Phosphorylation of DNA Topoisomerase I1 in Vivo and in Total Homogenates of Drosophila Kc Cells THE ROLE OF CASEIN KINASE 11*

The phosphorylation of DNA topoisomerase I1 in Drosophila Kc tissue culture cells was characterized by in vivo labeling studies and in vitro studies that examined the modification of exogenous enzyme in total homogenates of these embryonic cells. Several lines of evidence identified casein kinase I1 as the kinase primarily responsible for phosphorylating DNA topoisomerase 11. First, the only amino acyl residue modified in the enzyme was serine. Second, partial proteolytic maps of topoisomerase I1 which had been labeled with [S2P]phosphate by Drosophila cells in vivo, by cell homogenates in vitro, or by purified casein kinase I1 were indistinguishable from one another. Third, phosphorylation in cell homogenates was inhibited by pg/ml concentrations of heparin, micromolar concentrations of nonradioactive GTP, or anti-Dro-sophila casein kinase I1 antiserum. Fourth, cell homogenates were able to employ [y-”P]GTP as a phosphate donor nearly as well as [y-S2P]ATP. Although topoisomerase I1 was phosphorylated in homogenates under conditions that specifically stimulate protein kinase C, calcium/calmodulin-dependent protein kinase, or CAMP-dependent protein kinase, modification was al-ways sensitive to anti-casein kinase I1 antiserum or heparin. Thus, under a variety of conditions, topoisomerase

genic agents, such as concanavalin A (29) and epidermal growth factor (30), ' have been shown to stimulate topoisomerase I1 activity in mammalian cells. Although the activity of the enzyme remains relatively constant over the cell cycle of proliferating cells (31, 32), levels of the protein increase 1.5to 2-fold during mitosis (27,33). In addition, the intracellular location of topoisomerase I1 redistributes during the process of cell division. During interphase, the enzyme is a major polypeptide component of the nuclear matrix (23). During mitosis, a portion of the enzyme specifically associates with metaphase chromosomes (17-19), while the rest disperses throughout the cytoplasm (19, 23). Unfortunately, virtually nothing is known about the factors that modulate the in vivo activity or intracellular location of topoisomerase 11.
Since phosphorylation/dephosphorylation events have long been known to alter the functions of many enzymes (34) and structural proteins (35, 36), including topoisomerase I (37-40), the role of phosphorylation as a physiological regulator of topoisomerase I1 has been investigated. Indeed, the type I1 enzyme appears to exist in the eukaryotic cell as a phosphoprotein (41-43). In vitro, topoisomerase I1 is readily phosphorylated by casein kinase I1 (42) or protein kinase C (43, 44), and modification by either kinase stimulates enzyme activity by about %fold. Topoisomerase I1 also serves as an in vitro substrate for calcium/calmodulin-dependent protein kinase (44), but considerably higher levels of kinase are required for modification. In contrast, CAMP-dependent protein kinase shows no ability to phosphorylate the enzyme (39,44). While this last finding indicates that the cellular modification of topoisomerase I1 must be carried out by specific enzymes, the kinase(s) that phosphorylates the enzyme in viuo, the site(s) of modification, and the mechanism by which phosphorylation stimulates enzyme activity are completely unknown.
Clearly, before the physiological regulation of topoisomerase I1 can be fully defined, the cellular phosphorylation of the enzyme must be characterized. As a first step toward this end, the in vivo phosphorylation of DNA topoisomerase I1 in Drosophila Kc cells and the in vitro modification of exogenous Drosophila melanogaster enzyme in total Kc cell homogenates have been examined. On the basis of peptide mapping, enzymological and immunological studies, casein kinase I1 appears to be the kinase primarily responsible for phosphorylating topoisomerase I1 in this embryonic cell system.

Metabolic Labeling of DNA Topoisomerase II in Drosophila Kc
Tissue Culture Cells and Preparation of Nuclear Extracts-Cultures of Drosophila Kc cells (an embryonic line) were maintained at 25 "C in spinner flasks using D22 medium supplemented with penicillin and streptomycin (49). Cells (100-150 ml) were labeled during exponential growth (6-10 X lo6 cells/ml) by the addition of [32P]orthophosphate (2 mCi) to the medium. Cells were harvested 8-12 h later, and crude nuclei were prepared by the procedure of Shelton et al. (45). Nuclei were extracted with 3 ml of 30 mM Tris-C1, pH 7.9, 350 mM NaCl, 1 mM EDTA, 0.5 mM dithiothreitol, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 10 mM sodium metabisulfite (adjusted to pH 7.0 and added just prior to use) for 30 min on ice. Samples were centrifuged at 16,000 X g for 30 min at 4 "C. The nuclear extract supernatant was collected and used for the immunoprecipitation procedures described below.
Purification of in Vivo-labeled Topoisomerase 1 1 from Nuclear Extracts-Topoisomerase I1 was purified from the nuclear extracts described above by immunoprecipitation. The procedure of Soderquist and Carpenter (50) was employed. Briefly, nuclear extracts were incubated with %o volume of rabbit anti-Drosophila topoisomerase I1 antiserum (23, 45), and topoisomerase I1 was precipitated by the addition of formalin-fixed S. aureus cells (10% final concentration).
Immunoprecipitates were washed four times with 10 mM Tris-C1, pH 8.5, 1.0% Triton X-100, 1.0% sodium deoxycholate, 0.1% SDS? 1 mM EDTA, 0.02% sodium azide and were resuspended in approximately 100 pl of 5 mM Tris-C1, pH 7.4, 0.5 mM EDTA. Samples were mixed with an equal volume of 250 mM Tris-C1, pH 6.8, 8% SDS, 20% j3mercaptoethanol, 40% glycerol, 0.004% bromphenol blue and heated to 95 "C for 30 min. S. aureus cells were removed by centrifugation and topoisomerase I1 samples were subjected to electrophoresis on 7% acrylamide gels by the procedure of Laemmli (51). Metabolically labeled enzyme was located by autoradiography using Kodak XAR film with a Du Pont Lightning Plus screen and extracted from the gel as described under "Identification of Phosphorylated Amino Acyl Residues." Preparation of Kc Tissue Culture Cell Homogenates-Cells were maintained as described above and harvested during exponential growth (3-12 X lo6 cells/ml). For the preparation of homogenates, a volume containing approximately 5 x 10' cells was removed from stock cultures. Cells were pelleted at 8000 X g for 10 min at 4 "C and resuspended by gentle vortexing in 2 ml of 15 mM Tris-C1, pH 7.9, 5 mM KCl, 0.5 mM MgClZ, 0.05 mM EDTA, 0.05% Triton X-100, 350 mM sucrose, 1 mM phenylmethylsulfonyl fluoride, 10 mM sodium metabisulfite (adjusted to pH 7.0 and added just prior to use). The cell suspension was homogenized on ice using 10 strokes in a Dual1 glass tissue grinder (Kontes). As judged by light microscopy, this procedure disrupted the native morphology of both the cell and nuclear membranes. The average total protein content of cell homogenates was estimated to be 5 mg/ml by the method of Bradford (52) using bovine serum albumin as a standard. Freshly prepared Kc cell homogenates were employed for all the experiments described below.
Phosphorylation of DNA heating at 95 "C for 2 min. Reaction products were subjected to electrophoresis on 7% polyacrylamide gels as described by Laemmli (51). Protein phosphorylation was visualized by autoradiography. When required, the relative incorporation of radioactive phosphate into DNA topoisomerase I1 was determined from autoradiograms with a Biomed Instruments Model SL-504-XL scanning densitometer. The inhibition of phosphorylation by heparin was examined over a range of 0 to 20 pg/ml. The effect of nonradioactive GTP on modification was studied over a range of 0 to 500 p~. Experiments that examined the ability of Kc cell homogenates to utilize radioactive GTP as a phosphate donor employed 60 p~ [y3'P]GTP at a specific activity of 4 Ci/mmol. The effect of rabbit anti-Drosophila casein kinase I1 antiserum (53) and preimmune serum on phosphorylation was examined over a dilution range of 1:120 to 1:30.
Reactions designed to stimulate protein kinase C included 100 p~ CaClz and 33 pg/ml phosphatidylserine; those designed to stimulate calcium/calmodulin-dependent protein kinase included 100 PM CaC12 and 6 p~ calmodulin; and those designed to stimulate CAMP-dependent protein kinase included 10 p~ cAMP (44).
Reaction mixtures were prepared for the identification of phosphorylated amino acids and the generation of partial proteolytic maps by increasing the topoisomerase I1 concentration to 175 mM and the specific activity of the [Y-~'P]ATP to 30 Ci/mmol.
Phosphorylation of DNA Topoisomerase II by Purified Protein Kinases-Phosphorylation reactions that employed purified Drosophila casein kinase I1 rather than cell homogenate were carried out as described above except that vanadate was not used in the reaction buffer, the cell homogenate was replaced with 3 nM casein kinase 11, and incubation times were decreased to 2.5 min. Experiments that examined the ability of CAMP-dependent protein kinase to modify topoisomerase I1 were carried out in the phosphorylation buffer described above, except that vanadate was not included, the pH of the buffer was adjusted to 7.5, and the ATP concentration was increased to 100 p~. These latter two changes were made in an effort to optimize CAMP-dependent protein kinase activity (47). Purified CAMP-dependent protein kinase catalytic subunit (8 nM) was employed for all assays. In the absence of its regulatory subunit, cAMP is not required for kinase activity (47). Reactions were terminated by trichloroacetic acid precipitation onto Whatman GF/C filters (46). Phosphorylation was quantitated using a Beckman LS-7500 liquid scintillation counter and ACS scintillant.
Identification of Phosphorylated Amino Acyl Residues-In vivoand in vitro-modified enzymes were prepared as described above. Samples were subjected to electrophoresis on 7% polyacrylamide gels as described by Laemmli (51). Bands corresponding to phosphorylated topoisomerase I1 were located by autoradiography and excised. Labeled enzyme was extracted by electroelution in 25 mM Tris base, 192 mM glycine, 0.1% SDS for 8 h at 23 "C using a Bio-Rad Model 422 electroeluter. Alternatively, gel slices were washed twice for 2 h at 23 "C in 10% methanol to remove salts and SDS. Phosphorylated topoisomerase I1 was extracted at 23 "C overnight in several changes of 50 mM N-ethylmorpholine acetate, pH 8.2, with continuous rotation. With either procedure, supernatants were pooled, dialyzed overnight against 100 mM ammonium bicarbonate, and concentrated by centrifugation under reduced pressure.
Modified topoisomerase I1 was hydrolyzed in 6 N HCl for 2 h at 110 "C under reduced pressure. The hydrolysate was recovered, mixed with phosphorylated amino acid standards (1 mg/ml each of Ser (PI, Thr (P), and Tyr (P)), and analyzed by two-dimensional electrophoresis at pH 1.9 and pH 3.5 on Polygram Cel 300 thin layer plates (Brinkmann) as described by Hunter and Sefton (54). Standards were located by ninhydrin staining and [32P]phosphate was visualized by autoradiography as described above.
Partial Proteolytic Mapping of Phosphorylated Topoisomerase II-Phosphorylated topoisomerase I1 was isolated as described under "Identification of Phosphorylated Amino Acyl Residues." Enzyme samples were dissolved in proteolysis buffer (125 mM Tris-C1, pH 6.8, 0.5% SDS, 10% glycerol, 0.001% bromphenol blue) containing bovine serum albumin (0.5 mg/ml) and digested by the method of Cleveland et al. (55). Proteolytic enzymes were freshly dissolved in proteolysis buffer and added immediately to phosphorylated topoisomerase 11 samples. Final enzyme concentrations were a-chymotrypsin, 67 rg/ ml; elastase, 67 pg/ml; S. aureus V8 protease, 10 pg/ml; 8. griseus protease, 20 pg/ml; trypsin, 200 pg/ml. Mixtures were incubated at 37 "C for 30 min. Following the addition of 0-mercaptoethanol and SDS to final concentrations of 10 and 2%, respectively, samples were heated at 95 "C for 3 min and subjected to electrophoresis on 11% polyacrylamide gels by the procedure of Laemmli (51). Phosphorylated peptides were visualized by autoradiography as described above.

RESULTS
In Vivo Phosphorylation of Drosophila DNA Topoisomerase 11-DNA topoisomerase I1 has been reported to exist in Drosophila embryonic cells as a phosphoprotein (41,42). To provide direct evidence for in uiuo phosphorylation, nuclear extracts of Drosophila Kc tissue culture cells that had been metabolically labeled with [32P]orthophosphate were immunoprecipitated with rabbit anti-Drosophila topoisomerase I1 antiserum. Following electrophoresis on a 7% polyacrylamide gel (51), one predominant 32P-labeled protein band was observed in autoradiograms (Fig. 1, lane 2). This band was identified as topoisomerase I1 on the basis of its comigration with purified Drosophila type I1 enzyme (compare lanes 1 and 2 ) ) its positive reaction with anti-Drosophila topoisomerase I1 antiserum, and its absence when preimmune serum was employed in the immunoprecipitation procedures (not shown). These experiments provide conclusive evidence that DNA topoisomerase I1 exists in Drosophila embryonic cells as a phosphoprotein. A recent report draws a similar conclusion concerning the phosphorylation state of the enzyme in sponge (Geodia cydonium) cells (43).
Phosphorylation of DNA Topoisomerase 1 1 by Total Homogenates of Drosophila Kc Cells-In order to further characterize the cellular phosphorylation of the type I1 enzyme, the ability of total homogenates of Drosophila Kc cells to modify exogenous topoisomerase I1 was examined. As seen in the polyacrylamide gel of Fig. 2, the enzyme was phosphorylated by these cell homogenates. The labeled band in the autoradiogram was identified as topoisomerase I1 on the basis of 1) its complete absence when exogenous enzyme was not included in the assay mixture; 2) its dependence on the concentration of  added topoisomerase II; 3) its polypeptide molecular mass of approximately 166,000 daltons (45, 56); and 4) immunoblot analysis (not shown), in which this radioactive band was found to react specifically with anti-Drosophila topoisomerase I1 antiserum (23, 45). As determined by direct scintillation counting or by comparing densitometric scans of autoradiographs with those of casein kinase 11-modified topoisomerase I1 controls (see Fig. 7), approximately one phosphate group was incorporated per homodimer of topoisomerase I1 in standard homogenate-catalyzed reactions? Under the exposure conditions employed in Fig. 2, the phosphorylation of endogenous Kc cell topoisomerase I1 was not visualized. This is due to the fact that levels of the cellular enzyme (-0.1% total protein in Kc cells (45)) were approximately 2 orders of magnitude lower than those of the exogenous enzyme. Upon overexposure of the autoradiogram, bands representing phosphorylated endogenous proteins became visible (not shown).

29
To ensure that levels of [T-~~P]ATP remained relatively constant over the course of a 10-min phosphorylation reaction, 10 PM sodium vanadate, a potent inhibitor of cellular ATPases (57) was included in all samples. In the presence of vanadate, less than 0.5% of the added ATP was hydrolyzed to ADP and orthophosphate (as judged by thin layer chromatography (58)). When vanadate was not included in assays, approximately 60% of the added ATP was hydrolyzed during a 10-min reaction.
Identity of Phosphorylated Amino Acyl Residues-The amino acyl residue of Drosophila topoisomerase I1 that was phosphorylated by Kc cells in uiuo (Fig. 3, left panel) or by cell homogenates in uitro (right panel) was identified by partial acid hydrolysis and two-dimensional electrophoresis. As previously shown for the casein kinase 11-catalyzed reaction (42), only serine was modified. No phosphothreonine or Purified casein kinase I1 incorporates a maximum of 2 to 3 phosphate groups per homodimer of topoisomerase I1 (42) (see Table   111). Reaction times (i.e. 10 min) that yielded a submaximal -1:l modification of the enzyme were employed for the cell homogenate experiments described below in order to optimize the effects of added kinase inhibitors and stimulators. Higher levels of topoisomerase I1 phosphorylation were observed at longer reaction times.

In Vivo
In Vitro Five proteases with different cleavage specificities were employed for these experiments, including chymotrypsin (specific for aromatic amino acid residues), elastase (specific for neutral aliphatic residues), S. aureus V8 protease (specific for glutamic acid residues), S. griseus protease (nonspecific), and trypsin (specific for lysyl and arginyl residues). For each of the above proteases, unique banding patterns of phosphorylated peptides were observed following SDS-polyacrylamide gel electrophoresis and autoradiography. In all cases, corresponding peptide maps of in uiuo-, homogenate-, and casein kinase 11-phosphorylated topoisomerase I1 were identical to one another. Maps generated with chymotrypsin, elastase, or S. aureus V8protease are shown in Fig. 4 (upperpanel). Maps produced with S. griseus protease contained 7 major phosphorylated bands ranging in size from -20 to -3 kDa and those produced with trypsin contained 14 major phosphorylated bands ranging in size from -120 to <5 kDa (not shown).
A 15-fold overexposure of the low molecular mass portion of the partial proteolytic maps (Fig. 4, lower panel) revealed labeled peptides with masses in the 2.5to 3-kDa range. This limits the site of phosphorylation in topoisomerase I1 to a region no larger than 20 to 30 amino acyl residues and makes it likely that the kinase primarily responsible for modifying the enzyme in both Kc cells and cell homogenates is casein kinase 11. As described below, several additional experiments were carried out in order to confirm this identification. All of the following work was designed to characterize directly the enzymological nature of the kinase that is responsible for phosphorylating topoisomerase I1 in Kc cell homogenates.

Inhibition of Homogenate-catalyzed Phosphorylation by
Heparin-The modification of topoisomerase I1 by Kc cell homogenates was strongly inhibited by the inclusion of heparin in assay mixtures (Fig. 5). As determined from a semilog  (lanes 1 3 ) , elastase (lanes 4-6), or V8 protease (lanes 7-9). Digests (lanes 1-5) were subjected to electrophoresis on a 7% polyacrylamide gel. An autoradiogram of the gel is shown. Marker proteins are as in Fig. 1. 50% inhibition occurred at a heparin concentration of approximately 3 pg/ml.

29-
Most cellular kinases are totally insensitive to concentrations of heparin which are lower than 10 pg/ml (59, 60). In contrast, casein kinase I1 is strongly inhibited by heparin concentrations in the 0.1 to 1 pg/ml range (46, 59, 60).
Although this range is somewhat lower than that observed for the homogenate-catalyzed activity, the heparin inhibition of casein kinase I1 can be overcome by the presence of basic proteins, such as histones (not shown), which are present in high concentrations in the embryonic Kc cell line (49).
Inhibition of Phosphorylation by Nonradioactive GTP-The ability of nonradioactive GTP to inhibit the homogenatecatalyzed modification of topoisomerase I1 (when [Y-~~P]ATP was used as a phosphate donor) is shown in Fig. 6. As determined from a semilog plot (not shown), 50% inhibition was estimated to occur at 17 PM GTP. While most kinases require millimolar levels of GTP for inhibition, the Ki value of GTP for purified casein kinase I1 is 9 PM (61).

Utilization of [-y-32P]GTP as a Phosphate Donor by Kc Cell
Homogenates-A distinguishing characteristic of type I1 casein kinases is their ability to employ GTP as a phosphate donor (K, = 66 PM) nearly as effectively as they do ATP (K, = 17 pM) (60,61). Therefore, the ability of homogenates to utilize [y3*P]GTP was examined. As seen in Fig. 7 (lane 4 ) , homogenates could incorporate [32P]phosphate into Drosophila topoisomerase I1 using [y3*P]GTP as a phosphate donor. By normalizing levels of phosphorylation obtained with homogenates to those found with purified casein kinase I1 (i.e. homogenate-catalyzed phosphorylation/casein kinase II-catalyzed phosphorylation), it was possible to determine the relative abilities of cell homogenates to employ [ Y -~~P I G T P and [-y-32P]ATP as phosphate donors (Fig. 7). Under the conditions employed, the relative utilization of radioactive GTP by cell homogenates (39%) (compare lanes 3 and 4 ) was remarkably similar to their relative utilization of radioactive ATP (41%) (compare lanes 1 and 2).
To ensure that the [T-~~PIGTP that was included in assay mixtures was not converted to [Y-~~PIATP by the action of nucleoside diphosphate kinase or other nucleotide salvage enzymes (62), the fate of labeled GTP was followed over the course of a phosphorylation reaction. As determined by thin layer chromatography (58), less than 0.5% of the nucleoside triphosphate was hydrolyzed to GDP and, at most, less than 0.2% was converted to labeled ATP. Thus, the observed modification of topoisomerase I1 by cell homogenates must result from the direct utilization of GTP as a phosphate donor.  0,25,50, 100, or 500 p M (lanes 2-5) nonradioactive GTP (in the presence of 15 pM [y3*P]ATP) were subjected to electrophoresis on a 7% polyacrylamide gel. An autoradiogram of the gel is shown. Marker proteins are as in Fig. 2.  3 and 4 ) as a phosphate donor were subjected to electrophoresis on a 7% polyacrylamide gel. An autoradiogram of the gel is shown. Marker proteins are as in Fig. 1   Stimulate Protein Kinase Activities-While the enzymological and immunological studies presented above strongly suggest that topoisomerase I1 is modified by casein kinase I1 in Kc cell homogenates, they do not rule out possible phosphorylation of the enzyme by other cellular kinases. Therefore, phosphorylation assays were carried out under conditions specifically designed to stimulate protein kinase C or calcium/ calmodulin-dependent protein kinase, both of which have been shown to modify Drosophila topoisomerase I1 in vitro (44) or CAMP-dependent protein kinase, which shows no ability to modify the enzyme in vitro (39, 44) (see Table 111). Typical results are shown in Table 11. While the extent of modification in the presence of calcium/calmodulin was similar to that found under assay conditions lacking these compounds, the inclusion of calcium/phosphatidylserine (which stimulates protein kinase C) or cAMP increased topoisomerase I1 phosphorylation by approximately 50%. However, even in the presence of these specific kinase inducers, addition of anti-casein kinase I1 antiserum or heparin strongly inhibited phosphorylation (Table 11). Thus, phosphorylation by casein kinase I1 still appeared to be the event of importance. Several hypotheses are consistent with the above finding. First, topoisomerase I1 may become a more suitable substrate for protein kinase C or CAMP-dependent protein kinase only after it is first phosphorylated by casein kinase 11. Several examples of such synergistic phosphorylation have been reported (63-67), and, in all cases, the initial and critical phosphorylation event was mediated by casein kinase 11. Second, topoisomerase I1 may be phosphorylated exclusively by casein kinase 11, but the kinase's activity is stimulated by a direct   interaction (i.e. phosphorylation) with another activated protein kinase. In this regard, it should be noted that casein kinase I1 exists in the cell as a phosphoprotein (60). Third, topoisomerase I1 may be phosphorylated exclusively by casein kinase 11, but the kinase's activity is stimulated as an indirect result of protein kinase C or CAMP-dependent protein kinase activation. Such an explanation requires a cascade or second messenger mechanism.

166-
Studies were carried out in order to test the first two hypotheses. Experiments described below employed the catalytic subunit of CAMP-dependent protein kinase, since this kinase shows no ability to modify topoisomerase I1 in vitro (39,44) (see Table 111). The CAMP-dependent enzyme preparation used was highly active as judged by its ability to modify histone or Kemptide substrates (not shown).
The first experiment examined the possibility of synergistic phosphorylation. To this end, topoisomerase I1 was phosphorylated for 10 min with purified casein kinase 11. The kinase was then inhibited by the addition of either 250 p~ GTP or a combination of 250 p~ GTP and 10 pg/ml heparin. Purified catalytic subunit of CAMP-dependent protein kinase" was added to the reaction mixture to determine whether casein kinase 11-modified topoisomerase I1 became a substrate for the CAMP-dependent enzyme. As seen in Fig. 8, no synergistic phosphorylation was observed.
The second experiment examined the possibility that casein kinase I T was stimulated by the presence of the CAMPdependent enzyme. Accordingly, casein kinase I1 was incubated with the catalytic subunit of CAMP-dependent protein kinase prior to the addition of topoisomerase 11. As seen in Table 111, the time course for the phosphorylation of topoisomerase I1 by this kinase mixture was similar to that generated by casein kinase I1 alone. This result precludes a stimulatory interaction between the catalytic subunit of CAMPdependent protein kinase and casein kinase 11. Thus, it appears likely that the addition of cAMP to cell homogenates increases levels of casein kinase 11-mediated topoisomerase I1 phosphorylation by an indirect process.

DISCUSSION
Casein kinase I1 appears to be the enzyme primarily responsible for phosphorylating DNA topoisomerase I1 in Dro-' Neither 250 p~ GTP nor a mixture of 250 pM GTP and 10 pg/ ml heparin impaired the catalytic activity of the CAMP-dependent enzyme, as determined from experiments which employed either histone or Kemptide substrates. sophila Kc tissue culture cells in uiuo and in KC cell homogenates i n uitro. This conclusion results from several lines of evidence. First, as previously shown (42) for casein kinase 11catalyzed reactions, only serinyl residues in topoisomerase I1 were modified in cells or in cell homogenates. Second, peptide maps of in uiuo-, homogenate-, and casein kinase 11-phosphorylated topoisomerase I1 (generated with five different proteolytic enzymes) were indistinguishable from one another. Third, homogenate-catalyzed phosphorylation was strongly inhibited by pg/ml concentrations of heparin or micromolar concentrations of nonradioactive GTP. Sensitivity to these compounds are hallmark characteristics of type I1 casein kinases (46,(59)(60)(61). Fourth, homogenate could employ [r-"P] GTP as a phosphate donor nearly as effectively as it could [y3'P]ATP. Once again, this enzymological trait is unique to casein kinase 11 (46,60). Fifth, homogenate-catalyzed reactions were specifically inhibited by antiserum directed against Drosophila casein kinase 11. Sixth, even under reaction conditions that specifically stimulate protein kinase C, calcium/ calmodulin dependent protein kinase, or CAMP-dependent protein kinase, homogenate-catalyzed reactions were inhibited by anti-casein kinase I1 antiserum or heparin.
Two experimental models were used to examine topoisomerase 11-kinase interactions in the present study: metabolic labeling of topoisomerase I1 in intact Drosophila Kc cells and in uitro phosphorylation of exogenous enzyme in total homogenates of these embryonic cells. By coupling these two models, it was possible to characterize sites of phosphorylation in topoisomerase I1 as well as the enzymological and immunological properties of the modification reaction. This dualistic approach was favored over experiments designed to characterize only the modification of endogenous topoisomerase I1 in intact cells, because the metabolic approach relies totally on the identification of phosphorylation sites or the correlation between levels of phosphorylation and the presence of kinase effectors. When defining interactions between kinases and protein substrates, it is very important to couple enzymological studies with mapping experiments, because it is possible that more than one kinase can phosphorylate the same site on a given protein. As an example, serine 40 of tyrosine hydroxylase is modified by at least four different protein kinases in uitro, including CAMP-dependent protein kinase, cGMP-dependent protein kinase, calcium/calmodulin-dependent protein kinase, and protein kinase C (68). The necessity for enzymological studies is emphasized further by the kinase stimulation experiments shown in Table 11. Despite the fact that levels of topoisomerase I1 phosphorylation increased following the addition of protein kinase C or CAMPdependent protein kinase stimulators to cell homogenates, the enhanced enzyme modification clearly was mediated by casein kinase 11. To this point, it has recently been proposed that topoisomerase I1 is phosphorylated primarily by protein kinase C in sponge cells, because levels of enzyme modification increased following treatment of cells with 12-0-tetradecanoylphorbol-13-acetate, a protein kinase C stimulator (43).
Although DNA topoisomerase I1 is primarily a nuclear enzyme in interphase cells, total cell homogenates rather than isolated nuclei were employed for the i n uitro portion of the present work. The decision to use total cell homogenates was made for two reasons. First, during mitosis, the nuclear envelope dissolves (35) and topoisomerase I1 is exposed to the cytoplasm (19, 23). Second, in uitro phosphorylation of Drosophila topoisomerase I1 has been demonstrated with three different protein kinases: casein kinase I1 (42), protein kinase C (43,44), and calcium/calmodulin-dependent protein kinase (44). In the cell, casein kinase I1 and the calcium/calmodulin-dependent enzyme are located in both the nucleus and the cytoplasm (60,69,70), while protein kinase C is predominantly cytoplasmic (71,72). Thus, the use of total cell homogenates provided a system that was not biased toward or against any of the above kinases.
Drosophila topoisomerase I1 exists in the cell as a phosphoprotein (41-43) (Fig. l ) , but the physiological consequences of the modification are as yet unknown. Considering that 1) the activity of topoisomerase I1 increases markedly during the transition from cell quiescence to proliferation (25)(26)(27)(28)(29)(30), 2) the enzyme's intracellular location changes during the course of the cell cycle (17)(18)(19)23), and 3) phosphorylation/dephosphorylation events commonly affect both the enzymatic activities (34,(37)(38)(39)(40) and cellular locations (35, 36) of many protein systems, it is highly probable that phosphorylation plays a role in regulating at least some of the in uiuo functions of topoisomerase 11. This conclusion is strongly supported by the fact that serine phosphorylation stimulates the activity of the Drosophila enzyme in uitro (42)(43)(44). At the present time, virtually nothing is known about the biological events that control the cellular modification of topoisomerase 11. Clearly, identifying the kinase that mediates the in uiuo phosphorylation of the enzyme represents an important step toward elucidating these events. On the basis of the results presented above, it appears likely that casein kinase I1 is the enzyme that is primarily responsible for phosphorylating DNA topoisomerase I1 in the embryonic Drosophila Kc cell line.