Metastases and multiple myeloma generate distinct transcriptional footprints in osteocytes in vivo.

Osteocytes are the most abundant bone cells, playing important roles in tissue maintenance. Little is known of how they react in vivo to cancer stress. Here we present a comparative study of the effect of a bone-residing tumour (myeloma) and metastases of bone-remote cancers on osteocytes. While no differences in morphology of the bone are seen, the changes in the transcriptome of osteocytes are specifically related to the tumour stress present. Screening approximately 22 000 genes in osteocytes prepared from cryosections of native bone using laser-supported microdissection, we observed approximately 1400 and approximately 1800 gene expression differences between osteocytes dissected from normal bone compared with those associated with metastases and multiple myeloma, respectively. The genes up-regulated due to the stress exerted by metastases were repressed by multiple myeloma and vice versa, indicating stress-specific footprints in the transcriptome of osteocytes. Functionally, the stressors seem to impose selective pressures on signalling pathways such as that of TGFbeta, a major player in bone biology. Our data show for the first time that the transcriptome of osteocytes in vivo becomes strongly affected by cancer stress, generating gene expression footprints which, in contrast to comparable morphological changes, appear to relate to the nature of cancer and might thus become helpful in distinguishing different bone diseases.

Control of methionine biosynthesis genes by protein kinase CK2-mediated phosphorylation of Cdc34.

Methionine and metabolites such as S-adenosylmethionine (AdoMet) are of vital importance for eukaryotes; AdoMet is the main donor of methyl groups and is involved in expression control of the methionine biosynthesis genes (MET genes). Genome-wide expression profiling of protein kinase CK2 deletion strains of the budding yeast Saccharomyces cerevisiae has indicated a function for CK2 in MET gene control. Deletion of the regulatory CK2 subunits leads to MET gene repression, presumably due to an impaired phosphorylation of the ubiquitin-conjugating enzyme Cdc34, which controls the central MET gene transcription factor Met4. We show that CK2 phosphorylates Cdc34 at two sites and one of these, Ser282, has a significant impact on MET gene expression in vivo, and that high AdoMet levels inhibit CK2. The data provide evidence for a control of MET gene expression by protein kinase CK2-mediated phosphorylation of Cdc34, and appear to suggest a feedback control loop in which high AdoMet-levels are limiting CK2 activity and thus MET gene expression.

Osseous integration of calcium phosphate in osteoporotic vertebral fractures after kyphoplasty: initial results from a clinical and experimental pilot study.

INTRODUCTION: This study evaluated the radiological changes at the bone-cement interface of calcium phosphate cement (CPC) and polymethylmethacrylate (PMMA) 12 months after kyphoplasty. In a pilot experiment, we additionally performed a histomorphometric analysis in osteopenic foxhounds to analyze the process of osseous integration of CPC and PMMA. METHODS: Twenty postmenopausal female patients with 46 vertebral compression fractures (VCF) were treated by kyphoplasty, utilizing CPC (N=28) or PMMA (N=18) for intravertebral stabilization. After a 12-month follow-up, we measured the density changes of border voxels at the bone-cement interface by computed tomography (CT) using dedicated software algorithms. We defined the border-voxel density (BVD) as a parameter of cement resorption at the interface. We also investigated the bone-implant interface in three osteopenic foxhounds by histomorphometry 3, 6, and 12 months after cement implantation. RESULTS: Twelve months after kyphoplasty, only CPC showed a significant decrease of the BVD compared to PMMA (p<0.01), indicating a slow progress of resorption at the interface. Histomorphometry of the dog vertebrae showed near total bone coverage of CPC implants, whereas the PMMA surface exhibited only 30% direct bone contact (p<0.01). We also observed a time-dependent increase in the number of discernable osteons close to the interface of CPC, but no bone tissue within PMMA (p<0.01). CONCLUSIONS: The decrease of the BVD 12 months after kyphoplasty may indicate osseous integration of CPC by: (1) the ingrowth of bone tissue and (2) osteonal penetration close to the interface.

Ets1 is a common element in directing transcription of the alpha and beta genes of human protein kinase CK2.

Protein kinase CK2 is a conserved and vital Ser/Thr phosphotransferase with various links to malignant diseases, occurring as a tetramer composed of two catalytically active (CK2alpha and/or CK2alpha') and two regulatory subunits (CK2beta). There is balanced availability of CK2alpha and CK2beta transcripts in proliferating and differentiating cultured cells. Examination of the human CK2beta gene for transcriptionally active regions by systematic deletions and reporter gene assays indicates strong promoter activity at positions -42 to 14 and 12 to 72 containing transcription start sites 1 and 2 of the gene (positions +1 and 33), respectively, an upstream and a downstream enhancer activity at positions -241 to -168 and 123 to 677, respectively, and silencer activity at positions -241 to -261. Of the various transcription factor binding motifs present in those regions, Ets1 and CAAT-related motifs turned out to be of particular importance, Ets1 for promoter activation and CAAT-related motifs for enhancer activation. In addition, there are contributions by Sp1. Most strikingly, the Ets1 region representing two adjoining consensus motifs also occurs with complete identity in the recently characterized promoter of the CK2alpha gene [Krehan, A., Ansuini, H., Böcher, O., Grein, S., Wirkner, U. & Pyerin, W. (2001) J. Biol. Chem. 275, 18327-18336], and affects comparably, when assayed in parallel, the promoters of both CK2 genes, both by motif mutations and by Ets1 overexpression. The data strongly support the hypothesis that Ets1 acts as a common regulatory element of the CK2alpha and CK2beta genes involved in directing coordinate transcription and contributing to the balanced availability of transcripts.

Genes targeted by protein kinase CK2: a genome-wide expression array analysis in yeast.

Protein kinase CK2, a tetramer composed of two catalytically active (CK2alpha isoforms) and two regulatory (CK2beta isoforms) subunits, is suspected to have, among others, a role in gene transcription. To identify the genes targeted by CK2, the transcriptional effect of silencing the CK2 subunit genes in Saccharomyces cerevisiae (CK2alpha isoform genes: CKA1 and CKA2; CK2beta isoform genes: CKB1 and CKB2) was examined using genome-wide expression array analysis (oligonucleotide array chips). Silencing did not influence the overwhelming majority (5801) of the over six thousand open reading frames composing the yeast genome. Cells knocked-out for both CKA1 and CKA2 and plasmid-rescued by Cka1 affected specifically at 2-fold discrimination level the transcription of 57 genes, and when rescued by Cka2, the transcription of 118 genes. In CKB1/CKB2 double knock-outs, transcription of 54 genes was specifically altered. Interestingly, aside overlaps between the gene spectra affected by CKA1 and CKA2 silencing, there were overlaps also between those influenced by CK2alpha and CK2beta isoform silencing. The data indicate a distinct role of CK2 in gene transcription control, identify specific functional differences between the two catalytic subunits in gene targeting, and reveal independent effects by the regulatory subunits.

Transcriptional coordination of the genes encoding catalytic (CK2alpha) and regulatory (CK2beta) subunits of human protein kinase CK2.

Little is known of how protein kinase CK2 genes are regulated, and it is unclear whether there are mechanisms of transcriptional coordination. Response elements present in the promoter sequences of the human catalytic (CK2alpha) and regulatory (CK2beta) subunit genes have been examined for the significance in transcriptional control using reporter gene assays, electrophoretic mobility shift assays, site-directed mutagenesis, ectopic protein expressions, and transcript assessments. Most strikingly, in both promoters the regions of highest transcriptional activity contain two adjoining, completely identical and conserved Ets1 response elements, and both the mutation of motifs and the overexpression of Ets1 affect significantly transcriptional activity. Also in common are Sp1 response elements that cooperate with Ets1, and Sp1 is phosphorylatable by CK2 holoenzyme but not by individual CK2alpha, the phosphorylation negatively affecting DNA binding. CK2alpha and CK2beta transcript levels and stoichiometries of mRNA species turned out quite constant in cultured cells despite progressing through various stages of proliferation and differentiation. The data seem to indicate transcriptional coordination of the human genes encoding CK2alpha and CK2beta based on an Ets1 double motif common to both genes cooperating with Sp1 motifs and amenable to negative feedback control by the gene products which, following complexation into CK2 holoenzyme, could phosphorylate Sp1 (and Ets1?) and thus downregulate transcription and contribute to the observed constant cellular CK2alpha and CK2beta transcripts situation.

The protein kinase A catalytic subunit Cbeta2: molecular characterization and distribution of the splice variant.

Cbeta2, a 46 kDa splice variant of the Cbeta isoform, is the largest isoform so far described for catalytic subunits from cAMP-dependent protein kinase in mammals. It differs from Cbeta in the first 15 N-terminal residues which are replaced with a 62-residue domain with no similarity to other known proteins. The Cbeta2 protein was identified in cardiac tissue by MS, microsequencing and C-subunit-isoform-selective antibodies. The Cbeta2 protein has a very low abundance of about 2% of total affinity-purified C subunits from bovine cardiac tissue. This, and the similarity of its biochemical properties to Calpha and Cbeta, are probably some of the reasons why the Cbeta2 protein has escaped detection so far. The abundance of the Cbeta2 protein differs dramatically between tissues, with most protein detected in heart, liver and spleen, and the lowest level in testis. Cbeta2 protein shows kinase activity against synthetic substrates, and is inhibited by the protein kinase inhibitor peptide PKI(5-24). The degree of Cbeta2 removal from tissue extracts by binding to PKI(5-24) depends on the cAMP level, i.e. on the dissociation state of the holoenzyme. Two sites in the protein are phosphorylated: Thr-244 in the activation segment and Ser-385 close to the C-terminus. By affinity purification and immunodetection Cbeta2 was found in cattle, pig, rat, mouse and turkey tissue and in HeLa cells. In the cAMP-insensitive CHO 10260 cell line, which has normal Cbeta but is depleted of Calpha, stable transfection with Cbeta2 restored most of the cAMP-induced morphological changes. Cbeta2 is a ubiquitously expressed protein with characteristic properties of a cAMP-dependent protein kinase catalytic subunit.

Transcription factors ets1, NF-kappa B, and Sp1 are major determinants of the promoter activity of the human protein kinase CK2alpha gene.

CK2alpha 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 CK2alpha (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-kappaB 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-kappaB) were susceptible to phosphorylation by CK2 holoenzyme, a tetramer composed of two CK2alpha and two regulatory CK2beta proteins, but not by individual CK2alpha. Because the phosphorylation decreases promoter binding and repeated immunoprecipitation reveals presence of "free" CK2beta in cell extracts, it is tempting to speculate that the gene product CK2alpha might readily form CK2 holoenzyme and feed back onto gene transcription. The data represent the first promoter control analysis of a mammalian CK2alpha gene and provide a hypothesis of how the constant expression level of CK2alpha may be achieved.

Casein kinase 2 binds and phosphorylates the nucleosome assembly protein-1 (NAP1) in Drosophila melanogaster.

The nucleosome assembly protein-1 (NAP1) was originally identified in HeLa cells as a factor facilitating the in vitro assembly of nucleosomes. However, in yeast cells NAP1 is required in the control of mitotic events induced by the Clb2/p34(CDC28). Here, we show that Drosophila NAP1 is a phosphoprotein that is associated with a kinase able to phosphorylate NAP1. By using an in-gel kinase assay we found that this kinase displays a molecular mass of 38 kDa. Following purification and peptide microsequencing, we identified the kinase phosphorylating NAP1 as the alpha subunit of casein kinase 2 (CK2). With the help of a series of NAP1 segments and synthetic peptides, we assigned the CK2 phosphorylation sites to residues Ser118, Thr120, and Ser284. Interestingly, Ser118 and Thr120 are located within a PEST domain, while Ser284 is adjacent to the nuclear localization signal. Substitution of the identified phosphoresidues by alanine was found to reduce considerably the ability of CK2 to phosphorylate NAP1. The enhanced ability of CK2 to phosphorylate phosphatase-treated NAP1 extracted from Drosophila embryos and the similar tryptic phospho-peptide pattern of in vivo labelled NAP1 and in vitro labelled NAP1 with CK2 indicate that NAP1 is a natural substrate of CK2. Further analysis revealed that both CK2alpha and beta subunits are associated with NAP1 but we found that only the catalytic alpha subunit establishes direct contact with NAP1 on two distinct domains of this protein. The location of CK2 phosphorylation sites in NAP1 suggests that their phosphorylation can contribute to a PEST-mediated protein degradation of NAP1 and the translocation of NAP1 between cytoplasm and nucleus.

Heterogeneous nuclear ribonucleoprotein A2 interacts with protein kinase CK2.

The catalytic subunit of protein kinase CK2 (CK2alpha) was found associated with heterogeneous nuclear ribonucleoprotein particles (hnRNPs) that contain the core proteins A2 and C1-C2. High levels of CK2 activity were also detected in these complexes. Phosphopeptide patterns of hnRNP A2 phosphorylated in vivo and in vitro by protein kinase CK2 were similar, suggesting that this kinase can phosphorylate hnRNPA2 in vivo. Binding experiments using human recombinant hnRNP A2, free human recombinant CK2alpha or CK2beta subunits, reconstituted CK2 holoenzyme and purified native rat liver CK2 indicated that hnRNP A2 associated with both catalytic and regulatory CK2 subunits, and that the interaction was independent of the presence of RNA. However, the capability of hnRNP A2 to bind to CK2 holoenzyme was lower than its binding to the isolated subunits. These data indicate that the association of CK2alpha with CK2beta interferes with the subsequent binding of hnRNP A2. HnRNP A2 inhibited the autophosphorylation of CK2beta. This effect was stronger with reconstituted human recombinant CK2 than with purified native rat liver CK2.

Serum-stimulated cell cycle entry of fibroblasts requires undisturbed phosphorylation and non-phosphorylation interactions of the catalytic subunits of protein kinase CK2.

Protein kinase CK2 is a pleiotropic Ser/Thr kinase occurring as alpha2beta2, alpha'2beta2, or alphaalpha'beta2 tetramers. A requirement in serum-stimulated cell cycle entry in both the cytoplasm and the nucleus of human fibroblasts for phosphorylation(s) by CK2 has been concluded from stimulation inhibition by microinjected antibodies against the regulatory subunit (beta). We have now examined this idea more directly by microinjection-mediated perturbation of phosphorylation and non-phosphorylation interactions of the catalytic subunits (alpha and alpha'), and by verifying the supposed matching of the cellular partition of CK2 subunits in the fibroblasts employed. While immunostaining and cell fractionation indicate that the partitions of subunits indeed match each other (with their predominant location in the nucleus in both quiescent and serum-stimulated cells), microinjection of substrate or pseudosubstrate peptides competing for the CK2-mediated phosphorylation in vitro resulted in significant inhibition of serum stimulation when placed into the nucleus but not when placed into the cytoplasm. Also inhibitory were nuclear but not cytoplasmic injections of antibodies against alpha and alpha' that affect neither their kinase activity in vitro nor their complexing to beta. The data indicate that the role played by CK2 in serum-stimulated cell cycle entry is predominantly nuclear and more complex than previously assumed, involving not only phosphorylation but also experimentally separable non-phosphorylation interactions by the catalytic subunits.

Intermolecular contact sites in protein kinase CK2.

Chemical crosslinking and analysis of CNBr-digested fusion products by immunoblotting with sequence-specific antibodies identifies an interaction between positions 55-70 of subunit beta (beta55-70) and 65-80 of subunit alpha (alpha65-80). This has been supported by crosslinking of subunits with peptides alpha65-80 and beta55-70, by binding of subunits to immobilized peptides, and by the hindrance of coprecipitation with peptide-raised antibodies (anti-alpha65-80; anti-beta55-70). Functionally, beta55-70 is a negative regulatory region for the kinase activity of subunit alpha. The opposite, stimulatory property of subunit beta has been assigned to its C-terminal part. Subdivision of peptide beta155-181, that has stimulatory effect, into overlapping peptides and assaying for alpha binding and binding competition revealed a tight physical contact at beta162-175. This region, however, is non-stimulatory indicating binding a necessary but not sufficient quality for stimulation. A contact might exist to regions surrounding C147 and/or C220 at subunit alpha as indicated by crosslinking and peptide competition. The crosslinking data also confirm a beta-beta contact in CK2 holoenzyme. Effects by non-ionic detergents show hydrophobic interactions to play an important role in catalytic activity adjustment.

CK2alpha loci in the human genome: structure and transcriptional activity.

Two CK2alpha loci are present in the human genome. First, locus 20p13, that contains the CK2alpha coding gene. It spans around 70 kb, is composed of 13 exons and shows homology to the respective gene in the nematode Caenorhabditis elegans. The translation start site is located in the second exon, the stop codon in exon 13. Two transcription start sites were identified, the further 5' located site defines position 1 of the gene, the second site is located at position 50, respectively. The promoter region shows characteristics of a so-called house keeping gene: A high GC content, lack of a TATA-box and presence of several GC-boxes. By reporter gene assays, the promoter region of the CK2alpha gene could be located between position -256 and 144 relative to the first transcription start site. In the 3' noncoding region of the CK2alpha gene, six polyadenylation signals were identified. As indicated by Northern blot analysis and by comparison with expressed sequence tags from the EMBL databank, the most 3' located, active polyadenylation signal seems to be the fourth defining the end of the CK2alpha gene. The second CK2alpha locus is at 11p15. It contains a processed pseudogene, which shows all typical features of a processed sequence, such as absence of introns, short poly-A tail and direct flanking repeats. Interestingly, it contains a complete open reading frame and has potential promoter elements in its 5' region. Nevertheless, no promoter activity could be detected in reporter gene assays.

Searching interaction partners of protein kinase CK2beta subunit by two-hybrid screening.

To date, the intracellular regulation of protein kinase CK2 is unknown. However it was observed that the enzyme associates with several intracellular proteins and the formation of such molecular complexes may represent a mechanism for the control of CK2 activity. Using the Interaction Trap system in yeast, with the CK2beta as a bait, we looked for CK2 partners. We present the identification of new potential partners of CK2beta and it is hoped that their classification will help in understanding the physiological roles and the regulation of CK2 in the cell.

BTF3 is a potential new substrate of protein kinase CK2.

BTF3, initially discovered as a factor required for transcription inititation of RNA polymerase II, is expressed in two isoforms, termed a and b. BTF3b, the transcriptionally inactive isoform, was identified as an interaction partner of protein kinase CK2 subunit beta employing the interaction trap system for screening ofa HeLa cDNA fusion library. We report here on the interaction between the other isoform, BTF3a, and protein kinase CK2. The complete cDNA of BTF3a was cloned by RT-PCR and used for analysis in the two-hybrid system with a three-reporter yeast strain. Interaction of BTF3a with CK2 subunits alpha, alpha' or beta was detectable by one of three reporters, whereas the CK2beta - BTF3a interaction was activating two reporters. It was also shown that BTF3a is phosphorylated in vitro by the alpha2beta2 holoenzyme, but not by alpha or alpha' alone, indicating the requirement of beta for substrate recognition. Immunoprecipitations of GST-fused BTF3a carried out in vitro resulted in co-precipitation of beta. Similarly, GST-BTF3a, but not GST alone isolated with glutathione agarose beads from buffer containing recombinant CK2 subunits was found complexed with alpha and beta, likely representing alpha2beta2 holoenzyme. The data show a weak, nevertheless specific interaction of protein kinase CK2 via subunit beta with the putative transcription factor BTF3a in vitro and in vivo, and a role of BTF3a as a potential new substrate for CK2.

Protein kinase CK2alpha may induce gene expression but unlikely acts directly as a DNA-binding transcription-activating factor.

The gene-inducing property ofCK2alpha, a Ser/Thr protein kinase that appears normally to be complexed to a CK2beta protein controlling activity and substrate selectivity, has been unclear. We show here that CK2alpha induces in human JEG-3 cells the expression of Aromatase, an estrogen-synthesis key enzyme, which is regulated at transcriptional level. Electrophoretic mobility shift assays indicate that CK2alpha binds to the Aromatase gene promoter. To test for CK2alpha's transactivating ability as a DNA-binding protein, a CK2alpha binding site was cloned in front of indicator genes. The constructs were used to transform a yeast-based one-hybrid system. Overexpression of activation-domain fused CK2alpha in this system, i.e., CK2alpha in its native configuration, failed to activate the transcription machinery. The data indicate CK2alpha to affect gene expression at the level of transcription via an indirect as yet unknown mechanism rather than directly as a DNA-binding transcription-activating protein.

Multiple forms of protein kinase CK2 present in leukemic cells: in vitro study of its origin by proteolysis.

Human recombinant CK2 subunits were incubated for different times with the two main cytosolic proteases m-calpain and 20 S proteasome. Both, m-calpain in a calcium dependent manner and the 20 S proteasome, were able to degrade CK2 subunits in vitro. In both cases, CK2alpha' was more resistant to these proteases than CK2alpha. When these proteases were assayed on the reconstituted (alpha2beta2 holoenzyme), a 37 kDa alpha-band, analogous to that observed in AML extracts, was generated which was resistant to further degradation. No degradation was observed when the 26 S proteasome was assayed on free subunits. Studies with CK2alpha deletion mutants showed that m-calpain and the 20 S proteasome acted on the C-terminus end of CK2alpha. These results pointed to cytosolic proteases as agents involved in the control of the amount of free CK2 subunits within the cell, which becomes evident when CK2 is overexpressed as in AML cells.

Identification of structural elements of subunit beta of human protein kinase CK2 participating in tight physical alpha-beta intersubunit contacts directly adjacent to a surface-oriented region.

Sites essential for tight physical intersubunit interaction in protein kinase CK2, a tetramer composed of two catalytic (alpha and/or alpha') and two regulatory subunits (beta), have been assigned to the C-terminal part of subunit beta. Mutational analysis suggests region 171-181 of beta to be one of these but this is not consistent with the observation of coprecipitation of catalytic subunits by antibodies directed specifically to this beta segment which indicates that this region is accessible to antibodies even if the beta subunit is associated with the alpha subunit. In an attempt to clarify the apparent contradiction, we have subdivided beta-(155-181)-peptide, which includes the fragment of beta and that both binds to catalytic subunits and stimulates kinase activity, into six more or less overlapping peptides with a length of 9-16 amino acid residues and performed peptide competition and a subunit binding assays. The kinase-stimulating effect of beta-(155-181)-peptide was counteracted significantly by beta-(162-175)-peptide while the others had less or no effect. In an ELISA assay with the peptides covalently bound to wells via a C10 spacer arm, binding of a occurred besides beta-(155-181)-peptide only with beta-(162-175)-peptide and beta-(165-175)-peptide. The data provide the first evidence that the contact and the stimulation competences of subunit beta represent separable functions and suggest further that a tight physical contact to catalytic subunits is located at position 162-175 consisting of a hydrophobic stretch (position 162-171) and a hydrophilic binding motif, PEY (position 172-174), with the adjoining downstream part allowing for surface exposure and antibody binding. The presence of several Pro residues within that region might relate to the structural basis for both the alpha-beta interaction and the surface orientation.

Genomic organization and promoter identification of the human protein kinase CK2 catalytic subunit alpha (CSNK2A1).

The isolation and characterization of the complete gene coding for human protein kinase CK2 catalytic subunit alpha is described. The gene spans 70 kb and consists of 13 exons, and the exon/intron boundaries conform to the gt/ag rule. Exons range in size from 51 to 2960 bp, introns from 527 to around 34000 bp. The translation start site is located in Exon 2, the stop codon in Exon 13. Two transcription start sites were identified by primer extension analysis: The further 5'-located site defines position 1 of the gene, the second site is located at position 50. The 5' region of the CK2 alpha gene shows features of a housekeeping promoter, such as lack of a TATA box and presence of a CpG island and GC boxes. The region was analyzed by reporter gene assay, and promoter activity was detected within the region ranging from position -256 to 144. Six potential polyadenylation signals were identified in the 3' noncoding region of the CK2 alpha gene. As indicated by comparison with expressed sequence tags from the EMBL databank and by Northern-blot analysis, the most 3' located, active polyadenylation signal seems to be the fourth signal, defining the end of the gene.

Promoter of the gene encoding the bovine catalytic subunit of cAMP-dependent protein kinase isoform C beta 2.

Genomic sequences flanking the 5' end of the cDNA encoding isoform C beta 2 of the catalytic subunit of bovine cAMP-dependent protein kinase were cloned, sequenced and analyzed for promoter activity and transcription initiation sites. A region of 913 bp upstream the translation initiator ATG was amplified from genomic DNA by vectorette polymerase chain reaction. In primer extension reactions and RNase protection assays, residues C (at position -91), T (-71) and G (-70) were found to serve as transcription initiation sites of the gene. Amplification products and sub-fragments thereof were ligated upstream of the reporter gene chloramphenicol acetyltransferase to test for promoter activity. Constructs were transiently transfected into a Chinese hamster ovary cell line which was shown to express endogenous C beta 2 mRNA. The genomic sequence upstream the C beta 2 cDNA does have promoter activity. The region from position -51 to -292 proved sufficient to drive efficient transcription of the reporter gene. The promoter is AT rich (68%), does not contain a TATA box within 50 bp upstream of the first initiation site and possesses putative binding sites for several transcription factors such as PEA-3 and a glucocorticoid receptor.