Degradation of topoisomerase IIalpha during adenovirus E1A-induced apoptosis is mediated by the activation of the ubiquitin proteolysis system.

The human epithermoid carcinoma-derived cell line MA1, established by introduction of the adenovirus E1A 12 S cDNA linked to the mouse mammary tumor virus long terminal repeat, elicits apoptosis after induction of E1A12S in response to dexamethasone. The level of topoisomerase IIα begins to decrease steeply within 36 h preceding the onset of DNA fragmentation, whereas its mRNA level is unchanged (Nakajima, T., Ohi, N., Arai, T., Nozaki, N., Kikuchi, A., and Oda, K. (1995) Oncogene 10, 651-662). Topoisomerase IIα prepared by immunoprecipitation or extraction of the nuclear matrix was degraded much more efficiently in the S10 extract prepared from MA1 cells treated with dexamethasone for 42 h (the 42-h extract) than in the extract from untreated MA1 cells (the 0-h extract) in an ATP- and ubiquitin-dependent manner. The proteolytic activity for degradation of topoisomerase IIα was suppressed specifically by inhibitors for the proteasome and was much reduced in the 42-h extract prepared from MA1-derivative cell lines expressing E1B19k or Bcl-2. The proteolytic activity was lost after fractionation of the 42-h S10 extract into the S70 and P70 fractions by centrifugation at 70,000 × g for 6 h but partially recovered when these fractions were combined. Polyubiquitinated forms of topoisomerase IIα could be detected by incubating it in the S70 or S100 extract, which lacks most of the proteasome activity. The ubiquitination activity in S70 prepared from the 42-h extract was 4- to 5-fold higher than that prepared from the 0-h extract. These results suggest that a component(s) in the ubiquitin proteolysis pathway, responsible for ubiquitination and degradation of topoisomerase IIα, is activated or induced during the latent phase of E1A-induced apoptosis.

Human adenovirus early region 1 (E1) 1 comprises two distinct genes, E1A and E1B. Both genes cooperate efficiently to accomplish the productive infection of human cells and the transformation of rodent cells in vitro (1). The E1A gene has a function to induce apoptosis, and the E1B gene suppresses only the E1A function for apoptosis. E1A-induced apoptosis is dependent on the expression of wild-type p53 (2,3). The E1A gene of adenovirus types 2 and 5 generates two major species of mRNA with sizes of 13 S and 12 S that encode proteins of 289 amino acids (aa) (E1A 13S ) and 243 aa (E1A 12S ), respectively (4). E1A 12S has the identical amino acid sequence with E1A 13S , except for the internal 46-aa sequence unique to E1A 13S . The domain required for the induction of apoptosis has been recently mapped in the N-terminal region 5Ј to the conserved region 1 (5,6). The E1B gene also encodes two major proteins of 175 aa (19k) and 496 aa (55k) (4), both of which independently suppress apoptosis induced by E1A (5,(7)(8)(9). The roles of E1B55k and E1B19k have been ascribed to their abilities to modify p53 function. E1B55k inactivates wild-type p53 by forming a complex with it (10). Although the function of E1B19k is still unknown, it seems to be required for maintenance of the integrity of DNA, since the productive infection of adenovirus E1B19k mutants resulted in generation of an enhanced cytopathic effect and in degradation of both viral and cellular DNA (11)(12)(13).
To analyze the mechanism of apoptosis induced by E1A and its suppression by E1B19k, a derivative of human epithermoid carcinoma cell line KB, MA1, was established by introduction of the E1A 12 S cDNA linked to the hormone-inducible promoter of mouse mammary tumor virus long terminal repeat into KB cells (14). MA1 cells elicit apoptosis upon induction of E1A 12S in response to dexamethasone. MA1B17 and MA1B3 cell lines that are differently resistant to E1A-induced apoptosis, depending on the E1B19k expression levels which were also established by introduction of the E1B19k expression vector into MA1 cells (14). Analysis of the apoptotic process induced by E1A 12S by using these cell lines revealed the following. (a) The stabilization of p53 occurs shortly after E1A expression, reaching a maximal level of 10-fold higher than the original level. (b) Morphological changes characteristic of apoptosis and the fragmentation of DNA begin to be observed within 48 h. (c) Among topoisomerases I, II␣, and II␤, the level of topoisomerase II␣ decreases steeply within 36 h, preceding the onset of DNA fragmentation, while the level of its mRNA is unchanged. (d) The decrease in the topoisomerase II␣ level as well as the apoptotic process is suppressed by E1B19k, depending on its expression levels. These findings suggest that the decrease in the topoisomerase II␣ level is a key event to elicit apoptosis.
Topoisomerase II is a major component of the chromosomal matrix and is concerned with decatenation of double-stranded DNA for separation of daughter strands of DNA during repli-cation (15)(16)(17)(18). It binds selectively to matrix-attached regions of DNA (19,20) and is involved in complete condensation (21)(22)(23) and separation (24) of daughter chromosomes in mitosis and in a G 2 checkpoint that regulates entry into mitosis (25). Mammalian cells contain three topoisomerases, I, II␣, and II␤. Among these enzymes, topoisomerases I and II␤ are expressed with no cell cycle dependence, whereas the expression of topoisomerase II␣ is cell cycle-regulated, peaking in G 2 -M phase and declining to a minimal level at the end of M phase (26,27). This pattern of expression is reminiscent of the induction and degradation of cyclin B that occurs at the G 2 -M boundary and the end of metaphase, respectively (26,28,29). The degradation is mediated by the recognition of sequence, so called the destruction box of the ubiquitination system (29 -34).
In the present study, degradation of topoisomerase II␣ was investigated in the cell extracts prepared from apoptosis-induced and -uninduced MA1 cells. The results indicate that topoisomerase II␣ is polyubiquitinated and degraded preferentially in the apoptotic cell extract in an ATP-and ubiquitin-dependent manner, suggesting a component in the ubiquitin proteolysis pathway is activated or induced during the latent phase of the E1A-induced apoptosis. The proteolytic activity was much reduced in the extracts prepared from MA1-derivative cell lines expressing E1B19k or cellular anti-death gene product Bcl-2 (35)(36)(37).
Cell Culture and Viability Assay-The human epidermoid carcinoma cell line KB and its derivative cell lines, MA1, MA1B17, MA1B3, MA1C3, and MA1C22, were cultivated at 37°C in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. For the viability assay, MA1, MA1C3, and MA1C22 cells were cultivated in the presence of 1 M dexamethasone, and both floating and adherent cells were pooled at various times after treatment. Viability of the pooled cells were assessed by trypan blue exclusion. At least 1000 cells were counted for each determination using a Bü rker-Turk hemocytometer.
Plasmid Construction and Establishment of Bcl-2-expressing Cell Lines-The plasmid pB4 (35) carrying the human bcl-2␣ cDNA was cleaved at the NdeI site located 377 nucleotides downstream of the translation termination site, and the BglII linker was inserted. The DNA was then cleaved with BglII and SalI, which cleaves at 97 nucleotides upstream of the translation initiation site. The DNA fragment of 1.2 kilobases containing the bcl-2 cDNA was isolated and inserted between the SalI and BamHI sites located downstream of the human ␤-actin promoter of pH␤APr-1 (39) to generate pH␤APr⅐bcl-2.
Sparse cultures of MA1 cells were transfected with 16 g of pH␤APr⅐bcl-2 and 4 g of pSV2bsr (40) by the CaPO 4 coprecipitation procedure (1). After 48 h of transfection, the cells were replated and cultivated in the presence of 2.5 g/ml blasticidin S-hydrochloride. Blasticidin S-hydrochloride-resistant colonies developed were isolated and cultivated in four sets of 24-well culture plates. Two sets of cultures were incubated in the presence and absence of 1 M dexamethasone for 72 h and were used for microscopic examination to isolate clones refractory to apoptosis. The other two sets of cultures were used for preparation of RNA, and E1A and bcl-2 mRNAs synthesized in blasticidin S-hydrochloride-resistant clones in the presence and absence of dexamethasone were analyzed by RNA dot hybridization (41). Two clones designated MA1C3 and MA1C22 showing different degrees of suppression were selected.
Preparation of Cell Extracts-Crude cell extract (S10) was prepared essentially as described by Hegde et al. (42) with the following modifications. Subconfluent cultures of MA1 cells, treated or untreated with 1 M dexamethasone for 42 h, were washed twice in ice-cold phosphatebuffered saline (PBS; 0.14 M NaCl and 0.01 M potassium phosphate, pH 7.4), once in 10 volumes of ice-cold hypotonic buffer (20 mM Tris⅐HCl, pH 7.4, 5 mM MgCl 2 , 8 mM KCl, and 1 mM dithiothreitol (DTT)), and then resuspend in 10 volumes of the hypotonic buffer. After incubation on ice for 15 min, the swollen cells were collected in a 7.5-ml Dounce vessel and precipitated by centrifugation at 200 ϫ g for 2 min at 4°C to remove excess buffer. The cells were then disrupted by homogenization with a tight pestle for 40 times on ice. The homogenate was centrifuged twice at 10,000 ϫ g for 5 min, and the turbid supernatant (S10 extract) was collected. Aliquots of the S10 extract was centrifuged at 105,000 ϫ g for 6 h in the presence of 2 mM ATP to prepare the S100 extract, which lacks most of the proteasome activity. Protein concentration was determined by a dye-binding assay (43).
The precipitate fraction enriched with the 26 S proteasome was prepared essentially according to Ugai et al. (44). Briefly, 4 ml of the S10 extract containing about 20 -23 g/l of protein were centrifuged at Immunoprecipitation of Topoisomerase II␣-MA1 cells (approximately 1 ϫ 10 8 cells) were washed twice with ice-cold PBS and lysed for 30 min with 5 ml of immunoprecipitate lysis buffer (50 mM Hepes⅐KOH, pH 7.0, 1 mM EDTA, 420 mM NaCl, and 0.1% Nonidet P-40) containing protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml leupeptin, and 5 M calpain inhibitor I). The lysate was centrifuged at 12,000 ϫ g for 10 min to remove cell debris. The lysate was then precleared with 50 l of normal rabbit serum and 1 ml of 20% Staphylococcus aureus Cowan I suspension (45), and topoisomerase II␣ was immunoprecipitated with 400 l of mouse anti-human topoisomerase II␣ antibody (clone 8D2 hybridoma conditioned medium) and protein A-Sepharose FF beads (1 ml of 10% suspension). The immunoprecipitate was washed three times with immunoprecipitate lysis buffer containing protease inhibitors, twice with immunoprecipitate lysis buffer not containing protease inhibitors, resuspended in hypotonic buffer to give a 10% suspension, and stored at 4°C. The immunoprecipitate was resuspended in hypotonic buffer to a final concentration of 5% just before the degradation assay. Topoisomerase II␤ was similarly prepared from MA1 cells by immunoprecipitation with mouse anti-human topoisomerase II␤ antibody (clone 5A7 hybridoma conditioned medium).
Preparation of the Nuclear Matrix-The nuclear matrix was prepared essentially as described by Mirkovitch et al. (46) with slight modifications. MA1 cells (approximately 1 ϫ 10 8 cells) were washed twice with ice-cold PBS, once with NB (10 mM Pipes⅐KOH, pH 7.4, 10 mM KCl, 2 mM MgCl 2 , 1 mM DTT, 10 M cytochalasin B, 0.1 mM phenylmethylsulfonyl fluoride, plus 1 g/ml protease inhibitors (chymostatin, leupeptin, aprotinin, and pepstatin A)) (47), and collected in a 7.5-ml Dounce homogenizer vessel. The cells were resuspended in 10 volumes of NB and kept on ice for 20 min. The swollen cells were gently disrupted with a loose pestle by 20 strokes, and the resulting homogenate was layered over 30% sucrose in NB. The nuclei were pelleted by centrifugation at 800 ϫ g for 10 min, washed once with NB, resuspended in 10 volumes of NB, and then layered over 30% sucrose in NB and pelleted by centrifugation again. The precipitate of nuclei was resuspended in 2 ml of extraction buffer (3.75 mM Tris⅐HCl, pH 7.4, 0.05 mM spermine, 0.125 mM spermidine, 20 mM KCl, 1% (v/v) thiodiglycol, and 0.1% digitonin) and incubated at 42°C for 20 min. After incubation, 40 ml of low-salt extraction buffer (5 mM Hepes⅐NaOH, pH 7.4, 0.25 mM spermidine, 2 mM EDTA⅐KOH, pH 7.4, 2 mM KCl, 0.1% digitonin, and 25 mM 3,5-diiodosalicylic acid lithium salt) were slowly added at room temperature, and the suspension was incubated for 5 min. The histonedepleted nuclei were recovered by centrifugation at 2400 ϫ g for 20 min at room temperature and washed twice with 10 ml of digestion buffer (20 mM Tris⅐HCl, pH 7.4, 0.05 mM spermine, 0.125 mM spermidine, 20 mM KCl, 70 mM NaCl, 10 mM MgCl 2 , 2 mM CaCl 2 , 0.1% digitonin, 0.1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, and 10 g/ml leupeptin). Nine ml of DNase I in digestion buffer were then added at 44 units/ml, and digestion was allowed to proceed for 30 min at room temperature in a shaking water bath. After solubilization of DNA, the nuclear matrix was sedimented at 2400 ϫ g for 10 min at 4°C, washed four times with 10 ml of hypotonic buffer, and then resuspended in 500 l of hypotonic buffer.
Preparation of GST-Ub Fusion Protein-Preparation of GST-tagged ubiquitin protein (GST-Ub) was described in detail elsewhere. 2 Briefly, Arabidopsis thaliana ubiquitin cDNA was prepared by standard polymerase chain reaction protocol (48), and the resulting cDNA fragment was subcloned between BamHI and EcoRI sites of pGEX-2TK (49) located downstream of Schistosoma japonicum GST cDNA (38).
The expression of GST-Ub in Escherichia coli and purification on glutathione-Sepharose were performed as described by Kaelin et al. (50). All the steps were carried out at 4°C. Crude extract containing GST-Ub fusion protein prepared from 3 ml of E. coli precipitate was mixed and rocked for 1 h with 300 l of glutathione-Sepharose beads that was previously equilibrated with NETN buffer (20 mM Tris⅐HCl, pH 8.0, 100 mM NaCl, 1 mM EDTA, and 0.5% Nonidet P-40). The Sepharose beads (300 l) were then stuffed in a 1-ml disposable syringe after washing five times with 10 volumes of NETN buffer containing 1 mM phenylmethylsulfonyl fluoride and washed again with 20 volumes of NETN buffer. The fusion protein was eluted with eight volumes of 50 mM Tris⅐HCl, pH 8.0, containing 5 mM glutathione and precipitated by the addition of ammonium sulfate at 90% saturation. The precipitate was recovered by centrifugation at 15,000 ϫ g for 10 min and dissolved in 10 mM Tris⅐HCl, pH 7.6 (1/40 volume of the initial eluate), and then dialyzed twice against 1 liter of the same buffer. Solubilized GST-Ub fraction was stored at Ϫ80°C.
Degradation Assay-The reaction mixture (60 l) contained 300 g of protein of the cell extract, 5 l of immunoprecipitated topoisomerase II␣ or the nuclear matrix, 40 mM Tris⅐HCl, pH 7.6, 5 mM MgCl 2 , and 2 mM DTT. Where indicated, 2 mM ATP and/or 10 g of ubiquitin were added to the reaction mixture. ATP was added together with ATP regenerating system consisting of 10 mM creatine phosphate and 25 units/ml creatine phosphokinase. All reactions were carried out at 30°C and terminated by cooling on ice for 5 min and by boiling for 5 min immediately after the addition of 1/5 volume of 6 ϫ reduction buffer (360 mM Tris⅐HCl, pH 6.8, 12% SDS, 600 mM DTT, 60% glycerol, and 0.1% bromphenol blue), and the reaction mixture was subjected to electrophoresis. For detection of polyubiquitinated topoisomerase II␣, the reaction mixture was centrifuged at 8000 ϫ g at 4°C for 2 min after the termination of the reaction. The precipitates was washed in 1 ml of ice-cold immunoprecipitate lysis buffer for four times and boiled for 5 min after the addition of 20 l of 2 ϫ Laemmli sample buffer (45). The products were analyzed by Western blotting.
Western Blotting-Fifty g of protein in SDS-solubilized whole-cell extracts (14) were electrophoresed on 7-20% polyacrylamide gels with Laemmli running buffer (25 mM Tris⅐glycine, pH 8.3, and 0.1% SDS) as described by Harlow and Lane (45). Proteins were electrophoretically transferred to nitrocellulose membranes (BA85; Schleicher & Schuell) and incubated in immunoblotting diluent solution (5% skim milk (Snow brand, Sapporo, Japan), 100 ng/ml of goat IgG, and 0.1% Tween 20 in PBS) at room temperature for 1 h to minimize nonspecific binding of antibody. The membrane was incubated with primary antibody at an appropriate dilution as indicated in figure legends at room temperature for 1 h and washed three times in PBS containing 0.1% Tween 20 for 15 min. The membrane was then incubated with secondary antibody at dilution of 1:10,000 at room temperature for 1 h and washed three times in PBS containing 0.1% Tween 20 for 15 min. Immune complexes were detected by enhanced chemiluminescence (ECL) by treating the membrane with the ECL detection system according to the manufacturer's protocol (Amersham Corp.) and then exposed to x-ray film (Fuji-RX; Fuji-film, Tokyo, Japan).

RESULTS
Bcl-2 Suppresses the Decrease in the Topoisomerase II␣ Level during E1A-induced Apoptosis-The MA1 cell line, established from human epithermoid carcinoma KB cells by introducing the MMTV⅐LTR-linked E1A 12 S cDNA, elicits apoptosis upon induction of E1A 12S in response to dexamethasone. During the latent phase of apoptosis, the level of topoisomerase II␣ begins to decrease steeply, reaching a minimum of 1/50 of the original level. The decrease as well as the progression of apoptosis is inhibited in MA1-derivative cell lines expressing E1B19k, depending on its expression levels (14). It has been shown that E1A-induced apoptosis is also suppressed by enforced expression of Bcl-2, as is E1B19k (7). To see whether the decrease in the topoisomerase II␣ level in apoptotic MA1 cells is suppressed by Bcl-2, MA1 derivative cell lines expressing the exogenous bcl-2 gene were established by transfection of MA1 cells with pSV2bsr and pH␤APr-bcl2, which expresses the human bcl-2 cDNA constitutively from the human ␤-actin promoter (39). Blasticidin-resistant colonies that developed were transferred to 24-well tissue culture plates, and the clones refractory to E1A-induced apoptosis were screened after the addition of dexamethasone. Two clones designated MA1C3 and MA1C22 were selected, and the levels of Bcl-2 and topoisomerase II␣ were analyzed by Western blotting using anti-Bcl-2 and anti-topoisomerase II␣ antibodies. As shown in Fig. 1A, MA1 cells began to lose their viability within 48 h after treatment with dexamethasone, and only 15-20% of the original cell population was viable at 72 h. In constant, MA1C3 and MA1C22 cells lost their viability at a slower rate, and approximately 70 and 30%, respectively, of the cells were viable at 72 h. Reflecting the difference in the extents of the viability loss, the expression level of Bcl-2 in MA1C3 cells was near 10-fold higher than that in MA1C22 cells (Fig. 1B). The decrease in the level of topoisomerase II␣ was also reduced by Bcl-2 expression (Fig.  1C).  1. Bcl-2 suppresses E1A-induced apoptosis and the decrease in the level of topoisomerase II␣. In A, cells were seeded at 1 ϫ 10 5 cells/9-cm dish and cultivated at 37°C. The cells were treated with 1 M dexamethasone and harvested by trypsinization at the times indicated. Viable cell numbers were determined by trypan blue exclusion. In B and C, the cells were lysed in 2 ϫ Laemmli sample buffer at the times indicated. Aliquots of 50 g of protein per lane were electrophoresed on 12 and 7.5% polyacrylamide gels, and the amounts of Bcl-2 (B) and topoisomerase II␣ (C), respectively, were analyzed by Western blotting. Mouse anti-human-Bcl-2 monoclonal antibody (124; Dako) and mouse anti-human topoisomerase II␣ monoclonal antibody (8D2; hybridoma conditioned medium) were used as primary antibodies at dilutions of 1:2000 and 1:200. HRP-conjugated goat anti-mouse Ig antibody was used as a secondary antibody at a dilution of 1:10,000. The filters were treated with the ECL detection system and exposed to x-ray film for 30 min and 2 s, respectively. throughout the apoptotic process (data not shown). These results suggest that Bcl-2 protects the breakdown of topoisomerase II␣, depending on its expression levels, as does E1B19k.
Degradation of Topoisomerase II␣ Is ATP-and Ubiquitin-dependent-Topoisomerase II␣ is a component of the nuclear matrix as well as being a major component of the metaphase chromosome matrix. It is involved with condensation and separation of daughter chromosomal DNA during mitosis. In contrast to topoisomerases I and II␤, which are expressed at constant levels throughout the cell cycle, the expression of topoisomerase II␣ is cell cycle-regulated, peaking in G 2 -M phase and declining to a minimal level at the end of M phase. This transient induction is reminiscent of ubiquitin-dependent degradation of cyclin B during M phase. The cyclin B degradation is mediated by recognition of the sequence, RXXLXYIXN (29), so called the destruction box, by the ubiquitin-dependent proteolysis system (51). Computer search for the presence of the destruction box in human topoisomerase II␣ revealed the sequence RKQLISIWN in the N terminus region (14). To determine whether this sequence is also present in the topoisomerase II␣ cDNA, made from MA1 cell mRNA, the cDNA was sequenced, after amplification by PCR. The result showed that three amino acids at the N terminus side of the box were substituted by ENN. Whether the sequence ENNLISIWN (from amino acid residues 112 to 120) functions as a recognition signal for ubiquitin-mediated degradation is presently unknown. It has been recently shown that besides the destruction box, the other motifs are also recognized by the ubiquitin system, e.g. c-MOS is recognized via Pro-2 and Ser-3 (52) and c-Jun by the ␦ domain (53). The PEST sequence is also suspected as a recognition signal (54).
To characterize the degradation of topoisomerase II␣ in an in vitro system, the S10 extract was prepared from MA1 cells after treatment with dexamethasone for 42 h (the 42-h extract), when the level of topoisomerase II␣ had already decreased to near minimal level. The control extract was prepared from untreated MA1 cells (the 0-h extract). Topoisomerase II␣ was prepared by either immunoprecipitation or extraction of the nuclear matrix fraction from MA1 cells. As a control substrate, topoisomerase II␤ was prepared from MA1 cells by immunoprecipitation. The reaction for degradation of topoisomerase II␣ was performed by incubation of immunoprecipitated topoisomerase II␣ or the nuclear matrix in the 0-and 42-h extracts at 30°C for 1.5 h in the presence and absence of ATP and ubiquitin. The reaction was terminated by the addition of 1/5 volume of 6 ϫ reduction buffer, and aliquots of the samples were immediately subjected to electrophoresis on SDS-polyacrylamide gels. Quantitation of topoisomerase II␣ was performed by enhanced chemiluminescence (ECL) system after transfer to the nitrocellulose filter as described previously (14).
As shown in Fig. 2A topoisomerase II␣ prepared by immunoprecipitation was degraded completely in the 42-h extract, when both ATP and ubiquitin were added (lane 12). The addition of ubiquitin alone also resulted in the degradation of topoisomerase II␣ nearly completely (Fig. 2A, lane 11). This degradation was likely to be caused by residual ATP present in the extract, since the addition of apyrase that destroys ATP completely protected topoisomerase II␣ from degradation ( Fig.  2A, lane 13). A slight degradation was observed in the presence of ATP alone (Fig. 2A, lane 9). A significant but much less amount of immunoprecipitated topoisomerase II␣ was degraded in the 0-h extract in the presence of both ATP and ubiquitin ( Fig. 2A, lane 6), presumably reflecting the presence of mitotic cells in the population. The addition of ATP or ubiquitin alone ( Fig. 2A, lanes 3 and 5) had little effect, suggesting that the amounts of ATP and ubiquitin present in the extract were insufficient if any.
Topoisomerase II␣ associated with the nuclear matrix was less sensitive to degradation in the 42-h extract but degraded nearly completely in the presence of both ATP and ubiquitin (Fig. 2B, lane 9). The absence of either ATP or ubiquitin reduced the extent of degradation significantly (Fig. 2B, lanes 7  and 8). Topoisomerase II␣ was not degraded efficiently in the 0-h extract, even in the presence of both ATP and ubiquitin (Fig. 2B, lane 5). In contrast to topoisomerase II␣, topoisomerase II␤ was not degraded significantly in the 42-h extract in the presence of both ATP and ubiquitin (Fig. 2C). These results suggest that a component(s) in the ubiquitin proteolytic pathway required for degradation of topoisomerase II␣ is induced or activated during the latent phase of E1A-induced apoptosis. Western blotting was performed as described in Fig. 1. Lane 1, an initial amount of topoisomerase II␣ added to the reaction mixture. B, degradation of topoisomerase II␣ associated with the nuclear matrix. The nuclear matrix prepared from about 10 6 MA1 cells was incubated in the extracts as described in A. C, insensitivity of topoisomerase II␤ to degradation. Aliquots of immunoprecipitated topoisomerase II␤ were incubated in the 42-h extract as described in A. Detection of topoisomerase II␤ was performed with mouse anti-human topoisomerase II␤ monoclonal antibody (clone 5A7; hybridoma conditioned medium) as a primary antibody at a dilution of 1:200. indicated in Fig. 3A. Calpain inhibitor I and carbobenzoxy-Lleucyl-L-leucyl-L-norvalinal, which are known as proteasome inhibitors (55,56), suppressed the degradation of topoisomerase II␣ nearly completely, whereas leupeptin, E-64, the tetrapeptide aldehyde N-acetyl-L-tyrosyl-L-valyl-L-alanyl-L-aspart-aldehyde, and calpain inhibitor II, which inhibit serine protease, cysteine protease, interleukin 1␤-converting enzyme and calpains, respectively, had little or no effect. Although calpain inhibitor II is a tripeptide-aldehyde, it has been shown to have a little effect on the proteasome activity (57). The addition of dimethyl sulfoxide, the solvent of these inhibitors, had no effect. These results indicate that the degradation of topoisomerase II␣ is catalyzed by the proteasome.

Degradation of Topoisomerase II␣ Is Suppressed by Proteasome Inhibitors and in the Extracts from MA1 Derivative Cell
MA1-derivative cell lines expressing either E1B19k or Bcl-2 suppress the decrease in the topoisomerase II␣ level as well as the apoptotic process induced by E1A 12S , depending on their expression levels. If the in vitro cell free system for degradation of topoisomerase II␣ actually reflects the decrease in the topoisomerase II␣ level in the apoptotic cells in vivo, the extracts prepared from these MA1-derivative cell lines must have reduced activities to degrade topoisomerase II␣. The MA1 derivative cell lines MA1B17 and MA1B3 constitutively express E1B19k. The former expresses a high level of E1B19k and is considerably resistant to both E1A-induced apoptosis and the decrease in the topoisomerase II␣ level, whereas the latter expressing a low level of E1B19k is weakly resistant to both changes (14). MA1C3 and MA1C22 cells expressing high and low levels of bcl-2, respectively, showed similar patterns of resistance, depending on their expression levels (Fig. 1). The 42-h extract was prepared from these cells, and the degradation reaction was performed in the presence of ATP and ubiquitin at 30°C for 1 h. As shown in Fig. 3B, degradation of topoisomerase II␣ was much reduced in the extract from MA1B17 cells than in the extract from MA1B3 cells (Fig. 3B,  lanes 4 and 5), reflecting the difference in their expression levels of E1B19k. Although the degradation activities were reduced in the extracts from Bcl-2-expressing cells as compared with that observed in the extracts from E1B19k-expressing cells, degradation of topoisomerase II␣ was reduced more in the MA1C3 cell extract than in the MA1C22 cell extract (Fig. 3B,  lanes 6 and 7). These results suggest that the in vitro degradation assay of topoisomerase II␣ reflects the extent of the decrease in the topoisomerase II␣ level in vivo, and that the expression of E1B19k or bcl-2 in MA1 cells inhibits activation or activity of a component(s) required for topoisomerase II␣ degradation in the ubiquitin proteolysis pathway.
Polyubiquitination and Proteasome-dependent Degradation of Topoisomerase II␣-To see whether the degradation of topoisomerase II␣ is proceeded via polyubiquitination that is characteristic of the ubiquitin-dependent proteolysis pathway, S10 and S100 extracts were prepared from MA1 cells treated with dexamethasone for 42 h. The former extract contains both ubiquitination enzymes and proteasomes, whereas the latter lacks proteasomes. The immunoprecipitated topoisomerase II␣ was incubated in both extracts in the presence and absence of ATP and ubiquitin at 30°C for 1 h and was immediately subjected to electrophoresis. Western blotting revealed that degradation of topoisomerase II␣ occurred efficiently in the S10 extract but not in the S100 extract in the presence of ATP and ubiquitin, when the filter was exposed for a short time period (Fig. 4A, lanes 4 and 5). When the filter was exposed for a longer time period, slow migrating bands became visible above the unprocessed topoisomerase II␣ thick band of M r 170,000. These species of topoisomerase II␣ were formed much more in the S100 extract than in the S10 extract, suggesting that these species are polyubiquitinated topoisomerase II␣ escaped from degradation in the S100 extract, which lacks most of the proteasome activity. The formation of these slow-migrating bands was almost completely abolished in the S100 extract in the absence of added ATP and ubiquitin (Fig. 4A, lane 3). To confirm polyubiquitination of topoisomerase II␣, GST-Ub (58) was prepared and used for the formation of GST-Ub conjugates of topoisomerase II␣. Immunoprecipitated topoisomerase II␣ was incubated briefly at 30°C for 15 min in the S10 extracts prepared from MA1 cells (0 h) and MA1 cells treated with dexamethasone for 42 h (42 h), in the presence of ATP and GST-Ub. The products were similarly analyzed by Western blotting with antibodies specific to topoisomerase II␣ (T) and GST (G) (Fig.  4B). Anti-GST antibody revealed GST-Ub conjugates of topoisomerase II␣ in both 0-h and 42-h extracts (Fig. 4B, lanes 6 and  10). The amount of GST-Ub-topoisomerase II␣ conjugates formed was four to five times greater in the 42-h extract than in the 0-h extract. This difference was correlated with the extent of topoisomerase II␣ degradation. The amount of topoisomerase II␣ that decreased in the 0-h extract (Fig. 4B, lane 5) was very little, whereas the amount decreased to about onehalf in the 42-h extract (Fig. 4B, lane 9) during the brief incubation. The ubiquitination and degradation of topoisomerase II␣ were not observed in both 0-h (Fig. 4B, lanes 4 and 3) and 42-h (Fig. 4B, lanes 8 and 7) extracts, when GST-Ub was replaced by GST. These results indicate that the degradation of topoisomerase II␣ proceeds via polyubiquitination and that the activity of the ubiquitin conjugation to topoisomerase II␣ is elevated during the latent phase of the E1A-induced apoptosis.
The inability of the S100 extract to degrade topoisomerase II␣ efficiently suggests that the reaction requires a high molecular weight protease(s) such as 26 S proteasome. The enzyme is an ATP-dependent form of 20 S proteasome and the major component of the ubiquitin system. To demonstrate the requirement of the proteasome for degradation of topoisomerase II␣, the 0-h and 42-h extracts (S10) were fractionated into the precipitate (designated as P70) and the supernatant (des- ignated as S70) fractions by ultracentrifugation at 70,100 ϫ g as illustrated in Fig. 5. The proteolytic activity for topoisomerase II␣ present in the 0-h and 42-h S10 extracts (Fig. 5, lanes  2 and 6) was lost nearly completely after fractionation into S70 and P70 fractions (Fig. 5, lanes 3, 4, 7, and 8). The activity was partially recovered when S70 and P70 fractions were combined (Fig. 5, lanes 5 and 9). The combination of 0-h S70 and 42-h P70, and 0-h P70 and 42-h S70, had little effect on the recovery of the proteolytic activity. The result suggests that the final step of the degradation of topoisomerase II␣ is catalyzed by the 26 S proteasome.

DISCUSSION
The present study showed that topoisomerase II␣ prepared by either immunoprecipitation or extraction of the nuclear matrix is degraded efficiently in the apoptotic MA1 cell extract prepared after induction of E1A 12S expression. The following features of degradation suggest that a component(s) in the ubiquitin proteolysis pathway, responsible for degradation of topoisomerase II␣, is activated or induced during E1A-induced apoptosis. (a) The extract prepared from MA1 cells treated with dexamethasone for 42 h (the 42-h extract) had much stronger proteolytic activity for topoisomerase II␣ than that in the untreated cell extract (the 0-h extract). Under the same conditions, topoisomerase II␤ was unaffected in both extracts. (b) The degradation in these extracts was dependent on both ATP and ubiquitin. A little but significant degradation occurred in the 0-h extract. This residual activity, present in untreated MA1 cells, may be required for the cell cycle-regulated expression of topoisomerase II␣, which lowers the level at the end of M phase (27). (c) The degradation was inhibited by calpain inhibitor I and carbobenzoxy-L-leucyl-L-leucyl-L-norvalinal, which are known as proteasome inhibitors. Other protease inhibitors including N-acetyl-L-tyrosyl-L-valyl-L-alanyl-L-aspart-aldehyde for interleukin 1␤-converting enzyme, had little or no effect. (d) The proteolytic activity was much reduced in the 42-h extracts prepared from MA1 derivative cell lines expressing E1B19k or Bcl-2, depending on their expression levels, suggesting that these suppressors of apoptosis inhibit the activation or activity of a component(s) responsible for degradation of topoisomerase II␣.
The ubiquitin-mediated proteolysis of target proteins is ini- FIG. 4. Polyubiquitination of topoisomerase II␣. A, aliquots of immunoprecipitated topoisomerase II␣ were incubated in the 42-h extract, in the presence or absence of 2 mM ATP and 10 g ubiquitin at 30°C for 1 h, and subjected to electrophoresis. The products were analyzed by Western blotting. The filter was first exposed for a short time period to see the degradation of topoisomerase II␣ and then exposed for a longer time period to see the slower migrating species of topoisomerase II␣; lane 1, input immunoprecipitate; lanes 2 and 4, the immunoprecipitate was incubated in the 42-h extract (S10) in the presence and absence of ATP and ubiquitin, respectively; lanes 3 and 5, the immunoprecipitate was incubated in S100 prepared from the 42-h extract in the presence and absence of ATP and ubiquitin, respectively. B, aliquots of immunoprecipitated topoisomerase II␣ were incubated at 30°C for 15 min in the 0-h and 42-h extracts in the presence of 2 mM ATP and 10 g of GST-Ub or GST as indicated above each lane. After electrophoresis, the products were analyzed by Western blotting. The filter was first incubated with mouse anti-topoisomerase II␣ monoclonal antibody 8D2 (T) and rabbit anti-GST antibody Z-5 (G) at dilutions of 1:200 and 1:2,000, respectively, and then with HRP-conjugated goat anti-mouse Ig antibody at a dilution of 1:10,000. The filter was treated with the ECL detection system and exposed to x-ray film to monitor the degradation levels of topoisomerase II␣ (lanes 1, 3, 5, 7, and  9). The filter was then treated with 0.1% sodium azide to inactivate HRP and incubated with HRP-conjugated goat anti-rabbit Ig antibody at dilution of 1:10,000. The filter was treated with the ECL detection system and exposed to x-ray film again to detect the GST-Ub conjugates of topoisomerase II␣ (lanes 2, 4, 6, 8, and 10). Lanes 1 and 2, input immunoprecipitate (c).  5 and 10). The products were analyzed by Western blotting as stated. tiated by activation of ubiquitin by an ATP-hydrolyzing enzyme, E1, to which ubiquitin binds through a thiol ester bond. The activated ubiquitin is then transferred to one of a member of ubiquitin-conjugating enzymes, E2 n , that catalyze the polyubiquitination of target proteins at its specific lysine residue. This ubiquitination process usually requires a member of ubiquitin ligase, E3 n , which specifically binds to target proteins. Polyubiquitinated protein is then degraded by the ATP-dependent proteasome (59). Polyubiquitination of topoisomerase II␣ was well demonstrated in the S70 or S100 extracts prepared from the apoptotic MA1 cells, which are devoid of most of the proteasome activity. The enhancement of the ubiquitinconjugating activity in the 42-h extract as compared with the 0-h extract suggests that a component of the E2 and/or E3 families, responsible for ubiquitination of topoisomerase II␣, is activated or induced during E1A-induced apoptosis. The sequencing of the topoisomerase II␣ cDNA made from MA1 cell mRNA revealed the destruction-like box ENNLISIWN, which differs only one amino acid residue from the consensus sequence RXXLXYIXN (29) in the N-terminal side at positions from 112 to 120. A preliminary experiment for determination of the ubiquitinated site using the topoisomerase II␣ fragments prepared by in vitro transcription and translation suggested that the site of ubiquitination may reside in the N-terminal portion (data not shown). However, the ubiquitination could be detected with the N-terminal fragment, NRT (amino acid sequence from 1 to 198), but not with the N-terminal fragment NB (amino acid sequence from 1 to 711). In addition, the ubiquitin conjugates of the NRT fragment consisted of only monomer and dimer, and the polyubiquitinated forms could not be detected. These results suggest that the entire conformation of topoisomerase II␣ is required for polyubiquitination by E2 and E3 enzymes. The precise site of polyubiquitination and whether the sequence ENNLISIWN can be recognized by the ubiquitination system, therefore, remain to be solved.
Oncoproteins have been shown to have a function to modify the ubiquitin proteolysis pathway. Human papilloma viruses HPV-16-and HPV-18-derived E6 proteins stimulate degradation of p53 in crude reticulocyte lysates (60). E6-associated protein E6-AP functions as a ubiquitin-protein ligase to conjugate ubiquitin to p53 (58), along with a novel E2 enzyme (61). The stabilization of p53 during E1A-induced apoptosis (14) may also occur by modifying the function of a component in the ubiquitin proteolysis pathway by E1A, since the level of p53 mRNA was unchanged throughout the process. E1A may, therefore, act in a reverse way to modify the ubiquitin pathway for degradation of topoisomerase II␣.
The degradation of topoisomerase II␣ may alter the organization of the chromatin structure, since topoisomerase II␣ has been shown to be a component of the nuclear matrix (15,62) and binds selectively to the scaffold-attached region of chromatin DNA (19,46). Detachment of the chromatin loops from their anchorage sites may result in the abnormal condensation of chromatin, allowing the access of endonuclease. Similar alteration in the chromatin structure has been suggested during apoptosis in a variety of cell types in which proteases of the interleukin 1␤-converting enzyme/ced3 family are activated. One of the substrates cleaved early in the apoptotic process by a member of these enzymes is nuclear lamin (63)(64)(65). Lamins are components of the intermediate filament meshwork that play an important role in maintaining nuclear envelope integrity and in the organization of interphase chromatin (66). Lamins also bind to chromatin DNA, preferentially to the scaffold attachment region (67)(68)(69). The interaction between lamins and chromatin around the nuclear periphery is, therefore, destroyed after cleavage of lamins. The condensation of chromatin DNA and DNA strand breaks begins to be observed after breakdown of nuclear lamina by interleukin 1␤-converting enzyme-related proteases (63)(64)(65). The digestion of other nuclear proteins, poly(ADP-ribose) polymerase, topoisomerases I and II, and histone H1 has also been reported in an early stage of apoptosis (66, 70 -72). The nuclear matrix proteins such as topoisomerase II␣ and lamins seem to be the targets for specific proteases for execution of apoptosis.