Enhanced Casein Kinase I1 Activity in COS-1 Cells upon Overexpression of Either Its Catalytic or Noncatalytic Subunit*

Casein kinase I1 consists of catalytic (a) and regula- tory (@) subunits complexed into a heterotetrameric CY& structure. Full-length cDNAs encoding the a and @ subunits of human casein kinase I1 were subcloned into an expression vector containing the cytomegalovirus promotor, yielding the expression constructs pCMV-a and pCMV-@. Northern analyses of total cellular RNA prepared from COS-l fibroblasts 65 h after transfection with pCMV-a or pCMV-@ or with both expression constructs showed marked specific in- creases in corresponding a and @ subunit RNAs. Immunoblot analysis utilizing anti-casein kinase I1 anti- serum of cytosolic extracts prepared from COS-1 cells co-transfected with pCMV-a and pCMV-@ showed 2- and 4-fold increases in immunoreactive a and @ subunit protein, respectively, relative to vector-transfected cells. These same cytosolic fractions exhibited an average 5-fold increase in casein kinase I1 catalytic ac- tivity. COS-1 cells transfected with pCMV-a alone exhibited a %fold increase in immunoreactive a subunit protein and a nearly 2-fold increase in cytosolic casein kinase I1 catalytic activity. Transfection with the cDNA coding for the noncatalytic @ subunit alone also caused a near doubling of cytosolic casein kinase I1 catalytic activity. No increase in immunoreactive a subunit protein was observed in pCMV-@-transfected

Casein kinase I1 consists of catalytic ( a ) and regulatory (@) subunits complexed into a heterotetrameric CY& structure. Full-length cDNAs encoding the a and @ subunits of human casein kinase I1 were subcloned into an expression vector containing the cytomegalovirus promotor, yielding the expression constructs pCMV-a and pCMV-@. Northern analyses of total cellular RNA prepared from COS-l fibroblasts 65 h after transfection with pCMV-a or pCMV-@ or with both expression constructs showed marked specific increases in corresponding a and @ subunit RNAs. Immunoblot analysis utilizing anti-casein kinase I1 antiserum of cytosolic extracts prepared from COS-1 cells co-transfected with pCMV-a and pCMV-@ showed 2and 4-fold increases in immunoreactive a and @ subunit protein, respectively, relative to vector-transfected cells. These same cytosolic fractions exhibited an average 5-fold increase in casein kinase I1 catalytic activity. COS-1 cells transfected with pCMV-a alone exhibited a %fold increase in immunoreactive a subunit protein and a nearly 2-fold increase in cytosolic casein kinase I1 catalytic activity. Transfection with the cDNA coding for the noncatalytic @ subunit alone also caused a near doubling of cytosolic casein kinase I1 catalytic activity. No increase in immunoreactive a subunit protein was observed in pCMV-@-transfected cells, and no increase in immunoreactive @ subunit protein was observed in pCMV-a-transfected cells. These results indicate that a portion of the endogenous cellular casein kinase I1 protein is not fully active and that raising the concentration of the a or @ subunit stimulates this latent activity.
Casein kinase I1 is a ubiquitous serine/threonine protein kinase whose catalytic activity is independent of such effectors as cyclic nucleotides, calcium/calmodulin, and phospholipids (1-4). Casein kinase 11, identified in a wide variety of eukaryotic tissues, consists of a and a' subunits of molecular mass of -37-44 kDa and p subunits of molecular mass of -24-28 kDa, assembled into native heterotetrameric structures designated a& or act'(&. The existence of a and a' subunits has been demonstrated by deduced amino acid sequence from cDNAs encoding the subunit isoforms ( 5 ) and by direct primary sequencing (6, 7). Biochemical techniques such as affinity labeling with ATP analogs have identified the * This work was supported by National Institutes of Health Program Project Grant CA 39240 (to M. P. C) and by National Institutes of Health Training Grant DK 07302 (to R. A. H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. CY subunit as the catalytic subunit (8,9). The native heterotetramer has been shown to undergo autophosphorylation on the p subunit i n vitro (lo), but the site(s) and the physiological role for this autophosphorylation have yet to be determined. In addition, evidence has been recently presented that the subunit of casein kinase I1 is phosphorylated at site(s) other than the autophosphorylation site with concomitant increases in casein kinase I1 catalytic activity (11)(12)(13).
Renaturation studies by Cochet and Chambaz (14) have shown that optimal casein kinase I1 activity is achieved when a and /3 subunits are present in a 1:l molar ratio. Lin et al. (15) and Hu and Rubin (16) have both recently presented data which compare structural and functional properties of purified casein kinase I1 holoenzyme with purified casein kinase I1 a subunit expressed in Escherichia coli. Their data independently demonstrate that the isolated a subunit exhibits a low level of catalytic activity relative to native casein kinase 11, and their data strongly suggest that the p subunit plays a positive regulatory role in the expression of casein kinase I1 activity. The aim of the present study was to assess the effects of independently increasing the concentration of either a or p casein kinase I1 subunits on casein kinase I1 catalytic activity in intact cells. In this report, we demonstrate the successful transfection of human casein kinase I1 a and p subunit cDNAs into COS-1 fibroblasts. We find that the effect of overexpressing either of these individual subunits in COS-1 cells is to double total cytosolic casein kinase I1 catalytic activity.

Methods
Subcloning of the Human a and @ Casein Kinase II cDNAs into the Expression Vector pCMV-The vector pCMV contains a unique BglII restriction site located downstream from the cytomegalovirus promotor and upstream from a 3' human growth hormone DNA sequence and a polyadenylation site. For subcloning into this site, RamHI linkers were attached to the 1.6-kb human a subunit cDNA (18). This DNA was then digested with RglII to remove 109 nucleotides of 5' noncoding sequence, and the cDNA with BgllI-BamHI ends was inserted into BglII-digested pCMV. In order to subclone the human @ subunit cDNA into pCMV, the 3' 1.28-kb cDNA (19) was digested with DraI to remove 600 nucleotides of 3' noncoding sequence including the polyadenylation site. This 0.68-kb 3' fragment possessing EcoRI-DraI ends was ligated with the 0.16-kb 5' cDNA fragment possessing EcoRI-EcoRI ends. BamHI linkers were attached to this 0.84-kb cDNA product containing the entire open reading frame. The cDNA, now possessing BamHI ends, was inserted into pCMV at the BglII cloning site. The resultant expression constructs were designated pCMV-n and pCMV-@.
Transfection of COS-2 Cells-Transient transfection of pCMV-a and pCMV-@ into COS-1 fibroblasts was accomplished by calcium phosphate precipitation essentially as described by Gorman (20). Briefly, COS-1 cells were grown under an atmosphere of 5% CO, at 37 "C in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and 50 units/ml penicillin and 50 pg/ml streptomycin sulfate. Eighteen hours prior to DNA treatment, the cells were plated at a density of 5 X 10" cells per 100-mm tissue culture dish. Each dish of cells was transfected with 20 pg of either pCMV vector plasmid DNA, pCMV-a plasmid DNA, pCMV-@ plasmid DNA, or co-transfected with pCMV-a and pCMV-@ plasmid DNAs in buffer containing HEPES-buffered saline, pH 7.12, 250 mM CaCI?, and 0.04 mg/ml salmon testes DNA as carrier and incubated for 4 h. Cells were washed with complete medium 3 times and glycerol-shocked for 1 min. Cells were then washed 3 times with complete medium and incubated a t 37 "C for approximately 65 h prior to harvest.
RNA Preparation and Analysis-Total RNA from COS-1 fibroblasts transfected with pCMV vector, pCMV-a, or pCMV-@ alone or co-transfected with pCMV-a and pCMV-@ together, and total RNA from cultured HepG2 cells were prepared according to the method outlined by Davis et al. (21). Northern analyses were performed exactly as described by .
Preparation of Cellular Extracts-100-mm dishes of transfected COS-1 fibroblasts were washed 3 times with ice-cold phosphatebuffered saline consisting of 1.7 M NaC1, 33.5 mM KCI, 101 mM Na2HPOI, 18 mM KH,PO,, 0.68 mM CaCI,, 0.49 mM MgCI,. Cells were scraped using a rubber policeman into 0.5 ml of harvest buffer containing 80 mM @-glycerophosphate, 20 mM EGTA, 15 mM MgC12, 50 mM NaF, 100 p~ NaV04, 0.1 mM phenylmethylsulfonyl fluoride, 10 pg/ml aprotinin, pH 7.3, and then carefully drawn through a 22gauge needle 20 times. Lysates were centrifuged for 20 min a t 22,000 X g, and supernatants were transferred to a new tube and immediately frozen a t -70 "C. Protein was quantitated by the method of Bradford (22).
Protein Kinase Assays-3-10 pg of protein from transfected COS-1 fibroblasts prepared as described above were assayed at 30 "C in a total reaction volume of 25 pl in assay buffer containing 100 mM Tris-HCI, pH 8.0, 100 mM NaCI, 50 mM KC1, 20 mM MgC12, 100 p M NaV04, 1 mM synthetic peptide substrate RRREEETEEE (23) and 200 pM ATP ([r-"P]ATP). All reactions were performed in triplicate. Control reactions were run under identical conditions without the synthetic peptide substrate, and the values obtained were subtracted as background. Reactions were terminated after 10 min by adding an equal volume of 0.01 M ATP and 0.4 N HCI. Reactions were subsequently trichloroacetic acid-precipitated, incubated on ice for 5 min, and microcentrifuged for 5 min before 12.5 p1 were spotted onto 1.5cm' pieces of Whatman P81 paper. The pieces of paper were washed in 0.5% HzPO, 3 times, 10 min per wash, and counted in a liquid scintillation counter. As indicated, heparin a t a final concentration of 1 pg/ml was added to the assay buffer. Statistical analysis utilizing the Student's t test with a confidence interval of 95% was performed.
SDS-PAGE Electrophoresis and Immunoblotting-Cell fractions were thawed on ice, solubilized in sample buffer containing 2-mercaptoethanol with heating at 100 "C for 5 min, and resolved by SDS-PAGE using polyacrylamide gels as described by Laemmli (24). Proteins were transferred electrophoretically to nitrocellulose at 300 mA for 1 h, essentially as described by Towbin et al. (25). The nitrocellulose was blocked for 45 min a t room temperature in buffer containing 10 mM Tris-HCI, p H 7.4,150 mM NaCI, 2% bovine serum albumin, 0.1% Triton X-100 and was then incubated with gentle agitation for 2-4 h a t room temperature in rabbit anti-Drosophila casein kinase I1 antiserum diluted 1/100 in blocking buffer. The nitrocellulose was washed once in 10 mM Tris-HCI, pH 7.4, 150 mM NaCl for 5 min at room temperature and then 4 times in 10 mM Tris-HCI, p H 7.4, 150 mM NaCI, 0.1% Triton X-100 for 5 min a t room temperature. The nitrocellulose was incubated with gentle agitation in ""I-protein A in 10 mM Tris-HCI, pH 7.4, 150 mM NaC1, 0.1% Triton X-100 for 45 min at room temperature and then washed once in 10 mM Tris-HCI, pH 7.4, 150 mM NaCl for 5 min a t room temperature and then 4 times in 10 mM Tris-HCI, p H 7.4, 150 mM NaC1, 0.1% Triton X-100 for 5 min a t room temperature. Immunoreactive proteins were visualized by autoradiography using Kodak XAR film.

RESULTS AND DISCUSSION
Complementary cDNAs encoding the a and 0 subunits of human casein kinase I1 (18,19) were subcloned into a mammalian expression vector, pCMV (17), containing cytomegalovirus promotor sequences. The resultant expression constructs, pCMV-a and pCMV-B, were transfected individually or were co-transfected into subconfluent COS-1 fibroblasts. Cells were grown to confluence and total cellular RNA was prepared as described under "Experimental Procedures." reading frame. Following a 1-h exposure of the autoradiogram, three hybridizing species of RNA migrating at 1.8,4, and >6.5 kb were observed with RNA prepared from cells transfected with pCMV-a and cells co-transfected with pCMV-a and pCMV-P when probed with the 1.5-kb a subunit cDNA (Fig.  1, lanes 2 and 4). In contrast, no such hybridization signal was observed with RNA prepared from cells transfected with vector or with pCMV-P alone (Fig. 1, lanes 1 and 3 ) . After overexpressing casein kinase I1 subunits. Total cellular RNA was prepared from cells transfected with pCMV vector or with pCMVa and pCMV-@ constructs. RNAs (20 pg/lane) were electrophoresed on a formaldehyde gel, transferred to nitrocellulose, and the nitrocellulose was hybridized a t 42 "C in 50% formamide to the 1.5-kb HepG2 a subunit cDNA (lanes 1-5) or the 0.87-kb HepG2 @ subunit cDNA (lanes 6-10). Blots were washed at 62 "C in 0.1 X SSC and autoradiographed for either 1 or for 42 h. Lunes autoradiographed for 1 h: I, RNA prepared from cells transfected with pCMV vector; 2, RNA prepared from cells transfected with pCMV-a; 3, RNA prepared from cells transfected with pCMV-@; 4, RNA prepared from cells cotransfected with pCMV-a and pCMV-@; 5, RNA prepared from cells transfected with pCMV vector; 6, RNA prepared from cells transfected with pCMV-a; 7, RNA prepared from cells transfected with pCMV-@; 8, RNA prepared from cells co-transfected with pCMV-a and pCMV-@. Lunes autoradiographed for 42 h: HepG2 cell RNA hybridized with a and @ subunit cDNA probes as indicated.
1 HR approximately 42 h, two major RNA species migrating a t approximately 4.6 and 2.7 kb, which resulted from hybridization of the a probe with endogenous casein kinase I1 RNA, could be seen in these lanes (data not shown). Similarly, when RNAs prepared from transfected COS-1 cells were hybridized with the p subunit cDNA probe, only RNA from cells transfected with pCMV-@ or co-transfected with both pCMV-a and pCMV-P demonstrated a strong hybridization signal following a 1-h exposure (Fig. 1, lanes 7 and 8). These RNAs migrated at 1.2-1.5, 3.7, and >6.0 kb. After approximately 42 h, a single band migrating a t 4.2 kb, arising from hybridization of host cell casein kinase I1 RNA with the p cDNA probe, was observed in lanes containing RNA prepared from cells transfected with vector or pCMV-a. It is unclear why mutiple higher molecular weight RNA species which react with our a and p subunit-specific probes are so abundant in COS-1 cells transfected with pCMV-a or pCMV-P or with both expression constructs. Our data, presented in Figs. 2 and 3 and Table I in this paper, indicate that correctly processed human a and p subunits are being expressed in COS-1 cells. Hence we have not pursued the identification of these higher molecular weight species further.
In the above study, the 1-h exposure of the autoradiogram is shown to demonstrate the marked increase in expression of a and p subunit-specific RNAs in the respective transfected cells as compared with control transfected cells. To further emphasize the magnitude of increased expression of the a and subunit RNAs, a Northern blot of identical quantities (20 pg/lane) of RNA prepared from HepG2 cells and probed with the same a and / 3 subunit cDNAs is also shown (Fig. 1). The approximate sizes of the two major RNA species present in total HepG2 cell RNA which react with the a subunit cDNA probe are 2.7 and 4.6 kb. These sizes correlate well with the sizes of endogenous a subunit RNAs present in COS-1 cells.
std vector a p a+p TYPE:

FIG. 2. Immunoblot of protein fractions from COS-1 fibro-
blasts overexpressing casein kinase I1 subunits. COS-1 fibroblasts were transfected with pCMV vector or with pCMV-n and pCMV-8 constructs. Cell fractions were prepared for SDS-PAGE as described under "Experimental Procedures." 75 pg of protein per lane were boiled in SDS-PAGE sample buffer containing 2-mercaptoethanol and resolved by SDS-PAGE on a 12% polyacrylamide gel. Proteins were transferred to nitrocellulose which was subsequently blocked, incubated with rabbit anti-Drosophila casein kinase I1 antiserum, and finally incubated with '"I-protein A. Immunoreactive proteins were visualized by autoradiography. Molecular mass standards in kilodaltons are as follows: myosin heavy chain, 196.0; phosphorylase b, 105.7; bovine serum albumin, 71.0; ovalbumin, 44.2; carbonic anhydrase, 27.8; @lactoglobulin, 18.3. Lanes: std, bovine thymus casein kinase 11; vector, cytosolic fraction prepared from cells transfected with pCMV vector; a, cytosolic fraction prepared from cells transfected with pCMV-a; @, cytosolic fraction prepared from cells transfected with pCMV-P; n + 8, cytosolic fraction prepared from cells co-transfected with pCMV-a and pCMV-P. pCMV vector or with pCMV-n and pCMV-8 constructs as described under "Experimental Procedures" and in the legend of Fig. 2. 5 pg of total protein were assayed for 10 min a t 30 "C in a total reaction volume of 25 pl in assay buffer containing the synthetic peptide substrate RRREEETEEE and 0.2 mM ATP ([r-"YP]ATP). Reactions were terminated by adding an equal volume of 0.01 M ATP and 0.4 N HC1. Reactions were subsequently trichloroacetic acid-precipitated, microcentrifuged for 5 min, and 12.5 p1 were spotted onto 1.5-cm' pieces of Whatman P81 paper and counted in a liquid scintillation counter. As indicated, heparin at a final concentration of 1 pg/ml was added to the assay buffer.  Fig. 2 and catalytic activity measurements as described under "Experimental Procedures" and in the legend for Fig. 3 were performed for cells transfected with either pCMV-n or pCMV-P or both pCMV-n and pCMV-P. Data from immunoblot analyses were quantitated by densitometric scanning of autoradiograms and are expressed as -fold densitometric units over vectortransfected COS-1 fibroblasts. Catalytic activity is expressed as -fold activity over vector-transfected COS-1 fibroblasts. Statistical analysis utilizing the Student's t test with a confidence interval of 95% was performed. When HepG2 cell RNA was hybridized with the @ subunit cDNA probe, a major band migrating at 1.2 kb and a minor band migrating a t 4.2 kb were revealed. The differences in relative abundance of casein kinase I1 / 3 subunit RNAs observed in COS-1 cells and HepG2 cells may be due to transcriptional processing. It has been previously suggested (19) that in certain cell types, e.g. HepG2 cells, casein kinase I1 p subunit-specific RNAs undergo transcriptional processing resulting in a deletion of a portion of the 3"untranslated sequence flanking the termination codon and generating lower molecular weight RNA species such as the 1.2-kb RNA. As indicated, an exposure time of 42 h was required to achieve a hybridization signal that was approximately equal in intensity to that of the 1-h exposure of the autoradiogram in which RNAs were prepared from overexpressing cells. Hence, these results demonstrate that transcription of the human casein kinase I1 cDNAs is being driven by the cytomegalovirus promotor contained within the expression vector and that RNA levels of both the a and @ subunits are increased approximately 28-fold relative to controls.
In order to assess the presence of human casein kinase I1 a and @ subunit protein in the COS-1 cells transfected with pCMV-a and pCMV-@, immunoblot analysis was performed. Transfected cells were harvested and cytosolic fractions were prepared as described under "Experimental Procedures." Fig.  2 shows the results of a representative immunoblot utilizing a rabbit antiserum generated against Drosophila casein kinase I1 which cross-reacts with both the a and @ subunits of casein kinase I1 from a variety of species. Quantitation of immunoreactive a and @ protein was determined by densitometric scanning of autoradiograms from several experiments (Table  I). Cytosolic fractions prepared from cells transfected with pCMV-a or co-transfected with pCMV-a and pCMV-P showed approximately 2.8-and 2.0-fold respective increases in immunoreactive a subunit protein relative to a subunit protein present in cytosolic fractions prepared from vectortransfected cells. No increase in immunoreactive a subunit protein was observed in the cytosolic fraction prepared from cells transfected with pCMV-P alone ( Table I). As depicted in Fig. 2, the expressed human a subunit has a mobility identical to that of the COS-1 cell monkey a subunit. The mobility of both the human and monkey a subunits is decreased with respect to purified casein kinase I1 a subunit from bovine thymus (Fig. 2, std lane).
Cytosolic fractions prepared from cells transfected with pCMV-@ or co-transfected with pCMV-a and pCMV-P showed 9.7-and 4.1-fold increases in total immunoreactive p subunit protein, respectively, relative to P subunit protein present in cytosolic fractions prepared from vector-transfected cells (Table I). No increased expression in immunoreactive p subunit protein was detected in samples prepared from cells transfected with pCMV-a alone. Interestingly, the expressed human ( 3 subunit observed from cells either cotransfected with both pCMV-a and pCMV-@ or transfected with pCMV-@ alone migrated as a closely spaced doublet of molecular mass -24-26 kDa (Fig. 2). The lower band of this doublet co-migrates with the @ subunit of casein kinase I1 prepared from bovine thymus and with the COS-1 cell monkey @ subunit. As the predicted amino acid sequence of the human casein kinase I1 /3 subunit contains two potential N-linked oligosaccharide consensus sequences, the variation in the mobilty of the p subunit, resulting in the upper band of the doublet, might be due to the glycosylation of a population of the expressed p subunits. However, digestion of samples identical to those shown in Fig. 2 with endoglycosidase F did not alter the mobility of either band in the doublet (data not shown). The functionality of the expressed casein kinase I1 a and / 3 subunit proteins was examined by assaying fractions prepared from the transfected cells for phosphotransferase activity. Table I presents such results where casein kinase I1 catalytic activity was measured, utilizing the synthetic peptide substrate RRREEETEEE (23), in cytosolic fractions from cells transfected with pCMV vector, pCMV-a, pCMV-B, or cotransfected with pCMV-a and pCMV-b. Samples prepared from several experiments in which COS-1 cells were transfected with pCMV-a or pCMV-@ showed average 1.7-and 1.8fold increases, respectively, in casein kinase I1 catalytic activity relative to that in vector-transfected cells. The cytosolic fraction prepared from cells co-transfected with pCMV-a and pCMV-P showed a 5-fold increase in casein kinase I1 catalytic activity relative to vector-transfected controls (Table I). Heparin has been shown to be a potent negative effector of casein kinase I1 activity in vitro (26). At a concentration of 1 pg/ml, heparin virtually abolished the casein kinase I1 activity present in the cytosol of cells transfected with either pCMV-a or pCMV-/3 or with both expression constructs, as shown in a representative experiment (Fig. 3).
The marked &fold increase in casein kinase I1 activity observed in extracts of COS-1 cells co-transfected with pCMV-a and pCMV-P (Table I) is associated with increases in the cellular protein concentration of both subunits (Fig. 2). These data suggest that the heterologously expressed human a and b subunits assemble into native casein kinase I1 heterotetramers. This hypothesis is consistent with previously published results from Cochet and Chambaz (14) who demonstrated that increasing the amount of casein kinase I1 / 3 subunit in the presence of a constant amount of a subunit enhanced casein kinase I1 catalytic activity to a maximum of 5-fold relative to the catalytic activity of the a subunit alone. This maximum activity occurred when a and @ subunits were present in a 1:l stoichiometry. Likewise, the consistent increase in casein kinase I1 activity we observe in cytosolic fractions prepared from cells transfected with pCMV-a alone (1.7-fold greater than vector-transfected cells) is not surprising in light of the fact that the a subunit bears the ATPbinding site. These cells express approximately %fold more a subunit than vector-transfected cells (Table I). It has been previously shown by a number of investigators that isolated a subunit exhibits a low level of casein kinase I1 catalytic activity relative to native casein kinase 11. Hu and Rubin (16) and Lin et al. (15) have independently expressed the a subunit of casein kinase I1 in E. coli, and both have shown the kcat of the purified, expressed a subunit to be <lo% of the kcat of casein kinase I1 holoenzyme.
In contrast, the significant 1.8-fold increase in casein kinase I1 catalytic activity observed in cells transfected with pCMV-/ 3 alone (Table I) was unexpected because the p subunit does not contain an ATP-binding site. Furthermore, no increase in a subunit protein could be detected in extracts prepared from pCMV-@-transfected cells (Fig. 2). One possible explanation for the increase in casein kinase I1 activity when either pCMV-a or pCMV-P alone is transfected into COS-1 cells is that there are excess free a and p subunits present in these cells. Increases in a or @ subunit due to overexpression would result in an increase in casein kinase I1 holoenzyme and a consequent increase in catalytic activity. This hypothesis is consistent with the high ratio of a subunit$ subunit immuunoreactiviry in control COS-1 cells compared with that ratio observed with purified enzyme (Fig. 2). However, the discrepancy in the immunoreactivity ratio for a and @ subunits could also arise from differences in species reactivity with the casein kinase I1 antiserum employed. It would appear from our data that the availabilty of free a subunits must be limited, since COS-1 cells transfected with pCMV-P alone exhibited an almost 10-fold increase in p subunit protein yet displayed only a 1.8-fold increase in activity. The hypothesis is also consistent with our finding that increasing the expression of the a subunit alone resulted in a greater increase in activity than would have been predicted based on the above mentioned studies by Cochet and Chambaz (14), Hu and Rubin (16),and Lin et al. (15). In either case, our findings contrast with recent observations regarding the subunits of CAMP-dependent protein kinase. Uhler and McKnight (27) demonstrated that stable overexpression of the Ca and CP subunits of CAMPdependent protein kinase in NIH 3T3 cells resulted in increased levels of RI protein but not RII protein in these cells. Furthermore, it was shown that RI-specific mRNA levels were not increased in these cells suggesting that the degradation rate of RI had been altered by the presence of increased C subunit. The authors suggested that this was due to a stabilization of RI by its associating with C in the holoenzyme complex. We find no evidence for this phenomenon relating to casein kinase I1 since no detectable increase in casein kinase I1 a subunit protein was observed in cells transfected with pCMV-p in the four separate experiments we performed.
An alternative hypothesis relates to the possibility that a significant amount of cellular casein kinase I1 activity is present in an inhibited form. An endogenous casein kinase I1 inhibitor has been described (28). Such an inhibitor might normally be associated with p subunit in the holoenzyme complex and could be competitively bound by overexpressed 0 subunit. Such a mechanism would allow a population of inhibited holoenzyme to become active. Further experiments are required to test this hypothesis. Regardless, the data presented here strongly suggest that some portion of the endogenous COS-1 cell casein kinase I1 catalytic ( a ) subunits are less than fully active. This latent enzyme activity is significantly enhanced upon independent expression of p sub-