Structure-Function Analysis of Casein Kinase 2 with Synthetic Peptides and Anti-peptide Antibodies*

Casein kinase 2 (CK2) is a ubiquitous, multifunc- tional protein-seryl/threonyl kinase that has been implicated in cellular regulation. Synthetic peptides were patterned after three highly conserved regions in CK2: the N terminus (CKB-NT); the lysine-rich, kinase sub- domain I11 segment (CKB-111) (nomenclature of Hanks et d. (Hanks, S. K., Quinn, A. M., and Hunter, T. (1988) Science 241,42-52)); and a 10-residue segment located near kinase subdomain X that is shared be- tween CK2 and ~ 3 4 ~ ‘ ’ (CK2/cdc2). The CK2-I11 and CK2/cdc2 peptides markedly stimulated the autophos- phorylation of the a- and a’-subunits of purified CK2 from sea star oocytes, and they elicited up to 2-fold increases in its casein or phosvitin phosphotransferase activity. These peptides completely reversed nearly total inhibition of CK2 phosphotransferase activity toward itself, casein, and phosvitin by either heparin or poly(Glu,Tyr; 4:1), whereas CKP-NT was ineffective. Elution of CK2 from heparin-agarose with the CK2- I11 peptide indicated that this region of CK2 might mediate heparin binding to CK2. Affinity-purified rab- bit polyclonal

Casein kinase 2 (CK2) is a ubiquitous, multifunctional protein-seryl/threonyl kinase that has been implicated in cellular regulation. Synthetic peptides were patterned after three highly conserved regions in CK2: the N terminus (CKB-NT); the lysine-rich, kinase subdomain I11 segment (CKB-111) (nomenclature of Hanks et d. (Hanks, S . K., Quinn, A. M., and Hunter, T. (1988) Science 241,42-52)); and a 10-residue segment located near kinase subdomain X that is shared between CK2 and ~3 4~' ' (CK2/cdc2). The CK2-I11 and CK2/cdc2 peptides markedly stimulated the autophosphorylation of the a-and a'-subunits of purified CK2 from sea star oocytes, and they elicited up to 2-fold increases in its casein or phosvitin phosphotransferase activity. These peptides completely reversed nearly total inhibition of CK2 phosphotransferase activity toward itself, casein, and phosvitin by either heparin or poly(Glu,Tyr; 4:1), whereas CKP-NT was ineffective. Elution of CK2 from heparin-agarose with the CK2-I11 peptide indicated that this region of CK2 might mediate heparin binding to CK2. Affinity-purified rabbit polyclonal antibodies developed against both CK2-I11 and CK2/cdc2, but not CKP-NT, also produced up to 1.8-fold enhancements of the casein and phosvitin phosphotransferase activities of purified CK2. All three of the antipeptide antibody preparations immunoreacted with the a-and a'-subunits of CK2 on Western blots. These studies indicate that kinase subdomains I11 and X are involved in the modulation of CK2 phosphotransferase activity.
Casein kinase 2 (CK2)' appears to be universally distributed *This research was supported by an operating grant from the Medical Research Council of Canada (to S. L. P.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The abbreviations used are: CK2, casein kinase 2; CK2-NT, synthetic peptide patterned after the first, N-terminal 22 residues of Drosophila CK2; CK2-111, synthetic peptide patterned after a 20residue region located in catalytic subdomain I11 region of human CK2; CK2/cdc2, a synthetic peptide that includes common 10-residue segments found in both Drosophila CK2 and S. pombe ~34""'; p3pdc2, 34-kDa protein kinase encoded by the S. pombe cell division control-2 gene and its functional homologues in other species; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; TBS, Tris-buffered saline; Mops, MOPS, 4-morpholinepropanesulfonic acid.
in eukaryotes and has been detected in the cytosol, nucleus, mitochondria, and membranes of their cells (for reviews, see Refs. 1-3). CK2 occurs typically as a tetrameric complex (Mr = 130,000) with an a2P2 or aa'P2 configuration. For example, CK2 purified from sea star oocytes features a 44-kDa asubunit, a 40-kDa a'-subunit, and two 28-kDa ,%subunits (4). cDNA sequence analysis of the a-and a'-subunits from diverse species has confirmed that they are catalytic subunits encoded by distinct, but highly homologous genes (5)(6)(7)(8)(9). The extreme conservation of the primary structures of the a-and a'-subunits of CK2, as well as its &subunit (9-11) during evolution, together with its ubiquitous distribution, points to a fundamental role for this protein kinase in eukaryotic cells.
CK2 is likely a pleiotropic enzyme in vivo, since it can phosphorylate over 50 proteins in vitro (2). Studies with synthetic peptide substrates have indicated that either an aspartyl, glutamyl, phosphoseryl, or phosphotyrosyl residue at the third position from the C-terminal side of the target phosphorylatable residue is necessary and sufficient for recognition by casein kinase 2 (12)(13)(14)(15)(16). Recent findings have implicated CK2 in the control of such nuclear events as gene expression and oncogenic transformation. Among others, Fos (17), Myb (18), Myc (19), p53 tumor suppressor protein (20), adenovirus E1A protein (17), papillomavirus E7 protein (21,22), and SV40 large T antigen (1,23) are some of the putative targets of CK2. In this regard, it is significant that modest stimulations of CK2 activity have been reported in response to growth factor treatments of cells (24-28) or during cell proliferation (29). The mechanisms by which these stimulations of CK2 activity are achieved are obscure, but might reflect posttranslational modification of the enzyme by phosphorylation.
Bacterially expressed, recombinant a-subunits of casein kinase 2 from Caenorhubditis elegam (8), Drosophila (39), and human (40) have been shown to feature most of the characteristics of the tetrameric form, including catalytic activity with either ATP or GTP and an extreme sensitivity to inhibition by heparin (ICso = 0.1-0.3 rg/ml). However, co-expression of the P-subunit appears to be necessary for stimulation of the phosphotransferase activity of the a-subunit toward casein by greater than 10-fold, and it is required to mediate the stimulatory effects of polyamines and polylysine on CK2 activity.
Relatively little is known about the locations of the precise binding sites in CK2 for acidic inhibitors and basic activators of this protein kinase. In this study, synthetic peptides patterned after conserved regions in the primary structure of CK2 were evaluated for their effects on the phosphotransferase activity of CK2 in uitro. Two of these synthetic peptides, as well as antipeptide antibodies developed against these sequences, were particularly effective for activation of CK2 and reversal of nearly complete inhibition of CK2 activity by heparin and poly(Glu,Tyr; 4:l).
Casein K i m e 2 Activity and Protein Assays-CK2 activity in purified preparation of the enzyme or cytosolic extracts from sea star oocytes was assayed for 10 min at 30 "C in a final volume of 25 pl with 5 mg/ml partially dephosphorylated casein or 5 mg/ml phosvitin, 5 mM MgClz, 12 mM Mops, pH 7.2, and 50 p M [y-32P]ATP (2000 cpm/pmol). The reaction was terminated by spotting 20 pl onto a 1.5-cm2 piece of Whatman P-81 phosphocellulose paper. After the P-81 papers were washed extensively with 1% (w/v) phosphoric acid, they were transferred into 6-ml plastic vials containing 0.5 ml of Ecolume (ICN) scintillation fluid, and the radioactivity was quantitated in a Wallac (Pharmacia LKB Biotechnology Inc.) scintillation counter.
Protein was estimated by the method of Bradford (41) using bovine serum albumin as a standard (A%nm = 6.5).
Electrophoresis-SDS-PAGE was performed on 1.5-mm-thick gels, with acrylamide at 11% (w/v) in the separating gel and 4% (w/v) in the stacking gel, using the buffer system described by Laemmli (42). Samples were boiled for 5 min in the presence of SDS-PAGE sample buffer (125 mM Tris-HC1, pH 6.8, 4% SDS, 0.01% bromphenol blue, 10% mercaptoethanol, and 20% glycerol) and electrophoresed for -17 h at 10 mA. For autoradiography, gels were exposed to Kodak XAR-5 film at room temperature.
Immunological Studies-After SDS-polyacrylamide gel electrophoresis of purified CK2, the separating gel was soaked in transfer buffer (25 mM Tris, 192 mM glycine, 20% methanol) for 5 min and sandwiched with a nitrocellulose membrane, and then the proteins were transferred for 3 h at 250 mA.
Subsequently, the nitrocellulose membrane was blocked with TBS (Tris-buffered saline) containing 5% skim milk for 2 h at room temperature. The membrane was washed twice with TBS containing 0.05% Tween 20 (TTBS) for 5 min before incubation with CK2-specific anti-peptide antibodies (in 1% skim milk/TTBS; 1:500 dilution) overnight at room temperature. Next day, the membrane was washed twice with TTBS before incubation with the second antibody (goat anti-rabbit IgG coupled to alkaline phosphatase in 1% skim milk/TTBS; 1:3000 dilution) for 2 h at room temperature. The membrane was rinsed twice with TTBS followed by one wash with TBS before incubation with 5-bromo-4chloro-3-indolyl phosphate/nitro blue tetrazolium color development solution (mixture of 3% nitro blue tetrazolium in 1 ml of 70% dimethylformamide and 1.5% 5-bromo-4-chloro-3-indolyl phosphate in 1 ml of 100% dimethylformamide before adding to 100 ml of 0.1 M NaHC03, 10 mM MgClz, pH 9.8). The color reaction was terminated after 5-15 min by rinsing the membrane in a large volume of water.

RESULTS AND DISCUSSION
Regulation of CK2 Activity by Kinase Catalytic Domainbased Peptides-Hanks et al. (43) have identified approximately 25 residues, located in 8 of 11 subdomains, which are highly conserved within the -250-residue catalytic domain of nearly all sequenced protein-seryl/threonyl kinases (Fig. 1). We rationalized that other short sequences that are highly conserved in CK2 from diverse species, but not preserved in other members of the protein kinase family, might confer specific regulation or define the substrate selectivity of CK2. We focused our attention on three such uniquely conserved regions in CK2. These were the N-terminal22 residues of the kinase (CK2-NT), a lysyl-rich 20-residue segment that encompassed a highly conserved glutamyl residue in kinase subdomain I11 (CK2-111), and a 10-residue segment located at kinase subdomain X (CK2/cdc2) (Fig. 1). To examine the effects of these peptides on CK2, we employed an essentially homogeneous preparation of the enzyme obtained from maturing sea star oocytes. We recently established that the echinoderm CK2 displays many of the distinguishing characteristics of the mammalian forms of CK2 (4).
Incubation of purified CK2 with each of the synthetic peptides produced stimulation of its phosphotransferase activity toward either casein or phosvitin (Fig. 2). In the instance of casein, at 1 mg/ml concentrations of the peptides, the order of effectiveness of the peptides was CKS-III> CK2/ cdc2 > CK2-NT (Fig. 2 A ) . At 2 mg/ml, CK2-I11 increased the casein kinase activity by greater than 2-fold, whereas the stimulation by CK2-NT was -1.2-fold. With phosvitin as the phosphoacceptor, at 1 mg/ml concentrations, all of the peptides were equally effective in activation of CK2, but each of the stimulations were modest (-1.4-fold) (Fig. 2B). For both   (43). Note that for CK2 in most species where primary sequence information is available, the residues denoted with asterisks are instead S in subdomain I, V in subdomain 11, W in subdomain VII, and G in subdomain VIII. The boldfuce sequences in the synthetic peptides shown below are derived from CK2. casein and phosvitin, the increases in CK2 activity by CK2/ cdc2 was reversed when the concentration of the peptide was raised to -2 mg/ml. None of the peptides were found to be phosphorylated by sea star CK2 in control experiments (data not shown).
Stimulation of CK2 activity by each of the three peptides was surprising, although the CK2-I11 peptide was highly charged with 9 out of 23 residues derived from either arginine or lysine. At a hundredfold lower concentration, polylysine and polyarginine have been shown to potently enhance CK2 activity toward casein, and they stimulated a-subunit autophosphorylation at the expense of the @-subunit phosphorylation (38,44). As presented in Fig. 3B, CK2-I11 also enhanced a-and a'-subunit autophosphorylation, but without compromise of the @-subunit phosphorylation. Similar results were obtained with CK2-NT (Fig. 3B) and spermine (Fig. 3A). By contrast, CK2/cdc2, which was a very much less basic peptide, produced strong increases in a-and a'-subunit autophosphorylation at 0.4-1 mg/ml but a reduction of @-subunit phosphorylation (Fig. 3B). At ( l a n e 13) and presence of 0.2-1.0 pg/ ml poly(Glu,Tyr; 4 1 ) (lanes 1-4), 4-20 mM spermine (lanes 5-8), and 4-20 pg/ml heparin (lanes 9-12). Subsequently, SDS-PAGE and autoradiography were performed as described under "Experimental Procedures." Panel € 3 , prior to SDS-PAGE, purified CK2 was subjected to autophosphorylation in the absence (lane 24) and presence of 0.4-2 mg/ml CK2-NT (lanes I d ) , 0.4-2 mg/ml CK2 -111 (lanes 4-8), and 0.4-2 mg/ml CK2/cdc2 (lanes 9-13). Autoradiograms are shown. Migrations of marker proteins bovine serum albumin (67 kDa), ovalbumin (45 kDa), and carbonic anhydrase (31 kDa), as well as the a-, a'-, and @-subunits of CK2, are indicated. Similar results were obtained in two independent experiments. were reversed. Poly(Glu,Tyr; 4:l) and heparin also inhibited autophosphorylation of all three subunits of CK2 (Fig. 3A). Thus, there was a strong correlation between CK2 activity toward exogenous substrates and the a-and a'-subunit autophosphorylation but not with &subunit autophosphorylation in the presence of the various peptides and other agents.
Regulation of CK2 Activity by Anti-CK2 Peptide Antibodies-A simple model for the stimulatory effects of the synthetic peptides on CK2 activity is that they disrupted internal interactions within CK2 that negatively modulate the kinase. The catalytic domain of CK2 might be rendered more accessible to substrates by a partial opening of the kinase in subdomains I11 and X, upon which CKZ-I11 and CK2/cdc2 were based. Presumably, these peptides bound to those regions in the CK2 protein that normally interacted with subdomains I11 and X. Extremely high concentrations of the peptides could be necessary, rather than stoichiometric amounts, if the peptides only transitorily adopted the optimal conformations. It is more difficult to envision how CK2-NT exerted its marginal effects on CK2 activity.
To test the above hypothesis, polyclonal rabbit antibodies were raised against the three CK2 peptides, and each of these preparations was affinity-purified on the appropriate peptideagarose column. As shown in Fig. 4, each of these antibodies immunoreacted with the a-and a'-subunits of purified sea star CK2. They were also highly specific for these polypeptides in crude cytosolic extracts from sea star oocytes and various mammalian sources, with the exception of a -55-kDa polypeptide, which may be a close relative of CK2 and has aa' -P - 3 4 FIG. 4. Immunoblotting of purified CK2 with antipeptide antibodies. Following SDS-PAGE, purified sea star CK2 was immunoblotted with affinity-purified rabbit polyclonal antipeptide antibodies raised against the CK2-NT peptide (lane I ) , CK2/cdc2 peptide ( l a n e 2 ) , and CK2-111 peptide (lanes 3 and 4 ) . Lanes 2-3 correspond to samples where the antipeptide antibodies were incubated with the immunoblots in the absence of the corresponding immunogenic peptide, and the a and a' subunits of CK2 were detectable. In lane 4, immunoreactivity of the CK2 catalytic subunits with the anti-CK2-111 antibody was competed with 1 mg/ml of the CK2-111 peptide. Similarly, the binding of the anti-CK2-NT and anti-CK2/ cdc2 antibodies to CK2 were blocked by the appropriate synthetic peptides (data not shown). The migration positions of the prestained marker proteins bovine serum albumin (80 kDa), ovalbumin (50 kDa), carbonic anhydrase (33 kDa), soybean trypsin inhibitor (28 kDa), and lysozyme (19 kDa) are indicated. not yet been described (Ref. 4 and data not shown). Although the anti-CK2-NT antibodies had little or no effect on the casein phosphotransferase activity of purified CK2, both the anti-CK-I11 and anti-CK2/cdc2 antibodies facilitated almost 2-fold activations of CK2 (Fig. 5). This indicated that the binding of the antibodies to subdomains I11 and X did not interfere with the catalytic function of the kinase but apparently relieved negative regulation of the enzyme in keeping with the above model.

Relief of Heparin and Poly(Glu,Tyr; 4:l) Inhibition of CK2 by Synthetic
Peptides-Heparin and poly(Glu,Tyr; 4:l) are among the most potent inhibitors of CK2 that have been described. We considered the possibility that the sites of action of these inhibitors might involve the N terminus or kinase subdomains I11 and X and tested the effects of the CK2 peptides in concert with the heparin and poly(Glu,Tyr; 4:l) on the enzyme. CK2-I11 and CK2/cdc2 similarly reversed almost complete inhibition of the casein and phosvitin phosphotransferase activity of purified CK2 by 10 pg/ml heparin or 1 pg/ml poly(Glu,Tyr; 4:l) (Fig. 6). CK2-NT had little effect in this regard (Fig. 6 ) . Likewise, CK2-111, but not CK2-NT, reversed partial inhibition of the casein phosphotransferase activity of CK2 in cytosolic extracts from sea star oocytes (Fig. 7).
The ability of the CK2 peptides to reverse inhibition of CK2 autophosphorylation by 10 pg/ml heparin or 1 pg/ml poly(Glu,Tyr; 4:l) was also examined (Fig. 8). CK2-I11 and CK2/cdc2 both restored the autophosphorylation of all three subunits in the presence of either inhibitor, whereas CK2-NT only produced modest recoveries of the a'-and P-subunit phosphorylations. This restoration of the P-subunit phosphorylation by CK2/cdc2 was particularly intriguing, as this peptide reduced p-subunit phosphorylation in the absence of heparin or poly(Glu,Tyr; 4:l) (Fig. 3B).
Identification of Subdomain III as the Primary Heparinbinding Site on CK2"The high concentration of basic residues in kinase subdomain I11 of CK2 makes this an attractive site for interaction with the polysulfonated glycosaminoglycan heparin and poly(Glu,Tyr; 4:1), which are highly negatively charged at neutral pH. The ability of CK2-I11 to reverse inhibition by these polymers was consistent with this notion. T o further test whether this region of CK2 was important for heparin binding, the ability of all three CK2 peptides to elute purified CK2 bound to heparin-agarose beads was assessed. Only the CK2-I11 peptide effectively caused the release of Values are the means of duplicate incubations. Note that higher concentrations of heparin and poly(Glu,Tyr; 4:l) are required for inhibition of CK2 in crude extracts and that some of the casein phosphotransferase activity is contributed by other protein kinases. bound CK2; -33% of the adsorbed CK2 was eluted with 2.4 mg/ml CK2-I11 (Fig. 9).
The results of Hu and Rubin (8), in which two of the lysyl residues in subdomain I11 of C. elegans CK2 were converted to glutamyl residues by site-specific mutagenesis (i.e. K E KIKR) and the IC5o for heparin inhibition of CK2 was increased 7O-fold, strongly support the assignment of subdomain I11 as the heparin-binding site. The preponderance of basic residues in subdomain I11 could also facilitate nuclear localization of CK2, as well as substrate binding, since preferred CK2 targets contain acidic residues in the vicinity of the phosphoacceptor amino acid (2). However, CK2 from Saccharomyces cerevisiae contains two nonconservative re- Panel A, purified CK I1 was subjected to autophosphorylation in the absence (lane 11) and presence of 10 pg/ml heparin (lanes 1-10), 0.5-2 mg/ml CK2/cdc2 (lanes 1-3), 0.5-2 mg/ml CK2-NT (lanes 4-6), or 0.5-2 mg/ml CK2-I11 (lanes 7-9). Panel B, purified CK I1 was subjected to autophosphorylation in the absence (lane 11) and presence of 1 pg/ml poly(Glu,Tyr; 41) (lanes 1-10), 0.5-2 mg/ml CK2/ cdc2 (lanes 1 3 ) , 0.5-2 mg/ml CK2-NT (lanes 4-6), or 0.5-2 mg/ml CK2-111 (lanes 7-9). Autoradiograms of the gels following SDS-PAGE analysis are shown; the exposure time was 12 h for panel A and 36 h for panel B. Migrations of marker proteins bovine serum albumin (67 kDa), ovalbumin (45 kDa), and carbonic anhydrase (31 kDa), as well as the a-, a'-, and 0-subunits of CK2 are indicated. Similar results were obtained in two independent experiments. placements in 2 of these lysyl residues (i.e. KXKKIKR) (7). Likewise, the binding affinity for casein was not affected in the C. elegans K B K I K R CK2 mutant developed by Hu and Rubin (8). Therefore, in view of these observations, it seems that at least 3 of the lysyl residues in subdomain I11 of CK2 are nonessential for substrate recognition.
For at least one other protein kinase, i.e. ~34'~"', it has been shown that synthetic peptides modeled after subdomain I11 can exert biological effects. Doree and collaborators (45,46) have described the induction of sea star oocyte maturation and the augmentation of maturation-promoting factor-induced Xenopus oocyte maturation by a peptide that contains the sequence EGVPSTAIREISLLKE. Antibodies developed against this sequence have been relatively selective probes for p3PdC2 in different species (for review, see Ref. 47). The present findings with CK2 indicate that synthetic peptides and antipeptide antibodies based upon subdomain 111s of other protein kinases could provide useful structure-function information. We have already successfully developed antipep- tide antibodies based upon the subdomain I11 regions of ~4 3~'~' , p37Pi"' ' , p70 S6 kinase, p74'"*, and p105wee', which immunoblot the native forms of these protein kinases.2 Locations of Kinase Subdomains 111 and X in CK2"Recently, the three-dimensional structure of the catalytic subunit of murine cyclic AMP-dependent protein kinase was elucidated (48). Due to the high degree of conservation of protein kinases within their catalytic subdomains (43), this structural information has important predictive value for the architecture of CK2. Subdomains I11 and X are located relatively close together in cyclic AMP-dependent protein kinase at the interface between the two lobes of the kinase, which is also near the active site of the enzyme. If a similar arrangement exists in CK2, then it is possible that these subdomains may associate with each other and sterically hinder access to the active site. Disruption of subdomains I11 and X interaction by either CK2-I11 or CK2/cdc2 would be predicted to result in similar effects on the catalytic activity of CK2, as was observed in this study. One possible test of this model would be site-specific mutation of highly conserved residues in subdomain X of CK2, which could lead to an enhancement of the phosphotransferase activity of CK2.
Finally, it may be significant that ~3 4 '~"~ shares a very similar, highly conserved subdomain X region with CK2, despite the marked differences in their subdomain I11 regions. Unfortunately, we have failed to observe any effects of the CK2/cdc2 peptide on the histone H1 kinase activity of sea star ~3 4 '~'~, nor immunoreactivity of the echinoderm ~3 4 '~'~ with anti-CKP/cdcZ antib~dies.~ Nevertheless, subdomain X in ~3 4 '~~' is likely to play a regulatory role in this kinase, as the present study has implicated for CK2.