Protein Kinase CK2 Interacts with the Splicing Factor Hprp3p

S Lehnert, C Go¨ tz, S Kartarius, B Scha¨ fer and M Montenarh

Universita¨t des Saarlandes, Medizinische Biochemie und Molekularbiologie, Homburg, Germany

Numerous signalling pathways in cells are influenced by acidophilic targets, but they belong to different groups of the ubiquitous Ser/Thr protein kinase CK2. Protein kinases. CK2 belongs to the CMCG group whereas CK1 kinase CK2 is composed of two regulatory b-subunits belongs to the OPK XII group. A peculiar feature of and two catalytic a-ora0-subunits. Several of the known CK2 is that it not only acts as a phosphotransferase but CK2 substrates are proteins known to regulate transcrip- that it can also influence the biological activity of some tional events. Here, we describe that protein kinase CK2 proteins, like topoisomerase I, by pure interaction (Pinna, interacts with the splicing factor hPrp3p, which is 2002; Litchfield, 2003). Furthermore, there is ample important for the assembly of the spliceosome. In a two- evidence that CK2 not only appears as a tetrameric hybrid screen hPrp3p is exclusively bound to the catalytic holoenzyme in the cell but also as individual subunits. a-ora0-subunits of CK2 but not to the regulatory b- Free a-aswellasfreeb-subunits have been found in the subunit. The interaction was confirmed by coimmunopre- cell provided with important individual functions (Stalter cipitation experiments in vitro and in vivo. Moreover, both et al., 1994; Guerra et al., 1999). proteins colocalized in nuclear speckles which is typical Protein kinase CK2 is a highly conserved enzyme. It for splicing factor compartments within the nucleus. has been found in a wide variety of organisms from Phosphorylation experiments revealed that hPrp3p is also yeast to man and in all cells and tissues with a different a substrate of protein kinase CK2. The main phosphor- expression level. Although its true function is not yet ylation site was mapped to C-terminal residues. In vitro known, life without CK2 is not possible. These and in vivo splicing assays showed that the splicing activity observations together with the fact that CK2 activity is significantly influenced by the CK2–hPrp3p interaction. is enhanced in proliferating cells, underline the impor- Thus, these data showed that CK2 is involved in the tance of CK2 for cell growth. Among the more than 300 regulation of RNA processing. substrates discovered to this date, are a lot of proteins Oncogene (2008) 27, 2390–2400; doi:10.1038/sj.onc.1210882; that are implicated in cell cycle control, DNA replica- published online 19 November 2007 tion/repair, proliferation and transcription/translation (Meggio and Pinna, 2003). The C-terminal domain Keywords: protein kinase CK2; hPrp3p; splicing factor; (CTD) of the largest RNA polymerase II subunit phosphorylation; protein–protein interaction (Cabrejos et al., 2004), the TFIIF-dependent CTD phosphatase 1 (FCP1) (Palancade et al., 2002) and cyclin H (Faust et al., 2002; Schneider et al., 2002; Krempler et al., 2005) belong to the substrates that are known to regulate transcriptional events. Cyclin H is Introduction phosphorylated at threonine 315 by CK2 and this phosphorylation is needed for full activity of the cdk7/ More than 40 years ago Burnett and Kennedy isolated a cyclin H/Mat1 complex towards a CTD peptide of RNA phosphotransferase from liver extracts which did not fit polymerase II. The phosphorylation of the CTD is into the scheme of all other kinases found to that date necessary for the switch of the initiation to the (Burnett and Kennedy, 1954). It is a heterotetrameric elongation step of transcription. CK2 is also directly enzyme consisting of two catalytic a-ora0- and two involved in the splicing process as it phosphorylates regulatory b-subunits which prefers serine or threonine RNPS1, which is a splicing regulator whose activator residues within acidic consensus sequences as targets, and function is, at least in part, controlled by CK2 which can use both, guanosine 50-triphosphate (GTP) phosphorylation (Trembley et al., 2005). and adenosine 50-triphosphate (ATP) as phosphate hPrp3p is a human pre-mRNA splicing factor that, in donors. Both, CK2 and CK1 prefer to phosphorylate a mutant form, leads to autosomal dominant Retinitis pigmentosa (Chakarova et al., 2002). hPrp3p presumably plays an important role in recruiting other splicing Correspondence: M Montenarh, Universita¨ t des Saarlandes, Medizi- factors like hPrp4p to the U4/U6 small ribonucleopro- nische Biochemie und Molekularbiologie, Geba¨ ude 44, D – 66421 tein and docking the U4/U6/U5 tri-snRNP to the Homburg, Germany. prespliceosome (Liu et al., 2006). In the present study E-mail: [email protected] Received 2 October 2006; revised 28 September 2007; accepted 1 October we describe that hPrp3p interacts with protein kinase 2007; published online 19 November 2007 CK2 in vitro and in vivo and that is phosphorylated by CK2 is implicated in RNA splicing S Lehnert et al 2391 this kinase. The catalytic subunits of CK2 are colocalized experiments. Immunoprecipitation was performed with with hPrp3p in nuclear speckles during the gap phases of an antibody against the Xpress—tag of recombinant the cell cycle. Furthermore, the splicing activity of the hPrp3p. After sodium dodecyl sulfate–polyacrylamide spliceosome is influenced by protein kinase CK2. gel electrophoresis (SDS–PAGE) followed by a western blot analysis, CK2 was detected with specific antisera. As shown in Figure 2 hPrp3p coimmunoprecipated with Results the catalytic subunits of CK2 (Figure 2a, lane 3) but not with the CK2 holoenzyme consisting of catalytic a-and The catalytic subunits of protein kinase CK2 interact with regulatory b-subunits (Figure 2b, lane 2). These results showed that hPrp3p indeed bound to the catalytic the splicing factor hPrp3p 0 By performing a two-hybrid screen with a library from a- and a -subunits of CK2 but it failed to bind to the human testis and the catalytic a-anda0-subunits of CK2 holoenzyme. protein kinase CK2 as a bait, we found several For further analyses we generated a polyclonal serum interacting partners whose DNA was subsequently in rabbits (no. 771) against amino acids 608–618 of sequenced. Two clones out of 20 corresponded to hPrp3p as schematically shown in Figure 3a. The serum nucleotides 873–1092 (amino acids 291–360) of the was characterized by western blot analysis, immunopre- human splicing factor hPrp3p. The positive interaction cipitation and immunofluorescence. The antiserum was of one of these clones with both catalytic subunits (4and able to recognize a protein of about 90 kDa and its 5) in yeast is shown in Figure 1. Control experiments degradation products in HeLa cell extract (Figure 3b). with empty vectors and with CK2a and CK2a0 As shown in Figure 3c it is also able to immunopreci- constructs together with empty vectors were all negative. pitate the hPrp3p protein from HeLa cell extract To verify the interaction, hPrp3p was in vitro whereas the preimmune serum did not detect a translated and labelled with 35S-methionine, and the corresponding protein. Furthermore, immunofluores- CK2 holoenzyme and catalytic a-subunits were bacte- cence analysis revealed that it detects the protein in rially expressed and purified according to standard nuclear speckles of HeLa cells (Figure 3D, a–c). The protocols (Guerra et al., 1998). The CK2 subunits or the antigen–antibody interaction was successfully competed holoenzyme were mixed with the in vitro translated by the addition of increasing amounts of the peptide hPrp3p and incubated for 2–3 h. An interaction between used for immunization. Thus, the antiserum was suitable these proteins was analysed by coimmunoprecipitation for the following in vitro and in vivo experiments. In a next experiment we analysed the interaction of CK2a and hPrp3p in mammalian cells. We used human

1 123 hPrp3p

CK2α

CK2α‘

2 3 123 CK2α

CK2β

4 5 Figure 2 Co-immunoprecipitation of recombinant CK2 and hPrp3p. (a) hPrp3p was in vitro translated and incubated either with the recombinant CK2a- or CK2a0-subunit. hPrp3p was immunoprecipitated with an antibody, which was directed against the Xpress-tag of hPrp3p and 35S-labelled hPrp3p was detected by Figure 1 Two hybrid interaction between CK2 catalytic subunits fluorography. CK2 subunits were detected in a western blot and the hPrp3p clone (nt 873–1092). Yeast cells were transformed analysis with specific sera. Lane 1: input control; lane 2: control with the parental pGBKT7 plasmid or the same plasmid containing immunoprecipitation without hPrp3p; lane 3: immunoprecipitation one of the catalytic CK2aÀsubunits and the parental pAct2 with hPrp3p. (b) hPrp3p was in vitro translated and incubated with plasmid or the same plasmid containing the hPrp3p clone from recombinant CK2 holoenzyme. hPrp3p was immunoprecipitated nucleotide 873–1092. Yeast cells were grown under stringent with an antibody directed against its tag. CK2a- and CK2b- conditions for 3 days. (1) Parental pGBKT7 and parental pAct2. subunits were detected with specific sera in a western blot analysis. (2) pGBKT7-CK2a and parental pAct2, (3) pGBKT7-CK2 a0 and Lane 1: control immunoprecipitation without hPrp3p; lane 2: parental pAct2, (4) pGBKT7-CK2a and pAct2- hPrp3p (nt 873– immunoprecipitation with hPrp3p and CK2 holoenzyme; lane 3: 1092), (5) pGBKT7-CK2a0 and pAct2- hPrp3p (nt 873–1092). input control.

Oncogene CK2 is implicated in RNA splicing S Lehnert et al 2392 608 - 618

DAPI hPrp3p (#771) 1 683 a TSNTKGDDDEE

[kDa] 100 200 300 µg

116 b 90

c 49

hPrp3p

Figure 3 Characterization of the rabbit polyclonal serum 771 against hPrp3p. Rabbits were immunized with a peptide of amino acid 608–618 of human hPrp3p. The antiserum was characterized by western blot analysis, immunoprecipitation and immunofluorescence. (A) Localization of the immunizing peptide on the polypeptide chain of hPrp3p and its amino acid sequence. (B) Increasing amounts of total extract from HeLa cells were separated in a 12.5% SDS polyacrylamide gel, transferred to a PVDF membrane and incubated with serum 771 in a dilution of 1:500. After binding the secondary POD coupled antibody signals were visualized with the Lumilight system of Roche Diagnostics. (C) Total extract from HeLa cells was subjected to an immunoprecipitation procedure. The immunoprecipitated protein was separated through SDS–PAGE and transferred to a PVDF membrane. The detection of hPrp3p was performed as described in B. (a) control precipitation with preimmune serum, (b) immunoprecipitation with serum 771. (D) HeLa cells were grown on coverslips, fixed with formaldehyde and permeabilized using Triton-X100. Cells were incubated for 1 h at room temperature with serum 771 (a), with serum 771 which was preincubated with 0.5 mM immunizing peptide (b) or serum 771 which was preincubated with 2mM peptide (c). Subsequently, binding was visualized with a TRITC-coupled secondary goat anti rabbit antibody. In parallel nuclei were stained with DAPI. Magnification: 400 Â .

HeLa cells that turned out to have detectable amounts proteins during the gap phases of the cell cycle showing of hPrp3p. Cells were synchronized with hydroxyurea the strongest signals during the G1 phase. Thus, this data and hPrp3p immunoprecipitated from cell extracts of supports the results of the coimmunoprecipitation G1-, S- and G2/M-phase cells using serum 771. Figure 4a analysis and confirmed the cell-cycle-dependent interac- shows the cytofluorimetry analysis of the synchronized tion between the catalytic a-subunit of CK2 and hPrp3p. cells, which demonstrates that we indeed used G1-, S- and G2/M-phase cells. According to SDS–PAGE and western blot analysis, hPrp3p and the CK2a-subunit hPrp3p is phosphorylated by the CK2 holoenzyme were present in all cell extracts (Figure 4b, lane 4) So far, we have shown that hPrp3p is a binding partner throughout the cell cycle in similar amounts. However, of the CK2a- and a0-subunits but not of the holoenzyme. we only succeeded in coimmunoprecipitating detectable In order to test whether hPrp3p is a phosphoprotein amounts of the CK2a-subunit from cell extracts of both HeLa cells were labelled with 33P-orthophosphoric acid gap phases (Figure 4, lane 2). Lane 1 shows control for 4h. Cells were harvested and proteins subjected to an precipitates without serum 771, lane 3 shows an aliquot immunoprecipitation using the hPrp3p-specific antiser- of the supernatant after the immunoprecipitation um 771. After SDS–PAGE and autoradiography of the demonstrating that only a fraction of CK2a was immunoprecipitated proteins hPrp3p was detected as a coimmunoprecipitated. Thus, we have to conclude that phosphoprotein (Figure 6, lane 2). In the next step we the interaction between the catalytic subunit of CK2 and analysed whether phosphorylation of hPrp3p is due to hPrp3p exists in vivo, but only during the gap phases of CK2. Thus, in a parallel experiment we treated cells with the cell cycle. the CK2 inhibitor tetrabromobenzotriazole (TBB) dur- To show the interaction in vivo by another method we ing labelling. Figure 6 (lane 3) clearly demonstrates that subjected synchronized HeLa cells to an immunofluor- the phosphorylation of hPrp3p was strongly reduced by escence study with hPrp3p-specific and CK2a-specific this treatment. Therefore, we demonstrated that hPrp3p antibodies (Figure 5). In cells from all cell cycle phases is a phosphoprotein in vivo and that the phosphorylation hPrp3p was predominantly found in nuclear speckles, is probably done by protein kinase CK2. The following which is a typical distribution for splicing factors. In in vitro experiments served to further characterize the contrast, CK2a was found all over the nucleus and phosphorylation. somewhat weaker in the cytosol. When merging the two As the CK2 holoenzyme, consisting of catalytic and pictures we observed a strong colocalization of both regulatory subunits, as well as the individual a-subunits

Oncogene CK2 is implicated in RNA splicing S Lehnert et al 2393

G1 S G2/M

420 420 420 G G2/M G G2/M G G2/M S S S 280 280 280 Counts 140 Counts 140 140

0 0 0 0 200 400 600 0 200 400 600 0 200 400 600 FL2-A FL2-A FL2-A

G1 S G2/M

1234 12 341234 _ 117 _ 90 hprp3p

_ 49 CK2 α

Figure 4 Cell-cycle-dependent coimmunoprecipitation of hPrp3p and CK2a. HeLa cells were synchronized using hydroxyurea and harvested as G1,S-orG2/M-cells. (a) Cytofluorimetry (b) Nuclear extracts were treated with the hPrp3p-specific serum 771, separated by SDS–PAGE and analysed with the monoclonal CK2a antibody 1A5 and with the hPrp3p-specific antibody 4E3. Lane 1: control precipitate without serum 771; lane 2: immunoprecipitate with serum 771; lane 3: aliquot of the supernatant after immunoprecipitation; lane 4: aliquot of the nuclear extract.

hPrp3p CK2 α merge

G2/M

Figure 5 Immunofluorescence analysis of hPrp3p and CK2a in synchronized Hela cells. HeLa cells were synchronized using hydroxyurea. Cells of the corresponding cell cycle phases were fixed with formaldehyde and permeabilized with Triton-X100. Cells were stained with antibodies against hPrp3p (771) or CK2a (1A5) and the corresponding FITC- or TRITC-labelled secondary antibodies. Magnification: 400 Â . are known to phosphorylate proteins, we performed (Figure 7b) revealed that at the latest after 30 min the in vitro phosphorylation assays with both enzymes. hPrp3p protein was saturated with phosphate. Thus, hPrp3p was expressed as His-tagged protein in bacterial we chose this time point to determine the phosphate cells and purified by affinity chromatography with a incorporation into hPrp3p by CK2. We performed six Ni2 þ -NTA chelate resin. The purified hPrp3p was used independent experiments and determined the incorpora- for the phosphorylation with CK2. In Figure 7a (lane 2) tion rate to be 0.39±0.09 mol phosphate into 1 mol of we show that hPrp3p is phosphorylated by the CK2 hPrp3p. In order to compare the rate of phosphate holoenzyme. Time course studies of the phosphorylation incorporated in hPrp3p with a well-known and excellent

Oncogene CK2 is implicated in RNA splicing S Lehnert et al 2394 substrate of CK2 we used a C-terminal fragment vectors, then expressed and purified the protein fragments. of nucleolin (amino acids 269–710) (Schneider and Figure 7c shows the result of the phosphorylation by the Issinger, 1988). In this case we found an incorpora- CK2 holoenzyme. The N-terminal and the middle frag- tion rate of 2 mol phosphate into 1 mol of the nucleolin ment were not or only weakly phosphorylated (lane 1 and fragment (not shown). Thus, in comparison to nucleolin 2). In contrast, the C-terminal fragment (amino acids 457– hPrp3p seems to be a poor substrate. As hPrp3p 682) was significantly phosphorylated (Figure 7c, lane 3). obviously binds to the catalytic subunits of CK2 This C-terminal fragment harbours six out of thirteen independently of the regulatory b-subunit we were consensus sequences of hPrp3p for CK2. Therefore, we interested if the a-subunit is also able to phosphorylate have to conclude that hPrp3p is predominantly phos- the protein. However, the catalytic a-subunit alone was phorylated at one or several sites within the C-terminus. unable to phosphorylate hPrp3p (Figure 7a, lane 3 and Figure 7d shows a western blot analysis of aliquots of the 4) although the enzyme phosphorylated the CK2a hPrp3p fragments with an antibody against the His-tag. substrate mdm2 (Guerra et al., 1997) (Figure 7a, lane 5). In order to narrow down the phosphorylation site Next, we wanted to map the phosphorylation site on the more exactly we further used a peptide library contain- polypeptide chain of hPrp3p. We started to subclone three ing all the putative CK2 phosphorylation sites. The fragments of hPrp3p (H1–H3) into bacterial expression peptides consist of 11 amino acids with the serine or threonine to be phosphorylated in the middle of the sequence (bold letters in the table of Figure 8). The 32 123 peptide library was incubated with CK2 and PgATP; Figure 8 shows the result of the autoradiography. We reproducibly observed several spots which were present in the library but not in the control experiment without Figure 6 In vivo phosphorylation of hPrp3p. HeLa cells were CK2. These spots represent peptides containing serine labelled with 33P-orthophosphoric acid and treated with the CK2 or threonine residues S81 (peptide 2), T452 (peptide 7), inhibitor tetrabromobenzotriazole (TBB) or left untreated. La- S619 (peptide 10) and S642 (peptide 12). Besides the belled proteins were immunoprecipitated with the hPrp3p-specific serine residue at position 81, which was only weakly serum 771 and separated by SDS–PAGE. Labelled hPrp3p was detected by autoradiography. Lane 1: control precipitate without phosphorylated, and threonine 452, the other sites are antiserum, lane 2: immunoprecipitate of untreated cells, lane 3: located in the C-terminal fragment of hPrp3p (see immunoprecipitate from cells treated with 50 mM TBB. Figure 7).

140 1234 5 120 117 _ _ mdm2 100 90 hprp3p 80 60 49 _ 40 35 _ 20 relative CK2 activity [%] CK2 β 26 _ 0 0102030 t [min]

c 123 123

35 _ H3 26 _ CK2 β 19 _

H1 H2 H3 1 228 456 682 Figure 7 In vitro phosphorylation of hPrp3p. (a) Recombinant hPrp3p was subjected to a phosphorylation reaction using the CK2 holoenzyme (lane 2) or the CK2a subunit (lane 4) and 32PgATP. Proteins were analysed by SDS–PAGE and autoradiography. Lane 1: CK2 holoenzyme; lane 3: CK2a subunit; lane 5: CK2a subunit and mdm2. (b) Time course study of the phosphorylation of hPrp3p. hPrp3p was phosphorylated using the CK2 holoenzyme as in A. After 1, 2, 3, 5, 10, 15, 20 and 30 min the reaction was stopped by adding sample buffer. After SDS–PAGE and autoradiography the bands were subjected to a densitometric analysis. The relative intensity of the bands at the different time points is shown in the graph. The insert shows the autoradiography. (c) Fragments of hPrp3p H1–H3 were subjected to a phosphorylation reaction using the CK2 holoenzyme and 32PgATP and analysed by SDS–PAGE and autoradiography. Lane 1: H1 (hPrp3p 1–228), lane 2: H1 (hPrp3p 229–456); lane 3: H3 (hPrp3p 457–682); lane c: control with CK2 holoenzyme alone. (d) In parallel, aliquots of the fragments were analysed in a western blot using a His-tag speciﬁc antibody. Lane 1: H1, lane 2: H2, lane 3: H3. hPrp3p fragments are marked with an asterisk.

Oncogene CK2 is implicated in RNA splicing S Lehnert et al 2395 The splicing activity of the spliceosome is inﬂuenced by CK2 and hPrp3p. hPrp3p is a splicing factor involved in the CK2/hPrp3p interaction the processing of pre-mRNA. Whereby, we asked if the Next, we analysed the functional consequences of the splicing activity of hPrp3p is somehow compromised by interaction between the catalytic a-anda0-subunits of the interaction with CK2a. We treated HeLa cells with the CK2 inhibitor TBB and analysed the effect on the splicing activity of the nuclear extract. By the applica- control tion of 25 or 50 mM TBB the activity of CK2 was downregulated to 40 or 20% of the original activity, 1 2 3 4 5 6 7 8 9 10 11 12 13 respectively (Figure 9a). The expression of the catalytic + CK2 a-subunit remained the same during the treatment (Figure 9b). As pre-mRNA for the splicing assay we peptide No. amino acids sequence site used an E1A minigene which is under normal conditions 1 1 - 11 MALSKRELDEL serine 4 spliced to a 13S, 12S, 10S and 9S mRNA (Figure 9c, (0) 2 76 - 86 SRHSKSSSDRS serine 81 and Figure 9d). In the presence of TBB (50 mM) the

3 171 - 181 KTPSSSQPERL serine 176 splicing pattern is changed in favour of the unspliced form of the pre-mRNA. Thus, the inhibition of protein 4 263 - 273 RTVDATGKEIE threonine 268 kinase CK2 obviously compromises the splicing process. 5 304 - 314 DMESNTFFDPR threonine 309 In another approach to speciﬁcally target protein 6 437 - 447 LGVYTKKEQK threonine 441 kinase CK2 we downregulated the CK2 kinase activity 7 447 - 457 KLRRQTRREAQ threonine 452 by the application of kinase-dead dominant negative mutants CK2a K68M and CK2a0 K69M (Penner et al., 8 489 - 499 AVQDPTKVEAH threonine 494 1997). We transfected HeLa cells with these two mutants. 9 606 - 616 EQTSNTKGDDD threonine 611 In order to enhance the amount of CK2 we transfected 10 614 - 624 DDDEESDEEAV serine 619 HeLa cells with wild-type CK2a or CK2a0 cDNAs. 11 632 - 642 LVWEGTAKDRS threonine 637 HeLa cells transfected with the parental pRc/CMV

12 638 - 648 AKDRSFGEMKF serine 642 vector served as control. We harvested the cells 72 h after the transfection. As shown in Figure 10a the exogenous 13 672 - 682 LALSESVLEST serine 677 CK2 subunits were detected in a western blot analysis. Figure 8 Phosphorylation of a peptide library of putative CK2 The kinase activity of CK2 was enhanced up to 120% phosphorylation sites of hPrp3p. Eleven amino acids long peptides when expressing the two a-subunits; by transfecting the were synthesized on a cellulose membrane. The membrane was kinase dead CK2a-orCKa0-subunit the CK2 activity incubated with CK2 holoenzyme ( þ CK2) and 32PgATP or with 32PgATP alone as a control. The ﬁgure shows the autoradiography was reduced to 40 or 55%, respectively (Figure 10b). of the peptide ﬁlters as well as the amino acid sequences and the We used the nuclear extracts of transfected and positions of the peptides. control cells to perform a splicing assay as described

120 100 0 µM TBB µ 25 µM TBB TBB [ M]0 25 50 80 50 µM TBB

60 CK2α 40

CK2-activity [%] GAPDH 20

TBB [µM] 0 25 50 10S E1A minigene 13S unspliced 12S, 9S 13S 10S 12S 13S 10S 12S 9S 10S 9S

Figure 9 In vivo splicing assay. HeLa cells were transfected with the reporter construct pCMV-E1A and left untreated or 24h after transfection incubated with the CK2-specific inhibitor TBB. Forty-eight hours after transfection cells were harvested. (a) CK2 activity was determined by the incorporation of 32P phosphate into the synthetic peptide RRRDDDSDDD. The figure shows the relative CK2 activity of cells treated with 25 or 50 mM TBB compared to a solvent control. (b) Western blot analysis of untreated (0) cells and cells treated with 25 or 50 mM TBB using specific sera for CK2a and GAPDH. (c) mRNA was isolated from HeLa cells that were either left untreated (0) or treated with TBB and subjected to a reverse transcriptase reaction. E1A cDNA was amplified with specific primers and analysed by agarose gel electrophoresis in a 2.5% agarose gel. After staining with ethidium bromide the different E1A splicing products were visualized. (d) Schematic drawing of the full length RNA and the different splice products.

Oncogene CK2 is implicated in RNA splicing S Lehnert et al 2396 a (-) CK2α CK2α‘ _ α-KD α

Myc-CK2α 13 S 12 S 49 _

CK2α 10 S _ α‘-KD α‘ 9 S 49 _ Myc-CK2α‘ CK2α‘ 35 _

b (-) CK2α´ CK2α 140

120 mock

100 CK2α 80 CK2α‘ 60

40 CK2 activity [%] 20

0 wild type kinase dead

Figure 10 Overexpression of catalytic CK2a-subunits. Cos 1 cells were transfected with myc-tagged CK2a or CK2a0 either in its wild- type (a and a0) form or in its kinase-dead forms (a-KD or a0-KD). Figure 11 In vivo splicing assay in cells transfected with CK2a- subunits. Cos 1 cells were transfected with the pCMV-E1A reporter Mock-transfected cells (À) served as a control. Forty-eight hours 0 after transfection cells were harvested. (a) Western blot analysis plasmid and simultaneously with a CK2a or a plasmid. Mock- using specific sera against CK2a or CK2a0 and a POD labelled transfected cells (À) served as a control. Forty-eight hours after secondary antibody. Signals were visualized by enhanced chemilu- transfection cells were harvested; mRNA was isolated and minescence. (b) CK2 activity was determined by the incorporation transcribed into cDNA. E1A cDNA was amplified using specific of 32P phosphate into the synthetic peptide RRRDDDSDDD. The primers in a standard PCR reaction. Amplification products were figure shows the relative CK2 activity of cells transfected with analysed by agarose gel electrophoresis in a 2.5% gel and visualized CK2a or CK2a0 compared to the mock-transfected control. by ethidium bromide. (a) Overexpression of wild-type CK2a or a0-subunit. (b) Overexpression of kinase-dead CK2a or CK2a0- subunit. above. Figure 11 presents the results of the splicing assay. Compared to the control (À) the splicing pattern CK2 namely hPrp3p. Moreover, in addition to being a changed by the overexpression of CK2 in different ways. binding partner, hPrp3p is also a substrate for the CK2 By the overexpression of the CK2a-subunits, which holoenzyme consisting of two regulatory and two enhance the kinase activity in the cell, the amount of the catalytic a-ora0 subunits, but not for the catalytic spliced forms increased; thus, the splicing activity of the subunits alone. Binding partners for the individual spliceosome is activated. This is even more pronounced subunits of CK2 were already detected earlier including in the presence of the CK2a0-subunit (Figure 11a). By hsp90, PP2a, Pim-1, CKIP-1, cyclin H, NAP1 and Egr-1 transfecting the kinase-dead mutants of CK2 (Figure (Jain et al., 1996; Li et al., 1999; Bosc et al., 2000; Faust 11b) the activity of the spliceosome is reduced compared et al., 2002; Messenger et al., 2002). Some of these to the control. The splicing pattern is changed in favour binding partners are also substrates for the CK2 kinase. of the unspliced form, which is especially observed in the Regarding the function, hPrp3p is a new type of binding presence of the kinase-dead CK2a-subunit. partner, because it does not belong to the class of Therefore, we have to conclude that the kinase activity proteins involved in cellular signalling, cell cycle control of CK2 influences the splicing capacity of the spliceo- or intracellular trafficking, instead hPrp3p is implicated some. Moreover, both catalytic subunits obviously play in splicing. Together with hPrp4p hPrp3p is associated different roles in the regulation of the splicing process. with the U4/U6 small nuclear ribonucleoprotein parti- cle, which is essential for the assembly of an active spliceosome (Gonzalez-Santos et al., 2002). These Discussion snRNAs recognize conserved sequences of the pre- mRNA and assemble into a catalytically active spliceo- In the present study we identified a new binding partner some that catalyses the cleavage/ligation reactions of for the catalytic a- and a0-subunits of protein kinase RNA splicing. Mapping the binding site for the

Oncogene CK2 is implicated in RNA splicing S Lehnert et al 2397 interaction between hPrp3p and hPrp4p revealed that However, this kinase does not phosphorylate the CTD the interaction domain resides in the central region of of RNA polymerase II. But there was a kinase activity hPrp3p, within the same region that was found to be associated with cdk11, which later on turned out to be responsible for its interaction with CK2a or CK2a0 in protein kinase CK2 (Trembley et al., 2003). Thus, this the two-hybrid screen. hPrp3p has an RNA binding data together with the data presented here show that activity at least in vitro and it is believed that it CK2 plays an important role in the regulation of assembles proteins like hPrp4p as well as snRNAs at transcription and posttranscriptional processes. Here, the spliceosome. Thus, one might speculate that CK2 is we add a particular role in RNA splicing. targeted to the spliceosome by its interaction with hPrp3p. This assumption is further supported by the finding that the RNA binding site seems to be located Material and methods within the C-terminus of hPrp3p, which is distinct from the CK2a and a0 binding site. Yeast-two-hybrid analysis As shown by 33P-phosphate labelling of cells hPrp3p is Yeast-two-hybrid screens against CK2a and CK2a0 were a phosphoprotein. Phosphorylation of hPrp3p is at least performed using the Matchmaker-System from Clontech partially inhibited by treating the cells with TBB a according to the manufacturer’s instructions. Plasmids used specific inhibitor of protein kinase CK2 (Sarno et al., for the screens were pGBKT7 that contained the respective CK2-subunits and a human testis cDNA library subcloned 2001). Furthermore, bacterially expressed hPrp3p was into pACT2. Transformed yeast cells were grown on single readily phosphorylated by CK2 in vitro. Incorporation dropout media missing both leucine and tryptophan only or analysis revealed that 0.4mol phosphate were incorpo- leucine, tryptophan, histidine and adenine. Individual clones rated into 1 mol hPrp3p protein which means that about were sequenced and identified by a database search. 40% of hPrp3p is phosphorylated by CK2. In comparison to hPrp3p, nucleolin is a better substrate that In vitro translation incorporated 2 mol phosphate per mol protein. The use hPrp3p was subcloned into the pRSET A plasmid containing an of different fragments of hPrp3p indicated that the N-terminal Xpress-tag (Invitrogen, Karlsruhe GmbH, Germany). C-terminus of hPrp3p seems to harbour the main In vitro transcription was performed using the ‘TNT T7 coupled phosphorylation site(s). This initial result was confirmed reticulocyte lysate system’ (Promega GmbH, Mannheim, by a peptide filter assay which revealed that two Germany) according to the manufacturer’s protocol. different sites in the C-terminus, one in the middle part and one in the N-terminal part can be phosphorylated Cell culture and in vivo labelling of cells with by CK2. 33P-orthophosphoric acid Although hPrp3p is associated with CK2a it is not a HeLa and cos 1 cells were maintained in Dulbecco’s modified substrate for the catalytic subunit alone. A prototype of Eagle’s medium (DMEM) supplemented with 10% fetal calf serum (FCS). Cells were grown to 75% confluence in 10 cm a CK2a substrate is calmodulin, which is not phos- dishes in a 5% CO2 atmosphere. phorylated by the holoenzyme (Meggio et al., 1994). For in vivo labelling of proteins with 33P-orthophosphoric Instead hPrp3p is readily phosphorylated by the CK2 acid subconfluent cells were washed three times with phos- holoenzyme. Thus, binding of the a-subunit has a phate free DMEM and subsequently incubated with this function, which is clearly different from the interaction medium for 30 min at 37 1C. After this time 33P-orthopho- with the holoenzyme. One might speculate that hPrp3p sphoric acid (100 mCi in 1 ml medium) was added to the cells targets the CK2a subunit to the spliceosome where it and labelling of the proteins was performed for 2 h. After this may phosphorylate other substrates. time cells were washed three times with ice cold PBS and RNSP1, which was originally described as a pre- harvested. mRNA splicing activator and which is a component of an active spliceosome is bound by CK2a. Furthermore, Synchronization of cells and cytofluorimetry it is phosphorylated at serine 53 (Trembley et al., 2005) Subconfluent asynchronous HeLa cells were treated with 4m M 1 by CK2 at least in vitro. This phosphorylation stimu- hydroxyurea at 37 C for 24h and were thus blocked at the G 1/ S transition of the cell cycle. After removal of hydroxyurea, lated the splicing activity of RNPS1. Similar to RNSP1, cells reentered the cell cycle as a synchronous population for here we showed that an inhibition of the hPrp3p 1–2 complete cell cycles. After 30 min, 4and 8 h cells were phosphorylation also leads to an inhibition of the harvested as G1, S and G2/M population, respectively. An splicing. Several proteins, which are implicated in aliquot of the cells was subjected to a cytofluorimetry analysis. transcription and posttranscriptional processes are For cytofluorimetry analysis cells were detached from the substrates for protein kinase CK2 including RNA culture dish by using a trypsin/EDTA solution in PBS and polymerase II (Payne et al., 1989), cyclin H (Schneider transferred to a centrifuge tube. Cells were sedimented at et al., 2002) and FCP1 (TFIIF-dependent CTD phos- 400 Â g and the pellet was resuspended in 200 ml PBS. Cells phatase) (Trembley et al., 2003). Interestingly, RNPS1 were fixed by adding vigorously 2 ml ice-cold 70% ethanol and 1 also interacts with the largest subunit of RNA incubation at À20 C for 30 min. The cell suspension was centrifuged at 400 Â g and the sediment was resuspended in polymerase II. It was shown that the phosphorylation 800 ml PBS. Finally, cells were treated with 100 ml RNase A of the CTD of RNA polymerase also plays a role in (1 mg mlÀ1) and 100 ml propidium iodide (400 mgmlÀ1) to stain posttranscriptional mRNA processing (Proudfoot et al., the nuclei. Cells were analysed using a FACScan II cyto- 2002). The RNA polymerase multiprotein complex also fluorimeter (excitation: 488 nm, emission: 630 nm) and the harbours cdk11, which is a cyclin-dependent kinase. CellQuest software from Becton Dickinson.

Oncogene CK2 is implicated in RNA splicing S Lehnert et al 2398 Preparation of cell extracts pH 8.0, 7 mM 2-mercaptoethanol, 1 mM phenylmethyl sulfo- nylchloride (PMSF) and 1.5 M NaCl). Proteins were extracted Cells were harvested, washed 3 Â with phosphate buffered overnight, pooled with the supernatant from the first step, and saline (PBS), pH 7.4, and resuspended in lysis buffer (100 mM dialysed against buffer P300 (20 mM Tris-HCl, pH 8.0, 7 mM 2- Tris-HCl, pH 9.0, 100 mM NaCl, 0.5% (v/v) NP40, 1% mercaptoethanol, 1 mM PMSF and 300 mM NaCl). The lysate Trasylol). Proteins were extracted for 1 h on ice. Cell debris was loaded onto a P11 column preequilibrated with buffer P300 was eliminated by centrifugation (4 1C, 30 min, 13.000 Â g). and eluted with a linear gradient from 0.3 to 1.5 M NaCl. Fractions containing active holoenzyme were dialysed against Antibodies and western blot analysis buffer P300, concentrated and loaded onto a Superose-6 column For the detection of protein kinase CK2 we used polyclonal (Pharmacia, GE Healthcare, Freiburg, Germany). After gel sera 26 (a-subunit), 30 (a0-subunit) and 269 (b-subunit) (Faust filtration the peak fractions were collected. et al., 1999; Go¨ tz et al., 2005). Polyclonal rabbit serum 771 was raised against a peptide corresponding to amino acids 608–618 Expression and purification of bacterially expressed proteins (CTSNTKGDDDEE) of the hPrp3p protein. Alternatively, we Full-length hPrp3p, an N-terminal (amino acids 1–228), a used the commercially available antibody 4E3 from Biozol middle (amino acids 229–456) and a C-terminal fragment Diagnostica, Eching, Germany. For the detection of GAPDH (amino acids 457–682) of hPrp3p were subcloned into bacterial as housekeeping protein we used a polyclonal serum (Santa expression plasmid pQE30 containing an N-terminal histidine- Cruz, Heidelberg, Germany). In vitro translated hPrp3p was tag, transformed into E. coli strain M15. Clones were incubated detected using the mouse monoclonal Xpress-tag antibody overnight in 50 ml LB-medium. One litter LB-medium was SK2. The preparation of cell extracts, SDS–PAGE, and inoculated with the overnight-culture, grown to early log phase blotting procedure were described earlier (Schuster et al., and induced with 1 mM isopropyl-b-D-thiogalactoside at 30 1C 1999). For detection we used the Lumilight system (Roche for 6 h. Cells were harvested by centrifugation and resuspended Diagnostics). in 6 M guanidine hydrochloride, 0.1 M sodium phosphate, pH 8.0 and lysed at 4 1C overnight. The lysate was cleared by Immunofluorescence centrifugation and loaded onto a prewashed Ni2 þ -chelate Cells were grown on coverslips until they were 50–70% agarose column and incubated for 1 h at room temperature. confluent. Cells were fixed in 2% formaldehyde in PBS, pH The column was washed with 10 volumes of lysis buffer, 7.4, for 15 min at 20 1C and then washed with PBS, pH 7.4, for followed by 10 volumes of lysis buffer, pH 6.0, and lysis buffer, 3 Â 10 min. Cells were permeabilized with 0.5% Triton X-100. pH 8.0 containing 20 mM imidazole. Proteins were eluted with Cells were washed again with PBS three times for 10 min and lysis buffer containing 300 mM imidazole and subsequently then blocked with PBS containing 10% bovine serum albumin dialysed overnight against dialysis buffer A (20 mM Tris-HCl, (BSA), then incubated with a primary antibody in the pH 7.5, 100 mM KCl, 5 mM MgCl2, 0.1% Tween 20). appropriate concentration for 1 h at room temperature in a humidified chamber. Cells were washed and then incubated phosphorylation with the secondary antibody (FITC- or TRITC-conjugated) at In vitro room temperature for 1 h in a dark, humidified chamber. To analyse the CK2 activity in cell extracts 30 mg of the protein extracts were mixed with kinase buffer (50 mM Tris-HCl, pH 7.5, Finally, cells were washed again with PBS (4 Â 10 min). The coverslips were fixed with a drop of mounting media and 100 mM NaCl, 10 mM MgCl2,1mM DTT) to a total volume of analysed under a fluorescence microscope. 20 ml and incubated with the synthetic peptide RRRDDDSDDD (Kuenzel and Krebs, 1985) by adding 30 ml assay buffer (25 mM Tris-HCl, pH 8.5, 150 mM NaCl, 5 mM MgCl2,1mM DTT, Immunoprecipitation 50 mM ATP, 0.19 mM substrate peptide, 10 mCi 32PgATP/500 ml). For immunoprecipitation we used the polyclonal hPrp3p- The phosphorylation reaction was allowed to take place for 5 min serum (serum 771) and a monoclonal CK2a antibody (1A5). A at 37 1C. The mix was spotted onto a P81 ion exchange paper and protein A/G-sepharose mixture was preincubated for 1 h with washed three times with 85 mM H PO . After treatment with 1 3 4 30 ml of the serum or with 10 ml(1mg mlÀ ) of the monoclonal ethanol the paper was dried, and incorporated radioactivity was anti-CK2a antibody and washed three times with PBS, pH 7.4. determined in a scintillation counter. Four milligrams cell extract or 10 mg purified proteins was For the analysis of the phosphorylation of hPrp3p preincubated with a mixture of protein A- and protein recombinant hPrp3p proteins were mixed with recombinant G-sepharose to remove unspecifically bound proteins. The protein kinase CK2 in 20 ml kinase buffer (50 mM Tris-HCl, pH supernatant was applied to the preincubated sepharose– 7.5, 150 mM NaCl, 5 mM MgCl2,1mM DTT). The phosphor- antibody matrix and incubated for 1 h. The supernatant was ylation reaction was started by the addition of 32PgATP. For removed and the antibody-matrix washed five times with studying the phosphate incorporation we used 50 mM 32PgATP RIPA buffer (50 mM Tris/HCl, pH 8.0, 150 mM NaCl, 0.5% (1000–2000 cpm pmolÀ1). After 30 min at 37 1C the reaction sodium deoxycholate, 1% Triton X-100, 0.1% sodium dodecyl was stopped by the addition of 10 ml sample buffer. Proteins sulfate). The immune complex was subjected to SDS–PAGE, were separated in an SDS polyacrylamide gel and visualized by followed by western blotting. autoradiography. The incorporated radioactivity was mea- sured from the excised bands by scintillation counting. Purification of CK2 holoenzyme Undecapeptides representing the putative CK2 phosphor- Recombinant CK2 holoenzyme cloned in a bicistronic vector ylation sites of the hPrp3p sequence were synthesized and (Shi et al., 1994) and the three different subunits (a, a0 and b in immobilized on a cellulose membrane according to the Spot’s pT7-7) of the protein kinase CK2 were expressed in Escherichia method with a partially automated synthesizer (Abimed Auto- coli BL21 (DE3) and the proteins were purified according to a Spot Robot ASP 222) as recommended by the manufacturer protocol published by Grankowski et al. (1991). The bacterial (Abimed GmbH, Langenfeld, Germany). The membrane was pellet was lysed and after incubation on ice for 10 min, the lysate equilibrated overnight in kinase buffer supplemented with 1% was sonicated two times and centrifuged (10 000 Â g,step1). BSA. For phosphorylation of the peptides the membrane was The pellet was resuspended in buffer P1500 (20 mM Tris-HCl, incubated in 2 ml kinase buffer with recombinant CK2 in the

Oncogene CK2 is implicated in RNA splicing S Lehnert et al 2399 presence of 32PgATP. After 30 min at 37 1C the membrane was pCMV-E1A was transfected into HeLa cells. After transfec- washed three times with 1 M NaCl and subsequently with a tion cells were either treated with TBB or transfected with the urea-containing buffer (8 M urea, 1% sodium dodecyl sulfate, kinase dead CK2a-ora0- constructs (Penner et al., 1997). 0.5% 2-mercaptoethanol) to remove any bound proteins. Forty-eight hours after transfection of the reporter plasmid Finally, the membrane was washed with ethanol and dried. total RNAs were extracted and subjected to a Reverse Phosphorylated peptides were visualized by autoradiography. transcription–PCR reaction using exon 1- and 2-primers. PCR products were then analysed on a 2.5% agarose gel. Inhibition of protein kinase CK2 The activity of protein kinase CK2 in vivo was inhibited by the Acknowledgements specific inhibitor TBB. TBB was added to the cell-culture medium in concentrations of 25 or 50 mM. This study was supported by HOMFOR B 2004/11 to CG. We thank David Litchfield for providing us with the kinase dead In vivo splicing-assay CK2 mutants and Prof Ariga, Hokkaido University, Sapporo, In vivo splicing-assays were performed according to Maita Japan for the E1A-construct. Proofreading of the manuscript et al. (2004). Briefly, an adenovirus E1A reporter plasmid by Mrs. Jane C Crofts was greatly appreciated.

References

Bosc DG, Graham KC, Saulnier RB, Zhang CJ, Prober D, Gietz RD Li M, Strand D, Krehan A, Pyerin W, Heid H, Neumann B et al. et al. (2000). Identification and characterization of CKIP-1, a novel (1999). Casein kinase 2 binds and phosphorylates the nucleosome pleckstrin homology domain-containing protein that interacts with assembly protein-1 (NAP1) in Drosophila melanogaster. J Mol Biol protein kinase CK2. J Biol Chem 275: 14295–14306. 293: 1067–1084. Burnett G, Kennedy EP. (1954). The enzymatic phosphorylation of Litchfield DW. (2003). Protein kinase CK2: structure, regulation and proteins. J Biol Chem 211: 969–980. role in cellular decisions of life and death. Biochem J 369: 1–15. Cabrejos ME, Allende CC, Maldonado E. (2004). Effects of Liu S, Rauhut R, Vornlocher HP, Luhrmann R. (2006). The network phosphorylation by protein kinase CK2 on the human basal of protein-protein interactions within the human U4/U6.U5 components of the RNA polymerase II transcription machinery. tri-snRNP. RNA 12: 1418–1430. J Cell Biochem 93: 2–10. MaitaH,KitauraH,KeenTJ,InglehearnCF,ArigaH,Iguchi-ArigaSM. Chakarova CF, Hims MM, Bolz H, bu-Safieh L, Patel RJ, (2004). PAP-1, the mutated gene underlying the RP9 form of dominant Papaioannou MG et al. (2002). Mutations in HPRP3, a third retinitis pigmentosa, is a splicing factor. Exp Cell Res 300: 283–296. member of pre-mRNA splicing factor genes, implicated in auto- Meggio F, Boldyreff B, Issinger O-G, Pinna LA. (1994). Casein kinase 2 somal dominant retinitis pigmentosa. Hum Mol Genet 11: 87–92. down-regulation and activation by polybasic peptides are mediated by Faust M, Kartarius S, Schwindling SL, Montenarh M. (2002). Cyclin acidic residues in the 55–64region of the beta-subunit. A study with H is a new binding partner for protein kinase CK2. Biochem Biophys calmodulin as phosphorylatable substrate. Biochemistry 33: 4336–4342. Res Commun 269: 6–12. Meggio F, Pinna LA. (2003). One-thousand-and-one substrates of Faust M, Schuster N, Montenarh M. (1999). Specific binding of protein kinase CK2? The FASEB Journal 17: 349–368. protein kinase CK2 catalytic subunits to tubulin. FEBS Lett 462: Messenger MM, Saulnier RB, Gilchrist AD, Diamond P, Gorbsky GJ, 51–56. Litchfield DW. (2002). Interactions between protein kinase CK2 and Gonzalez-Santos JM, Wang A, Jones J, Ushida C, Liu J, Hu J. (2002). Pin1Evidence for phosphorylation-dependent interactions. J Biol Central region of the human splicing factor Hprp3p interacts with Chem 277: 23054–23064. Hprp4p. J Biol Chem 277: 23764–23772. Palancade B, Dubois MF, Bensaude O. (2002). FCP1 phosphorylation Go¨ tz C, Kartarius S, Schetting S, Montenarh M. (2005). Immunolo- by casein kinase 2 enhances binding to TFIIF and RNA polymerase gically defined subclasses of the protein kinase CK2b-subunit in II carboxyl-terminal domain phosphatase activity. J Biol Chem 277: prostate carcinoma cell lines. Mol Cell Biochem 274: 181–187. 36061–36067. Grankowski N, Boldyreff B, Issinger O-G. (1991). Isolation and Payne JM, Laybourn PJ, Dahmus ME. (1989). The transition of RNA characterization of recombinant human casein kinase II subunits a polymerase II from initiation to elongation is associated with and b from bacteria. Eur J Biochem 198: 25–30. phosphorylation of the carboxyl-terminal domain of subunit IIa. Guerra B, Go¨ tz C, Wagner P, Montenarh M, Issinger O-G. (1997). J Biol Chem 264: 19621–19629. The carboxy terminus of p53 mimicks the polylysine effect of protein Penner CG, Wang ZL, Litchfield DW. (1997). Expression and kinase CK2-catalyzed MDM2 phosphorylation. Oncogene 14: 2683– localization of epitope-tagged protein kinase CK2. J Cell Biochem 2688. 64: 525–537. Guerra B, Niefind K, Pinna LA, Schomburg D, Issinger O-G. (1998). Pinna LA. (2002). Protein kinase CK2: a challenge to canons. J Cell Expression, purification and crystallization of the catalytic subunit Sci 115: 3873–3878. of protein kinase CK2 from Zea mays. Acta Crystallogr D54: Proudfoot NJ, Furger A, Dye MJ. (2002). Integrating mRNA 143–145. processing with transcription. Cell 108: 501–512. Guerra B, Siemer S, Boldyreff B, Issinger OG. (1999). Protein kinase Sarno S, Reddy H, Meggio F, Ruzzene M, Davies SP, Donella-Deana CK2: evidence for a protein kinase CK2b subunit fraction, devoid of A et al. (2001). Selectivity of 4,5,6,7-tetrabromobenzotriazole, an the catalytic CK2a subunit, in mouse brain and testicles. FEBS Lett ATP site-directed inhibitor of protein kinase CK2 (‘casein kinase- 462: 353–357. 2’). FEBS Lett 496: 44–48. Jain N, Mahendran R, Philp R, Guy GR, Tan YH, Cao X. (1996). Schneider E, Kartarius S, Schuster N, Montenarh M. (2002). The Casein kinase II associates with Egr-1 and acts as a negative cyclin H/cdk7/Mat1 kinase activity is regulated by CK2 phosphor- modulator of its DNA binding and transcription activities in NIH ylation of cyclin H. Oncogene 21: 5031–5037. 3T3 cells. J Biol Chem 271: 13530–13536. Schneider HR, Issinger O-G. (1988). Nucleolin (C23), a physiological Krempler A, Kartarius S, Gu¨ nther J, Montenarh M. (2005). Cyclin H substrate for casein kinase II. Biochem Biophys Res Commun 156: is targeted to the nucleus by C-terminal nuclear localization 1390–1397. sequences. Cell Mol Life Sci 62: 1379–1387. Schuster N, Prowald A, Schneider E, Scheidtmann K-H, Montenarh Kuenzel EA, Krebs EG. (1985). A synthetic peptide substrate specific M. (1999). Regulation of p53 mediated transactivation by the b- for casein kinase II. Proc Natl Acad Sci USA 82: 737–741. subunit of protein kinase CK2. FEBS Lett 447: 160–166.

Oncogene CK2 is implicated in RNA splicing S Lehnert et al 2400 Shi Y, Brown ED, Walsh CT. (1994). Expression of recombinant Trembley JH, Hu DL, Slaughter CA, Lahti JM, Kidd VJ. (2003). human casein kinase II and recombinant heat shock protein 90 in Casein kinase 2 interacts with cyclin-dependent kinase 11 (CDK11) Escherichia coli and characterization of their interactions. Proc Natl in vivo and phosphorylates both the RNA polymerase II carboxyl- Acad Sci USA 91: 2767–2771. terminal domain and CDK11 in vitro. J Biol Chem 278: 2265–2270. Stalter G, Siemer S, Becht E, Ziegler M, Remberger K, Issinger O-G. Trembley JH, Tatsumi S, Sakashita E, Loyer P, Slaughter CA, Suzuki H (1994). Asymmetric expression of protein kinase CK2 in human et al. (2005). Activation of pre-mRNA splicing by human RNPS1 is kidney tumors. Biochem Biophys Res Commun 202: 141–147. regulated by CK2 phosphorylation. Mol Cell Biol 25: 1446–1457.

Oncogene