Journal of Molecular Medicine

, Volume 82, Issue 9, pp 621–628

Stimulation of c-Rel transcriptional activity by PKA catalytic subunit β

Authors

  • Shih-Hung Yu
    • Institute of BiochemistryNational Yang-Ming University
  • Wei-Chung Chiang
    • Institute of BiochemistryNational Yang-Ming University
  • Hsiu-Ming Shih
    • Division of Molecular and Genomic MedicineNational Health Research Institutes
    • Institute of BiochemistryNational Yang-Ming University
Original Article

DOI: 10.1007/s00109-004-0559-7

Cite this article as:
Yu, S., Chiang, W., Shih, H. et al. J Mol Med (2004) 82: 621. doi:10.1007/s00109-004-0559-7

Abstract

Nuclear factor κB (NF-κB) is a eukaryotic transcription factor which responds to different extracellular signals. It is involved in immune response, inflammation, and cell proliferation. Increased expression of c-Rel (or its viral homolog v-Rel), one component of the NF-κB factors, induces tumorigenesis in different systems. The activity of NF-κB can be regulated by protein kinase A (PKA) in a cAMP-independent manner. Our previous results showed that c-MYC induces the activity of PKA by inducing the transcription of the gene encoding the PKA catalytic subunit β (PKA-Cβ). Constitutive expression of PKA-Cβ in Rat1a cells induces their transformation. Here we show that CREB is unlikely to be a phosphorylation target of PKA-Cβ as characterized by different cell lines. Electrophoretic mobility shift assays showed that c-Rel is present as a significant component of the NF-κB factors in c-MYC overexpressing status. The transcriptional activity of c-Rel was significantly stimulated by PKA-Cβ. Coactivators p300/CBP are at least partially responsible for the enhanced activation mediated by c-Rel and PKA-Cβ. Interaction between c-Rel and PKA-Cβ was demonstrated using coimmunoprecipitation assays. Immunoprecipitation-in vitro phosphorylation assays showed the direct phosphorylation of c-Rel by PKA-Cβ. These results indicate that c-Rel is a reasonable phosphorylation target of PKA-Cβ, and that the transcriptional activity of c-Rel is stimulated by PKA-Cβ possibly through the interaction with p300/CBP.

Keywords

Nuclear factor kBc-RelProtein kinase A catalytic subunit βc-MYCTranscription

Abbreviations

CBP

CREB-binding protein

IκB

Inhibitor factor κB

NF-κB

Nuclear factor κB

PKA

Protein kinase A

SDS-PAGE

Sodium dodecyl sulfate polyacrylamide gel electrophoresis

TC

Tetracycline

Introduction

Nuclear factor κB (NF-κB) is a eukaryotic transcription factor that exists virtually in all cell types [1, 2, 3]. It is present in the cytoplasm and responds to extracellular signals such as cytokines, tumor necrosis factor, acute-phase proteins, bacterial and viral products, free radicals, UV light, and γ-irradiation [1, 2, 3]. NF-κB is composed of homodimers or heterodimers of Rel family proteins such as p65 (RelA), c-Rel, RelB, p52, and p50 [1, 2, 3]. After stimulation the regulatory partner of NF-κB, inhibitor factor κB (IκB), is phosphorylated, ubiquitinated, and degraded, and NF-κB translocates to the nucleus to carry out its transcriptional functions (for review see [1, 2]). Heterodimeric forms of p65-p50 or c-Rel-p50 are transcriptionally active after binding to the NF-κB binding sites in the promoters of different NF-κB responsive genes [1]. Recently NF-κB has emerged as important factors controlling cell growth and oncogenesis [4, 5]. Chromosomal amplification, overexpression and rearrangement of genes coding for Rel/NF-κB factors have been implicated in many human hematopoietic and solid tumors [5]. Among the family members of NF-κB, c-Rel (or its viral homolog v-Rel) was involved in the regulation of cell cycle [6, 7] and constantly altered in B-cell lymphoma cell lines, Hodgkin’s lymphoma, multiple myeloma, or other solid tumors [8, 9, 10, 11]. Overexpression of c-Rel was shown to transform primary chicken spleen cells [12]. Transgenic mice overexpressing mouse mammary tumor virus driven c-Rel develop mammary tumors [13]. Constitutive activation of NF-κB can contribute to oncogenesis by driving cell proliferation, enhancing cell survival, or promoting angiogenesis/metastasis, making NF-κB signal pathways important in the generation and treatment of human cancers [11, 14].

Although NF-κB usually becomes active after extracellular stimuli, it can also be activated by protein kinase A (PKA) after degradation of IκB in the NF-κB–IκB—PKAc complex in a cAMP-independent manner [15, 16]. Phosphorylation of p65 by PKAc regulates its transcriptional activity by changing its conformation to facilitate its interaction with coactivators CREB-binding protein (CBP)/p300 [15, 16]. Recently we have demonstrated that c-MYC, an oncoprotein, directly regulates the expression of the gene encoding the PKA catalytic subunit β (PKA-Cβ) [17]. This activation occurs in a cAMP-independent manner in the absence of extracellular stimuli [17]. PKA normally exists in the cytoplasm as an inactive tetrameric holoenzyme which is dissociated into two free catalytic and two regulatory subunits when cAMP is generated upon binding of hormones to their receptors and activation of adenylate cyclase [18, 19]. The PKA free active catalytic subunits (Cα, Cβ, or Cγ) phosphorylate serine and threonine residues on specific substrate proteins which include transcription factors involved in cell activation and proliferation [20]. The downstream targets of PKA include cAMP-response element binding protein (CREB), cyclic AMP-responsive element modulator, NF-κB, activator protein 1, and Raf-1 [15, 18, 19, 21, 22], controlling pleiotropic functions such as cell activation and DNA synthesis. PKA-Cα phosphorylates CREB or p65 to regulate their transcriptional activity [15, 21]. However, the phosphorylation targets of PKA-Cβ are unknown. Due to the role of c-MYC in tumorigenesis and its direct regulation of PKA-Cβ, it will be important to search for the downstream targets of PKA-Cβ to elucidate the mechanism of c-MYC mediated tumorigenesis.

This study demonstrated that c-Rel is a possible downstream target of PKA-Cβ. PKA-Cβ significantly stimulated the transcriptional activity of c-Rel. This enhanced activation may be mediated through the interaction of c-Rel with p300/CBP after its phosphorylation by PKA-Cβ. These results demonstrated that the transcriptional activity of c-Rel was regulated by PKA-Cβ.

Materials and methods

Cell lines, plasmids, and reagents

The lymphoblastoid cell lines, RatMyc, and EREB.TCMyc cells have previously been described [23, 24]. The pHeBOCMV-Cβ plasmid was made as described [17]. The pHeBOCMV-Cα plasmid was made by RT-PCR to generate a full-length PKA-Cα cDNA fragment, which was subsequently digested with NotI to facilitate its cloning into the NotI site of pHeBOCMV vector. The sequence of PKA-Cα was verified by sequencing. The primers used were: 5′-ACCGGCGCTAGCCGCCGCCGCCGCGATGGGCAA-3′ and 5′- ACCCATGCGGCCGCAGGCATGCCCCTAAAACTC-3′. c-Rel expression vector (pMT2T-c-Rel) and HIILuc plasmid containing two copies of MHC class II NF-κB binding sites were obtained from Dr. U. Siebenlist (NIAID, Bethesda, Md., USA) and Dr. D. Baltimore (California Institute of Technology, Calif., USA). pMT2T-c-RelMut expression vector was generated by site directed mutagenesis to mutate the amino acid 267 from serine into alanine. p300 and CBP expression vectors were obtained from Dr. R. Goodman (Oregon Health Science Center, Ore., USA).

Western blot analysis

We loaded 40 µg of cellular extracts (CB control, CBMyc.Max, CBMax, Rat1a control, RatMyc, Fig. 1A) and 80 µg of cellular extracts (EREB.TCMyc, Fig. 1B) onto a 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gel and transferred to nitrocellulose filters. The filters were probed with a polyclonal anti-phospho CREB antibody (Ser133 phosphorylation status) and a control CREB antibody (phosphorylation state independent; New England Biolabs) as a control for protein loading. Signals were revealed using a chemiluminescence kit (ECL, Amersham Pharmacia Biotech). Data shown here are representative of two or more experiments.
Fig. 1

Decrease or no change in phospho-CREB levels in different cell lines overexpressing c-MYC. A Significant decrease in phospho-CREB levels in CB33 or Rat1a cells overexpressing c-MYC. Western blot analysis of cell lysates from the CB control cells, CBMyc.Max cells, CBMax, Rat1a control, RatMyc cells using an anti-phospho-CREB antibody and an anti-CREB antibody as a control for protein loading. The anti-phospho-CREB antibody also recognizes phospho-ATF1. CREB extract was total cell extract prepared from SK-N-M cells treated with FGF, forskolin, and IBMX and provided by New England Biolabs. Control extract was from untreated SK-N-M cells. The amount of control extract and CREB extract loaded was 5 μg each, not enough to be recognized by the anti-CREB antibody. B No change in phospho-CREB levels in EREB.TCMyc cells induced to express exogenous c-MYC (24 h, lane −TC/−E2 vs. lane +TC/−E2)

Electrophoretic mobility shift assays

Electrophoretic mobility shift assays were performed as described [25]. Briefly, an oligonucleotide containing the NF-kB binding site was labeled and incubated with 10 µg nuclear extracts. The sequence of the oligonucleotide was: 5′-AGCTTGGGGACTTTCCAG-3′. Preparation of nuclear extracts was performed following the procedure of Andrews and Faller [26]. The binding reactions were carried out in 10 µl HMO buffer (25 mM HEPES pH7.5, 50 mM KCl, 5 mM MgCl2, 0.5 mM EDTA, 5% glycerol, 10 mM dithiothreitol, 0.1% NP-40, 0.5 mg/ml bovine serum albumin) in the presence of 0.5 µg poly(dI.dC). Products were analyzed on a 3.5% HEPES-acrylamide nondenaturing gel. Anti-c-Rel and anti-p65 antibodies (SC-71x, and SC-372x, Santa Cruz Biotechnology) were used to perform supershift experiments (0.5 µl for each reaction). Data shown here are representative of two experiments.

Transient transfection procedure and luciferase assays

Transfections were performed using the CaPO4 precipitation method as described [25]. HeBOCMV-Cβ, HeBOCMV-Cα, pMT2T-c-Rel, pMT2T-c-RelMut, p300, and CBP plasmids were transfected into 293T cells together with HIILuc reporter construct with the amounts indicated in the figure legends. Total amount of plasmids were kept constant in each transfection. The plasmid expressing the bacterial β-galactosidase gene (pCMVβgal, 0.5 µg) was also cotransfected in each experiment to serve as an internal control for transfection efficiency. Two days after transfection cells were harvested, and their luciferase activities were assayed. Each transfection was performed in triplicate and standard deviation bars were shown. Data shown here are representative of three or more experiments from independent transfections.

Coimmunoprecipitation assays

Coimmunoprecipitation assays were performed as described [27]. Briefly, 293T cells were transfected with equal amounts (5 µg for each plasmid) of PKA-Cβ and either pMT2T control vector, pMT2T-c-Rel wild type, or pMT2T-c-Rel mutant. Two days after transfection 500 µg of whole cell extracts were incubated with anti-PKA-Cβ for 3 h to perform coimmunoprecipitation. After incubation protein A beads were used to pull down anti-PKA-Cβ and the precipitate was washed with TNTG buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 0.1% Triton X-100, and 10% glycerol) three times, and final precipitate was mixed with protein loading buffer, boiled, and run on SDS-PAGE. After transfer the filter was probed with an anti-PKA-Cβ antibody (SC-904, Santa Cruz Biotechnology) and with an anti-c-Rel antibody (SC-71, Santa Cruz Biotechnology). Data shown here are representative of two experiments.

Immunoprecipitation-in vitro phosphorylation assays

Either c-Rel or PKA-Cβ expression vector was transfected into 293T cells and their extracts were harvested separately 2 days after transfection in a lysis buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, 1 mM trisodium vanadate, 20 mM sodium fluoride, and 1% Triton X-100). The extracts were immunoprecipitated with an anti-c-Rel antibody or an anti-PKA-Cβ antibody as mentioned above. The immunoprecipitate was eluted with 80 µl 50 mM glycine, pH 2.0 followed by neutralization with 20 µl of 1 M Tris, pH 7.5. The whole eluate was again immunoprecipitated with the same antibody and eluted, neutralized again using the same buffer. The in vitro phosphorylation assay was performed as described with some modifications [28, 29]. One-half of the eluted c-Rel immunoprecipitate was mixed with the eluted PKA-Cβ immunoprecipitate and the in vitro phosphorylation assay was performed. Control reaction contained only half of the eluted c-Rel immunoprecipitate. The reaction mixture contained 25 mM Tris, pH 7.5, 5 mM β-glycerophosphate, 2 mM dithiothreitol, 0.1 mM trisodium vanadate, 10 mM MgCl2, 10 µM ATP, and 10 µCi [γ-32P] ATP together with the eluted immunoprecipitate indicated. The reaction was carried out at 30°C for 30 min and the reaction mixture was run on SDS-PAGE and visualized by autoradiography.

Results

CREB is an unlikely target of PKA-Cβ in c-MYC overexpressing cells

We have demonstrated that PKA-Cβ was a direct transcriptional target of c-MYC and overexpression of PKA-Cβ in Rat1a cells induced transformation activity [17]. In order to search for the downstream targets of PKA-Cβ we tested whether CREB, the well characterized phosphorylation target of PKA-Cα [21], is a possible target of PKA-Cβ. We reasoned that the levels of phosphorylated CREB should be increased if it was phosphorylated by PKA-Cβ in c-MYC overexpressing status. We compared the levels of phosphorylated CREB vs. total CREB in three different cellular systems. The lymphoblastoid cell lines (CB33, CBMyc.Max, CBMax) are human Epstein-Barr virus immortalized B cell line (CB33) engineered to express high levels of either c-MYC/MAX (CBMyc.Max) or MAX/MAX (CBMax) complexes, representing the two regulatory situations of high and low c-MYC function, respectively [23]. The phenotype of these cells was previously characterized and shown to be consistent with their differential c-MYC expression levels since CBMyc.Max cells display a short doubling time, have clonogenic properties in vitro and cause tumors in vivo; whereas CBMax cells proliferate very slowly and lack any transformation-related phenotype [23]. Figure 1A shows that the levels of phosphorylated CREB were significantly decreased in CBMyc.Max cells, indicating that CREB is unlikely to be a phosphorylation target of PKA-Cβ. The levels of phosphorylated CREB also decreased in Rat1a cells overexpressing c-MYC (Fig. 1A). In addition, we exploited a previously described B cell line (EREB.TCMyc) in which proliferation and c-MYC expression could be independently controlled [24]. Since in these cells the EBV EBNA-2 gene (which induces B cell immortalization) is expressed as a chimeric fusion with the hormone binding domain of the estrogen receptor [30], estrogen (E2) removal leads to growth arrest (G0/G1) associated with complete downregulation of endogenous c-MYC expression (see Fig. 1B, +TC/−E2 lanes). Since these cells have also been transfected with a tetracycline (TC) repressed c-MYC vector [24], exogenous c-MYC expression can then be induced by TC withdrawal. The levels of phosphorylated CREB appeared similar in EREB.TCMyc cells before and after c-MYC induction (Fig. 1B, 24 h, lane +TC/−E2 vs. lane −TC/-E2). All these results showed that CREB is unlikely to be a phosphorylation target of PKA-Cβ since the levels of phosphorylated CREB is either significantly decreased or not changed in c-MYC overexpressing cell lines.

c-Rel is a significant component of the NF-κB factors in EREB.TCMyc cells induced to express exogenous c-MYC

To further search for the possible downstream phosphorylation target of PKA-Cβ, we used electrophoretic mobility shift assays to test whether NF-κB is a possible PKA-Cβ target since one of the NF-κB factors, p65, was shown to be phosphorylated by PKA-Cα [15, 16]. Electrophoretic mobility shift assays were used to demonstrate whether NF-κB was elevated in EREB.TCMyc cells induced to express exogenous c-MYC. Figure 2 shows that NF-κB activity is increased in EREB.TCMyc cells induced to express exogenous c-MYC (24 h, lane −TC/−E2 vs. lane +TC/−E2). In addition, the majority of the NF-κB activity could be supershifted by using a c-Rel specific antibody (+Ab, c-Rel lanes) but not by a p65 specific antibody (+Ab, p65 lanes). There is also an increase in c-Rel activity in EREB.TCMyc cells induced to express exogenous c-MYC (c-Rel, lane −TC/−E2 vs. lane +TC/−E2). This result indicated that c-Rel is a likely phosphorylation target of PKA-Cβ since c-Rel represents a significant proportion of increased NF-κB activity in EREB.TCMyc cells induced to express exogenous c-MYC.
Fig. 2

Electrophoretic mobility shift assays showed the increase in NF-kB activity in EREB.TCMyc cells induced to express exogenous c-MYC (24 h, lane −TC/−E2) and the predominant presence of c-Rel. Incubation with an anti-c-Rel antibody induced a supershift of NF-kB activity (c-Rel lanes). Incubation with anti-p65 antibody increased the intensity of shifted NF-kB activity (p65 lanes), which may be due to nonspecific enhancement

Enhanced activation of a reporter construct bearing NF-κB sites by c-Rel and PKA-Cβ

It has been demonstrated that PKA-Cα could enhance the transcriptional activity of p65 by phosphorylation of a consensus PKA phosphorylation site to increase its interaction with coactivators p300/CBP [15, 16]. In order to test the ability of PKA-Cβ to increase the transcriptional activity of c-Rel we used transient transfection assays to transfect a reporter construct (HIILuc) bearing two NF-κB binding sites into 293T cells. Expression vectors of c-Rel, PKA-Cβ, or both together were cotransfected with the reporter construct to test their transcriptional activity. Figure 3A shows that cotransfecting PKA-Cβ with c-Rel significantly increased the activity of the reporter construct (approx. 25-fold of reporter activity), indicating that c-Rel was likely to be phosphorylated by PKA-Cβ to exert the maximum transcriptional activity. The significant increase in reporter activity by PKA-Cα alone may be due to the activation of endogenous p65 by PKA-Cα [15]. Control experiments using c-Rel and PKA-Cα showed slightly increased activation by PKA-Cα, indicating that c-Rel is unlikely to be a target of PKA-Cα. To test whether this enhanced activation could be further stimulated by coactivators p300/CBP we used a p300 expression vector in the c-Rel and PKA-Cβ cotransfection experiments. Figure 3B shows that p300 alone is capable of increasing the transcriptional activity mediated by c-Rel, suggesting its interaction with c-Rel. Cotransfection using c-Rel, PKA-Cβ and p300 further enhanced the transcriptional activation mediated by c-Rel and PKA-Cβ (Fig. 3B). Similar results were obtained using coactivator CBP to enhance the activation mediated by c-Rel and PKA-Cβ (data not shown). These results indicated that PKA-Cβ enhanced the transcriptional activity of c-Rel possibly through the activity of coactivators p300/CBP.
Fig. 3

Enhanced activation of a reporter construct (HIILuc) bearing NF-kB sites by c-Rel and PKA-Cβ. A Stimulation of c-Rel transcriptional activity by PKA-Cβ but not by PKA-Cα. The quantities of 100 ng HIIluc, 0.25 µg c-Rel, and 3 µg PKA-Cβ or PKA-Cα were used separately or together in cotransfection experiments. B Further increase in the activation of c-Rel transcriptional activity by p300. The quantities of 100 ng HIIluc, 0.25 µg c-Rel, 3 µg PKA-Cβ, and 5 µg p300 were used separately or together in cotransfection experiments

The consensus PKA phosphorylation site of c-Rel is not required by PKA-Cβ to increase the transcriptional activity of c-Rel

It has been demonstrated that p65 was phosphorylated by PKA-Cα in a PKA consensus site to mediate maximal activation through interaction with p300/CBP [15, 16]. In order to test whether PKA-Cβ also phosphorylates similar PKA consensus site in c-Rel to increase its transcriptional activity, a point mutant changing the putative phsophorylation site from serine to alanine was used. Figure 4A shows the consensus sequence comparing the putative PKA phosphorylation site between p65 and c-Rel. We tested whether the c-Rel mutant has decreased activity alone, or whether its activity can be enhanced by PKA-Cβ. Figure 4B shows that c-Rel mutant had decreased activity (approx. 50%) compared to c-Rel wild type when it was transfected alone. However, together with PKA-Cβ, there was no difference in the enhanced activity between cotransfecting c-Rel wild-type +PKA-Cβ and c-Rel mutant +PKA-Cβ. These results indicated that the consensus PKA phosphorylation site of c-Rel is unlikely to mediate enhanced activation caused by PKA-Cβ.
Fig. 4

Consensus PKA phosphorylation site in c-Rel is not required for enhanced activation induced by PKA-Cβ. A Comparison of the consensus PKA phosphorylation site between p65 and c-Rel. B Mutation of the consensus PKA phosphorylation site in c-Rel did not decrease its enhanced activation induced by PKA-Cβ. The quantities of 100 ng HIIluc, 0.25 µg c-Rel wild type or mutant, and 3 µg PKA-Cβ were used separately or together in cotransfection experiments

Interaction between PKA-Cβ and c-Rel

In order to demonstrate whether PKA-Cβ can interact with c-Rel to induce its phosphorylations coimmunoprecipitation assays were performed. Cotransfection of PKA-Cβ and c-Rel (wild type or point mutant) into 293T cells was performed, and their extracts were harvested and incubated with an anti-PKA-Cβ antibody. Figure 5 shows that the samples including control, c-Rel wild-type, and c-Rel mutant all contain similar amounts of cotransfected PKA-Cβ. Ηοwever, only samples transfected with c-Rel wild type or mutants showed the presence of c-Rel (wild type or mutant). These results show that PKA-Cβ is able to pull down c-Rel, suggesting the interaction between PKA-Cβ and c-Rel. However, the sample of transfected c-Rel point mutant still yielded similar levels of immunoprecipitated c-Rel with the sample of transfected c-Rel wild type, suggesting that mutation of PKA consensus site did not decrease the interaction between PKA-Cβ and c-Rel. This result demonstrated the interaction between PKA-Cβ and c-Rel independent of the consensus PKA phosphorylation site of c-Rel (Fig. 4A).
Fig. 5

Interaction between c-Rel and PKA-Cβ. Coimmunoprecipitation assays showed the ability of PKA-Cβ to pull down both wild-type and mutant c-Rel. PKA-Cβ pulled down similar amounts of wild-type and mutant c-Rel proteins

Phosphorylation of c-Rel by PKA-Cβ

In order to demonstrate that the interaction between c-Rel and PKA-Cβ causes the subsequent phosphoryaltion of c-Rel an immunoprecipitation-in vitro phosphorylation assay was performed. c-Rel or PKA-Cβ expression vector was transfected into 293T cells and their proteins were immunoprecipitated separately using an anti-c-Rel or an anti-PKA-Cβ antibody. The eluates from immunoprecipitate were mixed and incubated in the in vitro phosphorylation reaction mixture. Figure 6 shows that the lane of eluted c-Rel immunoprecipitate alone did not show the band of phosophorylated c-Rel, whereas the lane of eluted c-Rel +PKA-Cβ immunoprecipitate mixture showed the band of significantly phosphorylated c-Rel. This result demonstrated the direct phosphorylation of c-Rel by PKA-Cβ and supported the in vivo increase in the transcriptional activity of c-Rel stimulated by PKA-Cβ.
Fig. 6

Phosphorylation of c-Rel by PKA-Cβ. Immunoprecipitation in vitro phosphorylation assays showed the presence of phosphorylated c-Rel in the lane of eluted c-Rel +PKA-Cβ immunoprecipitate mixture compared to the lane of eluted c-Rel immunoprecipitate alone which did not have any phosphorylated c-Rel

Discussion

NF-κB is involved in many different cellular processes such as inflammation, immune responses, and apoptosis [1, 2, 3]. It responds to extracellular stimuli and translocates to the nucleus to carry out its functions [1, 2, 3]. Although its activation usually requires outside signaling, recent studies have shown that it can be activated without extracellular signaling [15, 16, 17, 31]. Our results also showed that the stimulation of c-Rel activity by PKA-Cβ occurs in a cAMP-independent manner in the absence of extracellular stimuli, consistent with recent observations [15, 16, 17, 31]. Due to the regulation of PKA-Cβ by c-MYC and the ability of PKA-Cβ to transform Rat1a cells [17] it is important to delineate the downstream signaling pathways regulated by PKA-Cβ to induce transformation. Although different downstream targets of PKA have been defined (CREB, cyclic AMP-responsive element modulator, NF-κB, activator protein 1, Raf-1) [15, 18, 19, 21, 22], it is still not known which target is phosphorylated by PKA-Cβ. We demonstrated that c-Rel is a possible target of PKA-Cβ by the following criteria: (a) c-Rel levels are increased in EREB.TCMyc cells induced to express exogenous c-MYC; (b) PKA-Cβ stimulates the transcriptional activity of c-Rel; (c) There is interaction between c-Rel and PKA-Cβ; and (d) phosphorylation of c-Rel by PKA-Cβ using an immunoprecipitation-in vitro phosphorylation assay. This is the first demonstration of c-Rel as a PKA-Cβ downstream target protein. The enhancement of c-Rel by PKA-Cβ may be due to their interaction and subsequent phosphorylation of c-Rel by PKA-Cβ. It was interesting that PKA-Cα had minimal ability to increase the activity of c-Rel in spite of the high homology of PKA consensus site located in the Rel homology domain of p65 and c-Rel (Fig. 4A). In addition, the mutation of PKA consensus site in c-Rel has not changed the ability of PKA-Cβ to enhance its transcriptional activity; compared to the dramatic decrease in transcriptional enhancement provided by PKA-Cα to activate p65 mutant [15]. This result indicated that there are other phosphorylation sites in c-Rel which could be phosphorylated by PKA-Cβ to enhance its transcriptional activity. Recent results show the existence of serine rich region in the C-terminal transactivation domain of Rel [32, 33]. Changing these serines into alanines one by one decreased significantly its transcriptional activity [33]. It remains to be determined which site(s) could be phosphorylated by PKA-Cβ to mediate maximal transcriptional activation. Further mapping of the putative c-Rel phosphorylation site(s) is possible using the immunoprecipitation in vitro phosphorylation assays followed by phosphopeptide mapping [15]. The PKA-Cβ stimulated c-Rel transcriptional activity most likely occurred during the c-MYC overexpression status as implicated from our results. Although we have demonstrated the interaction between c-Rel and PKA-Cβ and it is possible that c-Rel also interacts with p300/CBP, we cannot rule out the possibility that c-Rel also interacts with other coactivators phosphorylated by PKA-Cβ to induce maximal transactivation.

c-Rel (or its viral homolog v-Rel) has been implicated in tumorigenesis including B-cell lymphoma, Hodgkin’s lymphoma, multiple myeloma, and other solid tumors [8, 9, 10, 11]. Constitutive activation of c-Rel is able to transform primary chicken spleen cells [12]. Transgenic mice overexpressing mouse mammary tumor virus driven c-Rel develop mammary tumors [13]. In addition, c-Rel is able to activate cell cycle progression through activation of cyclin D1 [7]. Other c-Rel target genes such as Bcl-xL, STAT5 target genes, MIP-1, and sca-2 may play important roles in mediating c-Rel induced antiapoptosis, transformation, and tumorigenesis [34, 35, 36]. More downstream target genes regulated by c-Rel to mediate transformation still remain to be defined. This result links the ability of PKA-Cβ to induce soft agar colony formation to the activation of c-Rel, which is capable of mediating transformation. Due to the requirement of the transactivation activity of c-Rel to induce transformation [32, 33], the ability of PKA-Cβ to enhance the transcriptional activity of c-Rel certainly played an important role in c-Rel mediated transformation and tumorigenesis.

PKA activity has been well characterized and some of their downstream phosphorylation targets (e.g., CREB, NF-κB, activator protein 1, Raf-1) [15, 18, 19, 21, 22] have been identified recently. However, these targets were limited to PKA-Cα. It has not been demonstrated whether different PKA subunits have different phosphorylation targets. It is possible that different PKA subunits exert their effects through the phosphorylation of different targets to mediate different functions. The ability of PKA-Cβ to enhance the activity of c-Rel links the ability of c-MYC to induce transformation to the downstream effects of phosphorylated c-Rel to mediate transformation. It has been demonstrated that NF-κB/Rel activity could regulate c-MYC expression by binding to c-MYC promoter in the WEHI B cell line or Ramos Burkitt’s lymphoma cell line [37, 38, 39]. Reciprocal regulation by c-MYC to regulate PKA-Cβ and subsequently activate c-Rel demonstrated the positive feed back loop in the c-MYC-c-Rel pathways. These results highlight the importance of cross talking and signal amplification between different signal transduction pathways such as PKA, c-Rel, and c-MYC.

Acknowledgements

We thank Dr. R. Dalla-Favera, in whose laboratory this work was initiated. We are grateful to Drs. D. Baltimore, U. Siebenlist, and R. Goodman for the gifts of HIILuc, pMT2T-c-Rel, and CBP/p300 plasmids. This work was supported by National Science Council NSC 91-2320-B-002-070 and NSC 92-2320-B010-078, and National Health Research Institutes NHRI-EX93-9329SI to K.J.W.

Copyright information

© Springer-Verlag 2004