Protein Kinase CK2α′ Is Induced by Serum as a Delayed Early Gene and Cooperates with Ha-ras in Fibroblast Transformation*

Protein kinase CK2 is an ubiquitous and pleiotropic Ser/Thr protein kinase composed of two catalytic (α and/or α′) and two noncatalytic (β) subunits forming a heterotetrameric holoenzyme involved in cell growth and differentiation. Here we report the identification, cloning, and oncogenic activity of the murine CK2α′ subunit. Serum treatment of quiescent mouse fibroblasts induces CK2α′ mRNA expression, which peaks at 4 h. The kinetics ofCK2α′ expression correlate with increased kinase activity toward a specific CK2 holoenzyme peptide substrate. The ectopic expression of CK2α′ (or CK2α) cooperates with Ha-ras in foci formation of rat primary embryo fibroblasts. Moreover, we observed that BALB/c 3T3 fibroblasts transformed with Ha-ras and CK2α′ show a faster growth rate than cells transformed with Ha-rasalone. In these cells the higher growth rate correlates with an increase in calmodulin phosphorylation, a protein substrate specifically affected by isolated CK2 catalytic subunits but not by CK2 holoenzyme, suggesting that unbalanced expression of a CK2 catalytic subunit synergizes with Ha-ras in cell transformation.

and cell differentiation. The nuclear proteins that are CK2 substrates includes: c-Myc (7), Max (7), c-Myb (8), serum response factor (SRF) (9), DNA ligase I (10), DNA topoisomerase 2 (11), p53 (12), and c-Fos (13). In mammalian cells phosphorylation of nuclear factors dependent on CK2 could be relevant for cell growth regulation and the progression into the cell cycle. A direct role of CK2 activity in cell cycle progression has been demonstrated by antibody-mediated CK2 depletion and by gene inactivation in Saccharomyces cerevisiae (14,15). Although hundreds of papers have been published on the subject, it is still unknown how the enzyme is regulated in vivo (4,5,16). CK2␤ undergoes stoichiometric autophosphorylation and both CK2␤ and CK2␣ (but not CK2␣Ј) are phosphorylated in vitro and in vivo by p34 Cdc2 kinase (17). However, these phosphorylations do not correlate with any regulation of activity. Moreover, it is not clear whether the holoenzyme represents an up-or a down-regulated form of the kinase, because some substrates are preferentially phosphorylated by the tetramer, but others, like calmodulin, are phosphorylated only by the free catalytic subunits (6).
In transgenic mice it was possible to demonstrate that in T cells the overexpression of the catalytic CK2␣ subunit enhanced the onset of lymphomas induced by either c-myc or tal-1 (18,19). These results shed new light on the previous observations that cattle infected by the parasite Theileria parva developed T cell lymphomas, because parasite-infected cells show increased CK2 activity (20 -22). Opposite results were obtained by the overexpression of CK2␣ in NIH 3T3 mouse fibroblasts. In these cells CK2␣ overexpression resulted in deactivation of the mitogen-activated protein kinase kinase and suppression of ras-dependent cell transformation (23).
To identify genes potentially involved in cell growth, we performed a differential screening for the isolation of transcripts induced by mitogenic stimuli within the G 1 phase. Here we report the identification and cloning of the murine lower molecular weight catalytic subunit, CK2␣Ј. We observed that in mouse fibroblasts CK2␣Ј is induced by serum treatment as a slow-early gene. Together with CK2␣Ј also CK2␣ and CK2␤ are induced. Cotransfection of an expression vector containing CK2␣Ј together with a vector expressing Ha-ras induced foci formation in rat primary embryo fibroblasts. Moreover, rastransformed BALB/c 3T3 fibroblasts overexpressing CK2␣Ј showed a faster growth rate than cells transformed with Ha-ras alone. ras-Transformed fibroblasts overexpressing CK2␣Ј also exibited increased phosphorylating activity toward calmodulin, which is a specific substrate of CK2 catalytic subunits. These findings suggest that unbalanced expression of either CK2␣Ј or CK2␣ plays a role in fibroblast cell transformation.

EXPERIMENTAL PROCEDURES
Cells and Cell Culture-NIH 3T3 and BALB/c 3T3 fibroblasts were grown at 37°C in DMEM 1 supplemented with 10% heat-inactivated FCS, penicillin-streptomycin, and glutamine. The cells were expanded by trypsin-EDTA treatment and subcultured at a ratio of 1:3 every 2-3 days. Rat embryo fibroblasts were isolated as described previously (24). Briefly, 14-day CDF(F344) rat embryos were sacrificed, rinsed, and trypsinized for 30 min at 37°C. DMEM containing 10% FCS was added, and the cells were centrifuged, dispersed, counted, and plated on 100-mm tissue culture dishes at a density of 2 ϫ 10 6 /dish. After 48 h the cells were trypsinized, and aliquots were frozen in liquid nitrogen.
Differential Display and Cloning of the Murine CK2␣Ј cDNA-To induce a relatively quiescent cell population, subconfluent NIH 3T3 fibroblasts were incubated for 48 h in DMEM plus 0.5% FCS. Cells were then treated for 2 and 4 h with DMEM supplemented with 10% FCS. Total cellular RNA was extracted using the guanidinium thiocyanate method (25) from quiescent and serum-treated fibroblasts and subjected to the differential display technique as described previously (26,27). The amplified cDNA fragments were compared in nondenaturing polyacrylamide gels. A serum-induced cDNA fragment, named L-0401, was excised, recovered by boiling, reamplified, and cloned into pGEM-T vector (Promega). The L-0401 cDNA fragment (230 bp), whose corresponding mRNA was homologous to human CK2␣Ј, was labeled with [ 32 P]dCTP by random primer labeling and used to screen a mouse fibroblast cDNA library (27). The positive clones, inserted into the pBluescript SK vector, were sequenced on both strands either automatically using a Perkin-Elmer model 373 DNA sequencer or manually using a Sequenase 2.0 kit (U. S. Biochemical Corp). A positive clone, named pBS38␣Ј (FS304), contained the full-length mouse CK2␣Ј cDNA.
Northern Blot Analysis-Total RNA (10 g) was run on denaturing formaldehyde-agarose gels and stained with ethidium bromide to verify that each lane contained similar amounts of undegraded rRNA. RNA was transferred onto nylon membranes and cross-linked by UV irradiation. Filters were hybridized with 32 P-labeled probes and washed as described (27). The mouse CK2␣Ј probe was obtained from the fulllength cDNA (pBS38␣Ј). The cDNA fragments of murine CK2␣ (base pairs 421-841) and CK2␤ (base pairs 912-1321) were obtained by cDNA amplification of a mouse fibroblast cDNA library (27). The sequences of the primers used for amplification (5Ј-GCTTCGATATGAC-CGTCACG-3Ј and 5Ј-GACTCAACTACTAAATCCG-3Ј for CK2␣ and 5Ј-GTACCAGCAGGGAGACTTTGGCTAC-3Ј and 5Ј-CATAGACTTCCTG-AAAGGGTGGCAG-3Ј for CK2␤), were obtained from EBI Nucleotide Sequence Data Base under accession number U17112 for CK2␣ and X56502 for CK2␤. The amplified cDNA fragments were sequenced, labeled with [ 32 P]dCTP, and used in Northern blot.
Construction of Expression Vectors-CK2␣Ј open reading frame (from P-2 to R-350) was amplified by PCR from pBS38␣Ј with a 5Ј primer containing a BamHI restriction site (5Ј-CGCGGATCCCGGCCCGGCC-GCG-3Ј) and a 3Ј primer containing a KpnI restriction site (5Ј-CGGG-GTACCTCATCGTGCTGCGGT-3Ј). The PCR product was digested with BamHI and KpnI and cloned into the BamHI/KpnI site of a bacterial expression vector (FS310) made by subcloning the XbaI/XhoI fragment of pQE-31 (Qiagen) into the XbaI/XhoI site of pBluescript SK to give pQEBS␣Ј (FS311). The construct expressing murine CK2␣Ј under the control of the cytomegalovirus promoter, pcDNA␣Ј, was made by subcloning a StuI/partial XhoI-digested fragment from pBS38␣Ј into the EcoRV/XhoI site of pcDNA3 (Invitrogen). To obtain murine CK2␣ under the control of the cytomegalovirus promoter, the 5Ј end of CK2␣ was amplified by PCR from pT7-7CKII␣ (28) with a 5Ј primer containing a HindIII restriction site and an optimal Kozak sequence (29) (5Ј-GAG-AAAGCTTCCACCGCCATGTCGGGACCCGTGCC-3Ј) and a 3Ј primer downstream from an XhoI restriction site (5Ј-CTTGATTTCCCCATTC-CACC-3Ј). The PCR product was digested with HindIII and XhoI and cloned into the HindIII/XhoI site of pcDNA3 to give pcDNA␣5Ј (MO339). A XhoI/blunt-ended HindIII fragment, corresponding to the 3Ј end of CK2␣, was released from pT7-7CKII␣ and subcloned into the XhoI/blunt-ended ApaI site of pcDNA␣5Ј to give pcDNA␣ (MO346). All constructs were sequenced on both strands.
Transfection of Cells and Foci Formation-Subconfluent cells were fed with culture medium 1-2 h before transfection. Cells were cotransfected with 5 g of activated Ha-ras and 10 g of each construct for 8 h by standard CaPO 4 precipitation procedures (24). Where necessary vector plasmid (pcDNA3) was added to reach the total amount of 15 g of DNA per transfected plate. At least two different cesium chloride DNA preparations of each construct were independently transfected. The cells were rinsed with PBS and re-fed with culture medium 15 h posttransfection. The cells were trypsinized 24 h after transfection and split 1:3. When the cells reached confluence they were re-fed with DMEM containing 2% FCS, and the medium was changed every 2 days. Stable cell transformants were visible after 7-10 days. Individual foci from BALB/c 3T3 cells transformed by Ha-ras or by Ha-ras plus CK2␣Ј were picked and examined for their growth properties. For growth rate analysis cells were plated in duplicate at 1 ϫ 10 5 cells on 60-mm Petri dishes in DMEM plus 2% FCS. Cells were counted every day using a hemacytometer and the medium was changed every 3 days. For tumor growth assays, 10 6 cells in midlog-phase growth were harvested, washed with PBS, resuspended in 200 l of PBS, and injected subcutaneously into the scapular region of BALB/c nude mice. After 2 weeks of growth, the mice were sacrificed, and the tumors were surgically removed and weighed.
Cell Extracts and Phosphorylation Assays-Whole-cell extracts were prepared by rinsing cultures grown on Petri dishes with PBS followed by harvesting with a rubber policeman. Cells were pelleted by brief centrifugation and resuspended in lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 2 mM EGTA, 0.05% Triton X-100) containing a protease inhibitor mixture (1 mM phenylmethylsulfonyl fluoride, 2 g/ml aprotinin, 0.5 g/ml leupeptin, 0.7 g/ml pepstatin A). The cell lysates were centrifuged at 40,000 ϫ g for 20 min at 4°C, and the supernatant was used in phosphorylation assays. Phosphorylation experiments were performed by incubating the substrate (200 M peptide RRRADDSD-DDDD or 10 M calmodulin) in a buffer containing 50 mM Tris-HCl, pH 7.5, 12 mM MgCl 2 , 100 mM NaCl, and 20 M [␥-32 P]dATP for 10 min at 37°C. NaCl was omitted when calmodulin was used as substrate. The reaction was stopped by cooling in ice followed, in the case of calmodulin, by SDS-polyacrylamide gel electrophoresis, staining with Coomassie Blue and either autoradiographed or directly scanned on Instant Imager Apparatus (Canberra-Packard). 32 P incorporated into the peptide substrate was evaluated by the phosphocellulose paper procedure (30).

Isolation and Characterization of the Murine CK2␣Ј
cDNA-In cultured mouse fibroblasts growth factor depletion leads the cell to exit from the cell cycle and become quiescent. Serum treatment induces re-entry into the cell cycle, which is likely because of the induction of early genes. To identify new serum-induced genes, we used the mRNA differential display technique (26,27). NIH 3T3 fibroblasts were serum-starved for 48 h and the RNAs were collected either from starved cells or from cells treated with serum at different time points. Several cDNAs, obtained only from serum-induced cells, were amplified and sequenced. Comparison of the cDNA fragments with the EBI Nucleotide Sequence Data Base revealed that a cDNA induced at 4 h after serum treatment was highly similar to the human protein kinase CK2␣Ј. To clone the full-length cDNA coding for CK2␣Ј, a mouse fibroblast cDNA library (27) was screened. The few positive clones were sequenced using internal sequencing primers, and the nucleotide sequence of one clone of 1877 base pairs in length revealed a single open reading frame coding for a putative protein of 350 amino acid residues. The mouse and human predicted protein products shared 98.9% amino acid identity. The strong similarity of the murine CK2␣Ј deduced protein sequence with its corresponding human homologue suggests a highly conserved function of CK2␣Ј. Murine CK2␣Ј protein shows a lower degree of identity with CK2␣. The two proteins share 82.4% identity over 347 amino acids overlap (Fig. 1). The greatest difference is in the C-terminal domains, because the deduced CK2␣Ј protein sequence is 41 amino acids shorter and thus lacks the p34 Cdc2 sites phosphorylated during the cell cycle (17). CK2␣Ј also lacks the HEHRKL amino acid residues (166 -171 of CK2␣) that have been implicated in the interaction with protein phosphatase 2A (23).
To confirm that the cDNA encoded a biologically active CK2␣Ј enzyme we expressed the cDNA in Escherichia coli.
Recombinant CK2␣Ј showed a molecular mass of about 41 kDa consistent with the predicted size of the protein, and the nondenatured soluble bacterial extract containing the immunoreactive CK2␣Ј was able to phosphorylate the CK2 peptide substrate RRRADDSDDDDD in vitro (not shown).
CK2 is an ubiquitously expressed protein kinase essential for cell growth. Northern blot analysis with a CK2␣Ј probe revealed two hybridizing transcripts of 2.2 and 4.2 kilobases, respectively. The expression of CK2␣Ј mRNA is relatively constant in all tissues with the exception of testis where a much stronger CK2␣Ј signal was detected (Fig. 2). These results contrast with CK2␣ expression, which is abundant in brain and barely expressed in testis (31,32). Thus, CK2␣ and CK2␣Ј are differentially regulated in these tissues.
CK2␣Ј Is Induced by Serum Treatment in Cultured Fibroblasts-The identification of CK2␣Ј in a screening for mRNAs induced by serum treatment of quiescent fibroblasts suggested that this gene is induced by mitogenic stimuli. To measure the induction of the CK2␣Ј transcript, we performed Northern blot analysis on CK2␣Ј mRNA in fibroblasts before and after serum treatment. As can be observed in Fig. 3A the CK2␣Ј mRNA was low in quiescent fibroblasts and increased about 50% in seruminduced fibroblasts, with a peak induction at 4 h. The quantitative analysis of the transcripts normalized to the glyceraldheyde-3-phosphate dehydrogenase mRNA levels is reported in Fig. 3B. The serum-induced increase in CK2␣Ј transcripts was not blocked by the presence of the protein synthesis inhibitor, cycloheximide, as shown by the superinduction of CK2␣Ј mRNA (Fig. 3, A and B). Thus, in mouse fibroblasts CK2␣Ј is induced with slow kinetics by serum treatment, and this induction is independent of protein synthesis. Because protein kinase CK2 is a tetramer containing two catalytic and two regulatory subunits, we also tested whether the mRNA corresponding to CK2␣ and CK2␤ were induced by serum treat-ment of quiescent cells. Northern blot analysis of CK2␣ revealed three hybridizing transcripts of 1.6, 3.1, and 4.6 kilobases, respectively. The CK2␣ mRNA showed a less pronounced but detectable increase at 2 h after serum treatment. Similar to CK2␣Ј the CK2␣ mRNA increase was not blocked by the inhibition of protein synthesis. Northern blot analysis with the probe for CK2␤ revealed a single transcript of 1.2 kilobases that increased with the same kinetic of CK2␣Ј, and cycloheximide treatment did not influence its induction.
The mRNA analysis showed a moderate but measurable induction of the CK2 subunits. To confirm the functional significance of this induction, we determined whether the increased mRNA levels corresponded to an increase of CK2 by testing the kinase activity of the whole-cell extracts from qui-escent and serum-induced cells, using the phosphorylation assay with the CK2 peptide substrate. The time course of the serum-induced CK2 activity showed a maximal increase of about 1.8-fold at 4 h compared with quiescent cells (Fig. 3C). To confirm that the activity monitored with the peptide substrate is because of CK2, the effect of heparin, a specific CK2 inhibitor, was examined. The phosphorylation of the peptide was entirely suppressed by 1 g/ml heparin as expected for CK2catalyzed phosphorylation (not shown).
CK2␣Ј Cooperates with Ha-ras in Rat Embryo Fibroblast Transformation-The above results showed that mouse fibroblasts respond to mitogenic stimuli with an increase of CK2 transcripts and kinase activity at the G 0 /G 1 phase transition. Although the induction observed is not dramatic, we measured a reproducible increase of CK2 activity of about 80%. Previous experiments showed that as little as a 10% increase in CK2␣ expression in lymphoid organs of transgenic mice accelerated the onset of lymphomas induced by either c-myc or tal-1 oncogenes (18,19). To test whether CK2␣Ј or CK2␣ play a direct role in tumor induction, we performed standard focus formation assay transfecting primary rat embryo fibroblasts with Ha-ras, CK2␣Ј, or CK2␣ alone; the combination of Ha-ras with each catalytic subunit. Neither Ha-ras alone nor the CK2 catalytic subunits transfected independently induced foci formation in primary cells (Table I). However, transformed foci were visible within 10 days in the plates cotransfected with Ha-ras and either CK2␣Ј or CK2␣. We therefore conclude that either CK2␣Ј or CK2␣ cooperate with oncogenic ras in primary cell transformation.
The Expression Level of CK2␣Ј Correlates with Increased Growth Rate of Transformed Clones-To study further the effect of CK2␣Ј expression on the growth rate of ras-transformed cells, we compared the growth behavior of mouse fibroblasts transformed either with Ha-ras alone or with Ha-ras and CK2␣Ј. For this experiment we chose immortalized mouse fibroblasts, because these cells can be transformed with Ha-ras alone, and therefore it is possible to obtain transformed clones either expressing or not ectopic CK2␣Ј. BALB/c 3T3 were used for this set of experiments, because the efficiency of NIH 3T3 transformation with Ha-ras alone was too high, making it difficult to quantitate the effects of CK2␣Ј. As shown in Fig. 4A in BALB/c 3T3 cells the cotransfection of CK2␣Ј and Ha-ras resulted in approximately a 3-fold higher number of foci compared with the number of foci induced by Ha-ras alone. Moreover, we observed that foci generated with Ha-ras alone were smaller than those from clones transformed with Ha-ras and CK2␣Ј, suggesting that overexpression of CK2␣Ј contributed to the cell growth of transformed cells (Fig. 4B). In parallel sets of experiments we observed a similar effect following cotransfection of Ha-ras with the CK2␣ catalytic subunit. The transfection of CK2␣Ј, or CK2␣, alone did not induce foci formation (not shown).

FIG. 3. Expression of CK2 mRNAs and CK2 kinase activity from cell extracts of quiescent and serum-induced fibroblasts.
A, Northern blots using the probes CK2␣Ј, CK2␣, and CK2␤ as indicated. Confluent NIH 3T3 fibroblasts were serum-starved for 48 h prior to the addition of medium plus 10% FCS alone or containing 10 g/ml of cycloheximide as indicated (CHX). At various time points, total cellular RNA was extracted and Northern blot analysis was performed. Numbers indicate the hours of serum induction of quiescent fibroblasts. The relative positions of 18 and 28 S rRNA are shown. Glyceraldheyde-3-phosphate dehydrogenase (gapdh) was used as a control for RNA loading. B, quantitative analysis of CK2 mRNA levels in cultured fibroblasts. The blots were analyzed using a PhosphorImager (Molecular Dynamics), and the values obtained were normalized to the glyceraldheyde-3-phosphate dehydrogenase mRNA levels. C, catalytic activity of the whole-cell extracts were determined using a specific CK2 substrate peptide. Confluent NIH 3T3 fibroblasts were serum-starved for 48 h and then induced with medium containing 10% FCS at the indicated time points. Cells were lysed, and whole-cell extracts were used in the phosphorylation assay. The CK2 activity is reported as picomoles of 32   To analyze in more detail the growth rate of the transformed clones, four ras-transformed clones (R-1 to R-4) and four ras-CK2␣Ј-transformed clones (R␣Ј-1 to R␣Ј-4) were chosen for further analysis. Fig. 5A shows the growth curves obtained by counting cells over a period of 5 days. The clones obtained by cotransfection of Ha-ras and CK2␣Ј showed a marked increase in their growth rates compared with ras-transformed clones. Eight other clones obtained from different transfections were analyzed, and those cotransfected with CK2␣Ј and Ha-ras also exhibited increased growth rates (not shown).
To examine whether the growth differences correlated with the increased expression of CK2␣Ј in the transformed clones, we tested the kinase activity mediated by CK2␣Ј by measuring the phosphorylation of calmodulin. Calmodulin is an ideal substrate for examining the contributions of the CK2 catalytic subunits, because in reconstitution experiments with recombinant subunits its phosphorylation is suppressed completely by adding CK2␤ (5,(33)(34)(35). Clones, transformed with both Ha-ras and CK2␣Ј, showed higher levels of calmodulin phosphorylation when compared with ras-transformed clones (Fig. 5B). This growth-related phosphorylation of calmodulin was inhibited (Ͼ90%) by addition of either a molar excess (0.5 mM) of the specific peptide substrate, or 1 g/ml heparin (not shown). These findings, in conjunction with the alkalilability of the phosphate incorporated into calmodulin, which rules out the possibility of tyrosine phosphorylation, show that calmodulin phosphorylation is entirely because of CK2 rather than to any other protein kinase(s). Thus, although we observed clonal variability, the enhanced growth correlated with increased CK2␣Ј-dependent calmodulin phosphorylation (Fig. 5C). Finally, in mice we tested the growth of Ha-ras versus Ha-ras plus CK2␣Ј-transformed clones. Exponentially growing transformed clones were collected and injected subcutaneously into the scapular region of nude mice, and the resultant tumors were removed surgically after 2 weeks of growth and weighed. Consistent with the results from the cell growth in vitro, CK2␣Ј was found to produce a significant enhancement of tumor growth (Fig. 5D). DISCUSSION Here we report the cloning of the murine CK2␣Ј subunit and show that its mRNA and kinase activity are induced in response to serum stimulation of quiescent fibroblasts. Furthermore, we show that expression of CK2␣Ј under the control of a constitutive promoter cooperates with Ha-ras in transformation of rat primary fibroblasts and increases cell growth of transformed cells both in vitro and in vivo.
Our study originated from a screening for serum-induced messages in quiescent mouse fibroblasts, which allowed us to identify the CK2␣Ј as an induced gene. We therefore cloned the murine full-length CK2␣Ј cDNA, analyzed its expression pattern in vivo, in vitro, and activity in cultured cells. Northern blot analysis showed a CK2␣Ј peak of induction at 4 h after serum treatment and that this induction does not require new protein synthesis. Therefore CK2␣Ј, like c-myc and MCP-1 (Refs. 36 -38 and references therein), belongs to a subset of early genes induced with a slow kinetic. The analysis of induction revealed a lower but still measurable increase of both CK2␣ and CK2␤ at the same time points, suggesting that newly assembled CK2 tetramers can be formed following serum induction. In accordance with this prediction we observed an increased CK2-dependent phosphorylation activity in protein extracts from serum-induced cells. An active role of CK2 in cell cycle progression has already been suggested both in mammalian cells and yeast (14,15,39). Our data show, for the first time, that CK2 activity is indeed increased at the boundary between G 0 and G 1 , suggesting that new CK2 synthesis is required at this stage of the cell cycle. Its specific induction could be necessary for several reasons. It is possible that the kinase already present in the cells is in a form which is not able to phosphorylate some critical substrates. Alternatively, because we observed a higher CK2␣Ј induction compared with the other catalytic subunit, a higher proportion of ␣Ј 2 ␤ 2 or ␣Ј␣␤ 2 tetramers could be formed. The formation of these two types of tetramer may lead to different substrate specificity. Despite the fact that the catalytic properties of isolated recombinant CK2␣ and CK2␣Ј are very similar (28,34), which is consistent with their high sequence homology in their catalytic domains, significant structural differences suggest divergent functional commitments. Thus, the C-terminal segment of vertebrate CK2␣, which lies outside the catalytic core and includes several phosphorylation sites affected both in vitro and in vivo by cyclin-dependent kinases, is absent in CK2␣Ј (17). Likewise a motif (HEHRKL) responsible for association of CK2␣ with protein phosphatase 2A (23) is substantially altered in CK2␣Ј (HQQKKL). This difference is especially remarkable as it occurs inside a region that is otherwise highly conserved between CK2␣ and CK2␣Ј. Suggestive of specific function(s) of CK2␣Ј in higher organisms is the observation that CK2␣Ј could not be detected in Drosophila, Xenopus, and Schizosaccharomyces pombe. In S. cerevisiae, however, a somewhat atypical CK2␣Ј subunit is found that exhibits functional differences from CK2␣ (39,40).
CK2␣Ј cooperates with oncogenic ras in the transformation of primary fibroblasts. Our data show that neither CK2 catalytic subunits nor Ha-ras alone induce foci formation when transfected in primary cells, whereas transformed foci become evident upon cotransfection with Ha-ras and either CK2␣Ј or CK2␣. Therefore, we can conclude that although the structural differences between CK2␣ and CK2␣Ј may reflect distinct functional roles, at least with respect to the cooperation with ras in cell transformation, these differences are not critical.
Recently, it was observed that exogenous expression of CK2␣ suppressed cell growth and inhibited foci formation induced by activated ras (23). Our results diverge from this observation. The reasons for such a discrepancy are presently unclear. It is possible that the inhibitory effect of CK2␣ previously observed (23) was dependent on the genetic background of the NIH 3T3 cells used in those experiments. Alternatively, the reduction of foci observed by Hériché and collaborators (23) may have resulted from a CK2␣ poisoning effect because of a too high expression of CK2␣. In a standard focus formation assay we observed cooperation between CK2 catalytic subunits and oncogenic ras both in primary and immortalized fibroblasts, suggesting we are observing a general phenomenon. Moreover, our results are in agreement with experiments in transgenic mice where the constitutive expression of CK2␣ accelerated the formation of lymphomas induced by c-myc or tal-1 (18,19). Comparison of the growth curves of fibroblasts transformed with Ha-ras alone versus fibroblasts transformed with Ha-ras and CK2␣Ј demonstrated that transformed clones, expressing constitutively CK2␣Ј, grew faster. The enhanced growth of Ha-ras and CK2␣Ј-transformed clones correlates with increased catalytic activity when monitored using calmodulin as a phosphoacceptor substrate, symptomatic of the presence of free catalytic subunit (33,34). Thus, these data suggest that unbalanced expression of CK2␣Ј or CK2␣ leads to phosphorylation of some critical target(s) necessary to accelerate the cell cycle progression of ras-transformed cells. Therefore, it seems likely that the transforming potential of CK2 in each experimental model is because of a fraction of catalytic subunits not combined with CK2␤ to form the canonical holoenzyme. This hypothesis is supported by the phosphorylation of calmodulin, because this protein is unaffected by the CK2 holoenzyme.
In contrast the hypothesis that following serum treatment CK2␣ and CK2␣Ј not combined with CK2␤ could be transiently present in untransformed dividing cells is not consistent with available experimental data. Indeed, we observed that after serum treatment of fibroblast, CK2␤ was induced with the same kinetics of the catalytic subunits. In addition, by using the specific substrate calmodulin or by titrating in with recombinant CK2␤, we did not detect free catalytic subunits in cell extracts (not shown). Therefore, it is likely that during the G 0 /G 1 progression of the cell cycle the newly synthesized CK2␣Ј and CK2␣ are rapidly assembled into tetrameric CK2. Thus, the assembly of newly synthesized CK2 subunits into a tetrameric enzyme could represent a mechanism for the modulation of the too reactive free catalytic subunits necessary to reprogram the CK2 kinase activity during the progression of the cell cycle. Future studies aimed at the identification of specific target(s) of the CK2 catalytic subunits may unveil some target(s) critical for cell transformation. A, growth curves of isolated foci. Cells were plated at a density of 10 5 cells/6-cm culture dish in medium containing 2% FCS and counted at daily intervals. The mean values of duplicate cultures are shown plotted against time. B, calmodulin phosphorylation by isolated foci. Confluent cells were lysed and total cellular extracts were used in phosphorylation assay using calmodulin as specific CK2 catalytic subunit substrate. The reaction products were run on SDS-polyacrylamide gel electrophoresis, and the blots were autoradiographed or directly scanned. All data are the mean of at least three separate determinations with a S.E. of less than 12%. C, correlation between calmodulin phosphorylation and growth curve rates. The cell number corresponds to the number of cells described in A at day 5 of growth. The values of calmodulin phosphorylation are the same of the experiment described in B. Clones transformed with Ha-ras are represented with open circles, whereas clones transformed with Ha-ras plus CK2␣Ј are represented with black squares. D, tumor formation in nude mice. Mice were injected with 10 6 log-phase cells from the ras-transformed cells (R-1 to R-4) and from the Ha-ras with CK2␣Ј-transformed cells (R␣Ј-1 to R␣Ј-4). Tumors were harvested after 2 weeks of growth and weighed. The error bars represent the S.D. for three mice used for each clone.