97

Molecular and Cellular Biochemistry 191: 97–104, 1999.© 1999 Kluwer Academic Publishers. Printed in the Netherlands.

Association of protein kinase CK2 with eukaryotictranslation initiation factor eIF-2 and with grp94/endoplasmin

Marta Riera, Nerea Roher, Francesc Miró, Carles Gil, Ramon Trujillo,José Aguilera, Maria Plana and Emilio ItarteDepartament de Bioquímica i Biologia Molecular, Universitat Autònoma de Barcelona, Campus de Bellaterra,Bellaterra, Barcelona, Spain

Abstract

Protein kinase CK2 forms complexes with some protein substrates what may be relevant for the physiological control ofthis protein kinase. In previous studies in rat liver cytosol we had detected that the trimeric form of eukaryotic translationinitiation factor 2 (eIF-2) co-eluted with protein kinase CK2. We have now observed that the ratio between eIF-2 and cytosolicCK2 contents in testis, liver and brain is quite similar, being eIF-2 levels about 5-fold higher than those of CK2. FurthermoreeIF-2 was present in liver samples immunoprecipitated with anti-CK2α/α′ antibodies, confirming the existence of complexescontaining both proteins. Nonetheless, these complexes would represent only a fraction of total cytosolic CK2 and eIF-2.

We had also observed that rat liver membrane glycoproteins obtained through chromatography on wheat-germlectin-Sepharose contain CK2 activity which copurifies with grp94/endoplasmin. We have now confirmed that this activitywas due to the presence of protein kinase CK2 as evidenced by immunodetection with antibodies against CK2α/α′. The fractionsenriched in grp94/endoplasmin and CK2 also contained another 55-kDa polypeptide (p55) phosphorylated by CK2 whichhas been identified as calreticulin by N-terminal sequencing. Calreticulin and grp94/endoplasmin could be partially resolvedfrom CK2 by chromatography on heparin-agarose and almost completely on ConA-Sepharose. However, phosphorylation ofimmunoprecipitated grp94/endoplasmin was enhanced by its preincubation with purified CK2 prior to immunoprecipitation,what confirms the easy reassociation between these proteins.

The association of protein kinase CK2 with eIF-2 and with grp94/endoplasmin may serve to locate the enzyme in the cellularmachinery involved in protein synthesis and folding, and reinforces the possible involvement of CK2 in these processes.(Mol Cell Biochem 191: 97–104, 1999)

Key words: protein kinase CK2, eukaryotic translation initation factor eIF-2, grp94/endoplasmin, calreticulin, proteinphosphorylation, rat tissues

Introduction

Protein kinase CK2 (formerly casein kinase 2) is a highlyevolutionary conserved enzyme that phosphorylates serineand threonine residues inserted in acidic regions of proteins,and uses GTP as efficiently as ATP as phosphate donor [1].Substrates for CK2 include proteins involved in the control

of cell metabolism, gene expression and protein synthesis.Furthermore, CK2 seems to be vital for cells, varies duringcell proliferation and differentiation [2], and becomesoncogenic when disregulatedly expressed [3].

The enzyme is composed of two types of catalytic subunits(CK2α and CK2α′) and a non-catalytic phosphorylatablesubunit (CK2β). CK2α and CK2α′ are encoded by different

Address for offprints: E. Itarte, Departament de Bioquímica i Biologia Molecular, Facultat de Ciències, Universitat Autònoma de Barcelona, E-08193 Bellaterra,Barcelona, Spain

98

genes but posses similar catalytic properties. Dimerizationof CK2β initiates the assembling of the enzyme that iscompleted with the binding of two of the catalytic subunits,either CK2α or CK2α′, giving rise to the classical hetero-tetrameric forms [4]. The existence of coordination betweenthe expression of CK2α, CK2α′ and CK2β is a matter ofdiscussion. A parallel biosynthesis of CK2α/α′ and CK2βoccurred in bovine adrenocortical cells incubated in thepresence of serum [5]. In contrast, the existence of a poolof CK2β in excess to that of CK2α has been observed inexponentially growing lymphoid cells [6]. The existence oftightly bound complexes between CK2α and intranuclearcomponents distinct from CK2β has been described [7].More recently, it has been reported that free CK2 subunitsmay influence cell signalling through the MAP kinasecascade by interacting with some of its components such asprotein kinases A-Raf and Mos, which bind to CK2β [8–10],and protein phosphatase 2A that binds to CK2α [11]. Thisindicates that these subunits may have other roles besidesforming part of CK2 holoenzyme.

The mechanisms for the physiological control of CK2 arestill uncertain since the enzyme is not responsive to anyknown second messengers [1]. It is well established thatassociation of CK2α with CK2β to form the heterotetramerincreases its activity on β-casein and some other substrates[1, 2]. Also, CK2 forms complexes with a diversity of proteinsubstrates which may influence its activity and/or subcellularlocalization. Substrates that interact with CK2 include hsp90[12], topoisomerase II [13], the tumour suppressor proteinp53 [14], nucleolin [15], the large T antigen of SV40 [16],cytoplasmic dynein [17], the nucleolar protein Nopp 140[18], and the cytoplasmic protein Dishevelled [19], amongothers. Studies from our group have shown the co-purificationof rat liver cytosol CK2 and a 49-kDa protein [20] which hasbeen characterized as composed by the β/γ-subunits ofeukaryotic translation initiation factor eIF-2 [21]. Also, wehad observed that rat liver membrane glycoproteins obtainedthrough chromatography on wheat-germ lectin-Sepharosecontain CK2 activity which copurifies with a 210-kDaprotein substrate identified as the dimeric form of grp94/endoplasmin [22]. The existence of complexes betweenCK2 and these proteins has been studied in more detail inthe present work.

Materials and methods

Materials

ATP, GTP, β-casein, heparin-agarose, and DEAE-Sepharosewere from Sigma Chemical Co. ConA-Sepharose was fromPharmacia and phosphocellulose P11 was from Whatman.[γ-32P]ATP and [γ-32P]GTP were from Amersham. Immobilon

P transfer membranes were from Millipore. Anti-CK2α/α′antibodies were from Upstate Biotechnology Incorporated,anti-CK2β and anti-eIF-2γ antibodies were raised in rabbitsas described previously [21]. Anti-eIF-2α and anti-eIF-2βserum was a gift of Dr. C.G. Proud (University of Kent,Canterbury). Monoclonal antibodies against chicken oviductgrp94/endoplasmin were obtained from StressGen Bio-technologies Corp. Purified rat liver CK2 and eIF-2(trimeric form) were obtained as described previously [20,21]. The sources of other reagents for electrophoresis,immunodetection and other chemicals have been indicatedpreviously [21, 22].

Preparation of extracts from rat tissues andchromatography on phosphocellulose

The different tissues were processed immediately after theirexcision from Sprague-Dawley rats (150–200 g of bodywt) which were sacrificed by decapitation after etheranaesthesia. Cytosolic extracts were prepared as describedpreviously for rat liver [21]. Aliquots of the 100,000 × gsupernatants were applied to centrifuge tubes that contained0.5 ml of phosphocellulose and stirred overnight at 4°C.The resin was sedimented by centrifugation at 5,000 × gfor 10 min and then washed with 1 ml of homogenizationbuffer supplemented with 0.3 M-KCl. Under these conditions,almost all of CK2 and eIF-2 are retained in the resin andeluted by washing it with 1.2 M-KCl in homogenizationbuffer.

To obtain endoplasmic reticulum membranes, rat liverhomogenates, prepared as described previously [22], werecentrifuged first at 600 × g for 30 min, then at 9,000 × g for30 min and finally at 100,000 × g for 90 min, all at 4°C. Thesediment from the last centrifugation was resuspended andsolubilized with 2% (v/v) Triton X-100 as described before[22]. The solubilized membrane proteins were applied to a10-ml DEAE-Sepharose column that was washed with 50 mlof the resuspension buffer and eluted with a 40 ml of a linearNaCl gradient (0.15–0.65 M). The pool of the relevantfractions eluted from DEAE-Sepharose was applied to a 4-mlheparin-agarose column and eluted with a 10 ml linear NaClgradient (0.15–0.65 M). The samples with the highestcontent of grp94/endoplasmin were applied to a 5-mlConA-Sepharose column that was eluted with 15 ml of alinear α-D-mannose gradient (0–1 M).

CK2 assays

CK2 activity was determined as described previously [21, 22]using either 125 µM [γ-32P]GTP or [γ-32P]ATP and 1 mg/mlβ-casein as substrates. One unit of CK2 activity is defined

99

as the amount that catalyses the transfer of 1 nmol ofphosphate from either [γ-32P]ATP or [γ-32P]GTP to theprotein substrate.

Gel electrophoresis and immunological assays

Polyacrylamide gel electrophoresis in the presence of SDS(SDS-PAGE), immunoblots and immunodetection withspecific antibodies were carried out as described previously[21, 22]. The intensity of the bands was quantified bydensitometry using a GS-700 Imaging densitometer(Bio-Rad).

Immunoprecipitation

Protein A-agarose beads were preincubated overnight in 1%BSA/PBS and then washed 3 times with TNX buffer (50 mMTriethanolamine/HCl, pH 7.4, 100 mM NaCl and 0.5% (w/v)Triton X-100). One hundred µl of the sample werepre-incubated overnight at 4°C with the specific antibodies.Then 50 µl of protein A-agarose were added to the tubes that

were supplemented with TNX buffer up to 500 µl andincubated for 2 h at room temperature. The immuno-precipitates were collected by centrifugation, washed 3times with TNX buffer, resuspended in 2 × Laemmli samplebuffer and subjected to SDS/PAGE electrophoresis.

Results

Tissue distribution of protein kinase CK2 and of eIF-2

The distribution of cytosolic CK2 subunits in the differenttissues was studied after partial purification and concentrationof the samples by chromatography on phosphocellulose asindicated in Materials and methods. Protein kinase CK2purified from rat liver cytosol was used as standard. A rangeof dilutions of the tissue samples was assayed and only thosethat fitted within the linear range of the calibration curvewere considered (Fig. 1a). Each individual sample wasanalysed at least twice and the mean of the values obtainedwere considered in the statistical calculations. Theanti-CK2α/α′ antibody recognized a set of 3–4 bands inpurified rat liver CK2, and the 37–42 kDa bands were the

Fig. 1. Quantitation of CK2αT content in rat brain, liver and testis. Purified rat liver CK2 was used as standard to construct the calibration curve. Aliquots ofthe phosphocellulose eluates obtained from each tissue, as indicated under Materials and methods, were used in the quantitation of CK2αT. In all cases, thesamples were subjected to SDS/PAGE under reducing conditions using 12% (w/v) polyacrylamide gels, and then transferred onto Immobilon-P, checked forreaction with antibodies against CK2α/α′ and the intensity of the bands was quantified by densitometry. (a) The optical density obtained with the indicatedamounts of purified rat liver CK2 is shown; (b) 20 µg of total protein present in the phosphocellulose eluates from brain (B), liver (L) or testis (T) were analysedas indicated above; (c) Different amounts (µg) of either purified rat liver CK2 or of total protein present in the phosphocellulose eluates from brain, liver andtestis were used in Western blots developed with antibodies against CK2α/α′.

100

most intensively stained (Fig. 1c). The samples from thedifferent tissues also showed several bands detected withthe anti CK2α/α′ antibody but the ratio between these bandsvaried between experiments for any given tissue (Figs 1b and1c). The origin of the different bands could correspond tothe presence of CK2α and CK2α′ and some proteolyticfragments derived from them. A complex pattern of catalyticsubunits has also been observed in our previous studies onpurified rat liver CK2 [20] but this was not reflected indifferences in the specific activity of the enzyme. Thus, allthese bands were considered together as CK2α

T. A similar

approach was used to detect CK2β, and a single band wasobserved in the samples analysed. Purified rat liver eIF-2was used as standard to quantify the amount of eIF-2β andeIF-2γ present in the tissue samples, using the samepreparations and following a similar approach to thatdescribed above for CK2 subunits.

A summary of the estimates of CK2 and eIF-2 subunitsin the tissues is indicated in Table 1. The specific contentof CK2α

T and CK2β in liver and brain were similar and lower

than those in testis. A similar fact was envisaged with eIF-2levels which in the three tissues studied were about 4–5-foldhigher than those detected for cytosolic CK2. The amountof protein recovered in the phosphocellulose eluate per gof wet tissue (gwt) was about 0.8 mg for brain, 1.2 mg forliver and 1 mg for testis. Considering these values, thetissue content of CK2 (expressed as CK2α

T) would be

about 0.07 nmol/gwt for brain, 0.12 nmol/gwt for liver and0.11 nmol/gwt for testis and the eIF-2 content (expressed aseIF-2γ) would be about 0.26 nmol/gwt for brain, 0.57 nmol/gwt for liver and 0.65 nmol/gwt for testis.

Co-immunoprecipitation of hepatic CK2 and eIF-2

In previous studies [20, 21] we had observed that CK2 samplespartially purified from rat liver through phosphocellulosecontained eIF-2 (at that time identified as pp49). In order tosee if both proteins might form complexes, these sampleswere immunoprecipitated with anti-CK2α/α′ antibody andthe presence of eIF-2 in the immunoprecipitates analysed

Table 1. Summary of the estimate of CK2 and eIF-2 in rat tissues

CK2 subunits eIF-2 subunits(pmol/mg prot.) (pmol/mg prot.)

CK2αT CK2β EIF-2β eIF-2γBrain 80.7 ± 6.3 59.4 ± 9.9 410 ± 42 331 ± 14Liver 97.5 ± 9.7 69.6 ± 16.6 457 ± 56 471 ± 23Testis 111.4 ± 15.4 151.8 ± 36.0 773 ± 36 615 ± 31

Extracts from each tissue were prepared and subjected to Western blotanalysis as indicated in the text. Values are pmols/mg of protein eluted fromphosphocellulose. The data are the mean ± S.E.M. of 3 different samples.

Fig. 2. Co-Immunoprecipitation of eIF-2α with protein kinase CK2. Thefraction eluted from a phosphocellulose chromatography which containedCK2 and eIF-2 was dialysed against 50 mM Tris/HCl, pH 7.5, 150 mMNACl and 10% glycerol. Aliquots of this fraction were immunoprecipitatedwith 4 µg of antibody against CK2α/α′ (lanes 2 and 5) or with non-immuneserum (lane 4) and analysed by Western blot. Lanes 1 and 2 were incubatedwith anti-CK2 antibody (1:1000) and lanes 3–5 with anti-eIF-2α (1:500)antibody.

in Western-blots using anti-eIF-2α antibodies. As shown inFig. 2, eIF-2 was detected in the immunoprecipitates,although the amount of eIF-2α recovered was much lowerthan that present in the immunoprecipitates obtained withanti-eIF-2α antibodies. This could reflect the fact that eIF-2is more abundant than CK2 and only a fraction of theinitiation factor would be bound to CK2.

Chromatographic resolution of rat liver membraneproteins

The crude preparation obtained by solubilization with 2%(v/v) Triton X-100 of rat liver membrane proteins as indicatedin Materials and methods was subjected to chromatographyon DEAE-Sepharose. Immunological analysis of the elutedproteins using commercial antibodies against grp94/endo-plasmin revealed the presence of this protein but also ofanother band with an apparent Mr of 55 kDa (p55) which wasstrongly detected by the antibodies (Fig. 3). In contrast togrp94/endoplasmin, that behaved as a dimer with an apparentMr of 210 kDa in SDS/PAGE under non-reducing conditions,

101

Fig. 3. Samples eluted from the column were subjected to SDS/PAGE under either (a) reducing or (b) non-reducing conditions using 7.5% (w/v) polyacrylamidegels, and then transferred onto Immobilon-P and checked for reaction with antibodies against grp94/endoplasmin. Grp94/endoplasmin and p55 were measuredby densitometry and the results are shown in the plot (c). The same samples were assayed for CK2 activity using [γ-32P]GTP and β-casein as a substrate andresults are also shown in the plot (c).

a

102

the mobility of p55 was not affected by the presence ofreducing agents. In order to discern if p55 derived fromproteolysis of grp94/endoplasmin or corresponded toanother protein the band was scised from the membrane andsubjected to N-terminal sequence analysis. This revealed thesequence DPAIYFKEQFLDGDA that coincides with that ofthe N-terminal end of mature rat calreticulin (Michalak,1992), a protein resident in the endoplasmic reticulumwhose C-terminal region shows homology with grp94/endoplasmin since it contains acidic stretches and a commonC-terminal KDEL sequence.

The elution profile of grp94/endoplasmin coincidedwith that of calreticulin (Fig. 3) and also with that of CK2activity detected using β-casein and GTP as substrates.Fractions with the highest content of grp94/endoplasmin werepooled and subjected to chromatography on heparin-agarose.CK2 and grp94/endoplasmin were retained by the columnbut part of grp94/endoplasmin was present in the 0.2-MNaCl wash that did not contain detectable CK2 activity,and the remaining co-eluted with CK2 in the 0.5-M NaClwash (data not shown). A more clear resolution of grp94/endoplasmin and CK2 was obtained when the pooledfractions were chromatographed on ConA-Sepharose (Fig.

4). The autoradiograph shown in Fig. 4a, demonstrates thatgrp94/endoplasmin and calreticulin were phosphorylatedby added CK2 and also that their elution overlapped withthat of another protein substrate for CK2 with and apparentMr of 76 kDa. No phosphorylation of any of these sub-strates was detected in the absence of added CK2 (datanot shown).

Reassociation and co-immunoprecipitation of CK2 andgrp94/endoplasmin

In order to see if CK2 re-associated easily with grp94/endoplasmin and if its presence was necessary for phos-phorylation of this protein, samples of grp94/endoplasmineluted from heparin-agarose with no detectable CK2activity were immunoprecipitated with antibodies againstendoplasmin after preincubation either in the absence or inthe presence of added CK2. As shown in Fig. 5, phos-phorylation of grp94/endoplasmin was detected only in thesamples that contained added CK2. This indicates that grp94/endoplasmin and CK2 had reassociated and coimmuno-precipitated with the anti-grp94/endoplasmin antibodies.

Fig. 4. Elution profile of solubilized membrane proteins from ConA-Sepharose. (a) Samples of the indicated column fractions were incubated with addedprotein kinase CK2 and [γ-32P]GTP, and then subjected to SDS/PAGE under reducing conditions using 7.5% (w/v) polyacrylamide gel and autoradiographed;(b) Aliquots of the samples were also subjected to SDS/PAGE under reducing conditions using 12% (w/v) polyacrylamide gel, transferred onto Immobilon-Pand checked for reaction with antibodies against CK2α/α′.

103

attack [27]. It is also interesting to remark the similarity inthe ratio of CK2 and eIF-2 content in the tissues but whetherthis was a mere coincidence or has any physiologicalmeaning is unknown and its ascertainment will requirefurther experiments. Our present data also demonstrate thatCK2 forms complexes with eIF-2. However, the amount ofeIF-2α immunoprecipitated with anti-CK2α antibodiesindicates that only a small fraction of eIF-2 and CK2 wouldbe complexed to each other. This is in contrast with thatreported for the interaction of CK2 with Nopp 140 [18]where a considerable proportion of the total cellular CK2protein was immunoprecipitated with Nopp 140 antibodies.Nonetheless, the increasing number of interacting partnersfor CK2 [7–22] suggests that the enzyme would be distri-buted in different complexes and thus the amount of each onewould be rather low.

The data obtained with membrane preparations has nowconfirmed that the CK2-like activity associated with grp94/endoplasmin [22] is due to presence in these preparationsof authentic CK2. This is in agreement with the recentobservation [28] that both the low ATPase and auto-phosphorylation activities reported for grp94/endoplasminwould be due to trace contamination with CK2. This is alsosupported by the fact that no autophosphorylation of grp94/endoplasmin was observed in the samples eluted fromConA-Sepharose, which contained no detectable CK2subunits, or in the immunoprecipitates with anti-grp94/endoplasmin antibodies, although phosphorylation of thisprotein was clearly evidenced when exogenous CK2 wasadded. On the other hand, the presence of CK2 activity inthe immunopreciptates of samples of grp94/endoplasminincubated with exogenous CK2 prior to immunoprecipitationwith anti-grp94/endoplasmin antibodies gives strong supportto its association.

The persistent presence of the p55 protein, identified ascalreticulin, in the samples of grp94/endoplasmin suggeststhe existence of heterocomplexes of both proteins, asindicated recently for mouse liver [29]. The ability ofcalreticulin to serve as substrate for CK2 is a matter ofcontroversy. Plant calreticulin is efficiently phosphorylatedby CK2 [30] and phosphorylation was also detected withmouse liver calreticulin [29]. In contrast no phosphorylationby CK2 was observed previously with rat liver calreticulin[31]. As indicated before [30] this could be due to the factthat the C-terminal region of calreticulin from differentplants contain potential phosphorylation sites for CK2,whereas rabbit liver calreticulin contains no serines orthreonines in this region. Rat liver calreticulin contains athreonine in the C-terminal region (EEDATGQAKDEL) andmouse calreticulin a serine (EEESPGQAKDEL) but theycannot be considered as potential phosphorylation sites forCK2. However, the primary structure of rat calreticulincontains the sequence 174VESGSLEDDW183 that is conserved

Fig. 5. Co-immunoprecipitation of grp94/endoplasmin and CK2. Grp94/endoplasmin was immunoprecipitated with 7.5 µg of antibodies againstchicken oviduct grp94 (lanes 1 and 2) or with control serum (lanes 4 and 5)either in the absence (lane 1 and 4) or in the presence (lanes 2 and 5) ofadded CK2. A sample of CK2 was also immunoprecipitated with anti-grp94antibody (lane 3) or with control serum (lane 6). Immunoprecipitated sampleswere incubated with [γ-32P]GTP and subjected to SDS/PAGE under reducingconditions using 7.5% (w/v) polyacrylamide gel and autoradiographed.

Discussion

The tissue distribution of protein kinase CK2 subunitsdetected in our present work agrees quite well with thatreported previously for cytosolic CK2 activity in rat tissues[23, 24]. However, it is interesting to note that the relativechanges in cytosolic CK2α

T content between tissues were

rather moderate and that the estimates of CK2β content werecomparable to those of CK2α

T in liver and brain, but higher

in testis. The possible functional meaning of this fact isunknown but it would favour the association between CK2αand CK2β to form the holoenzyme and may explain the highactivity detected in testis [23, 24].

Previous estimates of eIF-2 content indicated that it is amoderately abundant protein in HeLa cells [25] and a morerecent study using rabbit tissues has shown that its level inreticulocytes, liver, kidney and heart were relatively constantwhen expressed per g of wet tissue [26]. The estimates ofeIF-2 content (considered in terms of nmol of eIF-2γ/gwt)obtained in our present study show that they are relativelysimilar between liver and testis but lower in brain. Further-more, our data show that the eIF-2β/eIF-2γ ratios did notessentially vary between tissues, and the small differences canbe attributed to the high sensibility of eIF-2β to proteinase

104

in human, rabbit and mouse [32] and could be a potentialphosphorylation site for CK2. Our results indicate that ratcalreticulin is indeed a substrate for CK2 but the locationof the phosphorylation sites remains to be elucidated.

Summing up, our data show that in rat liver a fraction ofthe total protein kinase CK2 is associated with eIF-2 and alsoin part with grp94/endoplasmin. The existence of thesecomplexes may serve to locate the enzyme in the cellularmachinery involved in protein synthesis and folding, andreinforces the possible involvement of CK2 in theseprocesses [29, 33].

Acknowledgements

The authors are indebted to Dr. C.G. Proud (University ofKent at Canterbury, U.K.) for the generous gift of anti-eIF-2αand anti-eIF-2β antibodies, to Dr. F. Canals (ProteinChemistry Faculty/IBF, UAB) for N-terminal sequenceanalysis, and to Dr. S. Bartolomé (LAFEAL, UAB) forassistance in gel scanning and figure presentation. This workwas supported in part by grants PB95-0610 from DGICYT(Spain) and BMH4-CT96-0047 (BIOMED 2, E.U.).

References

1. Pinna LA: Casein kinase 2: An ‘eminence grise’ in cellular regulation?Biochim Biophys Acta 1054: 267–284, 1990

2. Issinger O-G: Casein kinases: Pleiotropic mediators of cellularregulation. Pharmac Ther 59: 1–30, 1993

3. Seldin DC, Leder P: Casein kinase IIα transgen-induced murinelymphoma: Relation to theileriosis in cattle. Science 267: 894–896, 1995

4. Gietz DR, Graham KC, Litchfield DW: Interactions between thesubunits of casein kinase II. J Biol Chem 270: 13017–13021, 1995

5. Filhol-Cochet O, Mackenbach PL, Cochet C, Chambaz EM: Caseinkinase 2 and the cell response to growth factors. Cell Mol Biol Res 40:529–537, 1994

6. Lüscher B, Litchfield DW: Biosynthesis of casein kinase II in lymphoidcells. Eur J Biochem 220: 521–526, 1994

7. Stigare J, Buddelmeijer N, Egyhazi E: A majority of casein kinase II αsubunit is tightly bound to intranuclear components but not to the βsubunit. Mol Cell Biochem 129: 77–85, 1993

8. Hagemann C, Kalmes A, Wixler V, Schuster T, Rapp UR: The regulatorysubunit of protein kinase CK2 is a specific A-Raf activator. FEBS Lett403: 200–202, 1997

9. Chen K, Li D, Krebs EG, Cooper JA: The casein kinase II beta subunitbinds to Mos and inhibits Mos activity. Mol Cell Biol 4: 1904–1912, 1997

10. Boldyreff B, Issinger O-G: A-Raf is a new interacting partner of proteinkinase CK2β subunit. FEBS Lett 403: 197–199, 1997

11. Hériché JK, Lebrin F, Rabilloud T, Leroy D, Chambaz EM, Goldberg Y:Regulation of protein phosphatase 2A by direct interaction with caseinkinase 2α. Science 276: 952–955, 1997

12. Myata Y, Yahara I: The 90-kDa heat shock protein, HSP90, binds andprotects casein kinase II from self-aggregation and enhances its kinaseactivity. J Biol Chem 267: 7042–7047, 1992

13. Bojanowski K, Filhol O, Cochet C, Chambaz EM, Larsen A-K: DNAtopoisomerase II and casein kinase II associate in a molecular complexthat is catalytically active. J Biol Chem 268: 22920–22926, 1993

14. Filhol O, Baudier J, Delphin C, Loue-Mackenbach P, Chambaz EM,Cochet C: Casein kinase II and the tumor suppressor protein p53associate in a molecular complex that is negatively regulated upon p53phosphorylation. J Biol Chem 267: 20577–20583, 1992

15. Li D, Dobrowolska G, Krebs EG: The physical association of caseinkinase 2 with nucleolin. J Biol Chem 271: 15662–15668, 1996

16. Götz C, Koenig MG, Issinger O-G, Montenarh M: A casein-kinase-2-related protein kinase is tightly associated with the largeantigen of simian virus 40. FEBS Lett 233: 327–334, 1995

17. Karki S, Tokito M, Holzbaur ELF: Casein kinase II binds and phos-phorylates cytoplasmic dynein. J Biol Chem 272: 5887–5891, 1997

18. Li D, Meier T, Dobrowolska G, Krebs EG: Specific interaction betweencasein kinase 2 and the nucleolar protein Nopp 140. J Biol Chem 272:3773–3779, 1997

19. Willert K, Brink M, Wodarz A, Varmus H, Nusse R: Casein kinase 2associates with and phosphorylates Dishevelled. EMBO J 16: 3089–3096, 1997

20. Molina E, Plana M, Itarte E: Heterogeneity of rat liver cytosol caseinkinase 2. Biochem J 277: 811–818, 1991

21. Gil C, Plana M, Riera M, Itarte E: Rat liver pp49, a protein that formscomplexes with protein kinase CK2, is composed of the β and the γsubunits of translation initiation factor eIF-2. Biochem Biophys ResCommun 225: 1052–1057, 1996

22. Trujillo R, Miró F, Plana M, José M, Bollen M, Stalmans W, Itarte E:Substrates for protein kinase CK2 in insulin receptor preparations fromrat liver membranes: Identification of a 210-kDa protein substrate as thedimeric form of endoplasmin. Arch Biochem Biophys 344: 18–28, 1997

23. Singh TJ, Huang K-P: Glycogen synthase (casein) kinase-1: Tissuedistribution and subcellular localization. FEBS Lett 190: 84–88, 1985

24. Hei YH, Chen X, Diamond J, McNeill JH: Distribution of MAP kinase,S6 kinase, and casein kinase II in rat tissues: Activation by insulin inspleen. Biochem Cell Biol 72: 49–53, 1994

25. Duncan R, Hershey JWB: Identification and quantitation of levels ofprotein synthesis initiation factors in crude HeLa cell lysates bytwo-dimensional polyacrylamide gel electrophoresis. J Biol Chem 258:7228–7235, 1983

26. Oldfield S, Jones BL, Tanton D, Proud CG: Use of monoclonal antibodiesto study the estructure and function of eukaryotic protein synthesisinitiation factor eIF-2B. Eur J Biochem 221: 399–410, 1994

27. Price NT, Nakielny SF, Clark SJ, Proud CG: The two forms of the β-subunit of initiation factor-2 from reticolocyte lysates arise fromproteolytic degradation. Biochim Biophys Acta 1008: 177–182, 1989

28. Wearsch P, Nicchitta CV: Interaction of endoplasmic reticulumchaperone GRP94 with peptide substrates is adenin nucleotide-independent. J Biol Chem 272: 5152–5156, 1997

29. Harada S, Karino A, Shimoyama Y, Shamsa F, Ohtuski K: Identificationof glycyrrhizin-binding protein kinase as casein kinase II andcharacterization of its associated phosphate acceptors in mouse liver.Biochem Biophys Res Commun 227: 102–109, 1996

30. Baldan B, Navazio L, Firso A, Mariani P, Meggio F: Plant calreticulin isspecifically and efficiently phosphorylated by protein kinase CK2.Biochem Biophys Res Commun 221: 498–502, 1996

31. Chen N, Davis AT, Canbulat E.C. Liu Y, Goueli S, McKenzie BA,Eccleston ED, Ahmed K, Holtzman J: Evidence that casein kinase 2phosphorylates hepatic microsomal calcium-binding proteins 1 and 2but not 3. Biochemistry 35: 8299–8306, 1996

32. Michalak M, Milner R, Burns K, Opas M: Calreticulin. Biochem J 285:681–692, 1992

33. Proud CG: Protein phosphorylation in translational control. Curr TopCell Reg 32: 243–369, 1992