Caspase-3 Is Required for DNA Fragmentation and Morphological Changes Associated with Apoptosis
1998; Elsevier BV; Volume: 273; Issue: 16 Linguagem: Inglês
10.1074/jbc.273.16.9357
ISSN1083-351X
AutoresReiner U. Jänicke, Michael L. Sprengart, Mas Rina Wati, Alan G. Porter,
Tópico(s)NF-κB Signaling Pathways
ResumoInterleukin 1β-converting enzyme-like proteases (caspases) are crucial components of cell death pathways. Among the caspases identified, caspase-3 stands out because it is commonly activated by numerous death signals and cleaves a variety of important cellular proteins. Studies in caspase-3 knock-out mice have shown that this protease is essential for brain development. To investigate the requirement for caspase-3 in apoptosis, we took advantage of the MCF-7 breast carcinoma cell line, which we show here has lost caspase-3 owing to a 47-base pair deletion within exon 3 of the CASP-3 gene. This deletion results in the skipping of exon 3 during pre-mRNA splicing, thereby abrogating translation of the CASP-3 mRNA. Although MCF-7 cells were still sensitive to tumor necrosis factor (TNF)- or staurosporine-induced apoptosis, no DNA fragmentation was observed. In addition, MCF-7 cells undergoing cell death did not display some of the distinct morphological features typical of apoptotic cells such as shrinkage and blebbing. Introduction of the CASP-3 gene into MCF-7 cells resulted in DNA fragmentation and cellular blebbing following TNF treatment. These results indicate that although caspase-3 is not essential for TNF- or staurosporine-induced apoptosis, it is required for DNA fragmentation and some of the typical morphological changes of cells undergoing apoptosis. Interleukin 1β-converting enzyme-like proteases (caspases) are crucial components of cell death pathways. Among the caspases identified, caspase-3 stands out because it is commonly activated by numerous death signals and cleaves a variety of important cellular proteins. Studies in caspase-3 knock-out mice have shown that this protease is essential for brain development. To investigate the requirement for caspase-3 in apoptosis, we took advantage of the MCF-7 breast carcinoma cell line, which we show here has lost caspase-3 owing to a 47-base pair deletion within exon 3 of the CASP-3 gene. This deletion results in the skipping of exon 3 during pre-mRNA splicing, thereby abrogating translation of the CASP-3 mRNA. Although MCF-7 cells were still sensitive to tumor necrosis factor (TNF)- or staurosporine-induced apoptosis, no DNA fragmentation was observed. In addition, MCF-7 cells undergoing cell death did not display some of the distinct morphological features typical of apoptotic cells such as shrinkage and blebbing. Introduction of the CASP-3 gene into MCF-7 cells resulted in DNA fragmentation and cellular blebbing following TNF treatment. These results indicate that although caspase-3 is not essential for TNF- or staurosporine-induced apoptosis, it is required for DNA fragmentation and some of the typical morphological changes of cells undergoing apoptosis. Apoptosis (programmed cell death) is typically accompanied by the activation of a class of death proteases (caspases) and widespread biochemical and morphological changes to the cell (1Nicholson D.W. Thornberry N.A. Trends Biochem. Sci. 1997; 22: 299-306Abstract Full Text PDF PubMed Scopus (2172) Google Scholar, 2Porter A.G. Ng P. Jänicke R.U. BioEssays. 1997; 19: 501-507Crossref PubMed Scopus (166) Google Scholar). These changes almost invariably involve chromatin condensation and its margination at the nuclear periphery, extensive double-stranded DNA fragmentation, and cellular shrinkage and blebbing (3Wyllie A.H. Kerr J.F.R. Currie A.R. Int. Rev. Cytol. 1980; 68: 251-306Crossref PubMed Scopus (6683) Google Scholar, 4Wyllie A.H. Nature. 1980; 284: 555-556Crossref PubMed Scopus (4129) Google Scholar, 5Earnshaw W.C. Curr. Opin. Cell Biol. 1995; 7: 337-343Crossref PubMed Scopus (509) Google Scholar). However, apoptosis can also occur in the absence of DNA fragmentation (6Ucker D.S. Obermiller P.S. Eckhart W. Apgar J.R. Berger N.A. Meyers J. Mol. Cell. Biol. 1992; 12: 3060-3069Crossref PubMed Scopus (234) Google Scholar, 7Oberhammer F. Wilson J.W. Dive C. Morris I.D. Hickman J.A. Wakeling A.E. Walker P.R. Sikorska M. EMBO J. 1993; 12: 3679-3684Crossref PubMed Scopus (1157) Google Scholar, 8Schulze-Osthoff K. Walczak H. Dröge W. Krammer P.H. J. Cell. Biol. 1994; 127: 15-20Crossref PubMed Scopus (411) Google Scholar, 9Sakahira H. Enari M. Nagata S. Nature. 1998; 391: 96-99Crossref PubMed Scopus (1409) Google Scholar). Recently, it has been demonstrated that a caspase activates an endonuclease (CAD) 1The abbreviations used are: CAD, caspase-activated DNase; ICAD, inhibitor of CAD; RT, reverse transcription; PCR, polymerase chain reaction; bp, base pair(s); Chx, cycloheximide; TNF, tumor necrosis factor. 1The abbreviations used are: CAD, caspase-activated DNase; ICAD, inhibitor of CAD; RT, reverse transcription; PCR, polymerase chain reaction; bp, base pair(s); Chx, cycloheximide; TNF, tumor necrosis factor. responsible for fragmentation of the DNA at the linker region between nucleosomes by specifically cleaving and inactivating ICAD (DFF45), the inhibitor of CAD (9Sakahira H. Enari M. Nagata S. Nature. 1998; 391: 96-99Crossref PubMed Scopus (1409) Google Scholar, 10Enari M. Sakahira H. Yokoyama H. Okawa K. Iwamatsu A. Nagata S. Nature. 1998; 391: 43-50Crossref PubMed Scopus (2785) Google Scholar, 11Liu X. Zou H. Slaughter C. Wang X. Cell. 1997; 89: 175-184Abstract Full Text Full Text PDF PubMed Scopus (1629) Google Scholar). There is evidence that caspases contribute to the drastic morphological changes of apoptosis by proteolysing and disabling a number of key substrates, including the structural proteins gelsolin, PAK2, focal adhesion kinase, and rabaptin-5 (12Kothakota S. Azuma T. Reinhard C. Klippel A. Tang J. Chu K. McGarry T.J. Kirschner M.W. Koths K. Kwiatkowski D.J. Williams L.T. Science. 1997; 278: 294-298Crossref PubMed Scopus (1027) Google Scholar, 13Rudel T. Bokoch G.M. Science. 1997; 276: 1571-1574Crossref PubMed Scopus (601) Google Scholar, 14Wen L.-P. Fahrni J.A. Troie S. Guan J.-L. Orth K. Rosen G.D. J. Biol. Chem. 1997; 272: 26056-26061Abstract Full Text Full Text PDF PubMed Scopus (308) Google Scholar, 15Cosulich S.C. Horiuchi H. Zerial M. Clarke P.R. Woodman P.G. EMBO J. 1997; 16: 6182-6191Crossref PubMed Scopus (70) Google Scholar). The most commonly activated caspase (caspase-3) can mediate the limited proteolysis of these proteins, as well as the cleavage inactivation of DNA fragmentation factor (DFF45; ICAD) (9Sakahira H. Enari M. Nagata S. Nature. 1998; 391: 96-99Crossref PubMed Scopus (1409) Google Scholar, 10Enari M. Sakahira H. Yokoyama H. Okawa K. Iwamatsu A. Nagata S. Nature. 1998; 391: 43-50Crossref PubMed Scopus (2785) Google Scholar, 11Liu X. Zou H. Slaughter C. Wang X. Cell. 1997; 89: 175-184Abstract Full Text Full Text PDF PubMed Scopus (1629) Google Scholar). Because there are several caspase-3-like proteases (1Nicholson D.W. Thornberry N.A. Trends Biochem. Sci. 1997; 22: 299-306Abstract Full Text PDF PubMed Scopus (2172) Google Scholar, 2Porter A.G. Ng P. Jänicke R.U. BioEssays. 1997; 19: 501-507Crossref PubMed Scopus (166) Google Scholar), it is not known if caspase-3 is required in vivo for breakdown of DNA or cleavage of any of the proteins involved in maintaining cellular architecture. Here we show that MCF-7 carcinoma cells, which can be killed by apoptotic stimuli without DNA fragmentation and many of the hallmarks of apoptosis (7Oberhammer F. Wilson J.W. Dive C. Morris I.D. Hickman J.A. Wakeling A.E. Walker P.R. Sikorska M. EMBO J. 1993; 12: 3679-3684Crossref PubMed Scopus (1157) Google Scholar), are devoid of caspase-3 owing to a functional deletion in the CASP-3 gene. This has enabled us to address the question of whether caspase-3 is essential for double-stranded DNA breaks and some of the morphological alterations typical of apoptotic cell death. The human Sy5y neuroblastoma cell line was provided by E. Feldman (University of Michigan), and the human breast carcinoma cell line MCF-7 was obtained from the ATCC. Cell lines were maintained in RPMI 1640 or Dulbecco's modified Eagle's medium (Sy5y) supplemented with 10% fetal calf serum, 10 mm glutamine, and 50 μg (each) of streptomycin and penicillin/ml (16Jänicke R.U. Lee F.H.H. Porter A.G. Mol. Cell. Biol. 1994; 14: 5661-5670Crossref PubMed Scopus (94) Google Scholar). The origin and cultivation of HeLa D98 and H21 cells were as described (16Jänicke R.U. Lee F.H.H. Porter A.G. Mol. Cell. Biol. 1994; 14: 5661-5670Crossref PubMed Scopus (94) Google Scholar). The protease inhibitors aprotinin, bacitracin, antipain, leupeptin, and phenylmethylsulfonyl fluoride as well as staurosporine were purchased from Sigma. The specific activity of TNF was 4 × 107units/mg protein. The monoclonal antibodies to CPP32 (caspase-3) were from Transduction Laboratories Inc. The monoclonal anti-actin antibodies were from Sigma. Cell extracts were prepared as described (17Jänicke R.U. Lin X.Y. Lee F.H.H. Porter A.G. Mol. Cell. Biol. 1996; 16: 5245-5253Crossref PubMed Scopus (55) Google Scholar). For detection of caspase-3, cell extracts were separated in 0.1% SDS, 12.5% polyacrylamide gels and subjected to Western blotting as described (16Jänicke R.U. Lee F.H.H. Porter A.G. Mol. Cell. Biol. 1994; 14: 5661-5670Crossref PubMed Scopus (94) Google Scholar). The protein was visualized by the Amersham ECL kit. The plasmid pcDNA3 containing the full-length Yama (CASP-3) cDNA was provided by V. Dixit. For the isolation of CASP-3 cDNA, total RNA was extracted by the guanidinium thiocyanate method as described (16Jänicke R.U. Lee F.H.H. Porter A.G. Mol. Cell. Biol. 1994; 14: 5661-5670Crossref PubMed Scopus (94) Google Scholar) or with a Qiagen RNA purification kit reverse transcribed with the SuperScript Kit (Life Technologies, Inc.) and amplified by PCR. The following primers were used for the PCR (BamHI and EcoRI sites are indicated by italicized letters): Y1, 5′-AAAGGATCCTTAATAAAGGTATCCATGGAGAACACT-3′ (nucleotides −15 to +12 relative to the AUG in the CASP-3 mRNA); and Y5, 5′-AAAGAATTCTTAGTGATAAAAATAGAGTTCTTTTGTGAG-3′ (nucleotides +834 to +805 of CASP-3 mRNA). The PCR products were cloned into the BamHI/EcoRI sites of pUC18. Genomic DNA was isolated from MCF-7 and Sy5y cells with a Qiagen purification kit. Based on the genomic organization of the mouse CASP-3 DNA (18Juan T.S.-C. McNiece I.K. Jenkins N.A. Gilbert D.J. Copeland N.G. Fletcher F.A. Oncogene. 1996; 13: 749-755PubMed Google Scholar) and published cDNA sequences of human CASP-3 (19Fernandes-Alnemri T. Litwack G. Alnemri E.S. J. Biol. Chem. 1994; 269: 30761-30764Abstract Full Text PDF PubMed Google Scholar), DNA fragments were amplified from both cell lines by using two different primer sets specific for exons 2/3 (Y1/Y2) and exons 3/4 (Y3/Y4), respectively (see Fig. 2 C). Primer sequences (BamHI and EcoRI sites are indicated by italicized letters) were: Y1, 5′-AAAGGATCCTTAATAAAGGTATCCATGGAGAACACT-3′ (−15 to +12: exon 2); Y2, 5′-AAAGAATTCCAGTGCTTTTATGAAAATTCTTATTAT-3′ (+178 to +152: exon 3); Y3, 5′-AAAGGATCCAAAGATCATACATGGAAGCGAATCAAT-3′ (+54 to +80: exon 3); and Y4, 5′-AAAGAATTCCATCACGCATCAATTCCACAATTTCTT-3′ (+307 to +281: exon 4). Amplification of MCF-7 or Sy5y DNA with the CASP-3-specific primer sets (Y1/Y2 and Y3/Y4) resulted in two different PCR products, both with a size of approximately 3200 bp, containing exons 2/3 or exons 3/4, respectively (see Fig. 2 C). The 5′ and 3′ ends of the PCR products were directly sequenced with a double-stranded DNA Cycle Sequencing System (Life Technologies, Inc.) to confirm that both fragments were derived from the human CASP-3 gene. Subsequently, DNA fragments from several independent PCR reactions obtained from MCF-7 and Sy5y DNA with the same primer sets (Y1/Y2 and Y3/Y4) were isolated and digested with BamHI and EcoRI. A ∼1800-bp BamHI/EcoRI fragment derived from digestion of the Y1/Y2 PCR product (comprising exon 3 and part of the 5′ flanking intron 2; see Fig. 2 C) and a ∼1900-bp BamHI/BamHI fragment derived from digestion of the Y3/Y4 PCR fragment (comprising exon 3 and part of the 3′ flanking intron 3; see Fig. 2 C) were isolated from both cell lines and cloned into pUC18. Sequence analysis of exon 3 and flanking intron sequences was performed with several clones derived from independent PCRs by using pUC18-specific forward and reverse sequencing primers. Apoptosis was induced with a combination of TNF (30 ng/ml) and cycloheximide (Chx; 10 μg/ml) or with staurosporine (1 μm) (Sigma). Cell death was measured with the standard TNF cytotoxicity assay as described (16Jänicke R.U. Lee F.H.H. Porter A.G. Mol. Cell. Biol. 1994; 14: 5661-5670Crossref PubMed Scopus (94) Google Scholar). TNF-treated or untreated cells were fixed in ice-cold 80% ethanol, washed with phosphate-buffered saline and stained with propidium iodide (50 μg/ml, Sigma) at 37 °C for 60 min in the presence of RNase (20 μg/ml, Sigma) and 0.1% Triton X-100. Cell cycle analysis was performed using a Becton Dickinson FACScan. For each determination, a minimum of 20,000 cells were analyzed. For DNA fragmentation analysis, cellular DNA was prepared using the Blood and Cell Culture Mini DNA kit (Qiagen, Germany). Purified DNA was incubated for 2 h at 37 °C with 200 μg/ml RNase and analyzed on 1.6% agarose gels. DNA was visualized by ethidium bromide staining. MCF-7 cells were stably transfected with caspase-3 or the expression vector alone (pcDNA3, Invitrogen) using the SuperFect Reagent (Qiagen, Germany). 40 h post-transfection, cells were trypsinized and reseeded in medium containing 800 μg/ml G418 (Life Technologies, Inc.). HeLa D98 and H21 carcinoma cells, which are highly or only marginally sensitive to TNF-induced apoptosis, respectively (16Jänicke R.U. Lee F.H.H. Porter A.G. Mol. Cell. Biol. 1994; 14: 5661-5670Crossref PubMed Scopus (94) Google Scholar), were treated with TNF or staurosporine, and the percentage of apoptosis was compared with the activation of pro-caspase-3. Because the caspase-3 antibody used for Western blot analysis did not detect the active 17-kDa subunit, the activation/processing of caspase-3 was judged by the disappearance of the 32-kDa precursor form. A 4-h treatment with TNF/Chx killed 78% of HeLa D98 cells (Fig. 1 A) and resulted in a complete loss of the 32-kDa caspase-3 precursor form (Fig. 1 B, compare lanes 1 and 2). In contrast, pro-caspase-3 was efficiently processed in HeLa H21 cells following a 4-h treatment with TNF/Chx (Fig. 1 B, lane 5), but only 18% of the cells were killed (Fig. 1 A). There was also no correlation between the extent of cell death and activation of caspase-3 when both HeLa cell lines were exposed to the apo-ptosis inducer staurosporine (20Bertrand R. Solary E. O'Connor P. Kohn K.W. Pommier Y. Exp. Cell Res. 1994; 211: 314-321Crossref PubMed Scopus (466) Google Scholar). Although both cell lines were efficiently killed after a 16-h staurosporine treatment (Fig. 1 A), extensive processing of pro-caspase-3 was observed only in HeLa D98 cells (Fig. 1 B, compare lanes 3 and 6). These results suggest that TNF and staurosporine may activate different apoptotic pathways and that caspase-3 activation is dependent not only on the death stimulus but also on the cell type. Our observation that caspase-3 may not be essential for all death scenarios was directly proven with the breast carcinoma cell line MCF-7. These cells were efficiently killed by both death stimuli, TNF/Chx and staurosporine (Fig. 1 A); however, we could not detect the expression of pro-caspase-3 (Fig. 1 B, lanes 7–9). To investigate why pro-caspase-3 was absent in MCF-7 cells, Northern blot analysis was performed with total RNA from MCF-7, HeLa D98, and H21 cells, as well as Jurkat cells that were shown to express functional CASP-3 mRNA (19Fernandes-Alnemri T. Litwack G. Alnemri E.S. J. Biol. Chem. 1994; 269: 30761-30764Abstract Full Text PDF PubMed Google Scholar). All cell lines showed similar expression of the CASP-3 mRNA, and no apparent size differences were observed (data not shown). Therefore, the lack of pro-caspase-3 in MCF-7 cells is not due to the failure to transcribe CASP-3 mRNA. To determine whether mutation(s) led to the loss of pro-caspase-3, total RNA prepared from MCF-7 cells was reverse transcribed and used for PCR amplification. The RT-PCR performed with specific primers flanking the entire coding region of the CASP-3 mRNA gave rise to a fragment of around 750 bp (Fig. 2 A, lane 2), which was shorter than the expected 867 bp observed in two control reactions performed with total RNA from HeLa D98 and Jurkat cells (Fig. 2 A, lanes 3 and 4). Sequence analysis revealed that the MCF-7 CASP-3 mRNA contained a deletion of nucleotides 54–178 (Fig. 2 B), whereas the mRNA from HeLa D98 cells matched the published sequence of the CASP-3 cDNA (19Fernandes-Alnemri T. Litwack G. Alnemri E.S. J. Biol. Chem. 1994; 269: 30761-30764Abstract Full Text PDF PubMed Google Scholar). The deletion of 125 nucleotides found in the MCF-7CASP-3 mRNA leads to a frame shift starting at codon 18, thereby creating a new in-frame stop codon 41 amino acids downstream of the initiation codon (Fig. 2 B), thus explaining the lack of pro-caspase-3 in MCF-7 cells. Six clones from two independent PCRs all contained the same deletion, and no DNA corresponding to the full-length CASP-3 cDNA was detected, suggesting that the PCR did not lead to the introduction of mutations in this gene and that no wild-type allele of CASP-3 is expressed. Furthermore, several MCF-7 clones obtained from different sources were all found to contain the same deletion in the CASP-3 mRNA, implying that the loss of pro-caspase-3 is not due to clonal variation. The genomic organization of the human CASP-3 gene is unknown. However, based on the genomic map of the mouseCASP-3 gene (18Juan T.S.-C. McNiece I.K. Jenkins N.A. Gilbert D.J. Copeland N.G. Fletcher F.A. Oncogene. 1996; 13: 749-755PubMed Google Scholar), it is obvious that exon 3 consists of exactly the 125 nucleotides deleted in the MCF-7 CASP-3 mRNA. Therefore, we concluded that the CASP-3 mRNA produced in MCF-7 cells does not contain exon 3. To identify the cause of the exon skipping, two different primer sets for the CASP-3 gene were used to amplify genomic DNA from MCF-7 and Sy5y cells (which express pro-caspase-3; data not shown). In both cell lines, primer set Y1/Y2 specific for exons 2 and 3 (Fig. 2 C) gave rise to a PCR product of the expected size of approximately 3200 bp, whereas the fragment obtained with primer set Y3/Y4 specific for exons 3 and 4 (Fig. 2 C) was much longer (∼3200 bp) than the corresponding intron 3 sequence reported for the mouseCASP-3 gene (∼1755 bp) (18Juan T.S.-C. McNiece I.K. Jenkins N.A. Gilbert D.J. Copeland N.G. Fletcher F.A. Oncogene. 1996; 13: 749-755PubMed Google Scholar). Sequencing of the amplified products revealed a deletion of nucleotides 84–130 within exon 3 of the MCF-7 CASP-3 gene, whereas the full-length exon 3 matching the known CASP-3 sequence was present in Sy5y cells. In addition, approximately 100 bp of adjacent intron sequences upstream and downstream of exon 3 (3′-intron 2 and 5′-intron 3, respectively) as well as intron consensus sequences downstream of exon 2 and upstream of exon 4 (Fig. 2 C) were found to be identical in both MCF-7 and Sy5y cells, indicating that the skipping of exon 3 in MCF-7 cells is not due to mutations in these important splicing regions. These results suggest that the deletion of 47 bp within exon 3 of the CASP-3 gene in MCF-7 cells leads to the skipping of this exon during splicing of the CASP-3 pre-mRNA. Recently, it was reported that MCF-7 cells do not express detectable levels of caspase-3 (21Li F. Srinivasan A. Wang Y. Armstrong R.C. Tomaselli K.J. Fritz L.C. J. Biol. Chem. 1997; 272: 30299-30305Crossref PubMed Scopus (331) Google Scholar). Here we extend this observation and show that the lack of caspase-3 in MCF-7 cells is due to a partial deletion within exon 3 of the CASP-3 gene. How does the deletion within exon 3 result in the skipping of the complete exon? Correct splicing of pre-mRNAs depends on conserved sequences around the 5′ and 3′ intronic splice sites, including a degenerate site of lariat formation (branchpoint) that is approximately 10–50 nucleotides upstream of the 3′ acceptor splice site (22Nilsen T.W. Cell. 1994; 78: 1-4Abstract Full Text PDF PubMed Scopus (168) Google Scholar, 23Valcárcel J. Green M.R. Trends Biochem. Sci. 1996; 21: 296-301Abstract Full Text PDF PubMed Google Scholar). In addition, splice site strength, RNA secondary structure around splice sites, exon length, and even purine-rich exon-specific enhancer sequences have been reported to be involved in exon recognition (23Valcárcel J. Green M.R. Trends Biochem. Sci. 1996; 21: 296-301Abstract Full Text PDF PubMed Google Scholar, 24Tanaka K. Watakabe A. Shimura Y. Mol. Cell. Biol. 1994; 14: 1347-1354Crossref PubMed Scopus (206) Google Scholar, 25Berget S.M. J. Biol. Chem. 1995; 270: 2411-2414Abstract Full Text Full Text PDF PubMed Scopus (855) Google Scholar). Except for the deletion, other sequences in the MCF-7 CASP-3 gene including those around the splice sites of exon 3 were identical to the CASP-3 gene of pro-caspase-3-expressing Sy5y cells. It is also unlikely that the skipping of exon 3 is caused by its small size, because exons shorter than 78 bp can be correctly spliced (25Berget S.M. J. Biol. Chem. 1995; 270: 2411-2414Abstract Full Text Full Text PDF PubMed Scopus (855) Google Scholar). Therefore, we conclude that the deleted internal exon sequences, which do not contain the consensus of the purine-rich exon-specific enhancer sequence GARGARGAR (where R is any purine) (26Xu R. Teng J. Cooper T.A. Mol. Cell. Biol. 1993; 13: 3660-3674Crossref PubMed Scopus (177) Google Scholar), are necessary for the correct splicing of the CASP-3 pre-mRNA. Additional studies are necessary to identify the critical sequences and interacting enhancer proteins involved in pre-mRNA splicing (23Valcárcel J. Green M.R. Trends Biochem. Sci. 1996; 21: 296-301Abstract Full Text PDF PubMed Google Scholar). MCF-7 cells treated with various apoptotic stimuli (e.g. transforming growth factor-β1 or etoposide) undergo cell death in the absence of DNA fragmentation (7Oberhammer F. Wilson J.W. Dive C. Morris I.D. Hickman J.A. Wakeling A.E. Walker P.R. Sikorska M. EMBO J. 1993; 12: 3679-3684Crossref PubMed Scopus (1157) Google Scholar). In addition, the apoptotic stimuli TNF/Chx or staurosporine efficiently killed MCF-7 cells (Fig. 1), but no DNA fragmentation was observed (Fig. 3 A). In contrast, treatment of HeLa D98 cells with TNF/Chx or staurosporine resulted in the appearance of the internucleosomal DNA laddering typical of cells undergoing apoptosis (Fig. 3 A, upper panel) (16Jänicke R.U. Lee F.H.H. Porter A.G. Mol. Cell. Biol. 1994; 14: 5661-5670Crossref PubMed Scopus (94) Google Scholar). Recently, it has been reported that caspase-3 activates the endonuclease CAD responsible for DNA fragmentation by specifically cleaving and inactivating ICAD, the inhibitor of CAD (9Sakahira H. Enari M. Nagata S. Nature. 1998; 391: 96-99Crossref PubMed Scopus (1409) Google Scholar, 10Enari M. Sakahira H. Yokoyama H. Okawa K. Iwamatsu A. Nagata S. Nature. 1998; 391: 43-50Crossref PubMed Scopus (2785) Google Scholar). However, because there are several caspase-3-like proteases, it is not known whether caspase-3 is sufficient or essential for DNA fragmentation. To investigate the hypothesis that caspase-3 is required for DNA fragmentation, CASP-3 cDNA was stably transfected into MCF-7 cells, which resulted in the generation of 26 individual caspase-3-expressing MCF-7 clones. In all of these clones, caspase-3 was not spontaneously activated, and there were no detectable morphological changes (data not shown). However, the caspase-3 activation pathway was fully functional in these transfectants, because TNF/Chx treatment of MCF-7.3.28 cells (one representative caspase-3-expressing clone out of three tested) resulted in the efficient activation of this protease as judged by Western blotting (Fig. 3 B, top panel). In addition, MCF-7.3.28 cells showed a slight increase in TNF sensitivity (data not shown), and more importantly, DNA from TNF/Chx-treated MCF-7.3.28 cells exhibited the internucleosomal DNA laddering typical of apoptotic cells (Fig. 3 B, middle panel). No internucleosomal DNA fragmentation was observed in MCF-7 cells transfected with the vector alone (Fig. 3 B, middle panel). Lighter exposures of both agarose gels revealed extensive double-stranded DNA breaks as judged by faster migrating DNA only in TNF/Chx- and staurosporine-treated HeLa D98 cells (Fig. 3 A, bottom panel) and TNF/Chx-treated MCF-7 cells stably expressing caspase-3 (Fig. 3 B, bottom panel). These data demonstrate that the components required for DNA fragmentation are present and fully functional in MCF-7 cells and, furthermore, indicate that caspase-3 is essential for their activation. Cell cycle analysis with propidium iodide-stained cells confirmed these results, because TNF/Chx induced the appearance of a sub-G1apoptotic peak characteristic of DNA fragmentation in MCF-7.3.28 cells (Fig. 4 B) but not in vector-transfected (Fig. 4 A) nor in parental MCF-7 cells (data not shown). Finally, MCF-7.3.28 cells but not MCF-7/vector cells showed the morphological changes typical of cells undergoing apoptosis such as shrinkage and blebbing following TNF/Chx treatment (Fig. 4,C and D). TNF/Chx treatment of two other individual caspase-3-expressing MCF-7 clones also resulted in DNA fragmentation and blebbing (data not shown). Taken together, these data provide strong evidence that caspase-3 is required for DNA fragmentation and the morphological changes during apoptosis. MCF-7 cells transfected to express the Fas antigen CD95 were shown to be sensitive to Fas ligand-induced apoptosis but were resistant to apoptosis induced by microinjection of cytochrome c (21Li F. Srinivasan A. Wang Y. Armstrong R.C. Tomaselli K.J. Fritz L.C. J. Biol. Chem. 1997; 272: 30299-30305Crossref PubMed Scopus (331) Google Scholar). In addition, we show that MCF-7 cells are as sensitive as HeLa D98 cells to TNF- or staurosporine-induced apoptosis despite the functional deletion of the CASP-3 gene, demonstrating that apoptosis induced by these agents does not require caspase-3. MCF-7 cells have been widely reported to die by apoptosis induced by TNF, Fas, or staurosporine (21Li F. Srinivasan A. Wang Y. Armstrong R.C. Tomaselli K.J. Fritz L.C. J. Biol. Chem. 1997; 272: 30299-30305Crossref PubMed Scopus (331) Google Scholar, 27Jänicke R.U. Walker P.A. Lin X.Y. Porter A.G. EMBO J. 1996; 15: 6969-6978Crossref PubMed Scopus (222) Google Scholar, 28Tewari M. Quan L.T. O'Rourke K. Desnoyers S. Zeng Z. Beidler D.R. Poirier G.G. Salvesen G.S. Dixit V.M. Cell. 1995; 81: 801-809Abstract Full Text PDF PubMed Scopus (2259) Google Scholar, 29Duan H. Chinnaiyan A.M. Hudson P.L. Wing J.P. He W.-W. Dixit V.M. J. Biol. Chem. 1996; 271: 1621-1625Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar). However, rounding and detachment of MCF-7 cells was accompanied by an altered chromatin structure but without a typical apoptotic morphology and, notably, in the absence of any DNA fragmentation (7Oberhammer F. Wilson J.W. Dive C. Morris I.D. Hickman J.A. Wakeling A.E. Walker P.R. Sikorska M. EMBO J. 1993; 12: 3679-3684Crossref PubMed Scopus (1157) Google Scholar). By expressing caspase-3 in MCF-7 cells, we have shown that caspase-3 is required for the surface blebbing that is so typical of a disintegrating apoptotic cell. Caspase-3 (or a related caspase-3-like protease) has been shown to cleave a number of structural proteins involved in maintaining cytoplasmic and nuclear architecture and integrity, such as gelsolin, actin, α-fodrin, PAK2, FAK, and rabaptin-5 (2Porter A.G. Ng P. Jänicke R.U. BioEssays. 1997; 19: 501-507Crossref PubMed Scopus (166) Google Scholar, 12Kothakota S. Azuma T. Reinhard C. Klippel A. Tang J. Chu K. McGarry T.J. Kirschner M.W. Koths K. Kwiatkowski D.J. Williams L.T. Science. 1997; 278: 294-298Crossref PubMed Scopus (1027) Google Scholar, 13Rudel T. Bokoch G.M. Science. 1997; 276: 1571-1574Crossref PubMed Scopus (601) Google Scholar, 14Wen L.-P. Fahrni J.A. Troie S. Guan J.-L. Orth K. Rosen G.D. J. Biol. Chem. 1997; 272: 26056-26061Abstract Full Text Full Text PDF PubMed Scopus (308) Google Scholar, 15Cosulich S.C. Horiuchi H. Zerial M. Clarke P.R. Woodman P.G. EMBO J. 1997; 16: 6182-6191Crossref PubMed Scopus (70) Google Scholar). It remains to be determined whether caspase-3 is required to cleave any of these important substrates for blebbing to occur. We have also demonstrated that caspase-3 is essential for fragmentation of MCF-7 cell chromosomal DNA during TNF-induced apoptosis. CAD was shown to be the DNA endonuclease responsible for cleaving chromosomal DNA specifically during apoptosis (9Sakahira H. Enari M. Nagata S. Nature. 1998; 391: 96-99Crossref PubMed Scopus (1409) Google Scholar, 10Enari M. Sakahira H. Yokoyama H. Okawa K. Iwamatsu A. Nagata S. Nature. 1998; 391: 43-50Crossref PubMed Scopus (2785) Google Scholar). The inhibitor of this endonuclease (ICAD; DFF45) can be cleaved and inactivated by caspase-3, releasing active CAD that translocates into the nucleus where it degrades the DNA (9Sakahira H. Enari M. Nagata S. Nature. 1998; 391: 96-99Crossref PubMed Scopus (1409) Google Scholar, 10Enari M. Sakahira H. Yokoyama H. Okawa K. Iwamatsu A. Nagata S. Nature. 1998; 391: 43-50Crossref PubMed Scopus (2785) Google Scholar, 11Liu X. Zou H. Slaughter C. Wang X. Cell. 1997; 89: 175-184Abstract Full Text Full Text PDF PubMed Scopus (1629) Google Scholar). Our results raise the possibility that ICAD degradation may be the caspase-3-dependent step in DNA fragmentation. If not, another as yet unknown caspase-3-dependent cleavage step may occur upstream or downstream of CAD activation. In conclusion, this is the first report showing that a caspase is functionally deleted in a cancer cell line and that caspase-3 is required for certain distinctive biochemical and morphological changes during apoptosis. Knockout mouse studies have established that caspase-3 is essential for proper brain development, but it is not known at what step it acts (30Kuida K. Zheng T.S. Na S. Kuan C-y. Yang D. Karasuyama H. Rakic P. Flavell R.A. Nature. 1996; 384: 368-372Crossref PubMed Scopus (1695) Google Scholar). In contrast, caspase-3 is apparently dispensable for programmed cell death in all other tissues and organs (30Kuida K. Zheng T.S. Na S. Kuan C-y. Yang D. Karasuyama H. Rakic P. Flavell R.A. Nature. 1996; 384: 368-372Crossref PubMed Scopus (1695) Google Scholar), which implies that apoptosis during development in caspase-3-deficient mice may occur without DNA fragmentation. We thank P. Singh and C. Pallen for critically reviewing the manuscript, E. Feldman for providing the Sy5y neuroblastoma cell line, and V. Dixit for the CASP-3 expression vector.
Referência(s)