26 S proteasome-mediated degradation of topoisomerase II cleavable complexes.

DNA topoisomerase II (TOP2) cleavable complexes represent an unusual type of DNA damage characterized by reversible TOP2-DNA cross-links and DNA double strand breaks. Many antitumor drugs and physiological stresses are known to induce TOP2 cleavable complexes leading to apoptotic cell death and genomic instability. However, the molecular mechanism(s) for repair of TOP2 cleavable complexes remains unclear. In the current studies, we show that TOP2 cleavable complexes induced by the prototypic TOP2 poison VM-26 are proteolytically degraded by the ubiquitin/26 S proteasome pathway. Surprisingly the TOP2beta isozyme is preferentially degraded over TOP2alpha isozyme. In addition, transcription inhibitors such as 5,6-dichlorobenzimidazole riboside and camptothecin can substantially block VM-26-induced TOP2beta degradation. These results are consistent with a model in which the repair of TOP2beta cleavable complexes may involve transcription-dependent proteolysis of TOP2beta to reveal the protein-concealed double strand breaks.

such as VP-16 (etoposide) and doxorubicin (11)(12)(13). In the presence of these TOP2-directed drugs (TOP2 poisons), TOP2 isozymes are trapped as their covalent reaction intermediates, the reversible TOP2 cleavable complexes in which each TOP2 subunit is covalently linked to the 5Ј-phosphoryl ends of the four-base staggered double strand breaks (14,15). While the double strand breaks within the TOP2 cleavable complexes are normally concealed by TOP2, many of the cellular effects of TOP2 cleavable complexes are clearly indicative of DNA damage. For example, TOP2 cleavable complexes induced by TOP2 poisons are known to induce DNA damage responses (e.g. G 2 arrest, elevation of sister-chromatid exchanges, NFB activation, and p53 stabilization) (16 -19). DNA repair mutant cells (e.g. ataxia telangiectasia, progeroid Werner's syndrome, and Rad52) are also known to be hypersensitive to TOP2 poisons (20 -22). However, how TOP2-concealed DNA strand breaks are converted to DNA damage signals is still unknown. Inhibitor studies have suggested that both DNA replication and RNA transcription may be important for processing TOP2 cleavable complexes into DNA damage signals (23)(24)(25).
Repair of topoisomerase cleavable complexes is conceptually challenging because of the bulkiness of the protein and the concealed nature of the breaks. However, recent studies on TOP1 cleavable complexes have suggested that both SUMO and ubiquitin pathways may be involved in repair of TOP1 cleavable complexes (26,27). While the role of SUMO-1 conjugation to TOP1 cleavable complexes is still unclear, the role of ubiquitin conjugation to TOP1 cleavable complexes appears to trigger degradation of TOP1 via the 26 S proteasome pathway (27). Proteolytic degradation of TOP1 cleavable complexes removes the protein bulk and presumably reveals the hidden strand breaks so that the normal DNA repair process can occur (27,28). Interestingly transcription inhibitors have been shown to block TOP1 degradation suggesting the involvement of RNA transcription in this particular repair process. 2 In the current study, we show that TOP2 cleavable complexes can also trigger ubiquitin conjugation to TOP2 resulting in 26 S proteasome-mediated degradation of TOP2. Surprisingly TOP2␤ is preferentially degraded over TOP2␣. In addition, transcription inhibitors can substantially block TOP2␤ degradation. These results are consistent with a model in which repair of TOP2␤ cleavable complexes may involve transcription-dependent proteolysis of TOP2␤ to reveal proteinconcealed double strand breaks.

EXPERIMENTAL PROCEDURES
Materials-ICRF-193 was purchased from ICN Biomedicals. VM-26 was kindly provided by Bristol Myers Squibb Co. Aphidicolin, cyclohex-* This work was supported by National Institutes of Health Grants GM27731 and CA 39662, Grant NSC 89-2320-B001-075 from the National Science Council of Taiwan, and Academia Sinica. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
imide, and 5,6-dichlorobenzimidazole riboside (DRB) were purchased from Sigma. Z-DEVD-FMK was purchased from Calbiochem. Staphylococcal S7 nuclease was purchased from Roche Molecular Biochemicals. Antiserum against hTOP1 was obtained from scleroderma patients. Rabbit antiserum against hTOP2␣ was raised against the C-terminal one-third of hTOP2␣ (29). The anti-human TOP2␤ antibodies was raised by immunization with an immunogen containing GST and seven linear repeats of the peptide fragment of human TOP2␤ from amino acid residues 1554 to 1565 (TOP2␤-(1554 -1565)). The construction of the DNA fragment encoding for seven repeats of TOP2␤-(1554 -1565) and the synthesis of the immunogen GST-TOP2␤-(1554 -1565) followed the published procedure (30). Briefly, the template-repeat polymerase chain reaction method was applied to the construct DNA fragment encoding multiple copies of TOP2␤-(1554 -1565). We designed two oligonucleotides, oligo A and oligo B. Oligo A, 5Ј-AAT GAA GGC GAT TAT AAC CCT GGC AGG AAA ACA TCC, encodes the target antigen (TOP2␤-(1554 -1565)), and oligo B, 5Ј-GTT ATA ATC GCC TTC ATT GGA TGT TTT CCT GCC AGG, is partly complementary to oligo A. To incorporate restriction sites for subcloning at both ends of the template-repeat polymerase chain reaction products (BamHI at the 5Ј-end and EcoRI at the 3Ј-end) as well as a stop codon at the 3Ј-end of the coding region, a second round of polymerase chain reaction (adapter polymerase chain reaction) with two adapter primers, primer A (5Ј-G ATC GGA TCC CCG GGA AAT GAA GGC GAT TAT AAC) and primer B (5Ј-A GCT TCT AGA ATT CTA GGA TGT TTT CCT GCC AGG) was performed. The DNA fragment encoding the seven repeats of TOP2␤-(1554 -1565) was subcloned into plasmid pGST-KG at the 3Ј-end of GST DNA. The resulting plasmid, pGST-TOP2␤-(1554 -1565), was introduced into XL-10 Gold, and the expressed fusion protein (GST-TOP2␤-(1554 -1565)) was purified by glutathione-Sepharose 4B affinity chromatography.
Cell Culture-The mouse mammary carcinoma cell line ts85 (temperature-sensitive for the ubiquitin-activating enzyme, E1) (31) was cultured in a humidified atmosphere of 5% CO 2 at 30°C in Dulbecco's minimum essential medium containing penicillin-streptomycin and 10% fetal bovine serum. FM3A, the parental cell line of ts85, was cultured under identical conditions at 37°C. Cells were shifted to the restrictive temperature by transferring the culture dishes to a 42°C incubator for 15 min and then maintaining at 39°C. HeLa, human breast cancer cell ZR75-1, human lung fibroblast WI-38 and its transformed subline 2RA, and leukemic CEM and U937 cells were cultured under similar conditions at 37°C.
Cell Lysis and Immunoblotting Analysis-Cells in subconfluency were treated with 100 M VM-26. At different times, cells were lysed by an alkali solution as described previously (27). Briefly, 50 l of alkali lysis buffer (200 mM NaOH, 2 mM EDTA) was added to each 35-mm dish, and cells were scraped by a rubber policeman. The lysate was neutralized by the addition of 8 l of neutralizing buffer (1 M HCl, 600 mM Tris, pH 8.0), and the pH was adjusted to 7-8. The neutralized lysates were mixed with 6.6 l of a 10ϫ S7 nuclease reaction buffer (50 mM MgCl 2 , 50 mM CaCl 2 , 5 mM dithiothreitol, 1 mM EDTA, 50 g/ml leupeptin, 50 g/ml aprotinin, 50 g/ml pepstatin A, 1 mM phenylmethylsulfonyl fluoride) and 60 units of staphylococcal S7 nuclease. The digestion was performed at room temperature for 15 min. After nuclease digestion, 30 l of a 3ϫ SDS loading buffer (150 mM Tris-HCl, pH 6.8, 45% sucrose, 6 mM EDTA, 9% SDS, 0.03% bromphenol blue) was added to the reaction. The samples were then analyzed by SDS-polyacrylamide gel electrophoresis and immunoblotted with anti-hTOP2␣ and anti-hTOP2␤ antibodies, respectively.

The TOP2 Poison VM-26 Induces a Decrease of the hTOP2␤
Level in HeLa Cells-VM-26 represents a prototypic TOP2 poison that traps both TOP2␣ and TOP2␤ into cleavable complexes (33). Previous studies have demonstrated that VM-26 induces rapid SUMO-1 conjugation to hTOP2 (32). In the current study, we show that prolonged treatment with VM-26 in HeLa cells resulted in a decrease of the hTOP2␤ level (Fig. 1A).
In the presence of VM-26, the hTOP2␤ level was reduced more than 50% in 2 h. A further decrease in the hTOP2␤ level was observed with increasing time of VM-26 treatment (at least up to 6 h) (Fig. 1A). The decrease in the level of hTOP2␤ was due to the presence of VM-26 since 1% Me 2 SO (solvent control) had no effect on the level of hTOP2␤ during the 6-h incubation (data not shown). Surprisingly the level of hTOP2␣ showed very little change over the entire 6-h period (Fig. 1A). One possible explanation for the preferential decrease of the hTOP2␤ level could be that hTOP2␤ was more efficiently trapped into cleavable complexes than hTOP2␣ by VM-26. To examine this possibility, a band depletion assay (27) was performed to monitor TOP2 cleavable complexes in HeLa cells treated with VM-26 for 1-15 min (the short time treatment was used to avoid the complication of proteolytic degradation of TOP2 over prolonged treatment). As shown in Fig. 1B, both hTOP2␣ and hTOP2␤ are trapped by VM-26 into covalent protein⅐DNA complexes with equal efficiency as evidenced by their disappearance in an SDS gel (band depletion) in the absence of nuclease (S7) treatment. Treatment of the lysates with staphylococcal nuclease S7 resulted in the reappearance of the TOP2 bands due to the release of the TOP2⅐DNA covalent complexes from DNA into free TOP2. These results suggest that VM-26 induces the formation of both TOP2␣ and TOP2␤ cleavable complexes with equal efficiency. However, only TOP2␤ cleavable complexes are proteolytically degraded over time. To confirm that the VM-26induced decrease of the TOP2␤ level is not a unique phenomenon in HeLa cells, we have examined many other cells lines including the human lung fibroblast cell WI-38 and its transformed subline 2RA, leukemic cell lines such as U937 and CEM, and breast cancer cell lines ZR75-1 (see Fig. 2). VM-26 was shown to induce a decrease of the TOP2␤ level in all these cells. However, the TOP2␣ levels appeared not to be significantly affected in all these cells.
VM-26 Increases the Rate of hTOP2␤ Degradation-The de- crease of the hTOP2␤ level induced by VM-26 could be due to an increase in the rate of degradation, a decrease in the rate of synthesis, or a combination of both. To test these possibilities, the degradation assay was performed in the presence of the protein synthesis inhibitor cycloheximide. As shown in Fig. 3  (compare lanes 5 and 6) in the presence of VM-26, hTOP2␤ was more than 80% degraded in 2.5 h. Co-treatment with cycloheximide had no effect on the VM-26-induced decrease of the hTOP2␤ level (Fig. 3, compare lanes 5 and 6 with lanes 7 and  8), suggesting that VM-26-induced down-regulation is not due to a decreased rate of TOP2 protein synthesis. As a control experiment, we showed that neither the hTOP2␣ nor the hTOP2␤ level was significantly affected by cycloheximide treatment alone, suggesting that neither isozyme was rapidly turning over in the absence of VM-26. Consequently the VM-26induced decrease in the level of TOP2␤ is primarily due to an increased rate of degradation of TOP2␤, and this process will henceforth be referred to as TOP2␤ down-regulation.
Involvement of 26 S Proteasome in hTOP2␤ Down-regulation-VM-26 is a potent inducer of apoptotic cell death (34). Therefore, VM-26-induced degradation of TOP2␤ could be due to the activation of caspases. To examine this possibility, we performed the degradation assay in the presence of the caspase inhibitor Z-DEVD-FMK (35). As shown in Fig. 4A, treatment of Z-DEVD-FMK had no effect on VM-26-induced hTOP2␤ downregulation. On the other hand, treatment of cells with the proteasome inhibitor MG132 nearly completely blocked the degradation of hTOP2␤ (Fig. 4B, lanes 5-8). This result suggests that the 26 S proteasome rather than an apoptotic caspase may be involved in the degradation of hTOP2␤. To further confirm the involvement of the ubiqutin/26 S proteasome pathway, we tested the degradation of TOP2␤ induced by VM-26 in a pair of mouse cell lines. The ts85 cells contain a temperature-sensitive E1 enzyme for ubiquitin conjugation. At the nonpermissive temperature, E1 enzyme is inactivated, and ubiquitin conjugation to substrates is inhibited (31). We performed the degradation assays in both ts85 cells and their parental cells (FM3A). The degradation of TOP2␤ was substantially blocked when cells were shifted to the nonpermissive temperature (Fig. 5A, compare lanes 1-3 with lanes 4 -6). By contrast, at both temperatures, VM-26 induced equally efficient degradation of TOP2␤ in FM3A cells (Fig. 5B).
The sensitivity to the 26 S proteasome inhibitor and the formation of ubiquitin-TOP2 conjugates provide support of the notion that VM-26-induced TOP2␤ down-regulation is mediated by a ubiquitin/26 S proteasome pathway.
Involvement of RNA Transcription in VM-26-induced hTOP2␤ Down-regulation-Previous studies have suggested that both DNA replication and RNA transcription may be involved in the processing of TOP2 cleavable complexes into DNA damage signals (23). To test whether DNA replication and/or RNA transcription is involved in TOP2␤ down-regulation, we have examined the effect of various inhibitors on VM-26-induced TOP2␤ down-regulation. As shown in Fig. 6A, inhibition of DNA replication by aphidicolin did not affect VM-26-induced hTOP2␤ down-regulation. However, inhibition of transcription by DRB substantially affected hTOP2␤ down-regulation (Fig.  6B, compare lanes 1-3 and lanes 4 -6). To confirm this result, we also tested another potent transcription inhibitor camptothecin (CPT) (36). As shown in Fig. 7A, treatment of the breast cancer ZR75-1 cells with either DRB or CPT efficiently blocked VM-26-induced down-regulation of hTOP2␤. To rule out the possibility that the effect is due to a reduced level of the cleavable complexes in the presence of DRB or CPT, the band depletion assay was performed. As shown in Fig. 7B, treatment of ZR75-1 cells with either DRB or CPT had no effect on the amounts of VM-26-induced TOP2␤ cleavable complexes. Since CPT can induce TOP1-mediated DNA damage, the blockage of A, U937 and CEM leukemic cells were treated with VM-26 for increasing times. Cell lysates were analyzed by immunoblotting as described under "Experimental Procedures." The antibody used in this experiment was anti-hTOP2␣␤ that recognizes both hTOP2␣ and hTOP2␤. B, human lung fibroblast WI-38 and its SV40-transformed subline 2RA cells were treated with VM-26. Cell lysates were analyzed by immunoblotting with anti-hTOP2␣ and anti-hTOP2␤ antibodies, respectively. C, ZR75-1 breast cancer cells were treated with VM-26. Cell lysates were analyzed by immunoblotting with anti-hTOP2␣ and anti-hTOP2␤ antibodies, respectively. The VM26 concentration used in these experiments was 100 M.
hTOP2␤ degradation by CPT could also be an indirect result of a DNA damage response rather than a direct result of transcription inhibition. To rule out this possibility, we have tested the effect of another DNA damage agent, cisplatin, on TOP2␤ down-regulation. As shown in Fig. 7C, treatment of ZR75-1 cells with cisplatin had no effect on VM-26-induced downregulation of hTOP2␤. DISCUSSION VM-26 is a potent inducer of TOP2 cleavable complex formation on chromosomal DNA. Different from other DNA damages, these TOP2 cleavable complexes are highly reversible. In addition, the DNA strand breaks are protein-linked and proteinconcealed (14,15). It has been suggested that the DNA-damaging effect of TOP2 cleavable complexes is dependent upon the cellular processing of TOP2 cleavable complexes into DNA damage (12). DNA replication, RNA transcription, helicase activity, and proteolysis have been speculated to be potentially capable of converting TOP2 cleavable complexes into DNA damage (23)(24)(25). However, the demonstration that any of these cellular processes is actually involved in processing TOP2 cleavable complexes into DNA damage in vivo is still lacking. Our current results suggest that both proteolysis and RNA transcription may be involved in the processing of hTOP2␤ cleavable complexes into DNA damage. In addition, proteolysis via the ubiquitin/26 S proteasome pathway and transcription appear to be coupled.
VM-26-induced down-regulation of TOP2␤ appears to be quite similar to CPT-induced down-regulation of TOP1. Both processes are dependent on the formation of topoisomerase cleavable complexes and involve ubiquitin/26 S proteasome (27). In addition, both processes appear to be transcription-dependent. 2 CPT-induced down-regulation has been suggested to be triggered by collision between TOP1 cleavable complexes and RNA polymerase elongation complexes. Down-regulation of TOP1 presumably reveals the TOP1-concealed single strand breaks so that repair can occur. 2 By analogy, we propose that TOP2␤ cleavable complexes can also collide with RNA polymerase elongation complexes resulting in transcription arrest (see Fig. 8). Proteolysis of TOP2␤ cleavable complexes by ubiquitin/26 S proteasome results in exposure of the protein-concealed double strand breaks (Fig. 8). The exposed double strand breaks can then be repaired by either homologous recombination or nonhomologous end joining (37,38). Alternatively unrepaired double strand breaks can trigger apoptotic cell death FIG. 3. VM-26-induced decrease of hTOP2␤ is not due to a decreased rate of protein synthesis. HeLa cells were treated with either 30 M cycloheximide or 1% Me 2 SO (DMSO) (solvent used for cycloheximide) for 0 and 2.5 h (lanes 1-4). Another set of cells was treated with 100 M VM-26 in the presence (lanes 7 and 8) or absence (lanes 5 and 6) of cycloheximide. All samples were analyzed by immunoblotting with antibodies against hTOP2␣ and hTOP2␤, respectively.  1-4) of 10 M MG132 for increasing times. Cell lysates were analyzed by immunoblotting with anti-hTOP2␣, anti-hTOP2␤, and anti-hTOP1 antibodies, respectively. (39). It has been reported that proteasome inhibitors can block VP-16-induced apoptosis (40,41). The inhibition of VM-26induced apoptosis by proteasome inhibitors is not due to inhibition of NFB activation (40,41). Our current model could explain why the 26 S proteasome inhibitor MG132 inhibits VM-26-induced apoptosis since the inhibitor blocks TOP2 down-regulation and hence the formation of the apoptosis-inducing double strand breaks. However, VM-26-induced apoptosis is likely to be mediated by multiple mechanisms, and degradation of TOP2␤ cleavable complexes may represent only one of these mechanisms.
The preferential degradation of TOP2␤ over TOP2␣ in VM-26-treated cells is intriguing. Apparently preferential degradation is not due to more efficient trapping of TOP2␤ cleavable complexes by VM-26. It appears that TOP2␤ cleavable complexes must be more efficiently recognized by the ubiqutin/26 S proteasome pathway than are TOP2␣ cleavable complexes. The preferential recognition by the ubiquitin/26 S proteasome pathway may reflect the functional difference between the two isozymes. These two isoforms are encoded by distinct genes but share about 72% identity in their primary sequences (42)(43)(44). Immunohistochemical studies have shown that TOP2␣ is only present in proliferating tissues including tumors, while TOP2␤ is present in all tissues including terminally differentiated tissues (45,46). The two isoforms are regulated very differently in cells. The TOP2␣ level peaks in G 2 /M phase, while the TOP2␤ level is not significantly changed throughout the cell  1-3) for 0, 2.5, and 5 h. Cells were lysed by the alkaline lysis procedure as described under "Experimental Procedures," and cell lysates were treated with staphylococcal nuclease S7 to free hTOP2␤ from covalent hTOP2␤⅐DNA complexes. B, DRB and CPT do not affect VM-26-induced formation of hTOP2␤ cleavable complexes. Cells were treated with VM-26 in the presence of DRB and CPT exactly the same as described in A except that the treatment step with staphylococcal nuclease S7 was omitted. In the absence of the nuclease, hTOP2␤ cannot be released from the hTOP2␤⅐DNA covalent complexes. Consequently by comparing A and B, the amounts of hTOP2␤ covalent complexes can be estimated. C, ZR75-1 cells were treated with 100 M VM-26 in the presence (lanes 5-8) or absence (lanes 1-4) of 40 M cisplatin (CDDP) for increasing times from 0 to 6 h. Cell lysates were analyzed by immunoblotting with anti-hTOP2␤ and anti-hTOP1 antibodies, respectively. cycle (47). TOP2␣ plays important roles in chromosome condensation, segregation, and sister chromatin separation (48,49). The cellular function of TOP2␤ is much less clear. However, recent studies have hinted at the possibility that TOP2␤, like TOP1, is involved in RNA transcription. TOP2␤ has been shown to be essential during mouse neuronal development (50). Studies in rat cerebellum have suggested the involvement of TOP2␤ in neuronal differentiation by regulating the transcription of neuronal genes (51). TOP2␤ has also been located in the transcribed regions of human rDNA repeats (52). One possible explanation for the preferential degradation of TOP2␤ over TOP2␣ is that TOP2␤ is preferentially located within the transcribed region, while TOP2␣ may be located in other regions. TOP2␤ cleavable complexes located within the transcribed region may block RNA polymerase elongation complexes leading to transcription arrest (see the model in Fig. 8). The ubiquitin/26 S proteasome pathway may be associated with or recruited to arrested RNA polymerase complexes to effect proteolytic degradation of TOP2␤ cleavable complexes.
TOP2 cleavable complexes have been associated with both antitumor as well as carcinogenic activities of TOP2 poisons. The use of TOP2-directed anticancer drugs such as etoposide (VP-16) and doxorubicin is associated with the high incidence of secondary leukemia, therapy-related acute myeloid leukemia (53,54). Therapy-related acute myeloid leukemia associated with the use of TOP2-directed drugs is characterized by translocation of the mixed lineage leukemia gene to its over 30 partner genes (55). All translocations of the mixed lineage leukemia gene occur within an 8.3-kilobase region termed the breakpoint cluster region (56,57). Recent studies have demonstrated that TOP2 can specifically form TOP2 cleavable complexes within the mixed lineage leukemia breakpoint cluster region, suggesting a direct role of TOP2 in mixed lineage leukemia gene translocations (58). Our current demonstration that TOP2␤ cleavable complexes can be efficiently degraded into translocation-competent double strand breaks within the transcribed regions may suggest the involvement of TOP2␤ cleavable complexes in chromosomal translocations. It remains to be determined whether TOP2␤ and TOP2␣ cleavable complexes may play different roles in their antitumor and carcinogenic activities.

FIG. 8. A working model for VM-26-induced down-regulation of TOP2␤.
In the presence of VM-26, TOP2␤ is trapped as covalent TOP2⅐DNA cleavable complexes on DNA. These TOP2␤ cleavable complexes block transcription and trigger ubiquitin/26 S proteasome-dependent degradation of TOP2␤. The consequence of TOP2␤ degradation is the revelation of the DNA double strand break. The double strand break can either be repaired by homologous recombination and/or nonhomologous end joining or induce apoptotic cell death. Ub, ubiquitin.