Cell cycle phase-specific phosphorylation of human topoisomerase II alpha. Evidence of a role for protein kinase C.

Type II topoisomerases are essential for faithful cell division in all organisms. In human cells, the alpha isozyme of topoisomerase II has been implicated in catalyzing mitotic chromosome segregation via its action as a DNA unlinking enzyme. Here, we have shown that the enzymatic activity of topoisomerase II alpha protein purified from HeLa cell nuclei was strongly enhanced following phosphorylation by protein kinase C. We have investigated the possibility that this kinase is involved in cell cycle phase-specific phosphorylation of topoisomerase II alpha in HeLa cells. Two-dimensional tryptic phosphopeptide mapping revealed that topoisomerase II alpha protein immunoprecipitated from metabolically labeled HeLa cells was differentially phosphorylated during the G2/M phases of the cell cycle. To identify sites of phosphorylation, and the kinase(s) responsible for this modification, oligohistidine-tagged recombinant domains of topoisomerase II alpha protein were overexpressed in Escherichia coli and purified by affinity chromatography. Phosphorylation of a short fragment of the N-terminal ATPase domain of topoisomerase II alpha by protein kinase C in vitro generated two phosphopeptides that co-migrated with prominent G2/M phase-specific phosphopeptides from the HeLa cell-derived topoisomerase II alpha protein. Site-directed mutagenesis studies indicated that phosphorylation of serine 29 generated both of these phosphopeptides. Our results implicate protein kinase C in the cell cycle phase-dependent modulation of topoisomerase II alpha enzymatic activity in human cells.

Type II topoisomerases are essential for faithful cell division in all organisms. In human cells, the ␣ isozyme of topoisomerase II has been implicated in catalyzing mitotic chromosome segregation via its action as a DNA unlinking enzyme. Here, we have shown that the enzymatic activity of topoisomerase II␣ protein purified from HeLa cell nuclei was strongly enhanced following phosphorylation by protein kinase C. We have investigated the possibility that this kinase is involved in cell cycle phase-specific phosphorylation of topoisomerase II␣ in HeLa cells. Two-dimensional tryptic phosphopeptide mapping revealed that topoisomerase II␣ protein immunoprecipitated from metabolically labeled HeLa cells was differentially phosphorylated during the G 2 /M phases of the cell cycle. To identify sites of phosphorylation and the kinase(s) responsible for this modification, oligohistidine-tagged recombinant domains of topoisomerase II␣ protein were overexpressed in Escherichia coli and purified by affinity chromatography. Phosphorylation of a short fragment of the N-terminal ATPase domain of topoisomerase II␣ by protein kinase C in vitro generated two phosphopeptides that co-migrated with prominent G 2 /M phase-specific phosphopeptides from the HeLa cell-derived topoisomerase II␣ protein. Site-directed mutagenesis studies indicated that phosphorylation of serine 29 generated both of these phosphopeptides. Our results implicate protein kinase C in the cell cycle phase-dependent modulation of topoisomerase II␣ enzymatic activity in human cells.
In order for chromosomes to be faithfully transmitted from mother to daughter cells, DNA must be fully replicated and segregated evenly. Chromosome segregation can be affected only when all covalent DNA interlinks between replicated sister chromatids have been removed. The enzyme that catalyzes the disentanglement of replicated chromosomes via its ability to decatenate covalently interlinked duplex DNA molecules is DNA topoisomerase II, a highly conserved, homodimeric nuclear protein (see Wang (1985), Osheroff et al. (1991), Holm (1994), and Watt and Hickson (1994) for reviews). Evidence of a role for topoisomerase II in mitotic chromosome segregation has been derived largely from studies in lower eukaryotes.
Yeast mutants defective in topoisomerase II activity fail to remove all chromosomal interlinks at mitosis and subsequently incur chromosomal breakage as cell division is attempted in the absence of proper segregation (DiNardo, et al., 1984;Holm, et al., 1985;Uemura and Yanagida, 1986;Uemura, et al., 1987;Holm, et al., 1989;Rose and Holm, 1993;Spell and Holm, 1994). This defective segregation leads to a rapid decline in cell viability (Goto and Wang, 1984;Uemura and Yanagida, 1984). Because yeast cells contain a single topoisomerase II gene, it has been possible to study topoisomerase II function using conditional lethal mutants defective in topoisomerase II function at the restrictive growth temperature. Similar studies in human cells have been hampered both by a lack of suitable mutants deficient in topoisomerase II activity and by the presence of two closely related topoisomerase II isozymes. The human isozymes are termed topoisomerase II␣ (170-kDa form) and topoisomerase II␤ (180-kDa form) (Tsai-Pflugfelder et al., 1988;Drake et al., 1989;Chung et al., 1989;Jenkins et al., 1992;Austin et al., 1993) and are the products of distinct genes encoded on different chromosomes (Tsai-Pflugfelder et al., 1988;Tan et al., 1992;Jenkins et al., 1992).
Regulation of mitotic events in mammalian cells may require the action of a number of different protein kinases. The cyclindependent protein kinase, p34 cdc2 , is regarded as the master controller of mitotic events and phosphorylates a number of nuclear/nucleolar proteins, including histone H1 and nucleolin (reviewed by Norbury and Nurse (1992), Murray (1992), Nigg (1993), and Morgan (1995)). However, recent studies have implicated both mitogen-activated protein (MAP) 1 kinase and protein kinase C (PKC) in the regulation of certain mitosisspecific functions. The role of MAP kinase has not been defined in detail, although this kinase is implicated in the generation of the mitosis-specific phosphorylated epitope recognized by the MPM-2 antibody (Kuang and Ashorn, 1993;Westendorf et al., 1994). This epitope is found in a number of nuclear proteins, including topoisomerase II␣ and ␤ (Taagepera et al., 1993). At least one mitotic role for PKC, triggering the depolymerization of the nuclear lamina, has been proposed (Goss et al., 1994).
Although little is known of the mechanisms by which the function of topoisomerase II is regulated in mammalian cells, a number of different protein kinases have been implicated in the modulation of topoisomerase II enzymatic activity. In general, dephosphorylation of eukaryotic topoisomerase II enzymes leads to loss of activity (Saijo, et al., 1990;Cardenas and Gasser, 1993), whereas phosphorylation by casein kinase II or PKC causes a mild stimulation of activity (Ackerman et al., 1985;Rottman et al., 1987;Ackerman et al., 1988;Cardenas et al., 1992;Corbett et al., 1993aCorbett et al., , 1993b II is particularly noteworthy in that Saccharomyces cerevisiae or mouse topoisomerase II proteins that have been inactivated by dephosphorylation can be reactivated by this kinase (Saijo et al., 1990;Cardenas and Gasser, 1993).
Several previous studies have suggested that topoisomerase II␣ protein from mammalian cells is phosphorylated in vivo on multiple sites (Saijo et al., 1990(Saijo et al., , 1992Kroll and Rowe, 1991;Burden et al., 1993;Ganapathi et al., 1993;Kimura et al., 1994;Wells et al., 1994;Wells and Hickson, 1995). At least some of these sites of phosphorylation correspond to recognition sequences for casein kinase II (Wells et al., 1994). Moreover, topoisomerase II␣ protein is hyperphosphorylated during the G 2 and/or M phases of the cell cycle (Saijo et al., 1992;Burden et al., 1993;Wells and Hickson, 1995). However, although casein kinase II appears to phosphorylate yeast topoisomerase II protein in a cell cycle phase-specific manner (Cardenas et al., 1992), no evidence has been presented that this particular kinase is implicated in the M phase-specific hyperphosphorylation of topoisomerase II proteins from mammalian cells.
In this paper, we have studied the cell cycle phase-specific phosphorylation of human topoisomerase II␣ protein. We have identified a serine residue in the N-terminal ATPase domain of topoisomerase II␣ protein, which is modified specifically during the G 2 /M phases of the HeLa cell cycle. We have shown that this residue is a target in vitro for PKC and that phosphorylation of topoisomerase II␣ protein purified from HeLa cells by PKC strongly stimulates enzymatic activity in vitro.

MATERIALS AND METHODS
Cell Lines-HeLa S3 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 3 mM L-glutamine, and antibiotics in a humidified atmosphere containing 5% CO 2 at 37°C.
Purification of Human Topoisomerase II␣ Protein-All procedures were carried out at 4°C. Buffers contained the following protease and phosphatase inhibitors: 1 mM phenylmethylsulfonyl fluoride, 2 g/ml leupeptin, 1 g/ml aprotinin, 1 g/ml pepstatin A, 1 g/ml chymostatin, 1 g/ml soybean trypsin inhibitor, 1 mM benzamidine, 1 g/ml antipain, 50 g/ml L-1-chloro-3-(4-tosylamido)-7-amino-2-heptonone hydrochloride, 0.1 mM ␤-glycerophosphate, 0.1 mM p-nitrophenylphosphate, 0.5 mM glucose-1-phosphate, and 10 mM 2-mercaptoethanol. 40 liters of exponentially growing HeLa cells in suspension (approximately 3 ϫ 10 10 cells, provided by the Imperial Cancer Research Fund Cell Production Unit) were used as the source of topoisomerase II␣ protein. A 1 M sodium chloride extract of purified nuclei was prepared and separated by hydroxylapatite and phosphocellulose chromatography according to the method of Miller et al. (1981). Peak topoisomerase II activity eluted from these columns at 400 mM potassium phosphate and 300 mM potassium phosphate, pH 6.8, respectively. This partially purified activity was further purified by chromatography on fast protein liquid chromatography monoQ, phenyl-superose, and mono-S columns, according to the methods of Drake et al. (1987) and Strausfeld and Richter (1989).
Metabolic Labeling, Nuclear Extraction, Immunoprecipitations, and Phosphoamino Acid and Phosphopeptide Analyses-These procedures were performed as described by Wells et al. (1994). The isozyme-specific antibody used for analysis of the topoisomerase II␣ protein was desig-nated CRB and has been validated previously (Smith and Makinson, 1989;Wells et al., 1994).
In Vitro Phosphorylation Reactions-The PKC reaction buffer contained 50 mM Tris-HCl, pH 7.4, 0.25 mM EDTA, 0.75 mM CaCl 2 , 12.5 mM MgCl 2 , and 0.005% (v/v) Triton X-100, 200 M ATP, and, where required, 1-5 Ci of [␥-32 P]ATP (3,000 Ci/mmol, Amersham Corp.). Reactions were initiated by the addition of 0.1 unit of specific PKC isotypes purified from bovine brain, kindly provided by Dr. P. J. Parker (Imperial Cancer Research Fund), in a reaction volume of 20 l and were allowed to proceed for 10 min at 37°C.
DNA Sequencing-Nucleotide sequencing was performed on doublestranded plasmid templates using the dideoxy chain termination method and Sequenase enzyme (U. S. Biochemical Corp.).
Purification of an N-terminal Recombinant Fragment of Topoisomerase II␣ Protein-The region of the topoisomerase II␣ cDNA (Jenkins et al., 1992) encoding residues Lys-25 to Lys-168 was amplified using the polymerase chain reaction with the following primers (both written 5Ј to 3Ј): 5Ј primer, AGAGAGCTCGAGAAGAAAAGACTGTCTGTTGAA-AGA, and 3Ј primer, AGAGAGCTCGAGTTATTTGGCTCCATAGC-CATTTCGA. The purified polymerase chain reaction product was digested with XhoI and ligated into XhoI-digested pET14b (Invitrogen), which contains the T7 promoter driving expression of fusion proteins linked to an oligohistidine leader peptide. The plasmid was transformed into Escherichia coli BL21 (DE3), and transformants were grown to A 600 of 0.5. After the addition of isopropyl-1-thio-␤-D-galactopyranoside (0.4 mM) to induce expression from the T7 promoter in pET14b, bacteria were grown for another 4 h. The 18-kDa recombinant fragment was purified from crude cell lysates using affinity chromatography on a nickel chelate column as recommended by the suppliers (Invitrogen).
Site-directed Mutagenesis-The substitution of alanine for serine-29 was achieved by amplifying the region of the topoisomerase II␣ cDNA identical to that described above, with the exception that the 5Ј primer (5Ј-AGAGAGCTCGAGAAGAAAAGACTGGCTGTTGAAAGA-3Ј) contained a GCT codon (encoding alanine) replacing the TCT codon (serine-29). Preparation of the mutant protein was as described above.
Western Blotting-Following protein gel electrophoresis, Western blotting was performed according to the method of Towbin et al. (1979). Electroblotting onto nitrocellulose filters (Hybond-C Super, Amersham Corp.) was performed at 30 V overnight in transfer buffer (50 mM Tris-HCl, pH 7.5, 380 mM glycine, 0.1% (w/v) SDS, and 20% (v/v) methanol) in a Bio-Rad transblot cell. The filters were incubated for 60 min in blocking buffer (20 mM Tris-HCl, pH 7.5, 0.9% (w/v) NaCl, 0.05% (v/v) Tween 20, and 1% (w/v) Marvel low-fat milk powder). Primary antibody reactions were carried out for 2 h in blocking buffer containing the CRB antibody at a dilution of 1:200. CRB is a rabbit polyclonal antiserum raised to a C-terminal peptide of topoisomerase II␣, supplied by Cambridge Research Biochemicals, and has been previously shown to specifically recognize the ␣ isozyme of topoisomerase II (Smith and Makinson, 1989). The filters were washed in blocking buffer for 60 min with at least three changes of buffer before incubation with 125 I-protein A (Amersham Corp.) for 60 min. Filters were then washed extensively with Tris-buffered saline and then analyzed by autoradiography.
Cell Synchronization-HeLa cells were synchronized at the start of S phase by the use of a double thymidine block. Exponentially growing cells were treated with 2 mM thymidine for 14 h and then released into thymidine-free medium. After 11 h, 2 mM thymidine was again applied to the cells for a 15-h period before release into thymidine-free medium.
Flow Cytometry-Cells were fixed for 30 min in ice-cold 70% ethanol/ 30% phosphate-buffered saline collected by centrifugation and were treated with RNase A (100 g/ml final concentration) and propidium iodide (40 g/ml) in phosphate-buffered saline for 30 min at 37°C. Cell cycle distribution was then determined using a Beckton Dickinson FACScan, and the data were analyzed using the Lysis II software.
Protein Gel Electrophoresis-Proteins were separated by the discontinuous SDS-polyacrylamide gel system described by Laemmli (1970).

Purification of Topoisomerase II␣ Protein from HeLa
Cells-It has been shown previously that the enzymatic activity of topoisomerase II proteins purified from lower eukaryotes can be stimulated by phosphorylation with PKC (Ackerman et al., 1985;Sahyoun et al., 1986;Rottman et al., 1987;Corbett et al., 1993aCorbett et al., , 1993b. To provide evidence that PKC is capable of influencing the enzymatic activity of human topoisomerase II␣, HeLa cell nuclear extracts were fractionated by conventional chromatography and fast protein liquid chromatography, and the topoisomerase II␣ protein was purified to near homogeneity. This method of purification has been used previously to separate the ␣ and ␤ isozymes of human topoisomerase II (Drake et al., 1987;Strausfeld and Richter, 1989). The purified topoisomerase II preparation contained a predominant 170-kDa protein that was recognized by the CRB antibody specific for topoisomerase II␣ (Fig. 1). Several antibodies specific for the ␤ isozyme failed to detect any topoisomerase II␤ in the purified protein preparation (data not shown).
Regulation of Topoisomerase II␣ Activity by PKC-The purified topoisomerase II␣ protein was tested as a substrate for PKC in vitro. Fig. 2 shows that the purified toposimerase II␣ preparation was free from contaminating kinases and had no intrinsic autophosphorylation activity. However, the 170 kDa topoisomerase II␣ protein was a substrate in vitro for phosphorylation by 3 isoforms of PKC. To study the effects of phosphorylation by PKC, the enzymatic activity of the topoisomerase II␣ protein was assayed using supercoiled plasmid DNA as a substrate. Fig. 3 shows that the rate of plasmid relaxation catalyzed by the PKC-phosphorylated topoisomerase II␣ protein was increased substantially over that catalyzed by the unmodified enzyme.
Topoisomerase II␣ Protein Is Hyperphosphorylated During the G 2 /M Phases of the HeLa Cell Cycle-The ␣ isozyme of topoisomerase II is a phosphoprotein in mammalian cells, and the level of its phosphorylation is regulated in a cell cycle phase-dependent manner. Studies with human and rodent cell lines have indicated that topoisomerase II␣ protein is hyperphosphorylated during the G 2 /M phases of the mammalian cell cycle (Saijo, et al., 1992;Kroll and Rowe, 1991;Burden et al., 1993;Wells et al., 1994;Burden and Sullivan, 1994;Kimura et al., 1994;Wells and Hickson, 1995). In order to identify the G 2 /M phase-specific sites of phosphorylation on human topoisomerase II␣ protein, as well as the kinases responsible for this modification, HeLa cells were labeled with [ 32 P]orthophosphate, and the topoisomerase II␣ protein was immunoprecipitated with the isozyme-specific CRB antiserum. The topoisomerase II␣ protein was then digested with trypsin, and the resulting phosphopeptides were separated in two dimensions on thin layer chromatography (TLC) plates. Fig. 4 shows a comparison of tryptic phosphopeptide maps for topoisomerase II␣ protein derived either from an asynchronous culture of HeLa cells or from a culture synchronized via a double thymidine block and released into fresh thymidine-free medium for 8 h. Flow cytometric analysis revealed that the culture released from the cell cycle blockade contained 91% G 2 /M phase cells. A number of phosphopeptides that were either specific for or greatly enriched within the G 2 /M phase sample were evident (arrows in Fig. 4b). Among these phosphopeptides are several that we have shown previously to be dependent upon phosphorylation by a proline-directed kinase (identified by open arrowheads in Fig. 4) and represent phosphorylation of serine residues in the C-terminal regulatory domain of topoisomerase II␣ protein (Wells and Hickson, 1995). However, two of the most prominent G 2 /M phase-specific phosphopeptides (indicated by arrows labeled A and B in Fig. 4b) have not been identified in previous studies. Thus, we sought to identify the kinase(s) responsible for phosphorylation of the serine or threonine residue(s) present in phosphopeptides A and B. To determine the identity of the phospho-acceptor residues, phosphopeptides A and B were excised from the TLC plate and subjected to phosphoamino acid analysis. This revealed, in each case, that serine was the sole phospho-acceptor

residue (not shown).
Purification of Recombinant Domains of Human Topoisomerase II␣ Protein-In order to map phospho-acceptor residues, oligohistidine-tagged recombinant domains of topoisomerase II␣ protein were overexpressed in E. coli, purified by nickel chelate affinity chromatography, and phosphorylated in vitro by different protein kinases. Phosphorylation of the central breakage/reunion domain from residues Lys-701 to Arg-1217 and the C-terminal domain that commenced at Glu-1178 and terminated at the natural stop codon (prepared as described by Wells et al., 1994) by protein kinase A, PKC, casein kinase II, p34 cdc2 , and MAP kinase failed to yield phosphopeptides that co-migrated with phosphopeptides A and B identified in Fig. 4 (data not shown). Consequently, a portion of the N-terminal ATPase domain of topoisomerase II␣ protein, representing residues Lys-25 to Lys-168, was expressed in E. coli and purified to homogeneity. An SDS-polyacrylamide gel of the recombinant polypeptide is shown in Fig. 5. This 18-kDa polypeptide had no intrinsic autophosphorylation activity but was a substrate in vitro for PKC ␤I isoform (Fig. 5) and for PKC ␣ and ␥ (data not shown).
The ATPase Domain of Topoisomerase II␣ Protein Is Phosphorylated in Vivo-A representative two-dimensional tryptic phosphopeptide map of the 18-kDa N-terminal fragment phosphorylated in vitro by PKC is shown in Fig. 6b. Four prominent phosphopeptides were evident, two of which (Fig. 6b, 3 and 4) had a mobility similar to that of phosphopeptides A and B from the HeLa cell-derived topoisomerase II␣ protein. To determine whether the PKC-specific sites in the 18-kDa fragment and the G 2 /M phase-specific sites were identical, two approaches were undertaken. First, the entire in vitro and in vivo phosphopeptide samples were mixed, and the mixture was separated in two dimensions on TLC plates. Fig. 6 (a-c) shows that phosphopeptides 3 and 4 from the in vitro sample appeared to co-migrate with phosphopeptides A and B from the HeLa cell-derived FIG. 4. Differential phosphorylation of topoisomerase II␣ protein during the G 2 /M phases of the cell cycle. Two-dimensional tryptic phosphopeptide maps of topoisomerase II␣ protein extracted from asynchronously growing HeLa cells (a) or a culture enriched for G 2 /M phase cells (b). Phosphopeptides were separated in the horizontal dimension by electrophoresis at pH 1.9 (anode on left) and in the vertical dimension by chromatography. The positions of the G 2 /M phase-specific phosphopeptides identified in previous studies (Wells and Hickson, 1995) 5. The 18-kDa N-terminal domain of human topoisomerase II␣ protein is a substrate in vitro for PKC. The E. coli-expressed 18-kDa fragment was electrophoresed on a 12% polyacrylamide gel and stained with Coomassie Blue (lane C). The recombinant protein was then incubated in the presence (lane 1) or absence (lane 2) of PKC␤I and electrophoresed on a 12% gel, which was exposed to x-ray film after drying. Lane 3 shows the PKC preparation in the absence of the 18-kDa topoisomerase II␣ fragment. The position of the topoisomerase II␣ protein fragment is indicated by the arrows. The sizes (in kDa) of molecular mass standards run in parallel are shown on the right. Note the selective increase in intensity of phosphopeptides A and B from the in vivo sample due to co-migration with phosphopeptides from the in vitro sample. d, a mix of the sample in a with phosphopeptide 3 extracted from a thin layer plate. Note the selective increase in intensity in phosphopeptide A due to co-migration with phosphopeptide 3. sample. To confirm this co-migration, single phosphopeptides from the in vitro sample were excised from the chromatography plate and mixed with the entire in vivo sample prior to separation of the mixtures on TLC plates as before. Fig. 6d shows that phosphopeptide 3 co-migrated with phosphopeptide A from the in vivo sample. The 18-kDa fragment was not phosphorylated efficiently in vitro by any of a range of other purified protein kinases tested, in particular those previously implicated in regulating topoisomerase II activity (Sahyoun et al., 1986;Ackerman et al., 1988;Saijo et al., 1990;Cardenas et al., 1992;Wells et al., 1994;Wells and Hickson, 1995), including casein kinase II, p34 cdc2 kinase, MAP kinase, and Ca 2ϩ /calmodulin-dependent protein kinase (data not shown).
Site-directed Mutagenesis of the cDNA Encoding the 18-kDa Fragment of Topoisomerase II␣ Protein-The amino acid sequence of the 18-kDa protein contains a number of potential serine phospho-acceptor residues, although only one, serine-29, lies in a sequence context closely matching the consensus for a recognition site for PKC ((R/K)(R/K)XS). Site-directed mutagenesis of the cDNA encoding the 18-kDa protein was carried out to confirm that serine-29 was the phospho-acceptor residue that gave rise to phosphopeptides 3 and 4. Fig. 7 shows that a two-dimensional tryptic phosphopeptide map generated by PKC-mediated phosphorylation of the purified 18-kDa protein containing a single amino acid substitution replacing serine-29 with alanine lacked both phosphopeptides 3 and 4.
To confirm that serine-29 is the only possible phospho-acceptor residue in peptide 3 and, therefore, to prove it is this residue that is modified in vivo, as well as in vitro, a synthetic peptide representing residues 24 -39 was synthesized. This peptide, which contains only one serine residue (serine-29), was phosphorylated in vitro by PKC␤I and then run on thin layer plates after trypsin digestion. Fig. 8 shows that a single strong tryptic phosphopeptide was generated following complete digestion, which was shown in mixing experiments to co-migrate with peptide 3 from the trypsin-digested recombinant N-terminal domain. To confirm the co-migration of these phosphopeptides, the level of radioactivity in each phosphopeptide was also quantified by volume integration using the ImageQuant program on a Molecular Dynamics PhosphorImager. The values obtained were as follows: peptide alone, 39,900; phosphorylated N-terminal recombinant domain alone, 84,600 (peptide 3) and 84,900 (peptide 2); and a mix of synthetic peptide and Nterminal recombinant domain, 126,000 (peptide 3) and 81,000 (peptide 2). These data show that the major phosphopeptide derived from the synthetic peptide co-migrates with phosphopeptide 3 from the recombinant N-terminal protein. This confirms that serine-29 is the target residue for PKC in vitro and is the only possible phospho-acceptor residue present in peptide A from the in vivo labeled topoisomerase II␣ protein. DISCUSSION We have identified serine-29 as a site of phosphorylation of topoisomerase II␣ protein from the human HeLa cell line and have shown that this residue is a substrate for PKC in vitro. Phosphorylation on serine-29 is increased greatly as HeLa cells traverse the G 2 /M phases of the cell division cycle. Further, we have shown that phosphorylation by PKC substantially increases the catalytic activity of purified topoisomerase II␣ protein in vitro.
Many mitosis-specific events in mammalian cells are regulated by the action of protein kinases. The most extensively characterized of the mitosis-activating kinases is the p34 cdc2cyclin B complex. Phosphorylation of target proteins by p34 cdc2cyclin B initiates many of the hallmark events in mitosis, such as nucleolar disassembly and chromatin condensation (reviewed by Norbury and Nurse, 1992;Murray, 1992;Nigg, 1993;Morgan, 1995). However, recent studies have indicated that kinases other than p34 cdc2 -cyclin B are intimately involved in the regulation of mitotic events. For example, MAP kinase appears to be at least one of the kinases responsible for the generation of the mitosis-specific phosphorylated epitope recognized by the MPM-2 antibody (Kuang and Asham, 1993;Westendorf et al., 1994). Moreover, although many isoforms of PKC are cell membrane-associated and unlikely, therefore, to be responsible for modulating nuclear events, certain PKC isoforms are clearly critical for cell cycle traverse and are strongly implicated in affecting an essential nuclear mitotic function (reviewed by Clemens et al. (1992)). Current evidence suggests that the translocation of the ␤II isoform from the cytoplasmic membrane to the nucleus and the subsequent phosphorylation of nuclear lamins at the G2/M phase transition are necessary for the depolymerization of the nuclear lamina (Goss et al., 1994). This was a role previously assigned to p34 cdc2 -cyclin B. Indeed, PKC is known to be required for the G 2 /M phase transition in at least some cell types, and the expression of the ␤II isoform has been shown to be essential for proliferation in a human leukaemic cell line (Usui et al., 1991;Levin et al., 1990;Murray et al., 1993). It is not unreasonable to assume, therefore, that PKC ␤II or another isoform of PKC that is located in the nucleus, such as PKC, is involved in the regulation of other nuclear factors that are required during mitosis. Our data are consistent with the proposal that activation of topoisomerase II␣ during mitosis is mediated, at least in part, by PKC. However, we cannot rule out the possibility that kinases other than or in addition to PKC modify serine-29 of topoisomerase II␣ in mitotic cells. Based upon the sequence context in which serine-29 lies and the data presented in this paper, it seems highly unlikely that any of the kinases previously implicated in modifying either lower eukaryotic topoisomerase II proteins in vivo (casein kinase II and p34 cdc2 kinase; Ackerman et al., 1988;Cardenas et al., 1992;Shiozaki and Yanagida, 1992), or the human topoisomerase II␣ protein in vivo (casein kinase II and a proline-directed kinase; Wells et al., 1994;Wells and Hickson, 1995) are directly responsible for the modification of serine-29.
A number of previous studies have suggested that PKC may be involved in the regulation of topoisomerase II functions. The most extensively characterized system for an analysis of the effects of phosphorylation on topoisomerase II activity to date has been Drosophila. Osheroff and colleagues have shown that phosphorylation by PKC enhances the activity of Drosophila topoisomerase II approximately 2.5-fold and that this activation is mediated via an enhancement in the rate of ATP hydrolysis (Corbett et al., 1993a(Corbett et al., , 1993b. Moreover, topoisomerase II phosphorylated by PKC has an altered susceptibility to inhibition by certain antineoplastic agents that interfere with the catalytic cycle of topoisomerase II . Similarly, Rottmann et al. (1987) showed that PKC phosphorylates topoisomerase II from the sponge, Geodia cydonium, in vivo and increases its activity around 2.5-fold in vitro. Our data on human topoisomerase II␣ protein are in agreement with those obtained with the Geodia and Drosophila topoisomerase II proteins. Moreover, the location of the PKC target serine residue in the N-terminal ATPase domain of human topoisomerase II␣ protein suggests an interesting possibility that this phosphorylation event is linked directly to alterations in ATP binding/hydrolysis. However, it should be noted that phosphorylation of Drosophila topoisomerase II by either casein kinase II or PKC (both of which apparently target the C-terminal domain) causes an increase in the rate of ATP hydrolysis (Ackerman et al., 1988;Corbett et al., 1992Corbett et al., , 1993aCorbett et al., , 1993b. Indeed, Ackerman et al. (1988) have provided evidence that casein kinase II is the predominate activity responsible for phosphorylating topoisomerase II in cultured Drosophila cells. Similarly, our previous data implicate casein kinase II as a key regulator of human topoisomerase II␣ (Wells et al., 1994). The data presented here indicate both that the phosphorylation of serine-29 of human topoisomerase II␣ is almost certainly not mediated by casein kinase II and that this strongly cell cycle-regulated modification would not necessarily have been detected in previous studies in which asynchronously growing cell cultures were employed.
There are a number of possible functions that phosphorylation of human topoisomerase II␣ protein might perform. Considering the G 2 /M phase-specific nature of the modification of serine-29, it would seem likely that this modification is required to affect a mitosis-specific function. Our data indicate that phosphorylation by PKC can substantially increase the catalytic activity of purified human topoisomerase II␣ protein. Indeed, the extent to which PKC can activate topoisomerase II␣ in vitro may be a significant underestimate of the true effect of PKC on activity in vivo, because the purified enzyme was likely to be already in at least a partially phosphorylated state. It is possible that this activation may be linked to a requirement for a highly efficient catalytic activity throughout the short time period in the G 2 and/or M phases in which chromosome segregation must be affected by topoisomerase II.
The data presented here indicate that phosphorylation of serine-29 gives rise to two tryptic phosphopeptides. We would suggest that this occurs via differential digestion of the protein by trypsin, which is known to occur at sites of adjacent lysine and arginine residues (Campbell et al., 1986). Indeed, serine-29 lies in a sequence that contains three consecutive target residues for trypsin (KKRLS 29 ; in the one-letter amino acid code), which would appear to provide the opportunity for differential digestion to occur.
Few of the previous studies on topoisomerase II phosphorylation in lower or higher eukaryotic cells have analyzed the location of phospho-acceptor residues. In those studies where sites have been mapped definitively or predicted from the mobility of phosphopeptides on TLC plates, it is the C-terminal domain that has been implicated as the major target for kinases. In budding yeast topoisomerase II protein, the C-terminal domain is proposed to be a target for multiple phosphorylations by casein kinase II (Cardenas et al., 1992). Moreover, several of these C-terminal sites appear to be hyperphosphorylated at mitosis. Similarly, we have shown previously that human topoisomerase II␣ protein is phosphorylated in vivo on two serine residues in the C-terminal domain by casein kinase II (Wells et al., 1994). The only previous data implicating the N-terminal ATPase domain of topoisomerase II as a target for phosphorylation have come from the work of Shiozaki and Yanagida (1992) on the fission yeast topoisomerase II enzyme. Their study indicated that phosphorylation was implicated in controlling nuclear localization. Clearly, therefore, PKC-mediated phosphorylation of the N-terminal domain of human topoisomerase II␣ protein could regulate enzymatic activity and/or nuclear localization. Studies are in hand to address these possibilities.
We have examined the possibility that an N-terminal site for PKC-mediated phosphorylation is conserved in other eukaryotic topoisomerase II enzymes (see Caron and Wang (1994) for a review of sequence homologies). In human topoisomerase II␣, serine-29 lies in a sequence context that comprises predominantly basic amino acids. This sequence motif is conserved in all mammalian topoisomerase II␣ enzymes. Moreover, this motif is also conserved in the ␤ isozyme of human topoisomerase II. Whether this indicates that topoisomerase II␤ protein is also a target for PKC in vivo will require further studies. Although accurate alignment of the human and lower eukaryotic topoisomerase II sequences is difficult due to general sequence divergence, there are potential target serine residues for PKC near to the N terminus of the yeast and Drosophila topoisomerase II proteins.
In summary, we have shown that PKC can modulate the enzymatic activity of human topoisomerase II␣ and have identified a serine residue in the ATPase domain of the protein that is phosphorylated specifically during the G 2 and/or M phases of the HeLa cell cycle. The challenge is now to identify the precise role of the cell cycle-regulated phosphorylations that occur on topoisomerase II␣ protein and that appear to require the action of at least two distinct kinases (Wells and Hickson (1995) and this work).