Ornithine decarboxylase expression leads to translocation and activation of protein kinase CK2 in vivo.

Ornithine decarboxylase (ODC) is the key initial enzyme in the biosynthesis of polyamines. Since polyamines have been shown to enhance protein kinase CK2 activity in vitro, ODC was overexpressed to examine the role of polyamines in CK2 regulation in vivo. Infection of Balb/MK cells with an ODC retrovirus to elevate ODC and polyamine levels increased overall protein phosphorylation as well as CK2 protein levels and enzyme activity in mimosine- or nocodazole- arrested cells. Immunofluorescence microscopy and enzyme analyses of subcellular fractions from ODC-overexpressing cells demonstrated translocation of CK2 from the cytoplasm to the nucleus with no apparent loss of cytoplasmic CK2 activity, suggesting polyamine activation of the remaining cytoplasmic enzyme. Similarly, K6/ODC transgenic mice exhibited higher ODC and CK2 enzyme activities than their normal littermates. ODC-immunostained cells in the transgenic skin also stained intensely for CK2 protein. Primary cultures of K6/ODC keratinocytes also exhibited increased ODC and CK2 enzyme activities compared with those from normal littermates. However, the addition of difluoromethylornithine, a specific ODC inhibitor, to the transgenic keratinocytes reduced both intracellular polyamine levels and CK2 enzyme activity. These results suggest that polyamines regulate the CK2 enzyme by affecting its cellular distribution as well as its enzyme activity and levels.

Polyamines are cellular cations essential for growth and differentiation (1,2). Their depletion results in deleterious biological effects, including growth inhibition and alteration of differentiation (3,4). While a great deal of evidence indicates that polyamines are involved in the regulation of proliferative events, their precise role(s) in biological processes remain poorly characterized.
Ornithine decarboxylase (ODC) 1 is the first and regulatory enzyme in the biosynthesis of the polyamines putrescine, sper-midine, and spermine. ODC expression in normal tissue is extremely low, yet highly inducible. In tumors, however, constitutively high levels of ODC enzyme activity, protein, and mRNA are observed (5)(6)(7)(8)(9). Whereas overexpression of ODC transforms NIH3T3 cells (10 -12), ODC overexpression does not transform normal diploid keratinocytes and fibroblasts (13). However, ODC overexpression cooperates with other genetic lesions such as activated c-Ha-ras to enhance tumor development in keratinocyte cell lines (13).
Reversible protein phosphorylation is one of the major mechanisms by which cells control metabolic and regulatory activities, especially in response to extracellular signals. Furthermore, it has become apparent that tumor progression involves both genetic and epigenetic disruptions in these pathways (14,15). A protein kinase that has been reported to be activated by polyamines in vitro is the highly conserved serine/threonine protein kinase CK2 (16 -18). A functional relationship between polyamines and CK2 may be inferred since polyamine biosynthesis and CK2 activity are both induced concurrently with stimulation of cell growth and proliferation (19,20).
CK2 is a serine/threonine protein kinase found in all mammalian tissues, both in the nucleus and cytoplasm (21,22). The enzyme exists as a heterotetramer composed of two and sometimes three subunits, i.e. as ␣ and ␣Ј subunits of 38 -44 kDa and a ␤ subunit of 24 -28 kDa, which associate to form the native ␣ 2 ␤ 2 , ␣Ј␣␤ 2 , or ␣Ј 2 ␤ 2 structures. The ␣ and ␣Ј subunits bear the catalytic site of the enzyme, while the ␤ subunit is thought to be regulatory since it confers optimal activity to the holoenzyme and influences substrate specificity. The identification of a large number of proteins that can be phosphorylated by CK2 supports the suggestion of a role for CK2 in signal transduction (23). Many of these proteins are involved in replication and transcription, and the phosphorylation of several of these has been demonstrated to cause a significant change in their biochemical activity (23). Moreover, transgenic mice overexpressing CK2␣ in lymphocytes exhibit increased susceptibility to lymphomas, and coexpression with a c-myc transgene results in neonatal leukemia (24). Thus, the increase in CK2 activity in transformed and proliferating tissues (25) and the apparent oncogenic activity of CK2␣ suggest the involvement of CK2 in both normal and unregulated cell proliferation.
The regulation of CK2 is still poorly understood. It does not appear to be regulated by any previously described second messengers. Increased kinase activity has been observed following stimulation by growth factors in selected cell types (26,27). Additional studies have implicated growth factors in the translocation of CK2 to the nucleus (28,29), and more recently, fibroblast growth factor-2 has been reported to directly interact with CK2 and to stimulate its activity (30). CK2 involvement in mitogenic signaling is supported by immunofluorescence microscopy studies that have demonstrated its translocation into the nucleus following mitogenic stimulation (31). Microinjection of antisense oligodeoxynucleotides and antibodies raised * This work was supported by National Institutes of Health Grant CA 68762 (to L. J. S.) and Grants CA 55066 and CA 70739 (to S. K. G.). 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 1 The abbreviations used are: ODC, ornithine decarboxylase; CDAP, N-cyclohexyl-1,3-propanediamine; EMEM, Eagle's minimum essential medium; EGF, epidermal growth factor; PBS, phosphate-buffered saline; HPLC, high performance liquid chromatography; dansyl, 5-dimethylaminonaphthalene-1-sulfonyl; PAGE, polyacrylamide gel electrophoresis; DFMO, difluoromethylornithine; ACTH, adrenocorticotropic hormone. against CK2␤ into the cytoplasm of cells at the time of mitogenic stimulation resulted in cell cycle arrest at the G 0 /G 1 and G 1 /S transition phases, further demonstrating the requirement for CK2␤ in proliferation (32,33).
In vitro, CK2 enzyme activity has been demonstrated to be stimulated by spermidine and, more notably, spermine (17). Recent evidence strongly suggests that these polyamines increase CK2 activity through allosteric regulation. For example, it has been reported that purified CK2 adopts a ring-like structure when spermine is added (34); further studies revealed this structure to be the most active polymeric conformation of CK2. Furthermore, a spermine-binding domain has been identified in the N-terminal region of the regulatory ␤ subunit of CK2 (35).
Previous reports demonstrated in separate studies that increased levels of polyamines (5-9) and CK2 (25,36,37) occur in solid tumors and in normal cells exhibiting high mitotic activity, although no connection between ODC and CK2 was inferred. While intriguing, studies to date have not addressed whether polyamines affect CK2 activity in vivo. We have taken two approaches to determine if increased intracellular levels of polyamines contribute to tumor development through alteration of CK2 activity. One method utilizes a replication-defective retroviral vector to overexpress murine ODC in mouse keratinocytes (13). The second approach involves the K6/ODC transgenic mouse, in which ODC expression is targeted to keratinocytes in the outer root sheath of the hair follicle (38). These tools have enabled us to increase the intracellular levels of polyamines and thus characterize any differences in CK2 activity and levels. The results of these studies suggest that polyamines regulate CK2 in vivo through enzyme activation as well as by nuclear translocation.
Retrovirus Infection-Balb/MK cells were grown to 50% confluence and infected with the control pLXSN virus or the ODC pLOSN virus (13) for 6 h with 4 g/ml Polybrene. The cells were then washed with PBS and refed with EMEM containing 8% Chelex-treated fetal bovine serum and 5 ng/ml EGF. Selection with G418 (120 g/ml) was begun 2 days after infection, and selected cells were maintained in 60 g/ml G418 thereafter. Cells were washed with PBS and presynchronized at G 0 by refeeding with EMEM containing 2% dialyzed serum and no EGF. Seventy-two hours later, cells were refed with EMEM containing 8% Chelex-treated serum, 5 ng/ml EGF, and either 0.2 mM mimosine (39) or 260 nM nocodazole. The mimosine-and nocodazole-treated cells were harvested for CK2 assays 20 h later. Flow cytometry analysis confirmed that the population of cells used in these experiments was predominantly in G 1 for mimosine-arrested cells and at the G 2 /M border for nocodazole-arrested cells.
For metabolic labeling with [ 32 P]orthophosphate, cells were equilibrated in low calcium and phosphate-free EMEM containing either 0.2 mM mimosine or 260 nM nocodazole for 30 min. Cells were then incubated in the same medium containing 0.02 mCi/ml [ 32 P]orthophosphate for 3 h prior to harvest.

Immunofluorescence
Virally infected cells were grown on chamber slides (Lab-Tek, Nalge Nunc) and fixed in 3% paraformaldehyde in PBS. The cells were then permeabilized with 0.5% Triton X-100 in 3% paraformaldehyde and PBS, blocked with normal goat serum, and reacted with polyclonal anti-CK2 antibody (1:1000 dilution) overnight. Bound antibodies were detected with a rhodamine-conjugated goat anti-rabbit secondary antibody.

Extract Preparation for CK2 Assays
For total soluble cell lysates, cells were collected by low speed centrifugation in PBS and lysed in twice the packed cell volume of a buffer containing 10 mM HEPES, pH 7.5, 1.5 mM MgCl 2 , 10 mM KCl, 0.5 mM dithiothreitol, protease inhibitors (1 g/ml each aprotinin, leupeptin, and pepstatin and 0.2 mM phenylmethylsulfonyl fluoride), and the phosphatase inhibitor NaF at 1 mM. The cells were mechanically lysed by 30 strokes of a tight-fitting Dounce glass homogenizer, and the supernatants were collected following centrifugation at 12,000 ϫ g for 15 min. The enzyme activities from these lysates are designated as total soluble CK2 activity. ODC activity and HPLC analyses of polyamine content were performed as described previously (13,38). Tissue extracts were prepared from skins of normal mice and their transgenic littermates by plunging excised skins into 55°C water for 20 s to allow for separation of the epidermis and dermis, which were then homogenized in 25 mM Tris-HCl, pH 7.5, 2.5 mM dithiothreitol, 0.1 mM EDTA, and 0.2 mM phenylmethylsulfonyl fluoride. Polyamine levels were determined by HPLC analysis of the dansylated products after overnight extraction in 0.2 N perchloric acid.
Extract preparation for subcellular fractionation was done essentially as described (31), except that 1 mM NaF was used in place of sodium orthovanadate. In addition, the nuclear and cytosolic pelleted fractions were homogenized twice and resuspended in their respective buffers and are designated as nuclear pellet and microsomal fractions, respectively. Completeness of cell breakage and purity of isolated nuclei were verified by light-phase microscopy.

CK2 Assay
Kinase reactions were carried out in a total volume of 30 l containing 50 mM HEPES, pH 7.5, 150 mM NaCl, 11 mM MgCl 2 , 1 mM peptide substrate RRRDDDSDDD (Research Genetics; purity Ͼ 96%) (40), and 1 mM [␥-32 P]ATP (specific activity ϭ 2500 cpm/pmol). The addition of 60 mM ␤-glycerophosphate to inhibit phosphatase activity (41) increased phosphate incorporation, but did not change the relative amounts of enzyme activity. The reaction was initiated by the addition of 0.5-1 g of protein, incubated at 34°C for 10 min, and stopped by spotting 15 l of the assay mixture onto Whatman P-81 cation-exchange paper discs. Standard kinetic analyses were performed to ensure that phosphate incorporation was linear under these conditions. The discs were washed five times with 75 mM phosphoric acid and twice with 95% ethanol and air-dried. The incorporated phosphate was quantitated using a Packard 1900 TR liquid scintillation counter. The radioisotope incorporation was corrected for phosphorylation of endogenous proteins by assaying reaction mixtures in the absence of peptide substrate. Assays were carried out in triplicate.

Extract Preparation for 32 P-Labeled Total Cell Lysates
Cells in 100-mm dishes were washed with PBS three times and lysed in 1 ml of modified radioimmune precipitation assay buffer (50 mM Tris-HCl, pH 7.5, 1% Nonidet P-40, 0.25% sodium deoxycholate, 0.25% SDS, 150 mM NaCl, 1 mM EGTA, and 1 mM NaF) containing the protease inhibitors at the concentrations listed above. The dishes were then rocked at 4°C for 10 min, followed by passage of the cell lysate through a 21-gauge needle several times to shear the DNA. The cell lysate was centrifuged at 12,000 ϫ g for 15 min. Equal amounts of cellular protein from the supernatants were analyzed by 12% SDS-PAGE and transferred to nitrocellulose, and the radiolabeled proteins were detected using a Molecular Dynamics PhosphorImager. Blots were stained briefly in Ponceau S (Sigma) to verify loading equality, and Western blotting was performed using monoclonal antibody 1AD9 (2 g/ml) to detect CK2␣, monoclonal antibody 6D5 (0.1 g/ml) to detect CK2␤, or polyclonal anti-ODC antibody (1:20,000 dilution). The CK2 monoclonal antibodies were kindly provided by Dr. Olaf-Georg Issinger (Odense Universitet, Odense, Denmark), and the polyclonal ODC anti-

FIG. 2. PhosphorImager analysis of control-and ODC-infected
Balb/MK cells. Cells were infected with either the control pLXSN virus or the ODC retrovirus pLOSN and then were arrested at G 1 /S and G 2 /M with mimosine and nocodazole, respectively. The cells were metabolically labeled with [ 32 P]orthophosphate for 3 h and then harvested. Equal amounts of total protein were separated by SDS-PAGE and transferred to nitrocellulose, and the radiolabeled proteins were detected using a PhosphorImager.

Histology and Immunocytochemistry
Tissues were fixed in Fekete's solution (60% ethanol, 3.2% formaldehyde, and 0.75 M acetic acid) overnight and embedded in paraffin. For immunolocalization of ODC and CK2, skin sections were incubated with a 1:500 dilution of polyclonal anti-ODC antibody or a 1:500 dilution of polyclonal CK2 antisera (kindly provided by Dr. Michael Dahmus, University of California), and specific staining was detected using the instructions supplied in the Elite ABC Vectastain kit (Vector Laboratories, Inc., Burlingame, CA).

Primary Keratinocyte Cultures
Primary cultures of epidermal cells were isolated from 2-3-day-old K6/ODC transgenic and normal littermate mice by a trypsin flotation procedure (42,43). Prior to keratinocyte preparation, newborn mice heterozygous for the K6/ODC transgene were distinguished from their normal littermates by polymerase chain reaction genotyping of tail DNA using the primers previously described (38). Isolated epidermal cells were cultured in low calcium EMEM supplemented with 0.05 mM calcium, 8% Chelex-treated fetal bovine serum, and 5 ng/ml EGF either in the presence or absence of the ODC inhibitor ␣-difluoromethylornithine (0.2 mM). Cells were harvested 3 days after the last refeeding.

RESULTS
Effect of CDAP on Total Soluble CK2 Activity-It has been previously reported that spermidine and spermine can increase CK2 activity in vitro (17). To confirm that the activation of CK2 by polyamines does occur in vivo, Balb/MK epidermal cells were incubated in the presence of the spermine synthase inhibitor CDAP to increase the intracellular spermidine content (44). CK2 enzyme assays were performed after the polyamines had returned to basal levels, when the polyamine levels were most different between the treated and control cells. Fig. 1 demonstrates the inverse correlation between CDAP concentration and intracellular spermine levels, with a concomitant increase in spermidine levels, thus confirming what has been found in other cell types (44). Moreover, CK2 activity parallels spermidine concentration. A 3-fold increase in intracellular spermidine resulted in a 2.5-fold increase in total soluble CK2 activity.
Effect of ODC Overexpression on in Vivo Protein Phosphorylation and CK2 Protein Levels and Activity-We utilized a replication-defective retroviral vector capable of overexpressing a truncated isoform of ODC in epidermal cells (13). Whereas full-length ODC protein typically has a half-life of 15-20 min within the cell, the truncated form is considerably more stable while still retaining full enzyme activity (45). Thus, high intracellular levels of ODC and polyamines can be achieved without the use of chemical inhibitors, thereby enabling examination of polyamine-mediated CK2 activation as observed in vitro, but using a more physiologically relevant system. Since ODC induction is associated with cellular proliferation, the effects of ODC overexpression were studied in cells blocked at the G 1 /S and G 2 /M borders with mimosine and nocodazole, respectively. This was done to prevent any enhanced growth potential acquired by the ODC-infected cells, thus assuring that the cell populations to be compared were identical with respect to cell cycle stage. After synchronizing the cells with either mimosine or nocodazole, they were labeled with [ 32 P]orthophosphate for an additional 3 h while still in the presence of the inhibitors. Proteins were extracted and separated by SDS-PAGE, transferred to nitrocellulose, and analyzed by a PhosphorImager (Fig. 2). In general, ODC-overexpressing cells demonstrated increased phosphorylation of many proteins; enhanced phosphorylation was especially evident for proteins with approximate molecular masses of 40, 50, and 70 kDa for the mimosinearrested cells and 16, 33, and 50 kDa for the nocodazolearrested cells. Probing with an anti-phosphotyrosine antibody failed to show any differences in phosphotyrosine content between the control-and ODC-infected mimosine-arrested cells (data not shown). Therefore, the increase in phosphorylation appeared to be due to predominantly serine/threonine phosphorylations. The immunoblot was then probed with anti-ODC, anti-CK2␣, and anti-CK2␤ antibodies (Fig. 3). Fig. 3 (A and B) shows that CK2 ␣ and ␤ subunit proteins are increased in the ODC-infected cells. Quantitation by densitometric scanning revealed a 4-fold increase in both subunits in the mimosinearrested cells and a 2-fold difference in the nocodazole-arrested cells. Only the ODC-infected cells expressed the truncated ODC protein, while the expression of endogenous ODC was below the level of detection in both control-and ODC-infected cells (Fig. 3C).
The effect of ODC overexpression on polyamine levels and total soluble CK2 activity in the mimosine-arrested cells is shown in Table I. Putrescine levels in the mimosine-arrested cells increased Ͼ150-fold, while spermidine and spermine levels increased only 2.2-and 0.75-fold, respectively. Similar values were obtained for the nocodazole-arrested cells. However, despite a 4-fold increase in immunodetectable CK2␣ and CK2␤ from total cell lysate of the mimosine-arrested cells and a 2-fold increase in the nocodazole-arrested cells, only a 1.5-fold increase in total soluble CK2 enzyme activity was observed for both.
CK2 undergoes translocation after mitogenic stimulus and has been demonstrated to associate with the nuclear matrix (46). Therefore, to determine the reason for the observed discrepancy between the total cellular CK2 protein levels and total soluble enzyme activity, the localization of CK2 in mimosine-arrested cells was examined by immunofluorescence (Fig. 4). CK2 has been reported to be localized in both the nucleus and cytoplasm (21). As expected, CK2 protein was present in the nucleus and cytoplasm both in control-infected (Fig. 4a) and in ODC-infected (Fig. 4b) cells. However, there was greater nuclear staining intensity in the ODC-infected cells. Furthermore, the ODC-infected cells also exhibited quite intense staining in structures lying just outside the nuclei. The identity of these structures has not yet been determined.
To further determine whether CK2 nuclear translocation is occurring in ODC-overexpressing cells, various subcellular fractions of the control-and ODC-infected cells were compared for CK2 enzyme activity (Table II). Enzyme activity in the microsomes and cytosol appeared to be unaffected by the increase in intracellular polyamines, whereas CK2 activity decreased 8-fold in the nuclear supernatant fraction and increased 4-fold in the nuclear pellet in response to elevated polyamine concentration. Subcellular fractions of control-and ODC-infected cells were also investigated for CK2␤ protein by immunoblot analysis. The known high affinity of the ␣ and ␤ subunits for each other, demonstrated by the fact that they are always purified as the holoenzyme and cannot be separated by harsh biochemical conditions, allows us to infer that translocation of the holoenzyme can be followed by analysis of the CK2 ␤ subunit alone (31,32). Taken together, the results of the CK2 activity assays and the immunoblot analysis of the cytosolic and nuclear pellet fractions (Fig. 5)   nucleus and activation of the remaining CK2 enzyme in the cytosol. Immunoblot analysis demonstrated 5-fold more CK2 protein in the nuclear pellet and a concomitant 3.5-fold reduction of CK2 protein in the cytosol of ODC-overexpressing cells compared with control-infected cells. Immunoblot analysis of the microsomal fraction detected CK2 protein in the controlinfected cells, but it was below the level of detection in the ODC-infected cells (data not shown). Surprisingly, this apparent translocation of CK2 protein from the cytoplasm to the nucleus, promoted by elevated intracellular polyamine concentration, did not result in the decrease of cytosolic or microsomal CK2 activity. Whereas CK2 protein in the cytosol decreased 3.5-fold, the increased intracellular polyamine levels appeared to activate the enzyme, resulting in the same absolute CK2 activity as existed before translocation. However, the stimulation promoted by increased polyamine levels does not appear to occur in the nucleus since the 4-fold increase in CK2 activity observed in the nuclear pellet is associated with a 5-fold increase in CK2 protein. This suggests that, upon translocation to the nucleus, CK2 either is no longer able to be activated by polyamines, perhaps due to competition with other macromolecules (47) such as DNA or p21 (48), or is already fully activated by the existing basal level of polyamines. The addition of spermine to the enzyme assay to a final concentration of 1 mM did not significantly increase CK2 activity (data not shown).
Effects of ODC Overexpression on CK2 in the Dermis of the K6/ODC Transgenic Mouse-In K6/ODC transgenic mice, a keratin-6 promoter is used to target high levels of ODC activity to the outer root sheath of the hair follicle (38). These ODC transgenic mice lose their hair at 2 weeks of age and simultaneously acquire large follicular cysts in the dermis of their skin. The ODC activity in the dermis of K6/ODC mice is elevated at least several 100-fold over that of their normal littermates, as shown in Table III. Similarly, CK2 activity is increased in K6/ODC transgenic mouse skin as well (Table III). Immunohistochemistry of skin from these K6/ODC transgenic mice reveals that ODC and CK2 colocalize in the cells lining the follicular cysts found in the transgenic dermis (Fig. 6, C and D). The observation of CK2 expression in the hair follicles of normal mouse skin (Fig. 6B) raises the question of whether the ODC/CK2 colocalization is polyamine-mediated or merely coincidental, and this is under active investigation. CK2 activity and protein were also detected in the epidermis of both normal and transgenic mice (Table III and Fig. 6, A and B), as has been previously reported (49). The increased CK2 activity of the transgenic epidermis over that of the normal littermates may be due to enzyme activation caused by diffusion of polyamines from the dermal follicular cysts.
To further establish an in vivo connection between high levels of ODC expression and CK2 activity, keratinocytes derived from transgenic mice and their normal littermates were assayed for CK2 activity following culturing in the presence and absence of ␣-difluoromethylornithine (DFMO), a specific inhibitor of ODC. As Table IV demonstrates, the primary keratinocytes of the transgenic mice displayed high ODC activity compared with keratinocytes harvested from normal littermates. Putrescine and, to a lesser extent, spermidine levels were also increased. CK2 activity in the transgenic keratinocytes was increased almost 4-fold. Furthermore, transgenic keratinocytes cultured with DFMO not only had reduced ODC activity and polyamine levels, but also total soluble CK2 activity was reduced to levels that resembled those in keratinocytes from normal littermates. The addition of DFMO to CK2 enzyme assays had no effect (data not shown), thus discounting any direct effect of DFMO upon CK2. DISCUSSION In this report, we demonstrate that, in intact cells, increased ODC expression and polyamine levels act not only to increase CK2 activity and protein levels, but also to redistribute CK2  e Relative levels of CK2␤ were determined by densitometric analysis of the immunoblot in Fig. 5.
f ND, not determined. g CK2␤ was below the level of detection in this fraction of the ODCinfected cells. protein within the cell. Furthermore, we present the first in vivo evidence of polyamine-mediated CK2 activation using transgenic mice that overexpress ODC. This polyamine-mediated increase in CK2 activity may be a contributor to the carcinogenic process by enhancing a tumor cell's growth potential. Many CK2 substrates are nuclear proteins involved in cell cycle regulation, DNA replication, and transcription. While the phosphorylation of c-Jun or c-Myb by CK2 reduces their binding activity to DNA, the phosphorylation of the large T antigen of SV40 causes an increase in its nuclear translocation, where it has been reported to function to sequester retinoblastoma protein and p53 (23). Furthermore, both topoisomerase II and nucleolin, recently identified as DNA helicase IV, are activated following CK2 phosphorylation (23, 50 -53).
Our studies focus on the in vivo effect of spermidine on CK2, in contrast to previous in vitro experiments utilizing spermine. Spermidine is probably the more physiologically relevant effector since its intracellular levels are increased to a greater extent in cycling cells as well as in epidermal papillomas (9). In fact, ODC overexpression leads to an increase in the putrescine and spermidine levels, with no change in the spermine levels. Moreover, our results with CDAP indicate spermidine as an in vivo activator of CK2 activity. The experiments with CDAP not only confirm that increases in CK2 activity occur with in-creases in spermidine levels, but also suggest that CK2 activation is, at least in part, mediated directly by polyamines, independent of a physical interaction between CK2 and ODC.
Reports demonstrating a spermine-binding site as well as a spermine-induced change in CK2 quaternary structure (34,35) are especially interesting in light of the recent evidence for mitogen-stimulated CK2 translocation from the cytoplasm to the nucleus and nuclear matrix (29,46). An increase in ODC activity is one of the first changes to take place when synchronized quiescent cells are induced to enter the G 1 phase of the cell cycle. While it may take up to 14 h before intracellular spermidine levels increase due to ODC activation (5,19), possible polyamine involvement in mitogen-stimulated CK2 translocation is suggested by recent studies in which bovine adrenocortical cells treated with ACTH exhibited rapid polyamine uptake and CK2 nuclear translocation in similar time spans (19). This was followed by a delayed secondary translocation of CK2 to the nucleus at 15 h, when spermidine levels peaked in response to mitogenic stimulation. However, this ACTH-induced CK2 nuclear translocation was inhibited by DFMO (19). Furthermore, CK2 uptake into purified nuclei was shown to be significantly increased upon addition of spermine (28). It is possible that polyamines mediate translocation of CK2 from the cytoplasm to the nucleus through a structural change in which a cryptic nuclear localization signal is unveiled.
However, equally plausible is a model in which structural changes induced by polyamines allow for a separate event to promote translocation. The observation of a nuclear translocation of CK2 upon ODC overexpression is consistent with both hypotheses, and we are currently investigating whether this polyamine-induced translocation is a direct or indirect event. In addition, a putrescine-mediated activation/translocation cannot be absolutely ruled out since the induction of putrescine is substantially higher than that of spermidine in ODC-overexpressing cells. Thus, putrescine is making a much greater contribution to the total effective concentration of positive charges in cells that overexpress ODC.
Polyamines may also affect CK2 stability through alteration of its quaternary structure. The regulatory subunit of CK2 contains a destruction box sequence similar to that found in cyclins (54). This raises the question as to whether the regulatory subunit functions in a similar manner to the cyclins, whose binding to cyclin-dependent kinases is controlled by their syn-  thesis and degradation at specific times in the cell cycle. At the primary sequence level, this destruction box resides in close proximity to the polyamine-binding site. Therefore, a polyamine-induced structural change might mask the destruction box, preventing ubiquitination and delaying degradation, ultimately leading to elevated CK2 levels. This could account, at least in part, for the increased total CK2 protein in ODCinfected cells as well as the colocalization of ODC and CK2 around the follicular cysts of K6/ODC transgenic mice. Furthermore, it may explain the reports of increased CK2 protein and activity in tumors (25,36,37).
Recent reports have demonstrated a differential response of cells to polyamines introduced by extracellular means as opposed to those produced by ODC within the cell. For instance, the transcriptional activity of promoter-reporter gene constructs was greater in cells overexpressing ODC than in cells with similarly elevated intracellular polyamine levels following uptake of polyamines added to the medium (55). In addition, overexpression of ODC in NIH3T3 cells resulted in tumorigenic transformation, while the addition of exogenous polyamines did not lead to transformation despite similarly high intracellular polyamine levels (56). This suggests that there may be polyamine compartmentalization inside the cells, meaning that polyamine pools that are synthesized within the cell are not the same or may not be interchangeable with those derived from extracellular sources (57,58). A localized high concentration of putrescine may then be responsible for CK2 regulation, as well as individual pools of spermidine or spermine that would accumulate as a result of ODC overexpression. At this time, there is no satisfactory method to definitively determine the existence, or localization, of individual polyamine pools. Therefore, the in vivo relevance of previous data demonstrating spermine as the better effector in vitro, as well as our data suggesting spermidine as the physiological effector, remains to be determined.
The data presented here demonstrate that polyamines are involved in regulating CK2 subcellular distribution and enzyme activity in vivo. Such regulation may place CK2 as a downstream effector for ODC action in normal as well as tumor cells. Fluctuations in polyamine levels as well as compartmentalization throughout the cell cycle may, among other roles, be involved in targeting CK2 activity to distinct areas within the cell. Sustained elevated polyamine levels throughout the cell cycle, as observed in tumor cells, would certainly be altering normal signal transduction pathways. If CK2 kinase activity acts constitutively to provide a basal level of phosphorylation of transcription factors and other proteins, then increased ODC expression and polyamine levels might serve to enhance the susceptibility of normal cells to an event promoting uncontrolled growth (22). Our models for elevating ODC and polyamine levels in various cell types provide the necessary tools for further elucidation of the role polyamines play in mediating CK2 signaling and tumor progression.