HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 24, 18327-18336, June 16, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/24/18327    most recent M909736199v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Krehan, A. Articles by Pyerin, W. Search for Related Content PubMed PubMed Citation Articles by Krehan, A. Articles by Pyerin, W. Social Bookmarking                      What's this?

Transcription Factors Ets1, NF-B, and Sp1 Are Major Determinants of the Promoter Activity of the Human Protein Kinase CK2 Gene^*

Andreas Krehan, Helenia Ansuini, Oliver Böcher, Swen Grein, Ute Wirkner, and Walter Pyerin^§

From the Biochemische Zellphysiologie (B0200), Deutsches Krebsforschungszentrum, 69120 Heidelberg, Germany and  Università di Perugia, 06100 Perugia, Italy

Received for publication, December 3, 1999, and in revised form, April 5, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CK2 is one of two isoforms of protein kinase CK2, a highly conserved, ubiquitous, and vital phosphotransferase whose expression is kept at constant cellular levels and whose dysregulated expression has been linked to malignant diseases. The upstream sequence of the gene coding for human CK2 (CSNK1A1, chromosomal location 20p13) has been examined for promoter location and transcription factor interactions using reporter gene assays (luciferase; HeLa cells), site-directed mutagenesis, electrophoretic mobility shift assays, super-shifts, UV cross-linking, Western blotting, and DNA affinity chromatography. Highest promoter activity has been found in a region comprising positions 9 to 46. Factors Sp1, Ets-1, and NF-B have been identified as interaction partners and, by mutation of individual sites and simultaneous mutations of two or more sites, shown to cross-talk to each other. At least two of the factors (Sp1; NF-B) were susceptible to phosphorylation by CK2 holoenzyme, a tetramer composed of two CK2 and two regulatory CK2 proteins, but not by individual CK2. Because the phosphorylation decreases promoter binding and repeated immunoprecipitation reveals presence of "free" CK2 in cell extracts, it is tempting to speculate that the gene product CK2 might readily form CK2 holoenzyme and feed back onto gene transcription. The data represent the first promoter control analysis of a mammalian CK2 gene and provide a hypothesis of how the constant expression level of CK2 may be achieved.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Protein kinase CK2¹ (also named casein kinase II) is a pleiotropic, ubiquitous, and conserved Ser/Thr kinase that is essential for viability of eukaryotes. CK2 occurs in two highly related isoforms, CK2 and CK2'. Both of these occur as tetrameric holoenzymes complexed stoichiometrically to regulatory CK2 proteins. This tetrameric structure is also highly conserved and required for appropriate control of substrate specificity. Although a considerable number of substrates has been documented, comprising proteins involved in processes such as transcription, replication, translation, and signaling, the exact physiological role of CK2 remains poorly understood. However, CK2 has been linked to proliferation, transformation, and cell cycle regulation (reviewed in Refs. 1-10).

The expression of CK2 proteins appears to be kept quite constant throughout tissues, and deviations have been related to diseased states. There is accumulating evidence that CK2 may, under certain cellular conditions, exert harmful effects; the targeted overexpression of CK2 in T cells of transgenic mice, for instance, results in the development of lymphomas, a situation paralleled by Theileria parva-infected bovine lymphocytes (11). Misregulation of CK2 works synergistically with oncogenes such as Myc and Tal-1 in lymphoma development (12, 13) and accelerates lymphomagenesis in mice deficient in functional p53 (14). Therefore control of CK2 expression is a subject of particular importance.

The human genome contains two CK2 loci, at chromosomes 20p13 and 11p15. Only the 20p13 locus seems to be transcriptionally active; the other appears to be a pseudogene remaining permanently silent despite the presence of potential promoter elements in the 5' region (15, 16). We have been successful in unraveling the structure of the active CK2 gene, a gene spanning roughly 70 kilobases (15). The promoter of the active gene has been located within a region comprising positions 256 to 144. The promoter shows features of a so-called housekeeping gene: no TATA box, presence of GC boxes, a CpG island around exon 1, and more than one, namely two, transcription start sites located at positions +1 and 50 (15, 16). Although the term housekeeping implies permanent expression with little regulation, even highly tissue-specific regulation has been shown to occur with this type of promoter. (e.g. with the promoter of peptidyl peptidase IV gene (17) or the promoter of platelet/endothelial cell adhesion molecule 1 (18)). In this context it was interesting that in addition to the potential cis-acting elements typical for housekeeping promoters, potential binding sites for transcription factors such as NF-B or AP2 are also present in the CK2 gene promoter. Whether these have a physiological significance has not yet been determined. We demonstrate here that three of those factors, Sp1, NF-B, and Ets1, bind to the promoter, are mutually interactive, and could be decisive in controlling the expression of the CK2 gene. In addition, the data seem to indicate the possibility for an autoregulatory loop of CK2 expression via phosphorylation of these transcription factors by CK2 holoenzyme formed upon complexation of the gene product CK2 to the regulatory protein CK2.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials

pGL3-vector system, luciferase and -galactosidase assay kits, and recombinant transcription factors were purchased from Promega. Stratagene was the provider of Pfu-turbo polymerase and the Quick Change site-directed mutagenesis kit; enzymes originated from Amersham Pharmacia Biotech, Stratagene, and MBI Fermentas. Chromatography material and Hybond-C membrane was from Amersham Pharmacia Biotech, Microcon filtration units were from Amicon, and polyvinylidene difluoride membranes were from Millipore. Radioactive labeled nucleotides were from Amersham Pharmacia Biotech. Antibodies were obtained from Santa Cruz Biotechnology. Plasmid kits, fetal calf serum, LipofectAMINE, and media were from Life Technologies, Inc. All other chemicals originated from Roth, Sigma, and Merck.

Methods

Construction of Luciferase Reporter Gene VectorsOligonucleotides derived from the sense and antisense strand of CK2 promoter regions (see Fig. 1) were synthesized (oligonucleotide synthesis group, DKFZ, Heidelberg) with additional terminal adapter sequences of XmaI, BglII, or KpnI/HindIII restriction sites and cloned into the multiple cloning site of pGL3 basic, enhancer, and promoter vectors (Promega). Correct insertion and orientation was confirmed by sequencing. Promoter sequences that were too long for synthesis (>100 base pairs) were created by deletion mutagenesis from  promoter insert 256/144.

Mutational Modification of CK2 Promoter Sequences-- The QuickChange^TM site-directed mutagenesis method of Stratagene was used for modifying defined nucleotides in CK2 promoter sequences previously inserted as wild type sequence into pGL3 vectors. Mutagenic exchange was confirmed by sequencing. Recloning of modified CK2 promoter inserts into pGL3 vectors eliminated any risk of random mutations in the vector sequence that might lead to altered reporter gene activity.

Transfection and Reporter Gene Assays-- HeLa cells were grown in minimal essential medium supplemented with 10% fetal calf serum in six-well plates to reach 60-80% confluence at the time of transfection. Before transfection LipofectAMINE (4 µg/well) was incubated with 4 µg of the respective luciferase vector and 1 µg of a -galactosidase reporter gene vector for 35 min in a volume of 50 µl of Opti-MEM and added to each well. Medium was adjusted to 10% fetal calf serum after 5 h and incubation for another 17 h followed. Cells were harvested in 120 µl of cell lysis buffer (Promega), and an ensuing 1-min centrifugation step (20,000 × g) yielded a luciferase-containing supernatant. In both cases aliquots of 20-µl supernatant were tested for luciferase activity (luciferase assay kit, Promega) and for -galactosidase activity (-galactosidase assay kit, Promega) to adjust for transfection efficiency.

Preparation of Nuclear Extracts-- Nuclear extracts were prepared according to Dignam et al. (19) and filtered (0.2 µm filters). Buffer exchange and desalting was achieved in one step with HiTrap desalting columns (Amersham Pharmacia Biotech). In the case of affinity chromatography, desalting columns were arranged in line with DNA affinity columns and removed immediately after proteins had passed through.

Affinity Chromatography-- N-Hydroxysuccinimide-activated fast protein liquid chromatography columns (Amersham Pharmacia Biotech) were used for coupling oligonucleotides of 4-fold repeats of CK2 promoter regions to be investigated, adapting the manufacturer's protocol for coupling of peptides. The coupling procedure was prolonged to 24 h at 4 °C. Finally, the columns were washed and equilibrated (19). Columns were arranged in line in an automated fast protein liquid chromatography system (Amersham Pharmacia Biotech) and HeLa S3 nuclear extracts (750 µg of total protein in 250 µl volume) were chromatographically separated. Poly(dI-dC) (50 µg/ml) was added to abolish nonspecific binding of nuclear proteins to DNA. The run-through was re-applied for at least 3 times before bound proteins were eluted by a linear gradient from 0 to 1 M KCl, followed by a step with 2 M KCl. Eluates were concentrated to equal volumes using Microcon concentrators (Amicon).

Gels and Western Blots-- Samples were applied on 12% SDS-PAGE (20) and either run at 200 V (constant) for 47 min or at 11 mA (constant) for 16-18 h. Gels of UV-cross-linking experiments were Coomassie-stained; all others were either dried and autoradiographed directly or used for semi-dry blotting on polyvinylidene difluoride membranes for immunological detection of separated proteins (21). In the latter case the membranes were either blocked with a solution containing 1 µg of polyvinyl alcohol/ml in phosphate-buffered saline or overnight with Tris-buffered saline containing 5% milk powder. Membranes were incubated for 2 h at room temperature with specific antibodies (diluted 1:4,000 for anti-329-343; 1:2,000 for anti-', and 1:10,000 for anti-171-186; anti-Ets1, anti-Sp1, and anti-NF-B-antibodies were diluted 1:1,500) treated with protein A-biotin (1:10,000) for 45 min at room temperature and followed by a 1-h incubation with peroxidase-coupled streptavidin. All steps were followed by 4 washings for 5 min (22, 23). Detection was achieved using 4-chloronaphthol as substrate (24).

Antibody Preparation-- Anti-peptide antibodies against CTCF were generated in a procedure described previously (22). For CTCF, amino acids 2-12 of human sequence EGDAVEAIVEES served as antigenic determinant.

Photochemical Cross-linking-- End-labeled double-strand oligonucleotides of 14-23 base pairs in length (40,000 cpm/reaction) shown to have a positive binding capacity for nuclear proteins in electrophoretic mobility shift assay (EMSA) technique were incubated for 20 min at room temperature with 12.5 µg of nuclear extracts prepared from HeLa S3 nuclei in a volume of 50 µl. The incubation mixture was identically composed as in EMSA. Samples cooled to 0 °C in an ice-water bath were exposed to 254-nm UV light for 4 times 15 min with 20-min intervals, 5-fold SDS sample buffer was added, and samples were heated to 95 °C for 5 min before separating cross-linked products on a 12% SDS-PAGE (11 mA constant for 18 h). Gels were Coomassie-stained for detection of marker bands, then dried and autoradiographed.

EMSA-- Oligonucleotides representing 14-23-mer CK2 promoter were synthesized in sense and antisense orientation, annealed, and end-labeled with T4-polynucleotide kinase (MBI Fermentas). Surplus radioactivity was removed by Amersham Pharmacia Biotech Microspin columns. Gel-shift assays were performed using 40,000 cpm of ³²P end-labeled oligonucleotides and 5 µg of HeLa S3 nuclear extracts per assay in a final volume of 20 µl. The presence of poly(dI-dC) prevented nonspecific protein-DNA binding. All other components were taken from Amersham Pharmacia Biotech gel-shift kit. An incubation for 20 min preceded electrophoretic separation on a native 6% polyacrylamide gel (100 V constant) at 4 °C. For super-shift EMSAs, native 3.5% polyacrylamide gels were run. Gels were dried and autoradiographed. For competition reactions with unlabeled oligonucleotides, the latter were added in 25- or 100-fold surplus depending on the nature of the oligonucleotide. Competitor oligonucleotides had the sequences: AP2, GATCGAACTGACCGCCCGCGGCCCGT; Ets1, GATCTCGAGCAGGAAGTTCGA; NF-B, AGTTGAGGGGACTTTCCCAGGC; CTCF, TCGGCCGCCCCCTCGCGGCGCGCC; GCF, TGGTGGGTGGTGAGGGGGCGGGGGTGG; Sp1, AATCGATCGGGGCGGGGCGAGC). For super-shift analyses appropriate amounts of antibody (as suggested by the supplier) were added to the samples and preincubated on ice for a minimum of 30 min before labeled oligonucleotide was added, and standard incubation for 20 min at room temperature followed.

Cell Culture-- HeLa cells were grown in modified Eagle's medium supplemented with 10% fetal calf serum (37 °C, 95% humidity, 5% CO₂). For immunoprecipitation experiments, near-confluent cells from 10-cm dishes were lysed in 190 µl of radioimmune precipitation buffer (supplemented with 150 µM aprotinin, 200 µM AEBSF (4-(2-aminoethyl)benzenesulfonyl fluoride), 1 mM benzamidine, 10 µM EDTA, 28 µM E-64, and 20 µM leupeptin), aspirated 10 × through an 18-gauge needle, and centrifuged at 20,000 × g (4 °C, 15 min). Supernatants were used for immunoprecipitations. SDS-PAGE and Western blot were performed as described above.

Immunoprecipitation-- 20 pmol of the respective antibody per assay and 40 µl of protein A-agarose beads were added to 500 µl of HeLa cell extracts obtained from 10 dishes. After agitation by rotation for 60 min (4 °C) and centrifugation (6000 × g, 1 min), the supernatant was used for three repeated immunoprecipitation reactions. In a subsequent fourth immunoprecipitation reaction (4 °C, overnight), an antibody against the respective other CK2 subunit was applied. Sedimented protein-A agarose beads were washed 6 times in radioimmune precipitation buffer at all steps, 35 µl of 5× SDS-PAGE sample buffer was added, and samples were incubated at 95 °C for 5 min.

Quantification of CK2 Immunoprecipitates-- Band intensities were determined by scanning films used for chemiluminescent detection. Different amounts of recombinant CK2 subunits (0.1 to 1.0 pmol) were used to calculate a standard curve for each individual blot. Quantification was supported by ImageQuant software (Molecular Dynamics).

Phosphorylation Assays-- Recombinant transcription factors (2 pmol each) were phosphorylated by 0.02 pmol of either CK2 or CK2 holoenzyme in a final volume of 10 µl for 30 min at 30 °C in a Tris-buffered solution (20 mM Tris-HCl, pH 7.2, 50 mM NaCl, 10 mM MgCl₂). 0.4 µCi of [-³²P]ATP per assay served as phosphoryl group donor. The reaction was stopped by adding an equal volume of 2× SDS-PAGE sample buffer and heating at 95 °C for 5 min. Samples were separated on a 12% SDS-PAGE. After Coomassie staining, the gel was dried and autoradiographed. For generation of CK2-phosphorylated Sp1 for EMSA analysis, 2 footprinting units of Sp1 (Promega) were phosphorylated in the presence of 100 mM MOPS, pH 6.9, 80 µM ATP, 10 mM MgCl₂ for 10 min at 30 °C.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Identification of Promoter-active Segments-- The region comprising positions 256 to 144 (256/144) has been shown to contain the promoter of the human CK2 gene (15). To identify the decisive segment, region 256/144 was sequentially deleted from either end. The resulting segments were cloned into pGL3 luciferase reporter gene vectors (pGL3 BV) and tested for promoter activity in HeLa cells. As shown in Fig. 1A, deletion of the 3' end to provide segment 256/65 had no significant effect; 92% of promoter activity remained. When deletion was extended to give segment 256/40, a complete loss of activity was observed. Deletion of the 5' end to provide segment 69/144 resulted in a roughly 20% decrease of promoter activity. The activity remained practically unaltered when deletion was extended to give segments 39/144 and 26/144. Stepwise further deletion then led to a moderate further decrease with segment 9/144 (66% activity), then dropped to half with segment 10/144 (32% activity) and to zero level with segment 47/144 (complete inactivity was previously observed with segment 45/144 (15)). Thus, the 3' deletion data seemed to locate promoter activity to positions 39 to 65, which may be narrowed down for a main activity segment comprising positions segment 9 to 46. In addition, the presence of enhancer activity upstream of position 39 (segment 256/70) was indicated.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Promoter-active segments of the CK2 gene promoter. A, 3' and 5' deletions of CK2 promoter were tested in luciferase reporter gene assay for promoter activity. Relative promoter activity is given in percent of CK2 segment 256/144 (set 100). Activity was determined by luciferase reporter gene assay from homogenates of HeLa cells transfected with the respective CK2 sequence cloned into luciferase reporter vector (pGL3BV). Given are mean values of three separate assays, each conducted in triplicates. S.D. of triplicates and of separate assays was below 15%. B, segments of CK2 promoter region 256/144 (cloned in pGL3EV) were tested in luciferase reporter gene assay for promoter activity. Given are mean values of three separate assays, each conducted in triplicates. S.D. of triplicates and of separate assays was below 15%.

To verify this assumption, region 256/144 was divided into five short segments, including segment 9/46, and the segments were cloned into a luciferase reporter gene vector containing an enhancer element (pGL3 EV) in order to detect even minor promoter activities. Highest activity was obtained with the construct containing segment 9/46, showing 91% of the activity of region 256/144 (Fig. 1B). The three segments 5' of 9/46 (69/10, 144/70, 145/256) were inactive, and the adjacent 3' sequence (segment 47/144) had only slight promoter activity (13% activity). In addition, two segments overlapping with segment 9/46 were tested, 39/13 containing transcription start site 1 and 10/65 containing transcription start site 2. They both showed promoter activity but were less than half as active as segment 9/46 (42% and 33%, respectively). It was concluded that the promoter activity of the CK2 gene is located within a region ranging from positions 39 to 65, that segment 9/46 was the most active part, and that there exist functional contexts between segments 39/13 and 10/65.

Proteins that Bind to Promoter-active Segments-- The upstream region of the human CK2 gene contains putative binding sites for various transcription factors, including sites for Sp1, GCF, Ets1, AP2 and NF-B (15), and CTCF (HUSAR program, DKFZ Heidelberg). When nuclear extracts (NEs) of HeLa cells were tested for their presence by Western blotting, signals were obtained for factors AP2, Ets1, NF-B (both forms, p50 and p65), and Sp1 (data not shown). (GCF and CTCF antibodies were not commercially available, and our own efforts to raise peptide antibodies were unsuccessful).

To examine its capability of binding proteins present in NEs, the upstream region was divided into nine overlapping segments of 14 to 23 base pairs in length synthesized as oligonucleotides and employed in EMSAs. Segments were selected according to the location of potential transcription factor binding sites, avoiding destruction of motifs (Fig. 2A). When incubated with NEs, strong protein binding occurred with oligonucleotides 26/8, 7/11, 12/27, and 18/40. Binding, although considerably weaker, was also obtained with oligonucleotides 7/22 and 28/45, whereas oligonucleotides 41/22 and 23/36 did not bind proteins (not shown), and oligonucleotide 46/65 only occasionally gave binding signals (hinting at DNA-protein complexes of variable stability) (Fig. 2B). Migration of the protein-DNA complexes differed significantly, indicating that different proteins had obviously interacted with these segments. The segments forming the most stable complexes contained transcription factor sites such Sp1, GCF, AP2, CTCF, NF-B, and Ets1. However, binding sites are known to interact frequently with different factors in different contexts. Therefore, various tests were carried out to pinpoint the nature of the binding proteins.

View larger version (62K): [in this window] [in a new window]   Fig. 2.   Binding of proteins present in nuclear extracts to CK2 gene promoter segments. A, promoter segments tested. Numbers indicate base pair positions relative to transcription start site 1 (Start 1). Positions of potential binding sites of transcription factors are indicated at the bottom. B, protein binding determined by EMSA. The indicated promoter segments were radioactively labeled and tested for their capability to bind proteins present in HeLa NE. Controls were run either without NE or with an additional 100-fold surplus of cold promoter segments as competitor DNAs (comp. DNA), + and  indicating EMSA in the presence and absence of competitor DNA, respectively.

A first clue came from cross-linking studies. DNA fragments incubated with NEs followed by UV-cross-linking and SDS-gel electrophoresis led, using segment 7/11, to DNA-bound proteins with molecular masses in the range of 80-100 kDa (Fig. 3). This segment possesses Sp1 sites, and Sp1 has a molecular mass of 97 kDa. By contrast, segment 7/22 and, even more pronounced, segment 12/27, both of which are devoid of Sp1 sites but contain a CTCF site, gave a number of bands ranging from 20 to 100 kDa, the higher molecular mass bands corresponding to the 70-, 73-, 80-, and 97-kDa isoforms of CTCF (25) and the anomalous migration behavior of recombinant human CTCF (26). In addition, with segment 12/27, a less intense signal in the 50-kDa range and a stronger one at around 65 kDa was obtained, corresponding to the p50 and p65 subunits of NF-B.

View larger version (50K): [in this window] [in a new window]   Fig. 3.   CK2 promoter-binding proteins by UV-cross-linking. HeLa NE was incubated with CK2 promoter oligonucleotides 7/11, 7/22, or 12/27, as given above each lane, and irradiated by ultraviolet light (254 nm). Separation of cross-linked products was performed in 12% denaturing SDS-PAGE before autoradiography. Left, molecular weight markers.

When oligonucleotides representing binding site consensus sequences of these factors were added to EMSAs, formation of radioactive complexes showed significant competitive inhibition (data not shown). In addition, DNA fragments in which consensus motifs had been mutated showed decreased or no protein binding with NEs, as exemplified for the consensus sequences of Sp1 (both binding fragments, 26/8 and 7/11), NF-B, and CTCF (Fig. 4A). When EMSAs were carried out in the presence of specific antibodies, complexes obtained with segments such as 26/8 or 7/11 containing Sp1 sites showed up-shifts in the presence of anti-Sp1 antibodies (Fig. 4B). Shifts were also obtained with NF-B and promoter segment 12/27 in presence of anti-NF-B antibodies (data not shown). Using commercially available recombinant transcription factors such as Sp1 or NF-B instead of NEs, binding was observed in EMSAs, with DNA fragments containing the respective consensus sites but not with fragments lacking those sites; results were negative for AP2 (data not shown).

View larger version (49K): [in this window] [in a new window]   Fig. 4.   Specificity of protein binding to CK2 gene promoter. A, effect of consensus motif mutation. Shown are autoradiographies of EMSA analyses. Oligonucleotides representing segments 26/8 and 7/11 with mutations of Sp1 consensus motifs and segment 12/27 with mutations of CTCF and NF-B consensus motifs were constructed, radioactively labeled, and compared with protein binding potential of identical oligonucleotides of wild type sequence. Oligonucleotides applied in each lane are given above. Modified sequences are indicated by the term in parentheses. B, characterization of binding proteins by super-shift analysis. Shown is an autoradiography of EMSA analysis. HeLa NE (5 µg) were incubated with oligonucleotides, as shown above each lane. The addition of NE, competitor DNAs (comp.), and/or specific antibodies (anti-Sp1; 1 µg/assay) is indicated.

Finally, affinity columns were prepared by immobilizing DNA fragments to solid supports and used for a chromatographic selection of proteins present in HeLa NEs. Specificity of affinity chromatography was ensured by adding excess poly(dI-dC) to NEs in order to compete for nonspecific binding, by increasing salt concentration (to 100 mM) to help dissociate unspecific protein complexes, and by employing pre-columns containing DNA with various consensus motifs to eliminate factors prone to nonspecific DNA binding, as for instance described for factor Ku (27). Eluates of columns prepared with 4-fold repeats of segments 7/11 or 26/8, i.e. segments with Sp1 sites, contained Sp1 as demonstrated by Western blotting (Fig. 5). Columns prepared with repeats of NF-B (p65) or Ets1 resulted in Western blots positive for NF-B or Ets1, respectively.

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Identification of CK2 promoter binding proteins by affinity chromatography. Proteins of HeLa NE were tested for specificity of DNA binding potential using affinity columns with CK2 promoter oligonucleotides. Oligonucleotides used for preparation of affinity columns are indicated at the top of each lane. Columns were arranged in line, and the order of arrangement is shown from left to right above each blot. Eluates from each column were separated by 12% SDS-PAGE, blotted, and detected in Western blot. Factors detected by specific antibodies are given above each blot. Molecular weight markers are indicated on the left side.

In summary, transcription factors Sp1, Ets1, NF-B, and CTCF appeared to be present in nuclear extracts and to interact specifically with DNA fragments that have promoter activity and contain the respective binding sites. These factors, therefore, were considered candidates for a role in transcriptional control of the CK2 gene.

Binding Motifs that Are Essential for Promoter Function-- The segment with the strongest promoter activity, segment 9/46, contains several transcription factor binding motifs, including a cluster of Sp1 motifs (4-fold; Sp1.2), a CTCF motif, two adjacent Ets1 motifs, and two overlapping NF-B motifs. These were examined by mutational analysis in combination with indicator gene assays (luciferase; HeLa cells) for their significance in determining promoter activity. Three or four bases each were exchanged within motifs by polymerase chain reaction-based mutagenesis to provide 9/46 segments Sp1.2mut, CTCFmut, Ets1mut, and NF-Bmut, respectively. Aside from a significant decrease or loss of protein binding (see Fig. 4A), this resulted in a gradiated effect on transcriptional activity (Fig. 6). Although activity was unchanged with Sp1.2mut and moderately decreased with CTCFmut, both Ets1mut and NF-Bmut were strongly affected; promoter activity was decreased to roughly 30% each.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Identification of factors relevant for CK2 gene promoter activity. The segment with highest promoter activity, 9/46, was tested for activation of reporter gene expression (pGL3 luciferase system) in HeLa cells. Activity of the wild type sequence was compared with that of sequences mutated in consensus motifs of Sp1, CTCF, Ets1, NF-B, or two of these simultaneously, as indicated (Sp1.1mut, etc.). Base exchanges are shown by vertical arrows, and consensus motifs are indicated on the left. Horizontal bars demonstrate relative activities of mutated sequences to wild type sequence (set 100%). Given are mean values of three separate assays, each conducted in triplicate. S.D. of triplicates and of separate assays was below 15%.

To detect functional linkages, double mutants were created with Ets1, NF-B, and Sp1 sites. Although showing no effect in isolation, the Sp1.2 site mutation appeared to amplify the effect of Ets1 site mutation (Sp1.2/Ets1mut), decreasing promoter activity to 8% and indicating a possible functional linkage of Ets1 and Sp1 (Fig. 6). By contrast, the Sp1.2 site mutation had little effect on NF-B site mutation; promoter activity of the double mutant (Sp1.2/NF-Bmut) was similar to that of the NF-B site mutant alone or slightly higher (42% activity). When Ets1 and NF-B sites were mutated simultaneously, promoter activity was decreased to 9%, i.e. the effect was significantly stronger than with either of the single mutations alone, indicating a possible functional cooperation between Ets1 and NF-B as well. The differential effects of Ets1 and NF-B mutations could be detected since particular attention was paid to the selectivity of base exchanges from adjoining motifs (28).

The question then was whether these mutations would also have effects when larger DNA fragments were investigated that more closely represent the in vivo situation. Thus, mutations analogous to those in 9/46 were introduced into the largest of the genomic fragments used in the present study, region 256/144. As shown in Fig. 7, mutation of Sp1 motifs had different effects. Although Sp1.1mut, a mutation within an 8-fold Sp1 cluster 20 base pairs upstream of Sp1.2, had no detectable effect on indicator gene expression, the Sp1.2mut decreased activity to 25%. CTCFmut had little impact on promoter activity, matching the moderate effect seen above with segment 9/46. Mutation of Ets1 and NF-B sites decreased activity to 46% and 61%, respectively. In the double mutants, Sp1.1/Sp1.2mut showed 54% activity, i.e. higher activity than Sp1.2mut but lower than Sp1.1mut. Double mutants Sp1.2/Ets1mut and Sp1.2/NF-Bmut showed 19% and 24% activity, respectively, i.e. were in the range of the mutation in Sp1.2 alone. Interestingly, the double mutant Ets1/NF-Bmut showed 15% activity and, thus, significantly less than either of the individual mutants Ets1mut and NF-B, supporting the above assumption of a functional cooperation between Ets1 and NF-B. Evidence for additional cooperation was obtained with triple mutations in which both of the Sp1 motifs were mutated; a Sp1.1/Sp1.2/Ets1mut was practically inactive (3% activity), and a Sp1.1/Sp1.2/NF-Bmut showed activity close to background (5% activity).

View larger version (25K): [in this window] [in a new window]   Fig. 7.   Impact on CK2 gene promoter activity and cross-talk of individual transcription factors. Reporter gene activity of HeLa cells transfected with CK2 promoter constructs (region 256/144) in pGL3 luciferase system. Effect of mutations of consensus motifs (Sp1, CTCF, Ets1, NF-B) on CK2 promoter activity. Shown are mutations of either individual sites or of two or three sites simultaneously, as indicated (Sp1.1 mut, etc.). Base exchanges are shown by vertical arrows. Horizontal bars demonstrate relative activities of mutated sequences to wild type sequence 256/144 set 100%. Given are mean values of three separate assays, each conducted in triplicates. S.D. of triplicates and separate assays was below 15%.

In summary, the data seem to indicate that three factors are of particular importance in determining the activity of the CK2 gene promoter: Sp1, Ets1, and NF-B. These factors seem to have functional linkages, i.e. to specifically cross-talk with one another.

Phosphorylation of Transcription Factors by CK2 Affects DNA Binding-- In a previous study (29), a set of proteins including transcription factors such as UBF, cAMP-response element-binding protein, and c-Jun was investigated in a comparative study for phosphorylation by individual CK2 and by CK2 holoenzyme. All of the factors were phosphorylated by the holoenzyme but to no significant extent by individual CK2. The phosphorylation of Sp1 and NF-B was tested under similar conditions (p50). Using identical molar amounts of kinase, phosphorylation of both of the transcription factors was observed when using the holoenzyme, whereas phosphorylation by CK2 remained at the limit of detection (Fig. 8A).

View larger version (48K): [in this window] [in a new window]   Fig. 8.   Phosphorylation of transcription factors and effect on DNA binding. A, phosphorylation of transcription factors. Selected transcription factors (Sp1; NF-B p50) were tested for phosphorylation by CK2 () and CK2 holoenzyme (). Phosphorylated samples were separated on 12% SDS-PAGE and autoradiographed. B, effect of phosphorylation on DNA binding. Recombinant transcription factor Sp1 was tested for site-specific binding to DNA (fragment 7/11) in CK2-phosphorylated and non-phosphorylated form by EMSA. The presence of CK2, Sp1, and ATP in the phosphorylation reaction before the presence of competitor DNA (comp.DNA) in EMSA is indicated above each lane.

Such phosphorylation affected DNA binding. As exemplified for Sp1, EMSAs carried out with segment 7/11 revealed that the interaction of DNA with Sp1 was significantly decreased after phosphorylation (Fig. 8B). This result opened the possibility that the gene product, CK2, might back-regulate the promoter of the gene via phosphorylation of factors such as Sp1. A prerequisite, however, would be availability of free CK2, i.e. CK2 not complexed into CK2 holoenzyme, that could associate with the newly generated CK2 to generate holoenzyme.

The availability of free CK2 was examined by repeated immunoprecipitation. HeLa cell extracts were stepwise precipitated four times with anti-CK2 antibodies followed by a precipitation with anti-CK2' antibodies, which could be expected to remove virtually all CK2 complexed into holoenzymes. The resulting supernatant was then treated with anti-CK2 antibodies. As a result, detectable amounts of CK2 were precipitated (Fig. 9). Similar results were obtained with supernatants of repeated anti-CK2' precipitations followed by anti-CK2 precipitations (data not shown). It was concluded that some extra pools of individual CK2, i.e. CK2 not complexed to CK2 or CK2', can exist in HeLa cells and, thus, be readily available for holoenzyme formation with newly synthesized CK2.

View larger version (12K): [in this window] [in a new window]   Fig. 9.   Presence of free CK2 in HeLa cells. Repeated immunoprecipitations were carried out with lysates of confluent HeLa cells using monospecific polyclonal antibodies against the individual CK2 subunits and precipitation with protein A agarose beads. Precipitates of each step were solubilized, separated by SDS-PAGE, blotted, and probed for the presence of the individual CK2 subunits with the respective antibodies. Quantification of the chemiluminescent signals was obtained by transilluminator scanning in comparison with defined amounts of recombinant subunits run in parallel on each gel. Four subsequent immunoprecipitation steps were conducted with anti-329-343 (anti-) followed by precipitation of supernatant with anti-'336-350 (anti-') followed by precipitation of supernatant with anti-171-186. Columns represent the means of at least three independent determinations ±S.D.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The promoter of the human CK2 gene contains potential binding sites for various transcription factors, including Sp1, Ets1, and NF-B (15, 16). Our data demonstrate that these factors are present in nuclear extracts of the human cells investigated, bind specifically to the predicted sites, and are major determinants of promoter activity. Their presence in nuclear extracts has been shown by Western blotting both with and without DNA affinity chromatography; their specific binding has been indicated by EMSA competition assays using either nuclear extracts or recombinant proteins and specification by antibody-mediated mobility shifts and mutation of binding sites; their effect on promoter activation has been measured by indicator gene assays employing cells that had also been used for the preparation of nuclear extracts and promoter segments with mutated individual binding sites and with combinations of two or more such sites. Analyzing a region comprising positions 256 to 144, the highest promoter activity could be localized to a segment covered by positions 9 to 46 (9/46), which contains consensus motifs for all three of the factors and of both of the transcription start sites (positions +1 and 50). No significant additional promoter activity has been found outside this region; a moderate enhancer activity seems to be present further upstream.

The 9/46 region contains several of the motifs for Sp1, Ets1, and NF-B binding. A pair of overlapping NF-B and a pair of immediately adjacent Ets1 sites are located in close proximity halfway between the two transcription start sites and a cluster of Sp1 sites located directly at start site +1 (Sp1.2; four overlapping motifs). Another Sp1 cluster (Sp1.1; eight overlapping motifs) is situated roughly 20 base pairs upstream in a region that still has some promoter activity (schematically summarized in Fig. 10). The 9/46 region and the adjacent upstream region lack a TATA box or an analogous AT-rich sequence, and no homologies are found with initiator elements (Inr) (30). Therefore, Sp1 and Ets1 are of particular interest, because both of these have been reported to participate in the formation of transcription initiation complexes (e.g. 31). Sp1 is a reported GC-box binding activator of transcription found in many promoters and enhancers (cytokeratin 18 (32), IGF II (33), K3 keratin (34), parathyroid hormone-related protein (35), and Waf/Cip1 (36)). As with the CK2 promoter, mutation of Sp1 site(s) reduces basal promoter activity. Usually more than one Sp1 site of functional relevance is present. Some of these may compensate for each other, as shown, for example, with the Waf1/Cip1 gene promoter that requires deletion of all consensus motifs in order to abolish induction by ectopically expressed Sp1 (36). Similarly, the effects seen with mutation of individual Sp1 sites of the CK2 promoter were enhanced upon simultaneous mutation of other Sp1 sites. Because it lacks a classical TATA box and other relevant transcription factor binding sites but shows a relatively high level of promoter activity, the results strongly indicate a role of Sp1 in the formation of the transcription initiation complex of the CK2 gene. As shown, for instance, for the terminal deoxynucleotidyltransferase gene and the carbamoyl phosphate synthase/aspartate carbamoyltransferase/dihydroorotase gene, Sp1 is capable of recruiting factors such as TFIID, also required by TATA-less promoters in order to activate RNA polymerase II (37). Furthermore, the finding that Sp1 site mutation effects depend significantly on the length of the DNA fragments investigated may relate to the known involvement of Sp1 in linking distant transcription control elements and, thus, in cross-talk between transcriptionally active proteins (31, 32).

View larger version (19K): [in this window] [in a new window]   Fig. 10.   Schematic representation of a transcriptional control model of the human CK2 gene. The human CK2 gene contains in the region with highest promoter activity (segment 9/46 and adjacently upstream) binding sites for transcription factors Sp1 (two clusters of eight and four overlapping sites, respectively), NF-B (two overlapping sites), and Ets1 (two immediately adjacent sites) symbolized by shaded areas within CK2 gene bar and circles below. There is cross-talk between these factors, as indicated by horizontal bars with double arrowheads. Cross-talk and/or binding to DNA is affected by phosphorylation by CK2 holoenzyme (22) generated due to complexation of the gene product CK2 to free CK2 (CK2 protein). As a result, a feedback control of gene transcription may occur.

Ets1 seems to be such a protein, cross-talking to Sp1 at the CK2 gene promoter and involved in initiation complex formation. Ets1 is a member of a family of transcriptionally active proteins found with Ap1 site(s) in polyoma enhancer and several other genes, and Ets proto-oncogenes often act in combination with Ap1. The pairwise presence of Ets1 motifs seen with the CK2 gene promoter seems to be a rather common feature. It has also been reported for genes such as the human T-cell receptor  gene (38) or all known HIV1 genes (39), and it seems to relate to a low Ets1 binding affinity of individual Ets1 sites and to DNaseI footprints, which are significantly larger than a minimal Ets-1 binding site (38). Interestingly, in a number of promoters that lack a TATA box, Ets motifs are located close to the transcription initiation sites (40), resembling the CK2 gene promoter situation (DNA polymerase  (41), DNA polymerase  (42), thymidylate synthase (43), CD3- (44), and CD4 (45). Furthermore, some of the Ets-family binding sites have been demonstrated to function as initiator elements for transcription in TATA-less genes, e.g. in cytochrome c oxidase subunit IV and Vb genes (46-48). Hence it was proposed that Ets-family binding sites in TATA-less promoters may function similarly to Sp1 in TFIID recruitment and preinitiation complex formation by the general mechanisms, resulting in transcriptional activity (49). Our analyses show that mutation of Ets1 motifs in any of the segments tested in indicator gene assays affects CK2 promoter activity significantly. This includes the activity of segment 10/65, which cannot be explained by TFIID binding to Sp1, since no Sp1 consensus motif is present within this segment, and no Sp1 binding could be detected with any of the methods applied (data not shown). In addition, the segment also lacks a TATA-like element. It does, however, contain the aforementioned double motif for Ets1 binding. The impact of Ets1 for CK2 expression has also been supported by antisense experiments. Treatment of HeLa cells by Ets1-antisense oligonucleotides resulted in a measurable decrease of CK2 protein.² Ets1 function is modulated by mitogenic signaling; phosphorylation of a conserved threonine (Thr-38) by Raf kinase causes activation of Ets1 (50), and Raf phosphorylation may relate to earlier reports on effects of mitogens on CK2 expression levels in cell culture experiments (51), which may indicate a direct link between mitogen stimulation and increases in CK2 levels and represent a link to transformation events (12).

Ets1 and Sp1 may function separately to initiate transcription. More likely, however, the mechanism is a concerted action of both factors as demonstrated for the megakaryocyte-specific IIb gene (35, 52, 53). The diminution of promoter activity of the CK2 gene due to Sp1 consensus motif mutation or due to Ets1 consensus motif mutation was increased when both of these motifs were mutated simultaneously (promoter activity practically ceased). Ets1-Sp1 interaction(s) seem to be particularly relevant under pathological conditions such as virus-induced transformation (54). A protein called tax, encoded by the T-cell leukemia virus type I (HTLV-1) genome, is capable of forming a ternary complex with Sp1 and Ets1 (as shown for the parathyroid hormone-related protein promoter), thereby facilitating Sp1-Ets1 interaction and contributing to transactivation (35).

A third important factor for the CK2 gene transcription seems to be NF-B. NF-B binding to the predicted sequence could be demonstrated in EMSA with purified protein, the failure to identify a retardation band in EMSA with HeLa nuclear extract most likely resulting from an extremely low NF-B level as reported previously (55). In fact, affinity chromatography with NF-B-containing sequence 12/27 pulled out both NF-B subunits (p50 and p65) from HeLa nuclear extracts. NF-B consensus mutations resulted in a significant loss of CK2 promoter activity. In contrast to Sp1 and Ets1, no evidence is available for a role of NF-B in TFIID binding and consecutive transcription initiation. The role of NF-B therefore seems rather to relate to interactions with Ets1 and Sp1. Sp1 and NF-B have been shown to act in synergy in regulating human immunodeficiency virus (HIV-I) gene expression (56, 57). Mutation of NF-B site has a moderate, although significant, effect on promoter activity of the 256/144 segment. In combination with Sp1 mutation (both sites, Sp1.1 and Sp1.2) or in combination with Ets1 mutation, promoter activity is strongly decreased. Thus, NF-B may play a role in the fine-tuning of CK2 expression. An indication for such an effect may be the moderate and transient increase of CK2 obtained when the cytokine interleukin 1 is being added to proliferating HeLa cells.² Interleukin-1 stimulation of HeLa cells (58) is known to modulate NF-B activity (59) and to persist for several hours (60). NF-B also offers an opportunity for tissue- and cell type-specific expression regulation. It was found to be abundant in nuclei of several cell types (e.g. corneal keratinocytes, lymphocytes, and monocytes) but low in nuclear extracts of kidney epithelial cells and fibroblasts (34). Among NF-B-responsive genes are genes encoding transcription factors such as Myc (56, 61). Interestingly, CK2 and Myc were reported to cooperate in cell transformation (12).

Cells seem to keep their CK2 level quite constant. It is not known how this is achieved. An explanation might be that a number of transcription factors can be phosphorylated by CK2 (29), including Sp1 and NF-B, as shown in the present study (Ets1 possesses several minimal CK2 phosphorylation consensus sites, but we have no evidence yet for a phosphorylation by CK2). Two features of the phosphorylation are of importance. First, it occurs with the CK2 holoenzyme but not with individual CK2. Second, the phosphorylation of factors such as Sp1 has been demonstrated to decrease its binding capacity to the CK2 gene promoter (our EMSA studies). This is supported by earlier data showing that phosphorylation of Sp1 by CK2 occurs at sites such as T579 and results in a reduced DNA binding (62). The phosphorylation of transcription factors by CK2 holoenzyme but not individual CK2 provides a basis for a working hypothesis of how the expression control might occur in order to achieve a quasi-constant cellular CK2 level. As schematically outlined in Fig. 10, it is tempting to suppose that due to the high affinity of CK2 to CK2 (63), newly generated CK2 would readily complex to free CK2 (availability shown by immunodepletion experiments) and form CK2 holoenzyme. As a consequence, Sp1 (and/or NF-B and Ets1), acting at the CK2 gene promoter and facilitating expression, might become phosphorylated. This may change either their binding to the promoter or their cross-talk to each other, resulting in a transcription decrease. Thus the gene product CK2 , indirectly, but following holoenzyme formation, could feed back on its own gene promoter to down-regulate transcription and by that keep CK2 at a certain cellular level.

Consistent with this hypothesis, situations in which the relation of CK2 level to CK2 availability has been disrupted are characteristic of certain diseased states. Infection of cattle by the parasite T. parva causes serious mortality in African lifestock and is characterized by extremely high CK2 levels and a lack of CK2 (11). The harmful effects of increased CK2 levels are also documented by various tumorigenesis experiments with transgenic mice (12-14). High expression levels of protein kinase CK2 have been reported for proliferative tissues as well as for tumors (64, 65). These are paralleled by increased catalytic activity. The finding that expression of CK2 may in part be regulated by the transcription factor Ets1 could provide an explanation; mitogenic signals are transmitted via the Ras/Raf-kinase pathway (66), leading to activation of Ets1. Since Ras and Ets1 are both proto-oncogene products, this could explain the increased expression levels of CK2. In addition, CK2 expression may be stimulated by NF-B-signaling pathways. On the other hand, the phosphorylation of Sp1 by CK2 may down-regulate Sp1 controlled genes, including the CK2 gene. Thus phosphorylation could be a means of adjusting precise cellular levels of CK2, ensuring the proper availability of CK2 for cell survival but avoiding cancer-prone CK2 overproduction.

	ACKNOWLEDGEMENTS

We are indebted to Dr. D. Kübler for providing us with nuclei from HeLa S3 cultures and the DKFZ oligonucleotide synthesis group for rapid and accurate work. The expert technical assistance of Andrea Waxman, the assistance of Michael Emmenlauer, and the secretarial support of A. Lampe-Gegenheimer are acknowledged.

	FOOTNOTES

^* This work was supported by the Commission of the European Communities (Biomed2) and the Deutsche Forschungsgemeinschaft.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ To whom correspondence should be addressed: Biochemische Zellphysiologie (B0200), Deutsches Krebsforschungszentrum, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany. Tel.: 49-6221-423254; Fax: 49-6221-423261; E-mail: w.pyerin@dkfz-heidelberg.de.

Published, JBC Papers in Press, April 12, 2000, DOI 10.1074/jbc.M909736199

² A. Krehan and W. Pyerin, unpublished data.

	ABBREVIATIONS

The abbreviations used are: CK2, protein kinase CK2 (also named casein kinase II); , isoform  of CK2; ', isoform ' of CK2; , protein  of CK2; EMSA, electrophoretic mobility shift assay; NE, nuclear extract; PAGE, polyacrylamide gel electrophoresis; MOPS, 4-morpholinepropanesulfonic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES