c and activation of caspase-3. These results suggest that the generation of 50-kb DNA fragments by DFF, together with the activation of PARP, mitochondrial dysfunction, and caspase-3 activation, contributes to an amplification loop in the death process.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Although apoptosis is a conserved and highly regulated mechanism of cell death, its inducers are diverse. They include proteins of the tumor necrosis factor (TNF)¹ family, genotoxic and cytotoxic agents such as anticancer drugs and -radiation, and antiinflammatory drugs such as aspirin and menadione (1-3). The intracellular mediators of apoptosis have been relatively well characterized (4). Regardless of the nature of the stimulus, the commitment to apoptosis occurs as a result of the activation of members of the caspase family of cysteine proteases (5). These enzymes are present in cells as inactive precursors that are proteolytically activated on induction of apoptosis. At least 14 members of the caspase family have been identified to date, all of which require an aspartic acid residue at the cleavage sites of their substrates (5, 6). These proteases can be divided into two major groups: initiator and executor caspases. The initiator caspases are responsible for converting the external death signal into an internal signal with the participation of death receptor-associated molecules and mitochondria. Once activated, the initiator caspases trigger the activation of the executor caspases, such as caspase-3, -6, and -7, which, in turn, results in the destruction of cellular components required for maintenance of structural and functional integrity. This destruction of cellular components leads to the phenotypic changes characteristic of apoptosis, including membrane blebbing, chromatin condensation, and internucleosomal DNA fragmentation.

DNA fragmentation is thought to be an important step in the disposal of genomic DNA from dying cells. Concomitant with cleavage of nuclear lamin, DNA is degraded into large fragments of 50-300 kb, which are then processed into internucleosomal repeats of ~200 base pairs that are responsible for generation of the characteristic ladder pattern apparent on electrophoresis (7). The initial cleavage of DNA into ~50-kb fragments is thought to be required for subsequent DNA fragmentation (7, 8) and may constitute an important step in the escalation of events leading to cell death. Several endonucleases have been implicated in the DNA fragmentation process (4, 9). For example, we recently described the cloning and expression of a human Ca²⁺- and Mg²⁺-dependent endonuclease, termed DNAS1L3 and its regulation by poly(ADP-ribosyl)ation (10, 11). DNA fragmentation factor (DFF), also known as caspase-activated DNase (CAD) or caspase-activated nuclease (CPAN), has also been suggested to contribute to this process (4, 12-14). DFF is composed of two subunits of 40 and 45 kDa termed DFF40 (CAD) and DFF45 (ICAD), respectively (12-14). The endonuclease activity of this enzyme, which is intrinsic to DFF40, is induced on cleavage of DFF45 by caspase-3. DFF45 functions as both an inhibitor of DFF40 activity and as a chaperone for this subunit, given that it is required for DFF40 expression (13-15). Both DFF45 and caspase-3 are therefore required for DNA fragmentation by DFF40.

Both DFF45-deficient (DFF45^/) mice and thymocytes isolated from these animals exhibit resistance to apoptosis induced by a variety of stimuli (16). The lack of DFF45 is associated with a reduced extent of chromatin condensation and the absence of internucleosomal DNA fragmentation in thymocytes (16). DFF has been implicated in the fragmentation of DNA into 50-kb pieces during apoptosis in part on the basis of the observation that expression of a caspase-3-resistant DFF45 mutant prevented the generation of such DNA fragments in Jurkat cells exposed to staurosporine (17). More recently, a role for DFF in the processing of genomic DNA into 50-kb fragments was further demonstrated using thymocytes derived from DFF45^/ mice (18).

The introduction of breaks into DNA molecules, including those generated by the production of 50-kb and internucleosomal DNA fragments, results in the activation of poly(ADP-ribose) polymerase (PARP), which has an absolute requirement for binding to the ends of DNA strands for activity. As a result of the poly(ADP-ribosyl)ation of nuclear proteins and automodification of PARP itself, the activation of this enzyme results in a marked decrease in the intracellular concentrations of NAD and ATP, creating an energy crisis that is thought to contribute to acceleration of the death process (19). We have now investigated the roles of DNA fragmentation into 50-kb pieces and PARP activation in the death process by examining the relation of these events to the resistance of DFF45^/cells to apoptosis. We show that DFF45^/ fibroblasts lack the ability to generate 50-kb DNA fragments in response to TNF. The absence of such DNA fragmentation protected the cells from excessive activation of PARP and thereby prevented depletion of intracellular NAD. The delay apparent in PARP activation correlated with delays both in caspase-3 activation and in proapoptotic mitochondrial events, including loss of mitochondrial membrane potential and the release of cytochrome c. We propose that the generation of 50-kb DNA fragments by DFF is not merely a passive step in DNA degradation but rather, together with PARP, mitochondria, and caspase-3, contributes to an amplification phase of apoptosis.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Isolation of Thymocytes and Fibroblasts, Cell Culture, Transfection, and Induction of Apoptosis-- Mice homozygous for DFF45 gene disruption were kindly provided by M. Xu (University of Cincinnati) through A. Faden (Georgetown University). For isolation of thymocytes, thymi from adult DFF45^/ and control (wild-type) mice were removed aseptically, washed with phosphate-buffered saline, and then passed through a wire screen to produce a single-cell suspension. The thymocytes were separated by centrifugation at 300 × g for 10 min and resuspended in RPMI 1640 medium supplemented with 10% fetal bovine serum and a mixture of penicillin and streptomycin. The cells were cultured in the same medium and immediately used for experiments. They were treated with TNF (10 ng/ml) in RPMI 1640 with serum. Primary fibroblasts were prepared from adult control or DFF45^/ mice according to standard protocols. They were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and penicillin-streptomycin and were used for experiments after several passages. PARP^/ and PARP^+/+ fibroblasts, kindly provided by Dr. Z. Q. Wang by way of Dr. S. Snyder, were maintained as described for DFF45^/ and DFF45^+/+ fibroblasts. Fibroblasts were treated with a combination of TNF (10 ng/ml) and cycloheximide (1 µg/ml) or, as indicated, in Dulbecco's modified Eagle's medium with serum. Transient transfection of a luciferase reporter plasmid under the control of a promoter containing several repeats of the consensus sequence for NF-B DNA binding, kindly provided by Dr. Giardina (University of Connecticut), was performed using MIRUS TransIT-LT1 (Panvera, Madison, WI) according to the protocol recommended by the manufacturer. Luciferase assays were performed using the standard luciferase assay system (Promega, Madison, WI) and normalized using pSV2-CAT plasmid.

Measurement of Cell Viability-- After treatment with TNF, cells cultured in 24-well plates were washed with Locke's solution (154 mM NaCl, 5.6 mM KCl, 3.6 mM NaHCO₃, 2.3 mM CaCl₂, 1.2 mM MgCl₂, 5.6 mM glucose, 5 mM Hepes-NaOH (pH 7.4)) and then incubated for 30 min at 37 °C with the same solution containing 2.5 µM calcein-AM (Molecular Probes, Inc., Eugene, OR). Fluorescence resulting from the deesterification of calcein-AM was monitored with a CytoFluor 4000 fluorometer (PerSeptive Biosystems, Framingham, MA) at excitation and emission wavelengths of 488 and 520 nm, respectively.

Analysis of DNA Fragmentation-- DNA was isolated from cells as previously described (20) and subjected to electrophoresis through 1.5% agarose gels in Tris borate-EDTA buffer. The gels were stained with ethidium bromide. Transverse alternating field electrophoresis was performed essentially as described (20).

Immunoblot Analysis-- Cells were washed with ice-cold phosphate-buffered saline and then lysed as described (20). A portion (30 µg of protein) of each lysate was fractionated by SDS-polyacrylamide gel electrophoresis on a 4-20% gradient gel, and the separated proteins were transferred to a nitrocellulose filter. The filter was stained with Ponceau S to confirm equal loading and transfer of samples and was then probed with antibodies to DFF45 (Biovision, Palo Alto, CA), to DFF40 (kindly provided by X. Wang, University of Texas), to poly(ADP-ribose) (10H-A) (20), or to PARP (Pharmingen, San Diego, CA). Immune complexes were detected with appropriate secondary antibodies and chemiluminescence reagents (Pierce).

Nuclear Extract Preparation and Electrophoretic Mobility Shift Assay (EMSA)-- Nuclear extracts were prepared, and EMSA analysis of DNA binding activity was performed as described previously (21). The NF-B binding oligonucleotide (Promega, Madison, WI) was end-labeled with T4 polynucleotide kinase using [-³²P]ATP (PerkinElmer Life Sciences).

Assay of Caspase-3 Activity-- Caspase-3 activity was measured essentially as described (20). In brief, cell extracts (25 µg of protein) were incubated for 30 min at 37 °C with 40 µM DEVD-aminomethylcoumarin (AMC) peptide substrate in a total volume of 200 µl. The fluorescence of free AMC, generated as a result of cleavage of the aspartate-AMC bond, was monitored continuously over 30 min with a CytoFluor 4000 fluorometer at excitation and emission wavelengths of 360 and 460 nm, respectively. The emission from each well was plotted against time, and linear regression analysis of the initial velocity (slope) for each curve yielded the activity.

Assay of NAD-- After treatment with TNF, cells cultured in six-well dishes at a density of 1.0 × 10⁶ cells/well were scraped into the culture medium, washed with ice-cold phosphate-buffered saline, and then extracted with 0.5 ml of ice-cold 0.5 M perchloric acid. After the addition of 0.47 ml of a solution prepared from 1 M KOH, 0.33 M K₂HPO₄, and 0.33 M KH₂PO₄ to adjust the pH of the extract to 7.3, cell debris and salt precipitates were removed by centrifugation. The amount of NAD in the resulting supernatant was determined with an enzymatic cycling assay as described (22).

Assay of Cytochrome c Release-- Cells were collected, washed with phosphate-buffered saline, suspended in 5 volumes of a hypotonic buffer (20 mM Hepes-KOH (pH 7.5), 10 mM KCl, 1.5 mM MgCl₂, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.1 mM 4-(aminoethyl)benzenesulfonyl fluoride, leupeptin (20 µg/ml), aprotinin (10 µg/ml), 250 mM sucrose), and incubated for 15 min on ice. They were then homogenized by 15-20 passages through a 22-gauge needle. The homogenate was centrifuged at 1000 × g for 15 min at 4 °C, and the resulting supernatant was then centrifuged at 12,000 × g for 15 min at 4 °C. The second (mitochondrial) pellet was suspended in the hypotonic buffer and then subjected to sonication with three 10-s pulses. Samples (15 µg of protein) of the mitochondrial and cytosolic (post-12,000 × g supernatant) fractions were subjected to immunoblot analysis with antibodies to cytochrome c (RDI, Flanders, NJ).

Measurement of Mitochondrial Membrane Potential (_mito)-- Thirty minutes before collection by exposure to trypsin, cells were stained with 5,5',6,6'-tetrachloro-1,1'3,3'-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) (Molecular Probes, Inc., Eugene, OR), a cell-permeable dye that aggregates and generates red fluorescence in mitochondria with a high _mito. This aggregation and consequent fluorescence of JC-1 dissipates as _mito is lost. Fluorescence was analyzed with a Becton Dickinson fluorescence-activated cell sorting flow cytometer.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Effects of DFF45 Deficiency on TNF-induced Internucleosomal DNA Fragmentation and Cell Death-- Disruption of the DFF45 gene has recently been shown to block dexamethasone-induced internucleosomal DNA fragmentation and cell death in isolated thymocytes as well as in vivo (16). We examined the effect of the apoptosis inducer TNF on internucleosomal DNA fragmentation in primary thymocytes isolated from DFF45^/ mice. Thymocytes isolated from control or DFF45^/ mice were exposed to TNF (10 ng/ml) for 12 h, after which DNA was extracted and analyzed by agarose gel electrophoresis. Whereas TNF induced marked internucleosomal DNA fragmentation in control thymocytes, it had no such effect in thymocytes from DFF45^/ mice (Fig. 1A), suggesting that DFF45 is required for the processing of genomic DNA in response to apoptotic stimuli. Similar results were obtained when primary thymocytes or splenocytes from DFF45^/ mice were treated with the apoptosis inducers etoposide or staurosporine (data not shown) (23), indicating that the requirement for DFF45 is not stimulus- or cell type-specific.

View larger version (47K): [in this window] [in a new window]   Fig. 1.   Effects of DFF45 deficiency on TNF-induced internucleosomal DNA fragmentation and cell death. A, effect of DFF45 deficiency on TNF-induced internucleosomal DNA fragmentation in primary thymocytes. Thymocytes isolated from DFF45/ and DFF45+/+ mice were incubated for 12 h in the absence (Con) or presence of TNF (10 ng/ml), after which total DNA was extracted and analyzed by agarose gel electrophoresis and ethidium bromide staining. B, expression of DFF45 and DFF40 in DFF45/ and DFF45+/+ fibroblasts. Fibroblasts and U937 cells were lysed and subjected to immunoblot analysis with antibodies to DFF45 (top panel) or to DFF40 (bottom panel). C, effect of DFF45 deficiency on TNF-induced cell death in fibroblasts. DFF45/ (closed circles) and DFF45+/+ (open circles) fibroblasts were incubated for the indicated times in the presence of TNF (10 ng/ml) and cycloheximide (1 µg/ml), after which cell viability was assessed by measurement of calcein fluorescence. Data are expressed as a percentage of the viability of untreated cells and are means ± S.D. of values from four wells from a representative experiment. D, effect of DFF45 deficiency on TNF-induced NF-B in fibroblasts. DFF45/ and DFF45+/+ fibroblasts were incubated for the 30 min in the presence of TNF (10 ng/ml) and cycloheximide (1 µg/ml), after which nuclear extracts were prepared, and their NF-B DNA binding activity was determined by EMSA.

To investigate further the role of DFF-mediated DNA fragmentation in apoptosis, we isolated fibroblasts from DFF45^/ and control mice. We confirmed by immunoblot analysis that the primary fibroblasts from DFF45^/ mice lacked DFF45, whereas the DFF45^+/+ fibroblasts expressed this protein at a level similar to that apparent in the human monocytic cell line, U937 (Fig. 1B). The abundance of DFF40 in DFF45^/ cells was also greatly reduced compared with that in DFF45^+/+ cells, consistent with the chaperone function of DFF45 in DFF40 expression (14, 15). These results are also consistent with the abundance of these proteins in the brain of DFF45^/ and control mice (23). We then compared the effects of the combination of TNF (10 ng/ml) and cycloheximide (1 µg/ml) on the viability of DFF45^/ and control fibroblasts. The DFF45^/ fibroblasts exhibited a marked resistance to TNF-induced apoptosis compared with the sensitivity of control cells (Fig. 1C). To exclude the possibility that DFF45^/ cells are resistant to TNF-induced cell death as a result of a dysfunctional TNF receptor, we examined the ability of this cytokine to activate the transcription factor NF-B, a known target of TNF signaling (24, 25). DFF45^/ and control cells were treated with TNF for 30 min, after which nuclear extracts were prepared and assayed for NF-B DNA binding activity by EMSA with an oligonucleotide probe containing a specific NF-B binding site. No difference was apparent between DFF45^/ and control fibroblasts in their ability to activate NF-B (Fig. 1D), indicating that TNF receptor is equally functional in both cell types. To further test TNF signaling, DFF45^/ and DFF45^+/+ cells were transiently transfected with a reporter plasmid encoding for the luciferase gene under the control of a promoter containing several repeats of the consensus sequence for NF-B DNA binding. Transfected cells were then treated with TNF in the absence of cycloheximide for 6 h, after which cell extracts were prepared and assayed for luciferase activity. No significant difference in luciferase activity was observed in TNF-treated DFF45^/ (2.3-fold) and DFF45^+/+ (2-fold) cells when compared with their respective controls. Additionally, no difference was observed in the expression of the TNF receptor protein as assessed by immunoblot analysis using antibodies against mouse TNF-R1 (data not shown).

Effect of DFF45 Deficiency on the TNF-induced Generation of 50-kb DNA Fragments-- The fragmentation of DNA into 50-kb pieces is thought to precede internucleosomal DNA degradation. DFF40 has been implicated in the cleavage of DNA into 50-kb DNA fragments in response to various apoptotic stimuli (17, 18). We therefore examined the effect of DFF45 deficiency on the generation of such DNA fragments in response to TNF. DFF45^/ and control fibroblasts were incubated with TNF and cycloheximide for various times, after which DNA was isolated and subjected to transverse alternating field electrophoresis. Whereas TNF induced the generation of 50-kb DNA fragments within 3 h in control cells, such DNA fragmentation was not detected in DFF45^/ fibroblasts even after 12 h (Fig. 2A). These results suggested that the resistance of DFF45^/ cells to apoptosis might be due to a lack of cleavage of DNA into 50-kb fragments. Excessive DNA damage that is beyond repair is thought to constitute a signal for triggering of apoptosis (19, 26).

View larger version (46K): [in this window] [in a new window]   Fig. 2.   Effect of DFF45 deficiency on DNA fragmentation in TNF-treated fibroblasts. DFF45/ and DFF45+/+ fibroblasts were incubated with TNF and cycloheximide for the indicated times, after which DNA was isolated and subjected either to transverse alternating field electrophoresis (A) or to agarose gel electrophoresis (B).

Fibroblasts from a variety of species have been shown to lack the ability to undergo internucleosomal DNA fragmentation (4, 11, 14). Consistent with these previous observations, primary DFF45^/ and DFF45^+/+ fibroblasts did not exhibit internucleosomal DNA fragmentation in response to TNF (Fig. 2B). The lack of internucleosomal DNA fragmentation in these cells provided us with a good model with which to examine the role of the cleavage of DNA into 50-kb fragments in TNF-induced apoptosis.

Effects of DFF45 Deficiency on TNF-induced Poly(ADP-ribosyl)ation and NAD Depletion-- Excessive DNA damage induces PARP activation, as manifested by poly(ADP-ribosyl)ation of nuclear proteins, including PARP itself, and a consequent depletion of NAD and of ATP (27), from which NAD is derived by nuclear NAD synthetase. We therefore examined the effect of the lack of cleavage of DNA into 50-kb fragments in DFF45^/ fibroblasts on TNF-induced PARP activation and NAD depletion. Immunoblot analysis of cell extracts with antibodies to poly(ADP-ribose) (PAR) revealed that TNF induced marked poly(ADP-ribosyl)ation of PARP (the predominant target for this modification under these conditions) in control cells that was maximal after 3 h (Fig. 3A), a time at which cleavage of DNA into 50-kb fragments was apparent (Fig. 2A). The extent of poly(ADP-ribosyl)ation of PARP had decreased by 6 h and was no longer apparent after 12 h. In contrast, poly(ADP-ribosyl)ation of PARP was not detected in DFF45^/ fibroblasts until 12 h after the initial exposure to TNF (Fig. 3A).

View larger version (37K): [in this window] [in a new window]   Fig. 3.   Effects of DFF45 deficiency on PARP activation and cleavage and on NAD depletion induced by TNF. DFF45/ and DFF45+/+ fibroblasts were incubated with TNF and cycloheximide for the indicated times, after which cell lysates were subjected to immunoblot analysis with antibodies to PAR (A) or to PARP (B); the bottom panel in B represents a longer exposure of the filter shown in the top panel. The positions of full-length (116 kDa) PARP and of a proteolytic fragment (89 kDa) are indicated. Alternatively, after treatment of DFF45/ (closed circles) and DFF45+/+ (open circles) fibroblasts with TNF and cycloheximide, cell extracts were assayed for NAD (C); data are expressed as a percentage of the value for untreated cells and are means ± S.D. of triplicates from a representative experiment.

To determine whether DFF45 deficiency affects the expression of PARP, we subjected the same cell lysates to immunoblot analysis with antibodies to PARP. The abundance of PARP was identical in both untreated cell types. Whereas the amount of PARP remained unchanged after exposure of DFF45^/ fibroblasts to TNF for 6 h, PARP was proteolytically processed to an 89-kDa polypeptide in control cells as early as 1 h after exposure to TNF, and this effect was more pronounced at 6 h (Fig. 3B). These results thus suggested that PARP cleavage might be related to the apparent decrease in the poly(ADP-ribosyl)ation of this protein in DFF45^+/+ fibroblasts 6 h after exposure to TNF.

We also examined the effect of TNF on the intracellular concentration of NAD in DFF45^/ and control fibroblasts. Whereas the intracellular abundance of NAD in DFF45^+/+ cells decreased on exposure to TNF in a time-dependent manner, the amount of NAD in DFF45^/ cells was largely unaffected by treatment with this cytokine (Fig. 3C). These results are thus consistent with the kinetics of the generation of 50-kb DNA fragments, of poly(ADP-ribosyl)ation, and of the loss of cell viability in the two cell types. They suggest that DFF45 deficiency protects cells, at least temporarily, from apoptosis by preventing the depletion of intracellular NAD, an effect that is achieved by blocking DNA fragmentation and the consequent activation of PARP.

Effect of DFF45 Deficiency on TNF-induced Activation of Caspase-3-- Given that PARP was cleaved early during incubation of control fibroblasts with TNF and that cleavage of this enzyme during apoptosis is thought to be mediated by caspase-3 (13, 15, 28), we measured caspase-3-like activity in lysates of control and DFF45^/ fibroblasts at various times after incubation with TNF and cycloheximide. Caspase-3-like activity was maximal at 6 h after exposure of DFF45^+/+ cells to TNF (Fig. 4A), a time at which the extent of PARP cleavage was marked (Fig. 3B). In contrast, DFF45^/ cells did not exhibit a substantial increase in caspase-3-like activity until 12 h after exposure to TNF (Fig. 4A). These activity measurements were confirmed by immunoblot analysis of the processing of procaspase-3 into the active subunits of the mature enzyme (data not shown). We also examined the effects of the PARP inhibitor 3-aminobenzamide (3-AB) on the increase in caspase-3-like activity (Fig. 4B), the depletion of NAD (Fig. 4C), and the decrease in cell viability (Fig. 4D) induced by TNF in DFF45^+/+ cells. Cells were incubated in the absence or presence of 10 mM 3-AB for 1 h and then exposed to TNF for 6 h (in the continued absence or presence of 3-AB). The PARP inhibitor partially inhibited each of these effects of TNF. These results establish a relation between PARP activation and both the depletion of intracellular NAD and cell death induced by TNF. They also suggest that events downstream of PARP activation may potentiate the activation of caspase-3 in TNF-treated DFF45^+/+ fibroblasts.

View larger version (36K): [in this window] [in a new window]   Fig. 4.   Effects of DFF45 deficiency on caspase-3 activation and of the PARP inhibitor 3-AB on caspase-3 activation, NAD depletion, and cell death in TNF-treated fibroblasts. A, effect of DFF45 deficiency on caspase-3-like activity. DFF45/ and DFF45+/+ fibroblasts were incubated for the indicated times with TNF and cycloheximide, after which cell extracts were prepared and assayed for caspase-3-like activity with the specific substrate DEVD-AMC. Data are expressed in arbitrary units and are means of duplicates from a representative experiment. B-D, effects of 3-AB on caspase-3 activation, NAD depletion, and death induced by TNF in DFF45+/+ fibroblasts. Cells were incubated for 1 h with or without 10 mM 3-AB and then for an additional 6 h in the additional absence or presence of TNF and cycloheximide. Cell extracts were then assayed for caspase-3-like activity (B) and NAD (C). Data in D and E are expressed as a percentage of the viability of untreated cells and are means ± S.D. of values from four wells from a representative experiment. Data in B and C are means ± S.D. of triplicates from a representative experiment.

Effects of DFF45 or PARP Deficiency on Cell Death Induced by N-Methyl-N'-nitro-N-nitrosoguanidine (MNNG) or TNF, Respectively-- In an attempt to reverse the apoptosis-resistant phenotype of DFF45^/ fibroblasts by reintroducing the DFF45 gene using conventional lipid-based transfection techniques, we were unsuccessful at expressing sufficient DFF40 protein in these cells to induce 50-kb DNA breaks after TNF treatment (data not shown). Alternatively, to prove that resistance of DFF45^/ cells to apoptosis was not due to a deficiency in responding to excessive DNA damage, we compared the sensitivity of DFF45^/ cells with that of ^+/+ cells to MNNG, a strong DNA strand-breaking agent and a well known PARP activator as well as an inducer of apoptosis (29-31). Cells were treated with 50 µM MNNG for 12 h, after which cell viability was assessed by calcein-AM staining. MNNG treatment caused a similar significant loss in cell viability in both DFF45^/ and DFF45^+/+ cells (Fig. 5A). This result demonstrates that DFF45^/ cells are able to react to DNA damage and that their resistance to TNF-induced apoptosis may reside in their inability to generate 50-kb DNA breaks. To further confirm the role of PARP in the onset of apoptosis, we examined the effect of PARP deficiency on TNF-induced cell death in fibroblasts. PARP^/ and PARP^+/+ cells were treated with various concentrations of TNF and 1 µg/ml cycloheximide for 18 h after which cell viability was assessed by calcein-AM staining. At low concentrations of TNF (0.5-5 ng/ml), the PARP^/ fibroblasts exhibited a marked resistance to TNF-induced apoptosis compared with the sensitivity of control cells (Fig. 5B), indicating that, at nonsaturating concentrations of TNF, PARP plays a significant role in the onset of apoptosis. These results are consistent with those of Halappanavar et al. (32) showing that PARP deficiency conferred a significant resistance to apoptosis in fibroblasts when induced by 2 ng/ml TNF.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Effects of DFF45 or PARP deficiency on cell death induced by MNNG or TNF, respectively. A, effects of DFF45 deficiency on MNNG-induced cell death in fibroblasts. DFF45/ and DFF45+/+ fibroblasts were incubated for 12 h in the presence of 50 µM MNNG, after which cell viability was assessed. Effects of PARP deficiency on TNF-induced cell death in fibroblasts. PARP/ (closed circles) and PARP+/+ (open circles) fibroblasts were incubated for 18 h in the presence of 0.5, 1, 5, 10, or 20 ng/ml TNF and cycloheximide, after which cell viability was assessed. Data in A and B are means ± S.D. of triplicates from a representative experiment.

Effects of DFF45 Deficiency on _mito and Cytochrome c Release in TNF-treated Fibroblasts-- The depletion of cellular NAD and ATP by PARP in response to DNA damage results in mitochondrial stress as a consequence of an increased demand for ATP synthesis (29). Mitochondrial dysfunction, including loss of _mito, contributes to apoptotic cell death (33-35). The loss of _mito results from the opening of the mitochondrial permeability transition pore. The opening of this pore is also thought to lead to the release of several proapoptotic factors, including cytochrome c, and to the subsequent activation of caspase-9 and, subsequently, of caspase-3 (5, 33-35). We therefore next examined the effects of DFF45 deficiency on mitochondrial integrity in TNF-treated fibroblasts. The loss of _mito induced by TNF was greatly delayed in DFF45^/ fibroblasts compared with that apparent in DFF45^+/+ cells, which was almost complete by 6 h (Fig. 6A).

View larger version (40K): [in this window] [in a new window]   Fig. 6.   Effects of DFF45 deficiency on the loss of mito and the release of cytochrome c into the cytosol induced by TNF. A, effect of DFF45 deficiency on the TNF-induced loss of mito. DFF45/ and control fibroblasts were incubated for the indicated times in the presence of TNF and cycloheximide; 30 min before each time point, cells were stained with JC-1. The mito was then assessed by flow cytometry. B, effect of DFF45 deficiency on TNF-induced release of cytochrome c. DFF45/ and control cells were incubated for the indicated times with TNF and cycloheximide, after which cytosolic and mitochondrial fractions were prepared and subjected to immunoblot analysis with antibodies to cytochrome c (top panels). Filters were also stained with Ponceau S to verify equal application of protein among lanes (bottom panels).

We also examined the kinetics of cytochrome c release into the cytosol of both DFF45^/ and control cells in response to TNF treatment. The release of cytochrome c into the cytosol of DFF45^+/+ cells was apparent as early as 3 h after exposure to TNF and was marked by 6 h, whereas only a small extent of cytochrome c release was detected in DFF45^/ cells even after 6 h (Fig. 6B). The appearance of cytochrome c in the cytosol of TNF-treated control cells occurred concomitantly with a loss of this protein from mitochondria (Fig. 6B). The kinetics of cytochrome c release in these two cell types correlated with those of the loss of _mito.

Effect of PARP Inhibition on Cytochrome c Release during TNF-induced Apoptosis-- To investigate further the relation between PARP activation and the release of cytochrome c into the cytosol, we examined the effect of 3-AB on cytochrome c release in DFF45^+/+ fibroblasts. This agent partially inhibited the release of cytochrome c that was apparent after incubation with TNF for 6 h (Fig. 7). This result is consistent with the partial inhibition by 3-AB of the TNF-induced increases in both caspase-3-like activity (Fig. 4B) and cell death (Fig. 4D), and it supports the notion that PARP activation contributes to the mitochondrial changes that result in caspase-3 activation.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Effect of 3-AB on TNF-induced cytochrome c release in DFF45+/+ fibroblasts. Cells were incubated with or without 10 mM 3-AB for 1 h and then in the additional absence or presence of TNF and cycloheximide for 6 h. Cytosolic and mitochondrial fractions were then prepared and subjected to immunoblot analysis with antibodies to cytochrome c.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We have investigated the resistance of DFF45^/ cells to apoptosis and the role of cleavage of DNA into 50-kb fragments in the apoptotic process. We have shown that the absence of DFF45 prevents the generation of 50-kb DNA fragments in response to TNF in fibroblasts, a process that is thought to be required for internucleosomal DNA fragmentation (7). Our results also indicate that the production of 50-kb DNA fragments by DFF is not a passive event in the processing of genomic DNA but rather constitutes an active step in the death process, playing a crucial role in the activation of PARP and the subsequent impairment of mitochondrial function and activation of caspases. Furthermore, mitochondrial dysfunction may result from the depletion of energy reserves caused by excessive PARP activation. These events thus appear to constitute an amplification phase of apoptosis that exists in addition to the initiation and execution phases (Fig. 8).

View larger version (31K): [in this window] [in a new window]   Fig. 8.   Proposed amplification loop in TNF-induced apoptosis. Interaction of TNF with its receptor (TNF-R) induces an initial activation of caspase-3 that is mediated either by caspase-8 or through mitochondria. Caspase-3 then activates DFF40 by cleaving DFF45, which results in the generation of 50-kb DNA fragments and the consequent activation of PARP. PARP activation results in the depletion of NAD and ATP, which, in turn, leads to the loss of mito and the release of mitochondrial cytochrome c into the cytosol. These events then trigger a second phase of caspase-3 activation, resulting in extensive degradation of target proteins, including DFF45 and PARP, and, ultimately, in cell death.

Fibroblasts provided a useful model for our investigation into the physiological function of 50-kb DNA fragments because these cells do not exhibit internucleosomal DNA fragmentation. The reason for the failure of these cells to process their DNA into such 200-base pair repeats is not known. Disruption of the DFF45 gene also resulted in a marked reduction in the expression of DFF40. The small amount of DFF40 present in DFF45^/ cells is probably not functional, given that no DNA fragmentation was observed in these cells and that the activity of this protein requires its initial heterodimerization with DFF45 (15). We were, however, unsuccessful in reversing the down-regulation of DFF40 by attempting to reintroduce DFF45 into DFF45^/ cells despite several different transfection techniques used. Currently, we are unable to explain the lack of DFF40 expression in DFF45^/ cells transfected with a DFF45 expression vector. This in itself is interesting, but we believe that DFF40-DFF45 interaction is beyond the scope of this study. The cleavage of DNA into 50-kb fragments is thought to constitute major damage that cannot be repaired and consequently is considered a critical step in the commitment of a cell to apoptosis (7). The generation of such fragments requires DFF endonuclease activity given that deletion of the DFF45 gene and the consequent reduction in DFF40 expression blocked this process in TNF-treated fibroblasts. Given that caspase-3 mediates the activation of DFF (13, 36), the activation of this protease is required for the endonuclease activity that generates the 50-kb DNA fragments. Our results are consistent with those of Zhang et al. (18) showing that DFF45^/ thymocytes do not generate 50-kb DNA fragments in response to etoposide, dexamethasone, or staurosporine as well as with those of Sakahira et al. (17) showing that expression of a caspase-3-resistant mutant of DFF45 in Jurkat cells blocked the generation of 50-kb DNA fragments in response to staurosporine.

DNA strand breaks, including those produced by fragmentation of DNA into 50-kb pieces, are potent activators of PARP. Indeed, PARP activity is totally dependent on its binding to such strand breaks (37, 38). With NAD as its substrate, PARP catalyzes the addition of long, branched chains of PAR to a variety of nuclear proteins, including PARP itself (19, 39). The activation of PARP results in depletion of NAD and ATP and, if unchecked, eventually leads to irreversible cytotoxicity and cell death (19). The PARP-induced depletion of ATP results from an effort by cells to resynthesize NAD (19, 40). TNF-induced cell death has previously been associated with PAR synthesis and depletion of NAD (41). We have also recently shown that an anti-Fas ligation induces a transient burst of PAR synthesis prior to internucleosomal DNA cleavage (42). Prevention of this early poly(ADP-ribosyl)ation, either by disruption of the PARP gene or by expression of PARP antisense RNA, blocked completion of the apoptotic program as assessed on the basis of various morphological and biochemical markers of cell death (42).

We have attempted to reverse the apoptosis-resistant phenotype of DFF45^/ fibroblasts by reintroducing the DFF45 gene; however, we were unsuccessful at expressing sufficient DFF40 protein in these cells to induce 50-kb DNA breaks after TNF treatment (data not shown). Alternatively, to prove that resistance of DFF45^/ cells to apoptosis was not due to a deficiency in responding to excessive DNA damage, we compared the sensitivity of DFF45^/ cells with that of DFF45^+/+ cells to MNNG. MNNG is considered to be a strong DNA strand-breaking agent as well as an effective PARP activator and also an inducer of apoptosis (29-31). Both DFF45^/ and DFF45^+/+ cells were equally sensitive to MNNG treatment. This demonstrates that DFF45^/ cells are able to react to DNA damage and that their resistance to TNF-induced apoptosis may reside in their inability to generate 50-kb DNA breaks. This result strongly supports the concept that DNA damage is pivotal in triggering cell death.

Extensive PARP activation has been associated with various pathological conditions, including energetic failure and vascular collapse in shock (43, 44), streptozotocin-induced diabetes (45), cerebral ischemia (40), and glutamate neurotoxicity (46). We have recently also demonstrated a role for PARP and poly(ADP-ribosyl)ation in drug-induced parkinsonism (47). Furthermore, with the use of comparative genomic hybridization and cells derived from PARP knockout mice, we have shown that PARP and poly(ADP-ribosyl)ation are important for the maintenance of chromosomal stability (48). In addition, with the use of DNA microarray analysis, we have shown that PARP contributes to regulation of the genes for various cytoskeletal and cell cycle proteins (49).

The depletion of cellular energy reserves results in mitochondrial stress and impairment of metabolism (50), effects that are thought to lead to cell death (51). Extensive activation of PARP has recently been shown to result in a loss of _mito (29). The loss of _mito is thought to result from the opening of the permeability transition pore, which also leads to the release of proapoptotic factors, such as cytochrome c, into the cytosol (34, 35). The released cytochrome c interacts with Apaf-1, and the resulting complex binds to and triggers the proteolytic activation of procaspase-9. Caspase-9 then directly cleaves and activates caspase-3, one of the principal executioner caspases. Caspase-3 is also thought to be directly activated by caspase-8, which associates with TNF receptors (2).

A variety of DNA-damaging agents, including the topoisomerase inhibitor etoposide (52) and the alkylating agent MNNG (29), negatively affect mitochondrial function and induce cell death. Inhibition of PARP by gene disruption or pharmacological inhibitors protects against cell death in both animals and cells (19, 29, 32, 53). The early appearance of 50-kb DNA fragments (3 h after exposure to TNF) and activation of PARP (as assessed by PAR synthesis and depletion of intracellular NAD) in DFF45^+/+ fibroblasts coincided with both the loss of _mito and cytochrome c release. In contrast, both the loss of _mito and cytochrome c release were markedly delayed in DFF45^/ cells, suggesting that the generation of 50-kb DNA fragments and PARP activation may be causally related to proapoptotic mitochondrial events. PARP inhibition by 3-AB or by gene disruption provided substantial protection against TNF-induced cell death, consistent with previous observations (32, 41, 53). This protection was associated with inhibition of NAD depletion, cytochrome c release, and caspase-3 activation. The balance between the activation of PARP and the subsequent cleavage of this enzyme by caspase-3 appears to be an important determinant of the type of cell death. We have previously shown that expression of a caspase-3-resistant PARP mutant in osteosarcoma or PARP^/ cells increases the rate of cell death as a result of severe NAD depletion (20). Herceg and Wang (54) showed that expression of a similar PARP mutant switches the mode of cell death induced by TNF from apoptosis to necrosis. To avoid excessive depletion of energy reserves and a switch to necrosis, cells exposed to inducers of apoptosis thus cleave PARP into inactive peptides (29, 54). Whereas nonmodified PARP is cleaved by caspase-3 (55, 56), automodified PARP is preferentially cleaved by caspase-7 (57), which emphasizes the essentiality of PARP cleavage during apoptosis. Although it has been suggested that PARP is cleaved predominantly to avoid futile DNA repair, the present results and those of previous studies (20, 42, 54) suggest an active role for PARP in apoptotic cell death.

Our observation that cell death was delayed but not prevented in DFF45^/ cells suggests that apoptosis can occur in the absence of both fragmentation of DNA into 50-kb strands and internucleosomal DNA degradation. The type of death undergone by TNF-treated DFF45^/cells was presumably apoptotic in nature, given that it was accompanied by activation of caspase-3, PARP cleavage, and cytochrome c release. A defect in cytochrome c has been associated with resistance to Fas-induced apoptosis in human melanoma cells (59). The slight activation of PARP observed later during TNF treatment could be due to DNA strand breaks caused by reactive oxygen species generated during treatment. Reactive oxygen species generation during TNF treatment is well established (60-63). Furthermore, reactive oxygen species can accumulate as a consequence of the low level release of cytochrome c and the concomitant decrease in _mito observed after 6 h of TNF treatment (Fig. 6). We have investigated the possibility that DFF45^/ cells are resistant to TNF-induced cell death as a result of a dysfunctional TNF receptor by examining the ability of this cytokine to activate the transcription factor NF-B, a known target of TNF signaling (24), and the expression of the luciferase gene under the control of NF-B. No difference was apparent between DFF45^/ and control fibroblasts in their ability to activate NF-B, as assessed by EMSA with an oligonucleotide probe containing a specific NF-B binding site, or NF-B-driven expression of the luciferase gene.

Cleavage of DFF45 and the resulting activation of the DFF40 endonuclease are not the only events required for DNA degradation during apoptosis. Proteolytic cleavage of lamin has been shown to be essential for nuclear breakdown and internucleosomal DNA fragmentation (58); expression of a caspase-resistant lamin mutant thus delayed DNA fragmentation and conferred resistance to apoptosis. It remains to be determined whether this resistance to apoptosis is due to a delay in PARP activation.

In conclusion, our results suggest that, in fibroblasts, TNF induces an initial activation of caspase-3, possibly mediated by caspase-8 (2) or through mitochondria, and that the activated caspase-3 then cleaves DFF45 and thereby activates DFF40. The endonuclease activity of DFF40 mediates the cleavage of DNA into 50-kb fragments, which results in the activation of PARP and the consequent depletion of NAD and ATP. This reduction in cellular energy reserves induces mitochondrial stress, which manifests as a loss of _mito and the release of cytochrome c and leads to a second phase of caspase-3 activation that results in extensive degradation of target proteins, including DFF45 and PARP, and, ultimately, in cell death.

	ACKNOWLEDGEMENTS

We thank Dr. A. Faden for DFF45^/ mice, Dr. X. Wang for antibodies to DFF40, Dr. S. Snyder for PARP^/and ^+/+ cells, and Dr. C. Giardina for NF-B-luciferase plasmid.

	FOOTNOTES

^* This work was supported in part by NCI, National Institutes of Health, Grants CA25344 and CA13195, United States Air Force Office of Scientific Research Grant AFOSR-89-0053, and the United States Army Medical Research and Development Command Grant DAMD17-90-C-0053.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, Georgetown University Medical Center, Basic Science Bldg., Rm. 351, 3900 Reservoir Rd. NW, Washington, D. C. 20007. Tel.: 202-687-1718; Fax: 202-687-7186; E-mail: smulson@bc.georgetown.edu.

Published, JBC Papers in Press, July 18, 2001, DOI 10.1074/jbc.M100629200

	ABBREVIATIONS

The abbreviations used are: TNF, tumor necrosis factor; DFF, DNA fragmentation factor; PAR, poly(ADP-ribose); PARP, PAR polymerase; AMC, aminomethylcoumarin; EMSA, electrophoretic mobility shift assay; _mito, mitochondrial membrane potential; JC-1, 5,5',6,6'-tetrachloro-1,1'3,3'-tetraethylbenzimidazolylcarbocyanine iodide; 3-AB, 3-aminobenzamide; kb, kilobase pair(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES