Casein Kinase II Catalyzes a Mitotic Phosphorylation on Threonine 1342 of Human DNA Topoisomerase IIα, Which Is Recognized by the 3F3/2 Phosphoepitope Antibody*

The 3F3/2 antibody recognizes a phosphoepitope that is implicated in the mitotic checkpoint regulating the metaphase-to-anaphase transition. Immunoprecipitation and Western blotting revealed that the 3F3/2 antibody binds to human DNA topoisomerase II α (HsTIIα) from mitotic but not interphase HeLa cells. Extracts from mitotic cells efficiently catalyzed the formation of the 3F3/2 phosphoepitope on fragments of HsTIIα expressed in bacteria. Expression and site-directed mutagenesis of various HsTIIα protein fragments mapped the 3F3/2 phosphoepitope to the region of HsTIIα containing phosphorylated threonine 1342. This threonine lies within a consensus sequence for phosphorylation by casein kinase II (CKII). CKII is present in cellular extracts and is associated with isolated mitotic chromosomes. The 3F3/2 phosphoepitope kinase present in mitotic cell extracts was able to create the epitope using GTP and was inhibited by heparin. A kinase associated with the isolated chromosomes also generated the 3F3/2 phosphoepitope on HsTIIα. Recombinant CKII catalyzed the formation of the 3F3/2 phosphoepitope on fragments of HsTIIα containing threonine 1342. These results indicate that the mitotic 3F3/2 phosphoepitope kinase activity is attributable to CKII. We suggest that the 3F3/2 phosphoepitope reflects a CKII-catalyzed phosphorylation of threonine 1342 that may regulate mitotic functions of HsTIIα.

DNA topoisomerase II (TII) 1 catalyzes the passage of one double-stranded DNA molecule through another doublestranded DNA molecule in an ATP-dependent manner by creating a transient DNA double strand break, passing the intact double strand through this break, and then ligating the break. This activity alters DNA topology during catenation/decatenation, enables knotting/unknotting, and relaxes DNA supercoils generated during DNA replication and RNA transcription (reviewed in Ref. 1). Eukaryotic cells require TII for survival. It is essential for proper chromosome condensation during mitosis and meiosis and for segregation of sister chromatids during anaphase (2)(3)(4). In addition, TII is thought to play a role in the maintenance of both interphase chromatin structure (5) and mitotic chromosome structure (6,7). SMC (stable maintenance of chromosomes) proteins and TII, the major components of the mitotic chromosome scaffold, have been shown to associate. It appears that TII and SMC proteins act collectively to maintain the structural integrity of the condensed chromosome (reviewed in Ref. 8). Many anti-cancer reagents inhibit the activity of TII. The majority of these therapeutic topoisomerase poisons act by blocking the break-ligation reaction, resulting in a trapped intermediate termed the cleavable complex (9 -11).
Two isoforms of topoisomerase II exist in higher eukaryotes, topoisomerase II ␣ (TII␣) and topoisomerase II ␤ (TII␤) (12)(13)(14)(15)(16). TII␣ is expressed in a cell cycle-dependent manner, and protein levels are greater in proliferating cells than in quiescent cells (17,18). During G 1 , the level of TII␣ is at its lowest. Levels begin to increase prior to S phase, remain relatively stable through S, and increase again and peak in late G 2 (19 -21). Decreased levels are due to marked reductions in mRNA stability and protein stability that result in the low levels of TII␣ observed in G 1 (22)(23)(24). In contrast, the level of TII␤ remains relatively constant throughout the cell cycle and accounts for 20 -30% of total TII in proliferating cells (19,20). Yeast and Drosophila possess one form of TII. Earlier studies indicated that TII protein levels remained similar throughout the cell cycle in these organisms (25,26), but subsequent studies have shown that yeast and Drosophila TII are expressed in a cell cycle-dependent manner similar to the expression of TII␣ in other eukaryotes (27,28).
TII␣ is also phosphorylated in a cell cycle-dependent manner. Phosphorylation of TII␣ is maximal in G 2 /M phase. Burden et al. (21) detected phosphoserine residues on Chinese hamster TII␣ in all phases of the cell cycle, whereas phosphothreonine residues were detected only in M phase cells. Several protein kinases including p34 cdc2 , protein kinase C, and casein kinase II (CKII) have been shown to phosphorylate TII␣. In G 2 /M phases, p34 cdc2 is thought to phosphorylate serines 1212 and 1246 of human TII␣ (HsTII␣) (29). Protein kinase C also phosphorylates HsTII␣ in G 2 /M phases on serine 29 (30). Taagepera et al. (31) found that mouse TII␣ is the major mitotic chromosomal protein recognized by MPM-2, an antibody that binds to certain mitosis-specific phosphorylations. TII is phosphorylated by CKII in yeast, Drosophila, and human cells (reviewed in Ref. 32). Wells et al. (33) reported that the major phosphorylation target sites for CKII on HsTII␣ were serine 1376 and serine 1524. CKII has also been reported to phosphorylate HsTII␣ threonine 1342 (34). In yeast, CKII differentially phosphorylates TII in a cell cycle-dependent manner. Some CKII phosphoacceptor sites are preferentially phosphorylated in mitosis, while others are preferentially phosphorylated in G 1 (27). CKII and TII form stable complexes in which both enzymes' respective catalytic activities are maintained even after the phosphorylation of TII by CKII (35). In addition, it has been reported that binding of CKII to HsTII␣ stabilizes enzymatic activity of HsTII␣ without phosphorylation (36).
The consequences of TII phosphorylation on the activity of the enzyme have been examined in several studies. In yeast, dephosphorylation of TII eliminates decatenation activity, which can be restored by CKII phosphorylation (37). CKII phosphorylation of Drosophila TII also enhances enzymatic activity 2-3-fold (38). In mammals, the correlation between TII phosphorylation and its activity is less clear. Phosphorylation of HsTII␣ by protein kinase C has been reported to increase its activity (28). Saijo et al. (39) have reported that dephosphorylated murine TII from Swiss 3T3 cells is almost completely inactive and that phosphorylation by a copurifying kinase increased its activity 8.6-fold over that of the dephosphorylated TII. In contrast, Redwood et al. (36) found that phosphorylation of HsTII␣ by CKII did not increase decatenation activity. Kimura et al. (40) also indicated that murine TII␣ from FM3A cells was unaffected either by phosphatase treatment or by phosphorylation with CKII. Phosphorylation of TII has been linked to chemotherapeutic drug resistance; phorbol 12-myristate 13-acetate induced hyperphosphorylation of HsTII by CKII in HL-60 cells treated with the anticancer drug etoposide, reduced the formation of cleavable complexes, and decreased cytotoxicity (41). The precise roles of the phosphorylations of mammalian TII and their effects upon its postulated activities have yet to be fully characterized.
The monoclonal phosphoepitope antibody, 3F3/2, was originally prepared by Cyert et al. (42) against Xenopus egg extracts that had been supplemented with ATP␥S. They found that the antibody recognized a large variety of thiophosphorylated proteins. Subsequently, we discovered that the antibody recognizes a small number of native phosphoproteins from mitotic cells that were never exposed to ATP␥S. In addition, immunofluorescence and immunoelectron microscopy revealed that the antibody bound to proteins in kinetochores and centrosomes (43). Li et al. demonstrated that levels of expression of the 3F3/2 kinetochore phosphoepitope were linked to tension created by kinetochore and spindle microtubule interactions (44). Moreover, when the 3F3/2 antibody was microinjected into living cells, the cells arrested at metaphase (45). These results suggested that a 3F3/2 phosphoepitope may participate in a tension-sensitive signaling pathway and is a component of the cell cycle checkpoint regulating entry into anaphase.
In this study, we have found that the 3F3/2 antibody binds to HsTII␣ extracted from mitotically arrested cells. We have identified the specific phosphorylation site on HsTII␣ required for the generation of the 3F3/2 phosphoepitope, investigated the cell cycle dependence of the phosphorylation, and characterized a mitotic kinase activity that can create this phosphoepitope.

Materials
Tissue culture reagents including Dulbecco's modified Eagle's medium and nonessential amino acids were purchased from Life Technologies, Inc. (Gaithersburg, MD). Demecolcine, leupeptin, pepstatin A, Chaps, Triton X-100, EGTA, Trizma (Tris base), ATP, GTP, heparin, penicillin G, streptomycin, and ␤-mercaptoethanol were purchased from Sigma. DNase and Pefabloc SC were purchased from Boehringer Mannheim. Coomassie Protein Reagent Assay and protein A trisacryl beads were purchased from Pierce. Immobilon-P Western blotting polyvinylidene fluoride membrane was purchased from Millipore (Bedford, MA). The Renaissance Chemiluminescence Kit was purchased from NEN Life Science Products. Oligonucleotide primers were synthesized by NBI (Plymouth, MN). The Pet30 System and S-Tag TM detection kit were purchased from Novagen (Madison, WI). The QuickChange TM site-directed mutagenesis kit was purchased from Stratagene (La Jolla, CA). Purified human recombinant casein kinase II was purchased from New England Biolabs (Beverly, MA).
Antibodies-The antiphosphoepitope antibody 3F3/2 was prepared as an ascites fluid in the Lymphocyte Culture Center at the University of Virginia. Anti-topoisomerase II ␣ mouse monoclonal antibodies SWT3D1 and SWR1C2, herein designated T3D1 and R1C2, respectively, were originally produced by Robinson et al. (46)  Peptides-The 16-amino acid peptide, FSDFDEKTDDEDFVPC, and the threonine-phosphorylated version were synthesized, purified, and characterized by mass spectroscopy by QCB (Hopkinton, MA). The first 15 residues of this peptide correspond to amino acids 1335-1349 of HsTII␣ with the phosphorylated threonine corresponding to residue 1342.

Cell Culture
HeLa S3 cells (ATCC CCL 2.2) were grown in 1-liter spinner flasks in Dulbecco's modified Eagle's medium supplemented with 5% (v/v) bovine calf serum, 20 mM Hepes, pH 7.2, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 0.05% (w/v) pluronic F68, 60 g/ml penicillin G, and 100 g/ml streptomycin sulfate. Cells defined as cycling cells were harvested from logarithmically growing cultures. Cells were arrested in S phase by the addition of 1 g/ml aphidicolin and subsequent culture for 18 h. To arrest cells in M phase, cells were incubated for 18 h in the presence of 0.15 g/ml demecolcine. The mitotic index, the percentage of cells containing condensed chromosomes, was determined for cycling, S phase, and M phase cells by propidium iodide fluorescent staining analysis. The mitotic index in cycling cells ranged from 4 to 10%, S phase cells from 0 to 1%, and M phase cells from 70 to 95%. Harvested cells were washed twice in 4°C phosphate-buffered saline by centrifugation prior to extraction and use in various assays. Cell pellets not immediately used in assays were frozen in liquid nitrogen and stored at Ϫ70°C.

Chromosome and Nuclei Isolation
HeLa S3 chromosomes were isolated from mitotically arrested cultures as described previously (47). Briefly, M phase-arrested cells were centrifuged, rinsed twice in swelling buffer (10 mM Hepes, pH 7.4, 40 mM KCl, 5 mM EGTA, 4 mM MgSO 4 , protease inhibitor mixture), and lysed by vigorous pipetting in 4°C extraction buffer (60 mM Pipes, 25 mM Hepes, pH 6.9, 10 mM EGTA, 4 mM MgSO 4 , 1 mM DTT, protease inhibitor mixture, and either 1% CHAPS or 0.5% Triton X-100). When indicated, the extraction buffer also contained 200 nM microcystin-LR, a serine-threonine phosphatase inhibitor. Protease inhibitor mixture additions result in final concentrations of 5 g/ml pepstatin A, 5 g/ml leupeptin, and 5 g/ml Pefabloc SC. Centrifugation at 200 ϫ g for 5 min at 4°C removed nonmitotic nuclei and other cellular debris from the suspended chromosomes. The exclusion of nuclei was confirmed by propidium iodide fluorescent staining analysis. Chromosomes were pelleted by centrifugation at 1600 ϫ g for 10 min at 4°C and washed three times in extraction buffer by repeated centrifugation. Nuclei were isolated from cycling cell cultures (described above) using identical buffers and techniques as used during the isolation of chromosomes with the exception that the nuclei were pelleted from cellular debris and extract by centrifugation at 64 ϫ g for 5 min at 4°C. The nuclei preparations contained greater than 99% nonmitotic nuclei as determined by propidium iodide fluorescent staining analysis. Chromosomes or nuclei not immediately used in assays were frozen in liquid nitrogen and stored at Ϫ70°C.

Immunoprecipitations
Immunoabsorbents-Protein A trisacryl beads were washed three times by centrifugation in 50 mM Tris, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, and 0.1% SDS (RIPA). 100 l of protein A bead slurry was then combined with 800 l of RIPA and either 1 l of 3F3/2 ascites or 150 l each of T3D1 and R1C2 anti-human topoisomerase II ␣ antibodies. The mixtures were agitated for 2 h at 4°C. Antibody-conjugated protein A beads were washed three times in RIPA containing 200 nM microcystin-LR and protease inhibitor mixture.
Chromosomal Extracts-Chromosomes were treated with 0.1 units/l DNase for 5 min in extraction buffer (described above) at 37°C. SDS was added to achieve a concentration of 1% (w/v), and the sample was vortexed. RIPA, without any SDS but containing 200 nM microcystin-LR and protease inhibitor mixture, was then added to adjust the final SDS concentration to 0.1%. The extract was centrifuged at 15,000 ϫ g for 15 min, and the supernatant was retained for immunoprecipitation.
Precipitations-The chromosomal extracts were combined with the antibody-conjugated beads and gently agitated at 4°C for 2-3 h. The immunoprecipitates were then washed three times by centrifugation with RIPA containing 200 nM microcystin-LR and protease inhibitor mixture. Supernatants were also retained for analysis. For SDS-PAGE and immunoblotting analysis, the immunoprecipitates and supernatants were treated with SDS-PAGE loading buffer and heated at 95°C for 5 min.

Endogenous Chromosome Kinase Assay
Chromosomes were isolated as described above in the absence of phosphatase inhibitors. The chromosomes were then suspended in extraction buffer with and without 200 nM microcystin-LR. ATP was added to 1 mM, and the chromosomes were incubated at 37°C for 20 min. Chromosomes were then treated with 0.1 units/l DNase for 5 min at 37°C, mixed 1:1 (v/v) with SDS-PAGE loading buffer, and heated at 95°C for 5 min for gel electrophoresis and immunoblotting.

Cell Extract Preparations
For SDS-PAGE analysis, whole cell extracts were prepared by dissolving cell pellets in SDS-PAGE loading buffer. Soluble cell extracts for SDS-PAGE were prepared by lysing cell pellets in 50 mM Tris-HCl, pH 7.5, 10 mM EGTA, 4 mM MgSO 4 (TEM) containing 0.5% Triton X-100, 200 nM microcystin-LR, and protease inhibitor mixture. The lysate was centrifuged at 15,000 ϫ g for 15 min at 4°C, and the resulting supernatant was mixed 1:1 (v/v) with 2ϫ SDS-PAGE loading buffer. Isolated chromosomes or nuclei were treated with 0.1 units/l DNase for 5 min in extraction buffer (described above) and mixed 1:1 (v/v) with 2ϫ SDS-PAGE loading buffer. All SDS-PAGE samples were heated at 95°C for 5 min prior to electrophoresis. Cellular extracts used in phosphorylation assays were prepared by lysing freshly isolated cells in TEM containing 200 nM microcystin-LR, 1% Chaps, 1 mM DTT, and protease inhibitor mixture or by thawing frozen cells in TEM containing 200 nM microcystin-LR, 1 mM DTT, and protease inhibitor mixture. These mixtures were intermittently vortexed and kept on ice for 5 min. The extracts were cleared by 15,000 ϫ g centrifugation at 4°C for 15 min. Protein concentrations for the extracts were determined using a Coomassie protein assay.

Gel Electrophoresis and Immunoblotting
Proteins were separated by electrophoresis using 5-20% gradient or 6% SDS-polyacrylamide gels and transferred to Immobilon-P membranes. For Western analysis, the membranes were blocked for 30 min with 10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.05% Tween 20 (TBST) containing 5% bovine serum albumin and 0.02% sodium azide. For detection of proteins recognized by the 3F3/2 antibody, 3F3/2 ascites was diluted 1:5000 in TBST and incubated with the membrane for 1-2 h. For detection of human topoisomerase II ␣, T3D1 and R1C2 mouse monoclonal antibodies from hybridoma culture supernatants were diluted 1:1000 in TBST and incubated with the membrane for 1-2 h. For detection of CKII, rabbit anti-CKII ␣ subunit antibody at 1 g/ml in TBST was incubated with the membrane for 1-2 h. After washing in TBST, membranes were incubated with peroxidase-conjugated goat anti-mouse IgG or goat anti-rabbit IgG antibodies diluted 1:20,000 to 1:30,000 in TBST for 1 h. Immunoblots were washed again in TBST and then developed using the Renaissance chemiluminescence system.
Immunoblotting inhibition assays for the 3F3/2 antibody were carried out by incubating 3F3/2 ascites at a 1:25,000 dilution in TBST with either the phosphopeptide or the corresponding nonphosphorylated peptide (described above). The peptide concentrations ranged from 0.1 to 100 g/ml. The 3F3/2 antibody was incubated with the peptides for 1 h at 4°C. This solution was then used for immunoblotting as described above.

Bacterial Expression of HsTII␣ Protein Fragments
A Bluescript II KS ϩ/Ϫ plasmid containing the complete coding sequence for HsTII␣ was obtained from the ATCC. This sequence was digested using appropriate restriction endonucleases, and the resulting DNA fragments were inserted into the pET30 vector (pET System, Novagen). pET30 vectors are bacterial expression vectors that encode an S-Tag TM and a 6ϫ histidine sequence on the N terminus of expressed proteins. When necessary, oligonucleotide primers containing restric-tion endonuclease sites were designed and used in conjunction with polymerase chain reaction methods to create sequences encoding portions of HsTII␣ suitable for insertion into pET30 vectors. Mutagenesis of HsTII␣ coding sequences was achieved using the QuickChange TM site-directed mutagenesis kit. Sequence analysis of mutagenized constructs was performed by the University of Virginia Biomolecular Research Facility using dye terminator chemistry and an Applied Biosystems 377 Prism DNA sequencer. BL21(DE3) Escherichia coli were transformed with pET30 vectors containing HsTII␣ constructs and induced to express HsTII␣ protein fragments. Synthesized proteins were separated by SDS-PAGE (described above), transferred to Immobilon-P membranes, and detected using the S-Tag TM alkaline phosphatase-conjugated detection system. In addition, these HsTII␣ fragments were used as substrates for the 3F3/2 kinase assay described below.

3F3/2 Phosphoepitope Kinase Assay on HsTII␣ Protein Fragments
Fragments of HsTII␣ protein were separated by gel electrophoresis and transferred to Immobilon-P membranes as described above. The membranes were then cut into strips. Nonspecific protein binding sites were blocked by incubating the strips with TBST containing 5% bovine serum albumin and 0.02% sodium azide for 15 min. Strips were rinsed twice in TEM and then equilibrated in TEM containing 1 mM DTT, 200 nM microcystin-LR, and protease inhibitor mixture. After equilibration, the strips were placed in either HeLa S3 cellular extracts (prepared above) or casein kinase II diluted in TEM containing 1 mM DTT, 200 nM microcystin-LR, and protease inhibitor mixture. When indicated, 1 mM ATP or GTP was added. Final protein concentrations of cellular extracts used in this assay ranged from 20 to 150 g/ml. To compare kinase activity between different cellular extracts, equivalent extract protein concentrations were used. Heparin at 50 g/ml was included in the assay to test its inhibitory effects. The reactions were incubated at room temperature or 37°C for 3-30 min. Multiple washes in TBST terminated the reactions. The strips were then immunoblotted with the 3F3/2 antibody as described above.

RESULTS
3F3/2 Anti-phosphoepitope and Anti-HsTII␣ Antibody Immunoblotting-Whole mitotic cells, extracts from mitotic cells, and mitotic chromosomal extracts isolated in the presence of the serine/threonine phosphatase inhibitor, microcystin-LR, were immunoblotted with the 3F3/2 anti-phosphoepitope antibody (Fig. 1). In whole cell and soluble cell extract preparations the major immunoreactive proteins were found at approximately 220-and 170-kDa. Whole cell and chromosome samples shared 220-, 170-, and 18-kDa immunoreactive proteins. The 170-kDa protein was concentrated in the isolated chromosome fraction that, along with the 220-and 18-kDa proteins, also contained immunoreactive proteins of approximately 100 and 55 kDa.
Because HsTII␣ is a 170-kDa protein known to be concentrated in mitotic chromosomes, we sought to determine whether the 170-kDa band identified by the 3F3/2 anti-phosphoepitope antibody might be HsTII␣. Immunoprecipitations from chromosomal extracts with 3F3/2 and anti-HsTII␣ antibodies were performed. Both immunoprecipitates and their supernatants were then immunoblotted with 3F3/2 and anti-HsTII␣ antibodies (Fig. 2). Immunoblotting with anti-HsTII␣ antibody shows that some but not all of the HsTII␣ was immunoprecipitated by the 3F3/2 antibody. Immunoblotting with 3F3/2 antibody demonstrates that all the 3F3/2immunoreactive 170-kDa band was immunoprecipitated by the anti-HsTII␣ antibodies. From these data, we conclude that HsTII␣ extracted from chromosomes isolated in the presence of microcystin-LR contains a phosphoepitope recognized by the 3F3/2 antibody. Since the 3F3/2 immunoprecipitates did not immunodeplete all HsTII␣ but did immunodeplete all detectable 3F3/2-reactive HsTII␣ from these extracts, a portion of HsTII␣ associated with isolated mitotic chromosomes does not appear to contain the 3F3/ 2-reactive phosphoepitope.
To investigate whether the 3F3/2-reactive phosphoepitope on HsTII␣ was present throughout the cell cycle, interphase nu-clei from cycling cells and chromosomes from mitotically arrested cells were isolated in the presence of microcystin-LR. The nuclear preparations contain nuclei from G 1 , S, and G 2 phase cells. Immunoblotting with anti-HsTII␣ antibodies revealed that the nuclear and chromosomal extracts contained approximately equal amounts of HsTII␣ (Fig. 3). Immunoblotting with 3F3/2 antibody showed that the phosphoepitope was only detected on the 170-kDa band from mitotic chromosomes. These data indicate that the phosphorylated form of HsTII␣ containing the 3F3/2 phosphoepitope is found in mitotic chromosomes but not in interphase nuclei.

3F3/2 Kinase Phosphorylation of HsTII␣ Protein Fragments-To identify the 3F3/2-reactive phosphoepitope on
HsTII␣, we developed an assay to monitor the generation of this epitope on bacterially expressed protein fragments of HsTII␣. HsTII␣ protein fragments were separated by gel electrophoresis and transferred to Immobilon-P membranes. Total expressed protein levels and confirmation of predicted molecular weights were facilitated by the detection of the N-terminal S-Tag TM portion of the HsTII␣ fragments. The membranes were cut into strips, and immobilized proteins were incubated with cellular lysates containing microcystin-LR and ATP. Generation of the 3F3/2-reactive phosphoepitope was then detected by immunoblotting with the 3F3/2 antibody. The HsTII␣ protein fragments used as substrates in this assay are depicted in Fig. 4A. This approach identified a region of the Y2C HsTII␣ protein fragment that, when treated with mitotic extract in the presence of ATP, generated an antigenic site that was recognized by the 3F3/2 antibody (Fig. 4B). This region includes the last 22 amino acids of the Y2C fragment and corresponds to amino acids 1341-1362 of HsTII␣ (Fig. 4A).
To investigate the cell cycle dependence of the kinase activity that creates the 3F3/2 phosphoepitope on HsTII␣, the abilities of cell extracts prepared from cells at various phases of the cell cycle were compared. Extracts prepared from cells arrested in mitosis were considerably more efficient at creating the 3F3/2 phosphoepitope than extracts prepared from cells arrested in S phase or from unsynchronized cells (Fig. 5).
To identify the phosphorylated residue required to create the 3F3/2 phosphoepitope, we used site-directed point mutagenesis to alter the coding sequences of selected serines or threonines within the last 22 amino acids of the Y2C HsTII␣ protein. These mutations resulted in the substitution of alanine residues in place of serine or threonine residues. Specifically, in four separate constructs, we individually substituted alanines for serine 1350, threonine 1342, threonine 1357, and threonine 1359 of HsTII␣ within the Y2C fragment. These point mutations were designated S1350A, T1342A, T1357A, and T1359A, respectively. Multiple bacterial colonies containing each of the mutagenized coding sequences described above were induced, and the subsequently expressed Y2C mutant HsTII␣ proteins were analyzed using the HsTII␣ 3F3/2 phosphoepitope kinase assay (Fig. 6). Mitotic extracts were able to create the 3F3/2 phosphoepitope on the S1350A, T1357A, and T1359A mutant protein fragments. However, the same extracts were unable to create the 3F3/2 phosphoepitope on the T1342A Y2C HsTII␣ protein fragment. Therefore, we conclude that phosphorylation of threonine 1342 of HsTII␣ is required to create the 3F3/2immunoreactive phosphoepitope.
Phosphorylated Peptide Inhibition of 3F3/2 Antibody Binding to Mitotic HsTII␣-The 16-amino acid peptide containing residues 1335-1349 of HsTII␣ and a C-terminal cysteine was synthesized in two versions, one without phosphorylated residues and one phosphorylated on threonine 1342. These peptides were tested for their ability to inhibit 3F3/2 antibody binding to 3F3/2-immunoreactive HsTII␣. The 3F2/2 antibody was incubated with varying concentrations of either peptide for 1 h prior to immunoblotting. The phosphorylated peptide, but not the unphosphorylated peptide, inhibited 3F3/2 antibody binding to mitotic HsTII␣ (Fig. 7). A phosphothreonine-containing peptide made to a sequence corresponding to a different region of HsTII␣ did not inhibit 3F3/2 antibody binding to mitotic HsTII␣ (data not shown). These results demonstrate that the 3F3/2 antibody recognizes a phosphorylated threonine residue in the context of linear adjacent amino acids identical to the amino acid residues surrounding threonine 1342 of HsTII␣.
Isolated Chromosomes Contain a 3F3/2 Phosphoepitope Kinase Activity-The proteins of mitotic chromosomes prepared in the absence of phosphatase inhibitor, particularly the 170-kDa HsTII␣ protein, show reduced labeling with the 3F3/2 antibody (Fig. 8A). Treating these chromosomes with ATP alone, or with ATP and microcystin-LR, regenerated the 3F3/2 phosphoepitope on HsTII␣ (Fig. 8B). This indicates that a kinase capable of creating the HsTII␣ 3F3/2 phosphoepitope, phosphorylated threonine 1342, co-purifies with the isolated chromosomes.
Cell Extracts and Isolated Chromosomes Contain CKII-CKII can phosphorylate serine or threonine residues adjacent to acidic residues and has been reported to form complexes with HsTII␣ (35). Moreover, Ishida et al. (34) recently reported that threonine 1342 could be phosphorylated by CKII. We immunoblotted cellular extracts and isolated chromosomes with antihuman CKII ␣ subunit (Fig. 9). CKII ␣ was present in unsynchronized, S phase-arrested, and mitotic cell extracts. It was also present in isolated chromosome preparations. The mitotic FIG. 4. Bacterially expressed HsTII␣ protein fragments and the 3F3/2 phosphoepitope kinase assay. A, multiple constructs containing sequences encoding HsTII␣ in a bacterial protein expression plasmid were prepared from a complete HsTII␣ DNA sequence. BL21(DE3) E. coli were transformed with these constructs and produced protein fragments of HsTII␣. The length in amino acids (aa) of the fragments and location of the amino acids with respect to full-length HsTII␣ are depicted. Subsequent analysis for the location of the 3F3/2 phosphoepitope on the protein fragments revealed that the phosphoepitope site is contained within the C-terminal 22 amino acids of the Y2C HsTII␣ fragment. B, bacterial lysate containing the expressed Y2C HsTII␣ protein fragment was transferred to polyvinylidene difluoride membrane. Extract from mitotically arrested HeLa S3 cells (mitotic index ϭ 91%) was incubated with the immobilized Y2C HsTII␣ protein with or without 1 mM ATP. The membranes were immunoblotted with the 3F3/2 antibody to detect the generation of the 3F3/2 phosphoepitope on the HsTII␣ fragment.
FIG. 5. Mitotic cell extracts contain high levels of the HsTII␣ 3F3/2 kinase. The 3F3/2 phosphoepitope kinase activities were compared in extracts prepared from mitotic, S phase, and cycling cells using the HsTII␣ 3F3/2 kinase assay. The mitotic indices of the mitotically arrested, S phase-arrested, and cycling cells were 70, 0, and 6%, respectively. Equivalent amounts of protein from each extract were used in the kinase assays. Immunoblotting demonstrates that 3F3/2 phosphoepitope kinase activity is greatest in the mitotic cell extracts.
FIG. 6. Mutation of threonine 1342 to alanine in the HsTII␣ Y2C protein eliminates the generation of the 3F3/2 phosphoepitope. Mutagenized plasmids were transformed into E. coli for HsTII␣ mutant protein induction. Bacterial lysates containing these expressed Y2C HsTII␣ mutant protein fragments were transferred to Immobilon-P membrane. Detection of the S-Tag TM portion of the S1350A, T1342A, T1357A, and T1359A mutant HsTII␣ protein fragments indicates that they were produced in approximately equal amounts. 3F3/2 immunoblotting of the HsTII␣ 3F3/2 kinase assay demonstrates that mitotic HeLa extracts could generate the 3F3/2 phosphoepitope on all mutant fragments except Y2C HsTII␣ T1342A. Therefore, phosphorylation of threonine 1342 of HsTII␣ is required to create the 3F3/2 phosphoepitope. cell extract and chromosomes contained a slower migrating form of CKII ␣, which reflects its mitotic phosphorylation (48).
Phosphorylation of Y2C HsTII␣ Protein Is Dependent upon CKII Activity-Purified recombinant CKII and mitotic extracts were assayed for their ability to phosphorylate threonine 1342 of HsTII␣ using the HsTII␣ 3F3/2 kinase assay. With the addition of ATP and microcystin-LR, both CKII and the mitotic extract were able to create the 3F3/2 phosphoepitope on the HsTII␣ Y2C fragment. Neither CKII nor the mitotic extract could create a 3F3/2 phosphoepitope on the mutant T1342A Y2C fragment (Fig. 10). Thus, CKII can phosphorylate the threonine corresponding to threonine 1342 of HsTII␣, thereby creating the 3F3/2 phosphoepitope. FIG. 7. A peptide containing phosphorylated threonine corresponding to threonine 1342 of HsTII␣ inhibits 3F3/2 antibody detection of mitotic HsTII␣. A, two peptides, containing the amino acid sequence identical to amino acids 1135-1349 of HsTII␣, were obtained in phosphorylated and nonphosphorylated forms. Sequences of the phosphorylated and nonphosphorylated peptides and the location of the threonine residue that corresponds to threonine 1342 of HsTII␣ are shown. B, 3F3/2 antibody was incubated with varying concentrations of either the phosphorylated or the nonphosphorylated peptide. These 3F3/2 antibody/peptide solutions were then used to immunoblot mitotic HeLa cell chromosome extracts prepared in the presence of microcystin-LR. Only the phosphorylated peptide inhibited 3F3/2 antibody detection of mitotic HsTII␣. Therefore, the 3F3/2 antibody's ability to bind a peptide sequence identical to HsTII␣ amino acid residues 1335-1349 is dependent upon the phosphorylated threonine residue corresponding to threonine 1342 of HsTII␣.

FIG. 8. A kinase activity associated with isolated HeLa S3 chromosomes can create the 3F3/2 phosphoepitope on HsTII␣.
A, when mitotic HeLa chromosomes are isolated without the serine/ threonine phosphatase inhibitor, microcystin-LR, 3F3/2 phosphoepitopes are diminished or lost. Immunoblotting demonstrates the loss of the 170-kDa HsTII␣ phosphoepitope in chromosomes isolated without microcystin-LR. B, chromosomes originally isolated in the absence of microcystin-LR were incubated at 37°C for 20 min in buffer alone, with ATP, or with ATP and microcystin-LR. Immunoblot analysis shows that the 3F3/2 phosphoepitope was regenerated on HsTII␣ when the chromosomes were incubated with ATP or ATP and microcystin-LR. Therefore, a kinase capable of creating the 3F3/2 phosphoepitope on HsTII␣ co-fractionates with mitotic chromosomes.
FIG. 9. Cellular extracts and isolated HeLa S3 chromosomes contain CKII ␣. Immunoblotting with anti-CKII ␣ subunit antibody reveals that CKII is present in cycling, S phase-arrested, and mitotically arrested HeLa S3 cells. CKII ␣ is also associated with isolated mitotic HeLa S3 chromosomes. Mitotic and chromosomal extracts contain CKII ␣ exhibiting reduced electrophoretic mobility, indicating its mitotic phosphorylation. Cellular extract lanes contain protein from 5 ϫ 10 4 cells. Immunoblotted chromosomal extract was derived from chromosomes isolated in the presence of microcystin-LR from 7.5 ϫ 10 5 mitotic cells.
CKII can utilize GTP to phosphorylate its substrates, and its activity is inhibited by heparin (49,50). In order to further characterize the kinase activity in the HeLa S3 mitotic extracts, the ability of mitotic extract and purified CKII to create the 3F3/2 phosphoepitope using GTP was assayed. With GTP in place of ATP, both mitotic extract and CKII phosphorylated the Y2C HsTII␣ fragment as shown by 3F3/2 immunoblotting (Fig. 11A). Heparin inhibited the phosphorylation of the Y2C HsTII␣ fragment by both the mitotic extract and by CKII (Fig.  11B). Together, these results suggest that the phosphorylation of the Y2C HsTII␣ fragment by the mitotic extract is dependent upon the activity of CKII. DISCUSSION We have found that the phosphorylation of HsTII␣ Thr 1342 creates an epitope recognized by the 3F3/2 phosphoepitope antibody. Thr 1342 lies within the C-terminal region of HsTII␣. Among the other known mammalian TII␣ sequences, this threonine is conserved in pig but is replaced by the acidic residue, aspartic acid, in rodents such as rat, mouse, and Chinese hamster. Although the region is not as highly conserved as the N-terminal portions of TII among species, it has been postulated to affect TII activity, dimerization, and nuclear localization in various organisms (27,51,52). Since many phosphorylations of TII have been reported in this C-terminal region, regulatory effects of these phosphorylations may be conserved although the phosphorylated residues may vary between species.
Because the phosphorylation of HsTII␣ Thr 1342 is mitosisspecific, this phosphorylation or corresponding phosphorylations of TII␣ in other species may participate in the regulation of its mitotic functions. These phosphorylations could induce chromosome condensation or decondensation or alter its association with the condensed chromosomal architecture. It is possible that an inhibitory mitotic phosphorylation of TII␣ may help maintain sister chromatid cohesion until the initiation of anaphase. In addition, the disappearance of TII␣ after M phase appears to involve regulated destruction of the existent protein.
Phosphorylation plays a key role in targeting proteins such as cyclin E for destruction by the ubiquitin-proteasome system (53). The mitosis-specific phosphorylation of TII␣ could play a similar role.
The phosphorylation of HsTII␣ Thr 1342 has been previously reported by Ishida et al. (34). In contrast to our study, these authors found that Thr 1342 was phosphorylated throughout the cell cycle. We do not detect phosphorylation of HsTII␣ Thr 1342 at stages other than M phase, and we are unable to reconcile the difference in the cell cycle occurrence of this phosphorylation reported between our studies. Ishida et al. (34) stated that the phosphorylation of this threonine residue was approximately 2-fold greater during the G 2 /M phase as compared with the G 1 or S phases. In 3F3/2 immunoblot experiments, the HsTII␣ 3F3/2 phosphoepitope was not detected on interphase HsTII␣ even when the amount of interphase HsTII␣ was greater than 10-fold above the amount necessary to detect the phosphoepitope on mitotic HsTII␣. This indicated that the level of phosphorylation of HsTII␣ Thr 1342 was at least 10-fold greater on mitotic HsTII␣ than on interphase HsTII␣. Since the HsTII␣ Thr 1342 phosphoepitope is restricted to mitotic HsTII␣, the activity that creates this phosphoepitope must be regulated. Indeed, we found that the kinase activity that creates the epitope was far greater in mitotic cell extracts than in cycling or S phase extracts. However, extracts prepared from cycling cells or from S phase-arrested cells could, upon extended incubation, catalyze the generation of the Thr 1342 HsTII␣ 3F3/2 phosphoepitope. These cycling or S phase cell extract activities in our assay may reflect a decrease in an inhibitory regulation of the 3F3/2 phosphoepitope kinase that would not occur under normal physiological conditions. Alternatively, the phosphoepitope kinase activity may simply be down-regulated although not completely eliminated during interphase. It is also possible that the low kinase activity could be due to the small population of mitotic cell extract present in the cycling (4 -10% mitotic component) or S phase-arrested (0 -1% mitotic component) cell extracts. If the 3F3/2 phosphoepitope kinase activity is not regulated, then the observed difference between the phosphorylation status of mitotic and interphase HsTII␣ Thr 1342 might be due to greater phosphatase activity in interphase extracts. However, this seems unlikely, since the inclusion of the serine/threonine phosphatase inhibitor, microcystin-LR, in our cell extracts and rephosphorylation assays did not reduce the differences in catalytic potential between FIG 11. Generation of the 3F3/2 phosphoepitope on the Y2C HsTII␣ protein by mitotic cell extract is dependent upon the activity of CKII. A, HeLa mitotic cell extract and purified CKII were compared for their ability to use ATP or GTP to create the 3F3/2 phosphoepitope on the Y2C HsTII␣ protein. Both CKII and the mitotic cell extract can create the phosphoepitope using either ATP or GTP. B, the CKII inhibitor, heparin, was included at 50 g/ml in the HsTII␣ 3F3/2 phosphoepitope kinase assay. 20 g/ml mitotic extract protein or 100 units/ml CKII was used in each reaction. Immunoblots show that heparin inhibited the 3F3/2 phosphoepitope kinase activity of both the mitotic cell extract and CKII. mitotic and interphase extracts.
The sequence of acidic amino acids surrounding Thr 1342 in HsTII␣ suggested to Ishida et al. (34) and to us that the relevant kinase might be CKII, a serine/threonine protein kinase. CKII has been implicated in the regulation of several proliferative events and plays a significant role in the regulation of many nuclear protein functions (reviewed in Ref. 54). Although we have not unequivocally eliminated the possibility that another enzyme is involved, our evidence is consistent with the interpretation that the mitotic 3F3/2 phosphoepitope kinase activity for threonine 1342 of HsTII␣ is due to CKII. In our assays, we found that the mitotically regulated kinase can use GTP to phosphorylate HsTII␣ Thr 1342 and that this activity is inhibited by heparin. In addition, we have shown that a portion of the mitotic CKII associates with condensed chromosomes and that a kinase activity associated with these isolated chromosomes can phosphorylate HsTII␣ and create the 3F3/2 phosphoepitope. In contrast to the previous report, we have found that the apparent CKII-dependent phosphorylation of HsTII␣ Thr 1342 occurs only in M phase.
The regulation of this CKII activity on HsTII␣ Thr 1342 is unclear. In most organisms, CKII is a tetrameric enzyme. It contains ␣, ␣', and ␤ subunits in the following configurations; ␣ 2 ␤ 2 , ␣␣'␤ 2 , and ␣' 2 ␤ 2 . In M phase, the ␣ catalytic subunit undergoes multiple phosphorylations and a consequent shift in electrophoretic mobility (48). We also observed the reduced electrophoretic mobility of CKII ␣ in mitotic and chromosome extracts (Fig. 9). The ␣Ј subunit does not appear to undergo specific M phase phosphorylations, nor does it contain the mitotically phosphorylated C-terminal residues present in the ␣ subunit (48). The accessory subunit ␤, thought to be a regulatory and/or targeting subunit, also exhibits a specific M phase phosphorylation on serine 209 (55). Mitotic phosphorylations of both the ␣ and ␤ subunits have been attributed to p34 cdc2 kinase activity (48,55). CKII has been activated by phosphorylation in vitro by p34 cdc2 and protein kinase C (56,57). While the mitotic phosphorylation of TII by CKII in yeast has been previously reported, to our knowledge the phosphorylation of HsTII␣ Thr 1342 by CKII represents the first reported, mitosisspecific CKII phosphorylation of TII␣ in mammals.
In summary, we have identified a specific mitotic phosphorylation of HsTII␣ that is recognized by the anti-phosphoepitope antibody 3F3/2. The preponderance of evidence suggests that the enzyme catalyzing this phosphorylation is CKII and that its enzymatic activity on HsTII␣ threonine 1342 is increased substantially in M phase. The consequences of this phosphorylation for the regulation of HsTII␣ in M phase remain to be determined through mutation of the substrate region and characterization of the consequences both in vitro and in vivo.