Phosphorylation of DNA Topoisomerase I by the c-Abl Tyrosine Kinase Confers Camptothecin Sensitivity*

DNA topoisomerase I (topo I) is involved in the regulation of DNA supercoiling, gene transcription, recom-bination, and DNA repair. The anticancer agent camptothecin specifically targets topo I. The mechanisms responsible for the regulation of topo I in cells, however, are not known. This study demonstrates that c-Abl-de-pendent phosphorylation up-regulates topo I activity. The c-Abl SH3 domain bound directly to the N-terminal region of topo I. The results demonstrate that c-Abl phosphorylated topo I at Tyr 268 in core subdomain II. c-Abl-mediated phosphorylation of topo I Tyr 268 in vitro and in cells conferred activation of the topo I isomerase function. Moreover, activation of c-Abl by treatment of cells with ionizing radiation was associated with c-Abl-dependent phosphorylation of topo I and induction of for topo I in RNA polymerase I-mediated transcription of topo I to nucleolin the cellular localization of topo Other that topo I a RING finger/ the factor factor RNA This study has shown that topo I associates with the nuclear c-Abl kinase. The results show that the c-Abl SH3 domain interacts with the N-terminal 1–250 amino acids of topo I. The c-Abl SH3 domain binds to a proline-rich sequence with the consensus X P XXXX P XX P (72, 73). A potential site for c-Abl SH3 domain binding is located at amino acids 225–233 (PV-FAPPYEP) in the conserved topo I core domain. The results also indicate that deletion of amino acids 1–210 interferes with c-Abl SH3 domain binding to topo I, perhaps by altering con-formation of the proline-rich motif. Nucleolin (55), splicing factor 2/alternate splicing factor (14), topors (71), and SV40 T antigen (71, 74) also bind to topo I within the N-terminal 210 amino acids, which are dispensable for topo I activity. Subdomains I and II of the topo I core form the top half or “cap” of the enzyme, bind to DNA, and are essential for the isomerase c-Abl also binds to DNA, prefers sequences containing AAC

Eukaryotic type I topoisomerases function in the relaxation of negatively and positively supercoiled DNA (1)(2)(3)(4). Topoisomerase I (topo I) 1 -mediated relaxation is accomplished by the introduction of a single-stranded nick in the phosphodiester backbone, rotation of the complementary DNA strand through the break and relief of torsional strain. Catalysis of the singlestranded nick involves the formation of a covalent bond be-tween the 3Ј-end of the DNA strand break and the active-site tyrosine in the C-terminal region of topo I (5,6). Following relaxation, reversal of the trans-esterification reaction results in religation of the phosphodiester bond and release of topo I. The reduction of torsional stress is essential to DNA replication (1,4). topo I has also been shown to participate in RNA polymerase II-mediated transcription (7,8), DNA repair (9 -11), and RNA splicing (12)(13)(14). Despite the importance of topo I to cellular functions, little is known about the mechanisms responsible for the regulation of this enzyme.
Human topo I is a nuclear protein with 765 residues, a predicted molecular mass of 91 kDa, and four domains. The N-terminal non-conserved domain from Met 1 to Lys 197 contains four putative nuclear localization signals, is sensitive to proteolysis, and can be deleted without effect on topo I activity (15)(16)(17). The conserved core domain extends from Glu 198 to Ile 651 and is followed by a short linker domain from Asp 652 to Glu 696 . The highly conserved C-terminal domain extends from Gln 697 to Phe 765 , contains the active-site tyrosine at position 723, and is required for topo I activity (6). Crystal structures of the core and C-terminal domains in covalent and noncovalent complexes with duplex DNA have shown that the enzyme clamps around B-form DNA (18). Moreover, the linker region assumes a coiled-coil configuration and protrudes away from the rest of the protein (19).
topo I is a cellular target for the plant alkaloid camptothecin (CPT) (3, 20 -22). The available evidence supports a model in which CPT binds to the topo I/DNA complex subsequent to DNA cleavage and covalent attachment of topo I to DNA (23,24). CPT then stabilizes the cleavage complexes by inhibiting topo I-mediated DNA religation (25,26). A CPT binding mode has been proposed in which CPT interacts with both DNA strands and topo I residues Arg 364 , Asp 533 , and Asn 722 (18). Stabilization of the cleavage complexes by CPT is believed to induce lethality as a result of collisions with proteins involved in DNA replication and transcription (25)(26)(27)(28). In addition, CPT-induced stabilization of the topo I/DNA complex is associated with conversion of single-stranded nicks to irreversible double-strand breaks (21, 29 -32). Thus, cells selected for resistance to CPT express decreased levels of topo I (33)(34)(35) or certain point mutations in the topo I gene (36). Post-translational modification of topo I has been shown to regulate topo I activity. Protein kinase C-dependent phosphorylation up-regulates topo I activity and thereby increases CPT sensitivity (37).
The activity of the ubiquitously expressed c-Abl proteintyrosine kinase is tightly regulated in cells (38,39). Nuclear c-Abl associates with the DNA-dependent protein kinase complex (40,41) and with the product of the gene mutated in ataxia telangiectasia, ATM (42,43). The DNA-dependent protein kinase catalytic subunit and ATM are members of a family of phosphatidylinositol 3-kinase-like enzymes involved in the regulation of the cell cycle, recombination, control of telomere length, and the DNA damage response (44). Importantly, c-Abl is activated by DNA-dependent protein kinase and ATM in cells exposed to ionizing radiation (IR) and other genotoxic agents (40,42,43,45,46). The available evidence indicates that activation of nuclear c-Abl contributes to DNA damage-induced growth arrest and apoptosis by mechanisms dependent in part on p53 and its homolog p73 (47)(48)(49)(50)(51). Other studies have demonstrated that nuclear c-Abl interacts with the human catalytic subunit of telomerase, TERT, and thereby regulates telomere length (52). These findings have supported a role for nuclear c-Abl in converting DNA damage into signals that control the genotoxic stress response.
Because c-Abl is activated by DNA strand breaks (45) and topo I is involved in the generation of these lesions (35), we investigated whether c-Abl interacts with topo I. The results demonstrate that c-Abl phosphorylates topo I and that c-Abl functions in the induction of topo I activity. We also show that c-Abl-mediated activation of topo I is important, at least in part, in imparting cellular sensitivity to CPT.
Immunoprecipitation and Immunoblot Analysis-Cell lysates were prepared for immunoprecipitation as described (45). Soluble proteins were incubated with anti-c-Abl (Santa Cruz Biotechnology) or anti-topo I (Topogen, Inc.) antibody as described (45). The immunoprecipitates were subjected to immunoblotting with anti-c-Abl or anti-topo I antibody. Antigen/antibody complexes were visualized by enhanced chemiluminescence (ECL detection system, Amersham Biosciences).
Fusion Protein Binding Assays-Glutathione S-transferase (GST) and GST-c-Abl-SH3 were purified by affinity chromatography using glutathione-Sepharose beads. Cell lysates were incubated with 2 g of immobilized GST or GST-c-Abl-SH3 for 2 h at 4°C. GST-Grb2-SH3 fusion protein was used as an extra control. The resulting protein complex were washed with lysis buffer and analyzed by immunoblot analysis with anti-topo I antibody. In the reciprocal experiment, GSTtopo I or fragments of topo I linked to GST were separately incubated with lysates, and the adsorbates were analyzed by immunoblot analysis with anti-c-Abl antibody.
Direct Interaction of c-Abl with topo I-Purified GST-c-Abl bound to glutathione-Sepharose beads (ϳ10 g) was incubated with 5 units of thrombin (Amersham Biosciences) for 4 h at 25°C. 5 g of c-Abl was incubated with 4 g of GST-topo I bound to glutathione-Sepharose beads at 4°C for 2 h. The beads were washed extensively with phosphate-buffered saline, and the adsorbates were eluted with 500 mM NaCl. The eluate was analyzed by SDS-PAGE followed by silver staining. The protein bands were excised and processed for trypsin digestion. The resulting peptides were analyzed by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) (Voyager DE-PRO, ABI, Framingham, MA) and electrospray ionization mass spectrometry (ESI-MS). ESI-MS and tandem mass spectrometry (MS/MS) were performed using an electrospray ion trap (LCQ-DECA, Thermo Electron). The tryptic peptides were fractionated on a capillary HPLC C 18 column coupled with a mass spectrometer. Tandem mass spectra were acquired using argon as the collision gas and sufficient collision energy to obtain complete sequence information on the precursor ion. MS and MS/MS data were then analyzed using the BioWorks 3.0 software package (Thermo Electron). In other experiments, c-Abl was incubated with GST-topo I, and adsorbates were analyzed by immunoblotting with anti-c-Abl antibody.
In Vitro Phosphorylation Studies-Purified GST-topo I (55) was incubated with c-Abl or c-Abl(K290R) in kinase buffer (20 mM Tris-HCl (pH 7.4), 10 mM MgCl 2 , and 10 mM MnCl 2 ) containing [␥-32 P]ATP or unlabeled ATP for 30 min at 30°C. Kinase-active c-Abl and kinaseinactive c-Abl(K290R) were purified from baculovirus-infected insect cells (56). The c-Abl proteins were purified by binding to glutathione beads and washing. Analysis of the c-Abl preparations by SDS-PAGE and protein staining demonstrated Ͼ90% purity. In certain experiments, c-Abl-phosphorylated GST-topo I was incubated with purified SHPTP1 protein-tyrosine phosphatase (57) for 5 min at 30°C. The reaction products were analyzed (i) by SDS-PAGE and autoradiography, (ii) in assays of topo I activity, and (iii) by mapping the phosphorylation site.
Preparation of Nuclear Lysates-Nuclear lysates were prepared as described (36,58). Protein concentration was determined by the Bio-Rad protein assay.
topo I Activity Assays-topo I activity was measured by DNA relaxation assays as described (36) in reaction buffer containing 10 mM Tris-HCl (pH 7.9), 1 mM EDTA, 150 mM NaCl, 0.1% bovine serum albumin, 0.1 mM spermidine, and 5% glycerol. The gels were imaged using an Alphaimager-2000 (Alpha Innotech). Percent supercoiled DNA was quantitated from the digital image using the ImageQuant software package.
Identification of in Vitro Tyrosine Phosphorylation Sites-Purified GST-topo I was incubated with c-Abl and [␥-32 P]ATP. The reaction products were subjected to SDS-PAGE. The topo I band was identified by Coomassie Blue staining and excised from the gel. In gel digestion with trypsin was performed as described (60,61). In brief, a corresponding slice of gel was cut into small pieces and dehydrated with acetonitrile. The content were rehydrated with 10 mM dithiothreitol in 100 mM ammonium bicarbonate and incubated at 56°C for 1 h. Following dehydration with acetonitrile, gel pieces were suspended in trypsin (12.5 ng/l) in 50 mM ammonium bicarbonate. In gel digestion was carried out at 37°C for 10 -12 h. The peptides were extracted in 50% acetonitrile and 5% formic acid.
For 32 P-labeled topo I, the trypsin-digested peptides were fractionated by reversed-phase HPLC. Aliquots of the fractions were assayed for 32 P. Positive fractions were subjected to Edman sequencing, and part of the digest was directly analyzed by MALDI-TOF-MS using the Voyager DE-PRO. Edman sequencing was performed for 14 cycles, and the eluate of every cycle was assayed for 32 P.
Identification of in Vivo Tyrosine Phosphorylation Sites-Nuclei isolated in nuclear buffer (20 mM KH 2 PO 4 , 5 mM MgCl 2 , 150 mM NaCl, and 1 mM EGTA) were resuspended in lysis buffer. The suspension was gently rotated for 1 h at 4°C and then centrifuged at 17,000 ϫ g for 10 min. The supernatant was incubated with pre-equilibrated nickel-nitrilotriacetic acid matrix (QIAGEN Inc., Valencia, CA). The matrix was washed with lysis buffer containing 25 mM imidazole. Proteins were eluted with lysis buffer containing 100 mM imidazole, concentrated, and separated by SDS-PAGE. The topo I band was identified by staining, excised, and subjected to in gel trypsin digestion. The masses of the trypsin-digested peptides were analyzed by MALDI-TOF-MS.
Site-directed Mutagenesis-Tyr 268 in topo I was mutated to phenylalanine using a site-directed mutagenesis kit (Stratagene).
Analysis of Protein-linked DNA Breaks (PLDBs) and DNA Fragmentation-topo I PLDBs were assayed as described (62). Quantification of DNA fragmentation by alkaline elution assays was performed as described (63).
Analysis of Sub-G 1 DNA Content-Cells were fixed with ethanol and stained with propidium iodide. DNA content was assessed using a FACScan (BD Biosciences).

RESULTS
c-Abl Associates with topo I-To determine whether c-Abl forms a complex with topo I, cell lysates were subjected to immunoprecipitation with anti-c-Abl antibody. Analysis of the immunoprecipitates with anti-topo I antibody demonstrated the coprecipitation of c-Abl and topo I (Fig. 1A). In the reciprocal experiment, immunoblot analysis of anti-topo I immunoprecipitates with anti-c-Abl antibody confirmed the association of c-Abl and topo I in cells (Fig. 1B). Based on the total amounts of topo I and c-Abl in the lysates subjected to immunoprecipitation, ϳ5% of the c-Abl pool and 4% of the topo I pool contribute to the formation of the c-Abl-topo I complexes. To further assess the interaction between c-Abl and topo I, cell lysates were incubated with GST, GST-Grb2-SH3, or GST-c-Abl-SH3. Analysis of the adsorbates on the glutathione beads by immunoblotting with anti-topo I antibody demonstrated binding of topo I to GST-c-Abl-SH3, but not to GST or GST-Grb2-SH3 (Fig. 1C). In other experiments, fragments of topo I linked to GST were incubated with cell lysates. Immunoblot analysis of the adsorbates with anti-c-Abl antibody demonstrated no detectable binding to topo I with a deletion of amino acids 1-210 (Fig. 1D). Localization of the c-Abl-binding site to the N-terminal region of topo I was confirmed by the demonstration that c-Abl bound to GST-topo I(1-250) (Fig. 1D). To assess direct interaction, purified recombinant c-Abl protein was incubated with GST-topo I linked to beads. After extensive washing, the adsorbates were analyzed by immunoblotting with anti-c-Abl antibody. Fig. 1E shows a direct interaction of c-Abl with topo I. These results demonstrate that c-Abl binds directly to topo I and that the c-Abl SH3 domain associates with the N-terminal region of topo I. topo I Associates Directly with c-Abl-To determine the direct interaction, purified c-Abl was incubated with GST-topo I linked to beads. After extensive washing with phosphate-buffered saline, the adsorbates were analyzed by SDS-PAGE and silver staining. Two topo I-interacting proteins (140 and 100 kDa) were visualized by SDS-PAGE ( Fig. 2A). Both proteins were trypsin-digested, and the resulting peptides were analyzed by MALDI-TOF-MS and ESI-MS/MS. The MALDI-TOF-MS-generated spectrum (data not shown) was subjected to a data base search using Protein Prospector. The protein was identified as the proto-oncogene tyrosine-protein kinase ABL1 (p150, c-Abl) (NCBI accession number 125135 and Swiss-Prot accession number P00519). A fraction of tryptic peptides was also analyzed by ESI-MS/MS. The peptides were separated in a C 18 capillary column, and a total ion chromatogram was recorded ( c-Abl Phosphorylates and Activates topo I in Vitro-To determine whether c-Abl phosphorylates topo I, purified c-Abl was incubated with GST-topo I in the presence of [␥-32 P]ATP. GST and the c-Abl substrate GST-Crk-(120 -225) (64) were included in similar reactions as controls. Analysis of the reaction products by SDS-PAGE and autoradiography demonstrated that, like Crk-(120 -225), topo I is subject to c-Abl phosphorylation (Fig. 3A). As shown previously (45), GST was not a substrate for the c-Abl kinase (data not shown). To quantitate the c-Ablmediated topo I phosphorylation, we next estimated the molar incorporation of ATP. Purified c-Abl was incubated with GSTtopo I in the presence of [␥-32 P]ATP, and the reaction product was analyzed by SDS-PAGE and Coomassie Blue staining. The samples were analyzed in duplicate by SDS-PAGE and autoradiography (Fig. 3A, lower panel). Coomassie Blue-stained GST-topo I and GST-Crk proteins were cut from the gel and assayed for 32 P. The results indicate that 2 pmol of c-Abl incorporated 10.6 pmol of ATP on 100 pmol of GST-topo I. GST-Crk-(120 -225) incorporated 4-fold more ATP. To assess the effects of c-Abl phosphorylation on topo I activity, assays measuring the effects of topo I on relaxation of supercoiled DNA were performed. In these experiments, GST-topo I was preincubated with c-Abl or kinase-inactive c-Abl(K290R) in the presence of ATP. topo I was then assayed at increasing concentrations for DNA relaxation activity. Fig. 3B (left panel) demonstrates that phosphorylation of topo I by c-Abl was associated with increased activity in converting supercoiled DNA to the relaxed form. As determined by image analysis of the activity gel, 2.7 ng of c-Abl-phosphorylated topo I relaxed nearly 90% of the supercoiled DNA, whereas 2.5 ng of control

FIG. 1. topo I interacts with c-Abl in cells and in vitro.
A, lysates from non-transfected 293 cells were subjected to immunoprecipitation (IP) with anti-c-Abl antibody or IgG. The immunoprecipitates were analyzed by immunoblotting (IB) with anti-topo I antibody. Lysate not subjected to immunoprecipitation was used as a control. The supernatant after anti-c-Abl immunoprecipitation was subjected to immunoblot analysis with anti-topo I antibody. The finding that ϳ95% of the topo I protein remained in the supernatant indicates that 1 of 20 topo I molecules associates with c-Abl. B, anti-topo I immunoprecipitates obtained from 293 cell lysates were subjected to immunoblotting with anti-c-Abl antibody. C, 293 cell lysate (800 g) was incubated with GST, GST-Grb2-SH3, or GST-c-Abl-SH3. The adsorbates were subjected to immunoblotting with anti-topo I antibody. Lysate (60 g) not incubated with GST proteins was included in the immunoblot analysis as a control. D, 293 cell lysate was incubated with GST-topo I-(221-765), GSTtopo I-(1-250), or full-length GST-topo I-(1-765). The adsorbates were analyzed by immunoblotting with anti-c-Abl antibody. E, purified c-Abl protein was incubated with GST-topo I, and adsorbates were analyzed by SDS-PAGE and immunoblot analysis with anti-c-Abl antibody.
c-Abl Activates DNA topo I FIG. 2. topo I associates with c-Abl directly. A, purified GST-c-Abl was treated with thrombin to cleave GST. c-Abl was then incubated with GST-topo I for 1 h. After extensive washing with phosphate-buffered saline, the adsorbates were analyzed by SDS-PAGE and silver staining. Thrombin-cleaved c-Abl showed multiple protein bands, whereas no proteins were observed in the GST-c-Abl lane. Two proteins with molecular mass of 140 and 100 kDa were observed in the GST-Topo1ϩc-Abl lane. B, two proteins visualized by SDS-PAGE (A) were excised and subjected topo I or topo I incubated with c-Abl(K290R) was associated with 20% relaxation (Fig. 3B, middle panel). Moreover, 4 ng of c-Abl-phosphorylated topo I conferred complete relaxation, whereas unphosphorylated topo I relaxed 40 -50% of the supercoiled DNA (Fig. 3B, middle panel). Similar findings were obtained in four independent experiments (Fig. 3B, right  panel). To confirm that tyrosine phosphorylation of topo I is responsible for activation of the isomerase function, topo I was incubated with c-Abl and [␥-32 P]ATP. Treatment of 32 Plabeled topo I with the SHPTP1 protein-tyrosine phosphatase was associated with a decrease in tyrosine phosphorylation (Fig. 3C, left panel). Similar studies were performed with topo I that had been preincubated with c-Abl and unlabeled ATP to assess the effects of SHPTP1 on topo I activity. Fig.  3C (middle and right panels) demonstrates that SHPTP1 attenuated c-Abl-mediated activation of topo I. These findings collectively demonstrate that c-Abl phosphorylates and activates topo I in vitro.
c-Abl-dependent Phosphorylation and Activation of topo I in Vivo-To determine whether topo I is phosphorylated by a c-Abl-dependent mechanism in cells, lysates from human MCF-7 breast carcinoma cells stably expressing the empty neo vector or kinase-inactive dominant-negative c-Abl(K290R) were subjected to immunoprecipitation with anti-topo I antibody. Analysis of the immunoprecipitates with anti-phospho-Tyr antibody demonstrated that tyrosine phosphorylation of topo I was decreased in MCF-7/c-Abl(K290R) cells compared with MCF-7/neo cells (Fig. 4A). Reprobing the membrane with anti-topo I antibody demonstrated equal amounts of topo I protein (Fig. 4A). Immunoprecipitation of the lysates with antiphospho-Tyr antibody and immunoblot analysis of the supernatants demonstrated a 25% decrease in the amount of topo I. These findings indicate that ϳ1 of 4 topo I molecules is subject to tyrosine phosphorylation and/or that topo I coprecipitates with other tyrosine-phosphorylated proteins. To assess topo I activity, nuclear extracts were assayed for relaxation of supercoiled DNA. Fig. 4B shows that topo I activity was decreased in MCF-7/c-Abl(K290R) cells compared with MCF-7/neo cells. Image analysis of the activity gel demonstrated that 4 g of nuclear extract from MCF-7/neo cells relaxed 70% of the supercoiled DNA, whereas the same amount of nuclear lysate from MCF-7/neo cells conferred 40% relaxation (Fig. 4B). To further assess involvement of c-Abl in the regulation of topo I, studies were performed on wild-type (c-Abl ϩ/ϩ ) and c-Abl Ϫ/Ϫ MEFs (54). Immunoblot analysis of anti-topo I immunoprecipitates with anti-phospho-Tyr antibody demonstrated that tyrosine phosphorylation of topo I in c-Abl Ϫ/Ϫ cells was decreased compared with that in wild-type cells (Fig. 4C, left panels). Reprobing the blots with anti-topo I antibody demonstrated equal amounts of topo I in both cell types (Fig. 4C, left panels). Of note, expression of topo I proteins in MEFs differs from that in MCF-7 cells, particularly in terms of the isomerase-active 68-kDa form (16,65,66). Densitometric analysis of the anti-phospho-Tyr antibody signals in four separate experiments demonstrated that tyrosine phosphorylation of topo I was decreased by ϳ80% in c-Abl Ϫ/Ϫ cells (Fig. 4C, right panel). Analysis of nuclear extracts for relaxation of supercoiled DNA showed that topo I activity was decreased in c-Abl Ϫ/Ϫ cells compared with wild-type cells (Fig. 4D). Image analysis of the activity gels in four separate experiments demonstrated that 9 g of nuclear extract from c-Abl Ϫ/Ϫ cells relaxed ϳ40% of the supercoiled DNA, whereas that amount of nuclear lysate from c-Abl ϩ/ϩ cells was associated with 90% relaxation (Fig. 4D). These findings demonstrate that c-Abl phosphorylates topo I in vivo and that c-Abl-dependent phosphorylation up-regulates topo I activity.
c-Abl-dependent Phosphorylation and Activation of topo I in Response to Genotoxic Stress-Exposure of cells to IR and other genotoxic agents is associated with activation of the c-Abl kinase function (45). To assess the effects of IR on the interaction of c-Abl and topo I, lysates from irradiated wild-type and c-Abl Ϫ/Ϫ cells were subjected to immunoprecipitation with antitopo I antibody. Immunoblot analysis of the immunoprecipitates with anti-phospho-Tyr antibody demonstrated an IR-dependent increase in tyrosine phosphorylation of topo I in wildtype cells (Fig. 5A). By contrast, IR had little effect on tyrosine phosphorylation of topo I in c-Abl Ϫ/Ϫ cells (Fig. 5A). These findings indicate that IR induces tyrosine phosphorylation of topo I by a c-Abl-dependent mechanism. Analysis of topo I activity indicated that IR treatment of wild-type cells was associated with increased relaxation of supercoiled DNA (Fig.  5B). By contrast, IR had less of an effect on topo I activity in c-Abl Ϫ/Ϫ cells (Fig. 5B). Image analysis of the gel confirmed that IR induced topo I activity by a c-Abl-dependent mechanism (Fig. 5C). Of note, the change in tyrosine phosphorylation of topo I in response to IR remained more pronounced in comparison with the basal tyrosine phosphorylation. With higher basal tyrosine phosphorylation (Fig. 4C), the immunoprecipitation technique does not allow one to see the difference, as only 1-2% of cellular topo I is immunoprecipitated.

Identification of the c-Abl-mediated topo I Tyrosine Phosphorylation Site-To identify the in vitro phosphorylation site, purified GST-topo I was incubated with c-Abl and [␥-32 P]ATP.
The reaction products were subjected to SDS-PAGE. The topo I band was identified by Coomassie Blue staining and excised from the gel. In gel digestion with trypsin was performed, and the peptides were fractionated by reversed-phase HPLC on a cation exchange column. The fractions were assayed for 32 P. Of 60 fractions, nine showed significantly higher levels of 32 P. All nine fractions were analyzed by MALDI-TOF-MS, and fraction 26 was further analyzed by Edman sequencing (Fig. 6A). Amino acids 263-275 of topo I were sequenced (Fig. 6B). Further analyses of the sequence eluates showed higher activity in the sixth sequence cycle, representing topo I Tyr 268 (Fig. 6B).
To define the topo I sites phosphorylated by c-Abl in wildtype MEFs, topo I was purified by binding of the histidine-rich N-terminal region to nickel-conjugated beads. Purified topo I was subjected to SDS-PAGE, excised, and then analyzed by MALDI-TOF-MS. Tyr 268 was identified as phosphorylated by MALDI-TOF-MS and Edman sequencing (data not shown). To confirm these findings, c-Abl was incubated with wild-type topo I or mutant topo I(Y268F) in the presence of [␥-32 P]ATP. Analysis of the reaction products demonstrated that c-Abl phosphorylation was decreased as a result of the Y268F mutation (Fig.  7A). Analysis in DNA relaxation assays demonstrated that mutant topo I(Y268F) exhibited less isomerase activity compared with wild-type topo I (Fig. 7B). In addition and unlike wild-type topo I (Fig. 3B), incubation of topo I(Y268F) with c-Abl and ATP had no detectable effect on DNA relaxation activity (Fig. 7C). These findings demonstrate that topo I Tyr 268 is phosphorylated by c-Abl and that this site is functionally important for topo I activity.  Functional Significance of the Interaction between c-Abl and topo I Tyr 268 -As there is less topo I activity and no detectable phosphorylation of topo I Tyr 268 in c-Abl Ϫ/Ϫ MEFs, the response of these cells to CPT was compared with that of wildtype cells. CPT induces PLDBs that are stabilized intermediates of topo I activity (62). Using an in vivo KCl/SDS coprecipitation assay, analysis of PLDBs demonstrated a CPT concentration-dependent increase in wild-type MEFs (Fig. 8A). By contrast, the induction of PLDBs was attenuated in c-Abl Ϫ/Ϫ cells (Fig. 8A). To extend the analysis of c-Abl Ϫ/Ϫ cells, CPT-induced DNA fragmentation was measured by quantitating the formation of DNA single-strand breaks (SSBs) in alkaline elution assays. Fig. 8B demonstrates that CPT treatment of c-Abl Ϫ/Ϫ cells was associated with decreased formation of DNA SSBs compared with that in c-Abl ϩ/ϩ cells. Moreover, the finding that CPT treatment of c-Abl ϩ cells was associated with formation of DNA SSBs at a level similar to that in c-Abl ϩ/ϩ cells (Fig. 8B) provides support for the involvement of c-Abl in the response. To determine whether decreased formation of protein-linked DNA SSBs affects sensitivity to CPT-induced apoptosis, cells were analyzed for sub-G 1 DNA content. Fig. 8C shows that c-Abl Ϫ/Ϫ cells were significantly less sensitive to the apoptotic effects of CPT. Thus, 25% of the CPT-treated c-Abl ϩ/ϩ cells contained sub-G 1 DNA compared with only 0.5% of the c-Abl Ϫ/Ϫ cells exposed to this agent (Fig. 8C). The finding that the apoptotic response of c-Abl ϩ cells to CPT was similar to that of c-Abl ϩ/ϩ cells (Fig. 8C) provides further support for a c-Abl-dependent mechanism. These findings collectively demonstrate that the c-Abl-dependent phosphorylation of topo I is of functional importance to inhibition of topo I activity by CPT.

DISCUSSION
Interaction between topo I and Nuclear c-Abl-Proliferating mammalian cells require topo I for viability (1,3). However, the functions of topo I that are indispensable for growth are not yet known. Moreover, little is known about the interactions between topo I and other nuclear proteins, particularly the significance of such interactions to topo I function and the sensitivity of this enzyme to CPT. topo I is associated with actively transcribed regions of chromatin (68) and functions as a coactivator of transcription (7,8). Furthermore, topo I associates with the TATA-binding protein (7) and enhances transcription factor IID⅐IIA complex assembly during activation of RNA polymerase II-mediated transcription (69). The physical interaction between topo I and nucleolin has also underscored a c-Abl Activates DNA topo I potential role for topo I in RNA polymerase I-mediated transcription (55). In addition, binding of topo I to nucleolin is involved in the cellular localization of topo I (70). Other studies have shown that human topo I interacts with a RING finger/ Arg-Ser protein named topors (71) and with the splicing factor 2/arginine-serine-rich splicing factor RNA splicing factor (12)(13)(14).
This study has shown that topo I associates with the nuclear c-Abl kinase. The results show that the c-Abl SH3 domain interacts with the N-terminal 1-250 amino acids of topo I. The c-Abl SH3 domain binds to a proline-rich sequence with the consensus XPXXXXPXXP (72,73). A potential site for c-Abl SH3 domain binding is located at amino acids 225-233 (PV-FAPPYEP) in the conserved topo I core domain. The results also indicate that deletion of amino acids 1-210 interferes with c-Abl SH3 domain binding to topo I, perhaps by altering conformation of the proline-rich motif. Nucleolin (55), splicing factor 2/alternate splicing factor (14), topors (71), and SV40 T antigen (71,74) also bind to topo I within the N-terminal 210 amino acids, which are dispensable for topo I activity. Subdomains I and II of the topo I core form the top half or "cap" of the enzyme, bind to DNA, and are essential for the isomerase function (18). c-Abl also binds to DNA, prefers sequences containing an AAC motif, and exhibits a higher affinity for bent strands (75,76). However, our in vitro binding studies demonstrated that the interaction between topo I and c-Abl is not dependent on the presence of DNA. Direct binding experiments with purified c-Abl and GST-topo I demonstrated that the interaction is not mediated by DNA or any other protein (Figs.  1E and 2, A and B). In contrast to the findings with c-Abl, we have not detected complexes of topo I and c-Abl-related Arg kinase in cells. One potential explanation is that Arg is expressed predominantly in the cytosol and thereby would exhibit little interaction with nuclear topo I.
c-Abl Phosphorylates and Activates topo I-The present results further demonstrate that the interaction between topo I and c-Abl involves c-Abl-mediated phosphorylation of topo I. In vitro and in vivo studies identified topo I Tyr 268 as a c-Abl phosphorylation site. topo I Tyr 268 is located in the ␣2 helix of subdomain II, which extends away from the body of the molecule as it wraps around the DNA duplex (18). As such, the c-Abl phosphorylation site is present on the surface of topo I in a complex with DNA. By contrast, sites of mutations in topo I that confer resistance to CPT have been localized to areas of the core subdomains and the C-terminal domain that interact with DNA (18). These findings support a model in which topo I and c-Abl can interact when topo I is wrapped around the DNA duplex or when unbound to DNA. In this context, in vitro c-Abl Activates DNA topo I studies demonstrated that c-Abl bound and phosphorylated topo I in the absence or presence of DNA (data not shown).
The functional significance of c-Abl-mediated topo I phosphorylation is supported by the finding that this event is associated with stimulation of topo I activity. In vitro studies showed that c-Abl-phosphorylated topo I was more active in DNA relaxation assays. Moreover, studies in (i) MCF-7 and MCF-7/ c-Abl(K290R) cells and (ii) wild-type and c-Abl Ϫ/Ϫ MEFs demonstrated that topo I was phosphorylated at tyrosine by a c-Abl-dependent mechanism and that topo I activity was decreased in the absence of c-Abl activity. Of note, the demonstration that tyrosine phosphorylation of topo I was decreased by ϳ80% in c-Abl Ϫ/Ϫ MEFs and MCF-7/c-Abl(K290R) cells compared with controls supports phosphorylation of topo I in part by other tyrosine kinases. The findings also support a role for the Tyr 268 site in the regulation of topo I activity. Mutation of the Tyr 268 site was associated with the inhibition (but not complete abrogation) of c-Abl-mediated tyrosine phosphorylation in vitro and activation of topo I. These findings clearly indicate the possibility of multiple c-Abl-dependent phosphorylation sites in topo I. Taken together, these results demonstrate that topo I is subject to post-translational modification that affects the activity of topo I.
c-Abl-dependent Activation of topo I in the DNA Damage Response-The nuclear form of c-Abl is activated in the cellular response to DNA damage (45). Previous work has shown that c-Abl is activated by DNA-dependent protein kinase and ATM in cells exposed to IR and other genotoxic agents (40,42,43,45,46). The results of the present study demonstrate that IR induced c-Abl-dependent tyrosine phosphorylation of topo I. Thus, IR treatment of wild-type (but not c-Abl Ϫ/Ϫ ) MEFs results in phosphorylation of topo I at tyrosine. The results also demonstrate that phosphorylation of topo I by c-Abl in the IR response resulted in the stimulation of topo I activity. These findings support a model in which induction of c-Abl activity by DNA lesions transduces signals that confer activation of topo I.
Studies have linked topo I to the recognition of DNA lesions, including mismatched bases, abasic sites, cyclopyrimidine dimers, and deaminated cytosines (77)(78)(79)(80)(81). Moreover, genotoxic agents that cause DNA strand cross-links, such as mitomycin C, have been shown to induce topo I activity (82). Notably, mitomycin C-induced activation of topo I has been attributed to interactions between topo I and the p53 tumor suppressor (82)(83)(84). Other studies have demonstrated that topo I is subject to SUMO-1 (small ubiquitin-like modifier-1) modification in response to DNA damage (85). The association between topo I and wild-type p53 occurs transiently in response to genotoxic stress, whereas binding of topo I to mutant p53 is constitutive (84). c-Abl also binds to p53 in response to genotoxic stress and results in stabilization of the p53 protein by a mechanism involving Mdm2 (47,86). A proline-rich sequence in the C terminus of c-Abl is necessary for the binding of c-Abl to p53 (47,87). Other studies have demonstrated that c-Abl binds to the C terminus of p53 (amino acids 363-393) (88), whereas the topo I-binding site is located between amino acids 302 and 321 (84). Taken together, these findings and those in the present study indicate that both c-Abl and p53 contribute to the activation of topo I. c-Abl-dependent Activation and Sensitivity of topo I to CPT-Yeast cells devoid of topo I are resistant to the lethal effects of CPT (89,90). Moreover, cells that overexpress topo I are hypersensitive to CPT, and cells selected for resistance to CPT exhibit decreased levels of topo I expression (79, 90 -92). Other findings have demonstrated that certain mutations in topo I, particularly those involved in interactions with duplex DNA (18), confer CPT resistance (59). There is no available information, however, regarding the regulation of topo I activity as a factor in the sensitivity of cells to CPT. The present results demonstrate that c-Abl-mediated phosphorylation of topo I increases the isomerase activity. In concert with these observations, the results further show that c-Abl Ϫ/Ϫ cells are resistant to CPT-induced PLDBs. The c-Abl Ϫ/Ϫ cells also exhibit decreased levels of topo I activity as a consequence of loss of tyrosine phosphorylation and are less sensitive to CPT-induced apoptosis. These findings collectively demonstrate that c-Ablmediated phosphorylation is functionally important to topo I activity and sensitivity to topo I poisons. The results are expressed as percent PLDBs (means Ϯ S.E.) compared with that in control cells as determined from three separate experiments, each performed in duplicate. B, c-Abl ϩ/ϩ , c-Abl Ϫ/Ϫ , and c-Abl ϩ cells were labeled with [ 3 H]thymidine, treated with 10 M CPT for 1 h, and then assayed for DNA SSBs by alkaline elution. The results are expressed as SSB frequency in rad equivalents (73). C, c-Abl ϩ/ϩ , c-Abl Ϫ/Ϫ , and c-Abl ϩ cells were incubated with 10 M CPT for 30 min and harvested at 24 h. Control and treated cells were analyzed for sub-G 1 DNA by flow cytometry. Sub-G 1 DNA content was as follows: control c-Abl ϩ/ϩ cells, 1.7%; CPT-treated c-Abl ϩ/ϩ cells, 25.3%; control c-Abl Ϫ/Ϫ cells, 0.1%; CPT-treated c-Abl Ϫ/Ϫ cells, 0.5%; control c-Abl ϩ cells, 0.3%; and CPT-treated c-Abl ϩ cells, 25.0%.