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

J. Biol. Chem., Vol. 277, Issue 50, 48664-48676, December 13, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/50/48664    most recent M207029200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sheridan, C. M. Articles by Gardner, P. Search for Related Content PubMed PubMed Citation Articles by Sheridan, C. M. Articles by Gardner, P. Social Bookmarking                      What's this?

Protein Kinase A Negatively Modulates the Nuclear Accumulation of NF-ATc1 by Priming for Subsequent Phosphorylation by Glycogen Synthase Kinase-3^*

Colleen M. Sheridan^§, E. Kevin Heist^¶, Chan R. Beals^**, Gerald R. Crabtree^**, and Phyllis Gardner^§§

From the  Program in Immunology, Department of Molecular Pharmacology, Department of Medicine, ^** Department of Pathology and Developmental Biology, Howard Hughes Medical Institute, and ^¶ Department of Neurobiology, Stanford University, Stanford, California 94305

Received for publication, July 13, 2002, and in revised form, September 24, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The nuclear localization and transcriptional activity of the NF-ATc family of transcription factors, essential to many developmental, differentiation, and adaptation processes, are determined by the opposing activities of the phosphatase calcineurin, which promotes nuclear accumulation of NF-ATc, and several kinases, which promote cytoplasmic accumulation. Many reports suggest that protein kinase A (PKA) negatively modulates calcineurin-mediated NF-ATc activation. Here we show that overexpression of PKA causes phosphorylation and cytoplasmic accumulation of NF-ATc1 in direct opposition to calcineurin by phosphorylating Ser-245, Ser-269, and Ser-294 in the conserved serine-proline repeat domain, and that mutation of these serines blocks the effect of PKA. Activation of endogenous PKA is similarly able to promote phosphorylation of these sites on NF-ATc1 in two lymphoid cell lines. We further show that a complete block of NF-ATc1 nuclear localization by PKA requires a second kinase activity that can be supplied by glycogen synthase kinase-3 (GSK-3), and that mutation of either the PKA phosphorylation sites or the upstream GSK-3 sites prevents the effect of PKA. Thus, we propose that PKA functions cooperatively as a priming kinase for further phosphorylation by GSK-3 to oppose calcineurin-mediated nuclear accumulation and transcriptional activity of NF-ATc1 and that, through this mechanism, PKA may be an important modulator of many NF-ATc-dependent processes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Many developmental, differentiation, and activation processes, such as dorsoventral axis formation (10), cardiac valve formation (11, 12), muscle differentiation (13), skeletal myocyte and cardiac hypertrophy (14, 15), and T cell activation (16, 17), utilize calcium influx and calcineurin activation pathways leading to the nuclear localization of the cytoplasmic components of the NF-AT¹ transcription complexes (2). The cytoplasmic components of these complexes (NF-ATc), first identified in T cells (18, 19), are encoded by four genes,² NF-ATc1 (NF-ATc) (20), NF-ATc2 (NF-ATp/NFAT1) (21), NF-ATc3 (NF-AT4/NF-ATx) (22-24), and NF-ATc4 (NF-AT3) (22), and each has several RNA splice variant isoforms. The NF-ATc family members exhibit several properties, such as distinct cell-type and differentiation expression patterns, different DNA binding affinities, the varied requirement for heterologous DNA binding partners, and complex regulation of their subcellular accumulation, which may enable them to be involved in such a diverse array of cellular processes (1, 25).

An hypothesis of the regulation of NF-ATc subcellular accumulation is that it undergoes continuous rapid cycling between the cytoplasm and the nucleus, where a dynamic equilibrium between its import and export rates determines its steady-state nuclear or cytoplasmic accumulation (4, 26, 27). These rates are regulated through the relative activity levels of the calcium/calmodulin-dependent protein phosphatase calcineurin (28, 29) and various opposing NF-ATc kinases. Dephosphorylation of NF-ATc by calcineurin increases NF-ATc nuclear accumulation and NF-AT-mediated gene transcription, whereas phosphorylation by the opposing NF-ATc kinases increases NF-ATc cytoplasmic accumulation and the cessation of transcriptional activity (1, 3).

This kinase/phosphatase interplay determines the phosphorylation state of the ~20 conserved serines found in the amino-terminal domain termed the NF-AT homology region (NHR) (30). The NHR is necessary and sufficient to confer calcium-responsive localization to a green fluorescent protein (GFP) fusion partner and carries several distinct amino acid motifs: the serine-rich region (SRR), three serine-proline (SP) repeats with the sequence spXXSPXXSPXXSPrXsXX(D/E)(D/E), an amino-terminal nuclear localization sequence (N-NLS), and a nuclear export sequence (NES) (31, 32). For the NF-ATc1 protein, dephosphorylation of critical serine residues in the SRR or the first SP repeat (SP1) is thought to induce a conformational change that unmasks its two functionally redundant NLSs, resulting in subsequent nuclear accumulation (31). Others have found that phosphorylation is required for nuclear export, suggesting that phosphorylation, although masking the NLS, also unmasks an NES (4, 33).

Regulation of NF-ATc subcellular localization through this multisite phosphorylation mechanism adds an opportunity for complex signal integration (16, 25). One report describes the requirement for a threshold concentration of nuclear NF-AT to activate NF-AT-dependent gene transcription (34), suggesting that a stimulus which serves to counteract the drive for NF-ATc nuclear import, even if only to a moderate degree, may have dramatic consequences on NF-AT-mediated gene transcription. Many NF-ATc kinases have been reported to phosphorylate from one to many of the serines in the NHR, and several of these kinases, once overexpressed, can oppose calcineurin-mediated nuclear accumulation of specific NF-ATc family members (35-39). Thus, the large number of possible phosphorylation sites creates the potential for graded modulation of NF-ATc activity through the activation of these various kinase signaling pathways (33).

Mounting evidence suggests that the cAMP/PKA pathway can modulate the activity of NF-ATc. The membrane-permeable cAMP analog dibutyryl-cAMP (Bt₂cAMP) inhibits the calcium-mediated transcriptional activity of the distal interleukin-2 (IL-2) NF-AT DNA element (6). Overexpression of either constitutively active calcineurin or of NF-ATc can overcome IL-2 promoter sensitivity to this cAMP/PKA opposition (8, 9). Furthermore, NF-AT DNA binding activity, dependent on the T cell source (9, 40-42), is decreased when the cells are stimulated in the presence of cAMP-elevating agents (5, 7, 43, 44).

We have previously reported that PKA is able to phosphorylate NF-ATc1 in vitro, providing the necessary priming phosphates for further rephosphorylation by the NF-ATc kinase, glycogen synthase kinase-3 (GSK-3) (35). GSK-3 is able to oppose ionomycin-induced NF-ATc1 nuclear accumulation and transcriptional activity, as well as to increase the rate of NF-ATc1 nuclear export. In addition, GSK-3 is thought to be a strong candidate as a physiological NF-ATc kinase (16). However, the in vivo relevance and mechanistic details of the in vitro ability of PKA to prime for GSK-3 phosphorylation of NF-ATc1 have not yet been demonstrated.

Here we show that PKA phosphorylates NF-ATc1 in vivo and has the ability to oppose both calcineurin-mediated NF-ATc1 dephosphorylation and nuclear accumulation. It does this by acting as a priming kinase for subsequent rephosphorylation of the SP2 and SP3 repeat motifs of NF-ATc1 by GSK-3, and by synergizing with limiting amounts of GSK-3 to oppose activated calcineurin. Thus, activation of PKA may enable GSK-3 to more effectively oppose the calcineurin-mediated activation of NF-ATc1 during simultaneous stimulation of both the calcium and cAMP pathways.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA Constructs and Recombinant Proteins-- All of the mammalian expression constructs used in this study were under control of the SR promoter. The DNA constructs included the NFATZH -galactosidase NF-AT reporter gene construct (45); the X₃ZH -galactosidase AP-1 reporter gene construct (45); the catalytic subunit of protein kinase A (PKAc) (46); the constitutively active truncated form of calcineurin CaM-AI (cCaN) (47); the constitutively active mutant of multifunctional calcium/calmodulin-dependent protein kinase II (cCaMKII) (48); the constitutively active truncated form of protein kinase C, PKC (cPKC) (46); the wild-type full-length cDNA of nuclear factor of activated T cells c1 (NF-ATc1), either hemagglutinin-tagged (pSH102) (20) or FLAG-tagged (pSH160c) (23); the wt NF-ATc1-GFP fusion construct pSH160E-J/C655-GFP made by cloning GFP cDNA from the pEGFP-1 (Clontech) in frame into the EcoRI site of pSH160c, resulting in a truncation of NF-ATc1 at amino acid 655 fused to the amino terminus of enhanced GFP; the single S245A, S269A, and S294A NF-ATc1 mutants and the triple S245A/S269A/S294A m3 NF-ATc1 mutant in the pSH102 construct made with the Transformer^TM site-directed mutagenesis kit (Clontech); the triple mutant m3 NF-ATc1-GFP construct made by subcloning the MluI-BstXI fragment of m3 NF-ATc1 into the MluI-BstXI complementary site of wt NF-ATc1-GFP; and the mGSKSP23 NF-ATc1-GFP mutant made by blunt-end ligation of nucleotides 731-996 PCR fragment (wt SRR, wt SP1, and S233A/S237A/S241A SP2) amplified from m233 NF-ATc1 mutant (SP2 serines 233, 237, and 241 mutated to alanine using the Stratagene QuikChange^® site-directed mutagenesis kit) and of nucleotides 997-1875 PCR fragment (wt N-NLS, S278A/S282A/S286A/S290A SP3) amplified from NF-ATc1A S/A SRRSP123 mutant (SRR, SP1, SP2, and SP3 serines mutated to alanine using the Stratagene QuikChange^® site-directed mutagenesis kit), then digested with MluI/BstXI and ligated into complementary MluI/BstXI sites in wt NF-ATc1-GFP. All of the bacterial expression constructs used in this study were cloned into pGEX-3X (Amersham Biosciences). These expression constructs included: pGSP (wt NF-ATc1-GST) (35) and m3 GSP (m3 NF-ATc1-GST), serine-to-alanine triple mutant of pGSP at S245A, S269A, and S294A made by cloning the MluI-BstXI fragment from m3 NF-ATc1 into the complementary MluI-BstXI site in pGSP. All mutants were verified by sequencing.

Cell Culture and Reagents-- TAg Jurkat and TAB cells were grown in RPMI 1640 media (Life Technologies, Inc.) containing 10% heat-inactivated defined fetal bovine serum (HyClone), 100 units/ml penicillin (Life Technologies, Inc.), and 100 µg/ml streptomycin (Life Technologies, Inc.) in humidified incubators at 5% CO₂ and 37 °C. HEK 293T cells were grown in high glucose DMEM (Life Technologies, Inc.) containing 10% heat-inactivated defined fetal bovine serum (HyClone), 100 units/ml penicillin (Life Technologies, Inc.), and 100 µg/ml streptomycin (Life Technologies, Inc.) in humidified incubators at 10% CO₂ and 37 °C. The reagents used this study were: Phorbol-12-myristate-13-acetate (PMA; Calbiochem); ionomycin, free acid (Calbiochem); cyclosporin A (CsA; Calbiochem); N⁶,O^2'-dibutyryl-adenosine 3',5'-cyclic monophosphate (Bt₂cAMP; Calbiochem); 3-isobutyl-1-methylxanthine (IBMX; Calbiochem); cycloheximide (Sigma); staurosporine (Sigma); lithium chloride (Sigma); indirubin-3-monoxime, alsterpaullone, and roscovitine (Calbiochem); the peptide PKA inhibitor (Upstate Biotechnology); Complete^TM Mini protease inhibitor mixture tablets, EDTA-free (Roche Molecular Biochemicals); leupeptin, aprotinin, and pepstatin A (Sigma); isopropyl-1-thio--D-galactoside (Life Technologies, Inc.); mouse anti-HA monoclonal antibody 12CA5 (Roche Molecular Biochemicals); mouse anti-NF-ATc1 monoclonal antibody 7A6; rabbit anti-NF-ATc1 polyclonal antibody (gift from Dr. Tim Hoey, Tularik); rhodamine-conjugated goat IgG fraction to mouse IgG, peroxidase-conjugated goat affinity-purified anti-mouse IgG, and peroxidase-conjugated goat affinity-purified anti-rabbit IgG (Cappel Research Products, Organon Teknika Corp.); biotin-SP-conjugated AffiniPure goat anti-mouse IgG (H+L) and FITC-conjugated streptavidin (Jackson Immunoresearch Laboratories, Inc.); and TO-PRO-3 (Molecular Probes).

Transfection of HEK 293T and TAg Jurkat Cells-- HEK 293T cells were plated at 50% confluence in 35-mm tissue culture dishes the night before transfection. 2 µg of indicated plasmid DNAs were mixed with 100 µl of Opti-MEM medium (Life Technologies, Inc.) then mixed with 8.5 µl of LipofectAMINE^TM reagent (Life Technologies, Inc.) in 100 µl of OPTI-MEM medium, incubated 15-45 min at room temperature, mixed with 800 µl of serum-free DMEM (Life Technologies, Inc.), placed on cells pre-washed once with serum-free DMEM, incubated at 37 °C for 5 h, and transferred to cell culture medium for 24-48 h. 10⁷ TAg Jurkat cells were resuspended in cytomix (49) (120 mM KCl, 0.15 mM CaCl₂, 10 mM K₂HPO₄/KH₂PO₄, pH 7.6, 25 mM HEPES, pH 7.6, 2 mM EGTA, pH 7.6, 5 mM MgCl₂, pH adjusted with KOH, 2 mM ATP, pH 7.6, 5 mM glutathione), and DNA was added to a final volume of 295 µl. Cells were electroporated at 250 volts, 960 microfarads, room temperature, then transferred to 10 ml of cell culture medium for 48 h.

Reporter Gene Assays-- TAg Jurkat T cells were electroporated as described above with 10 µg of indicated reporter plasmid, 2 µg of pGL3 (transfection control), ± 20 µg of PKAc and/or pSR as indicated to 37 µg of total DNA, incubated 48 h, and then stimulated overnight with 50 ng/ml PMA and 2 µM ionomycin or Me₂SO vehicle. Cells were lysed in Reporter Lysis Buffer (Promega) and gene reporter assays were performed as described in the Galacto-Light Plus^TM kit (Tropix) and the Luciferase Assay System kit (Promega) protocols and read with a Monolight 2001 luminometer (Analytical Luminescence Laboratory). Protein assays were performed as described in the Bio-Rad Protein Assay Reagent protocol. The transcriptional activation values were calculated as "-galactosidase/luciferase" = (-galactosidase/µg of protein)/(luciferase/µg of protein). The PMA + ionomycin value was considered to be 100% stimulation; therefore, using CA-Cricket^®Graph III^TM (Computer Associates International, Inc.), each value was calculated relative to this giving "relative -galactosidase activity" = (-galactosidase/luciferase) × 100/(value of PMA + ionomycin).

Immunostaining-- HEK 293T cells were fixed in either 3.75% formaldehyde or 4% paraformaldehyde in PBS for 10 min at room temperature, permeabilized with 20 °C methanol for 2 min, rehydrated with PBS, blocked in 1% BSA in PBS for 1 h, incubated 2 h to overnight with 5 µg/ml anti-HA (12CA5) monoclonal antibody and then 1:250 rhodamine-conjugated goat anti-mouse IgG for 45 min in a humidified chamber at 37 °C and viewed with water-immersion 40× objective on Zeiss Axiophot fluorescence microscope and photographed with Elite Chrome 400 ASA 35-mm film for color slides (Eastman Kodak Co.). HEK 293T cells expressing GFP constructs were stimulated as indicated, then fixed in 4% paraformaldehyde in PBS for 10 min, washed, and photographed. TAB cells were cytospun (700 rpm, 7 min) onto poly-L-lysine glass slides (Sigma) after indicated stimulation, briefly air-dried, and fixed in 2% paraformaldehyde in PBS at room temperature for 10 min, then permeabilized in 20 °C 100% methanol for 2 min, and rehydrated in PBS. Cells were blocked in 2% fetal calf serum, 1% BSA, PBS for 1-2 h at room temperature, probed with 1:1000 7A6 antibody for 2 h, then 1:500 biotin-conjugated goat anti-mouse for 1 h, then 1:500 FITC-conjugated streptavidin for 1 h. The red nuclear stain TO-PRO-3 (Molecular Probes) was diluted 1:1000 in PBS and incubated on cells for 5 min and washed. The Slow-Fade-Light kit (Molecular Probes) was used to prevent fading, and then cells were covered with glass coverslip and sealed. Cells were viewed on a Amersham Biosciences MultiProbe 2010 confocal laser scanning microscope with 60× oil-immersion objective located in the Cell Sciences Imaging Facility (Stanford University, Stanford, CA). Data were collected with the Amersham Biosciences ImageSpace (version 3.2) software.

Stimulation of TAg Jurkat and TAB Cells-- Cells were pre-incubated with 20 µg/ml cycloheximide for 15 min, then stimulated at 37 °C with either 1 µg/ml CsA, 1 mM Bt₂cAMP + 1 mM IBMX, or Me₂SO vehicle () for 30 min before addition of 1 µg/ml ionomycin or Me₂SO () for 30 min as indicated.

Cell Lysis and Western Blot of HEK 293T Cells-- 150 µl of either 1% Nonidet P-40 buffer (1% Nonidet P-40, 150 mM NaCl, 50 mM Tris, pH 8.0) or radioimmune precipitation assay buffer (1% Nonidet P-40, 0.5% deoxycholic acid, 0.1% SDS, 150 mM NaCl, 50 mM Tris, pH 8.0) plus protease inhibitors (1 mM EDTA, 10 µg/ml each of leupeptin and aprotinin, 1 µg/ml pepstatin A, 2 mM phenylmethylsulfonyl fluoride) and phosphatase inhibitors (1 mM sodium orthovanadate, 0.4 mM sodium molybdate, 10 mM sodium pyrophosphate) were added to HEK 293T cells, incubated for 30 min on ice, sonicated 15 s, centrifuged, and the supernatant was collected. 10-20 µl of supernatant were run on 7.5 or 6% separating, 4% stacking SDS-PAGE Bio-Rad mini-gels at 150 V for 1 h, then transferred to nitrocellulose, blocked overnight in 5% BSA, 0.2% Tween 20, 1× Tris-buffered saline, incubated with primary antibody as indicated 1.5 h at room temperature, washed, incubated with horseradish peroxidase-conjugated secondary antibody as indicated 45 min at room temperature, washed, developed with the Amersham ECL kit, and exposed to film.

Isolating NF-ATc1 Peptides for Sequencing-- Purified wt NF-ATc1-GST protein was used to isolate the peptides phosphorylated by PKA. GST fusion proteins were purified by lysing bacteria in PBS, 1% Triton X-100, 0.1% -mercaptoethanol plus protease inhibitors (1 µg/ml each of leupeptin and aprotinin, 0.1 µg/ml pepstatin A, and 174 µg/ml phenylmethylsulfonyl fluoride) and sonicated on ice three times for 1 min each and centrifuged for 30 min at 4 °C. Glutathione-Sepharose (Amersham Biosciences) was added to supernatant, rotated at room temperature 45-60 min, washed with lysis buffer, then washed three times with cold Tris-buffered saline, and resuspended in 20 mM Tris, pH 7.5. In vitro PKA phosphorylation was performed with 4 µg of wt NF-ATc1-GST/beads, one reaction with [³²P]ATP and one with cold ATP, with 50 mM PIPES, pH 7.5, 10 mM MgCl₂, 20 µM cold ATP (plus 1 µCi of [³²P]ATP for hot reaction only), 10 µl of immunoprecipitate, 0.2 µg of PKAc (Sigma), and incubated for 30 min at 30 °C. Samples were run on SDS-PAGE gel, transferred to PVDF, rinsed with water, and exposed to film. ³²P-Labeled wt NF-ATc1-GST and corresponding cold-labeled bands were cut out. Phosphotryptic peptide mapping protocol was followed as described below, except that 1000 cpm of ³²P-labeled sample was combined with all of the cold-labeled sample before the last wash in pH 1.9 buffer to ensure that sufficient peptide was available for both sequencing and radio-isotope detection. After mapping, ³²P-labeled peptides were located on the cellulose plate, scraped off with a razor blade, and collected in a microcentrifuge tube, then washed with water. Eluted peptides were lyophilized. The sample was then resuspended in 50 µl of water for Edman degradation peptide sequencing at the PAN Facility at Stanford University.

Phosphotryptic Peptide Mapping-- Either GST fusion proteins (see above) or NF-ATc1 proteins immunoprecipitated from HEK 293T cells were used in phosphotryptic peptide mapping. Cell lysis was performed in 1% Nonidet P-40 buffer as described above. NF-ATc1 was immunoprecipitated with 7A6 anti-NF-ATc1 antibody and Protein G-Sepharose at 4 °C overnight, then washed three times in lysis buffer. NF-ATc1 had been either in vivo orthophosphate-labeled (see below) or in vitro phosphorylated with PKA as GST fusion proteins above. Phosphotryptic peptide mapping was performed as described (50). Thin-layer electrophoresis dimension was run in pH 1.9 buffer for 30 min at 1000 volts, and the thin-layer chromatography dimension was run in phosphochromatography buffer (37.5% n-butanol, 30.4% pyridine, 7.5% glacial acetic acid) overnight.

[³²P]Orthophosphate in Vivo Labeling-- Cells were starved for 1 h in phosphate- and serum-free medium (RPMI 1640 for TAB cells, and DMEM for 293T cells), then incubated for 3 h in 500 µCi/ml [³²P]orthophosphate in phosphate-free medium with 10% fetal calf serum (dialyzed against HEPES to remove excess ATP). Cells were washed in ice-cold PBS and lysed in radioimmune precipitation assay buffer plus protease inhibitors (Complete^TM Mini protease inhibitor mixture tablets, EDTA-free (Roche Molecular Biochemicals), 1 mM EDTA, 1 µg/ml pepstatin A) plus phosphatase inhibitors (1 mM sodium orthovanadate, 1 mM sodium fluoride, 500 ng/ml cyclosporin A, 100 µM -glycerol-phosphate, 1 mM sodium pyrophosphate, 30 µM sodium molybdate) plus kinase inhibitors (20 µM peptide PKA inhibitor, 2 µM staurosporine, 10 mM lithium chloride) and immunoprecipitated with 7A6 (anti-NF-ATc1) and Protein G-Sepharose (Amersham Biosciences). For mapping, samples were run on SDS-PAGE gels and transferred to PVDF for further analysis as described above. For quantitation of phosphorylation and comparison with protein levels, samples were run on SDS-PAGE gels and transferred to nitrocellulose. Membranes were wrapped in Saran and exposed to a phosphorimager screen for 12-24 h, developed on a STORM 840 Phospho Screen phosphorimager and analyzed with ImageQuantMac version 1.2 software (Amersham Biosciences). The membrane was assayed by Western blot with 7A6 (anti-NF-ATc1) and horseradish peroxidase-conjugated goat anti-mouse IgG as described above. The film was scanned into Adobe Photoshop 5.0 and analyzed using NIH Image 1.58 software. The data were calculated by (phosphate quantitation  background)/(protein quantitation  background) relative to Me₂SO control levels (, ) and graphed with CA-Cricket^® Graph III^TM (Computer Associates International, Inc.).

Site-directed Mutagenesis-- Mutagenesis was performed as described either in the Transformer^TM site-directed mutagenesis kit (Clontech) protocol or in the QuikChange^® site-directed mutagenesis kit (Stratagene) protocol as indicated.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

PKA Inhibits NF-AT Transcriptional Activity but Enhances AP-1 Transcriptional Activity-- Several studies have indicated that cAMP and the cAMP-dependent protein kinase PKA inhibit NF-AT-mediated transcriptional activity at a level distal to the calcium signal (6-8). Consistent with these results, we found that the catalytic subunit of PKA (PKAc; Ref. 46) was able to inhibit the transcriptional activity of an NF-AT reporter gene construct in TAg Jurkat T cells stimulated with the T cell activation mimetics PMA and ionomycin (Fig. 1). The NF-AT reporter gene, NFATZH (19, 45), consists of a trimer of the region from 257 to 286 of the IL-2 enhancer linked to the IL-2 minimal promoter (72 to +47). This construct requires both PMA and ionomycin stimulation for full transcriptional activity (51) and is a composite DNA element for the cooperative binding of the calcium-responsive NF-ATc and PMA-inducible AP-1 transcription factors (52). To discern whether PKAc was affecting the NF-ATc component, the AP-1 component, or both, we then tested whether PKAc could affect an AP-1 reporter gene construct. This construct, X₃ZH (45), consists of three copies of the AP-1 binding site from the metallothionein gene enhancer linked to the SV40 minimal promoter. We found that, in the absence of PMA and ionomycin, PKAc enhanced AP-1 transcriptional activity 6-fold over basal activity, and in the presence of PMA and ionomycin, enhanced it to an even greater degree (Fig. 1). This is consistent with reports demonstrating that PKA can increase AP-1 activity (53, 54). These results suggest that PKA is likely to be acting at the level of the cytosolic NF-ATc component, and not the AP-1 component, to down-regulate NF-AT transcriptional activity.

View larger version (9K): [in this window] [in a new window]   Fig. 1.   PKA inhibits NF-AT transcriptional activity but enhances AP-1 transcriptional activity. TAg Jurkat T cells were transfected with the indicated plasmids, and reporter gene transcriptional assays were performed as described under "Experimental Procedures." Stimulation with 50 ng/ml PMA and 2 µM ionomycin (black columns) or Me2SO vehicle (white columns) is indicated, and values are averages of three independent experiments ± standard deviation shown relative to PMA + ionomycin activity level, which was given the arbitrary value of 100%.

PKA Opposes Calcineurin-mediated Nuclear Accumulation and Dephosphorylation of NF-ATc1-- A transcription factor can be regulated in three main ways: localization, DNA binding, and transactivation. The primary known form of NF-ATc regulation is its cytoplasmic localization in a resting cell, thus sequestering it from access to its target genes until given the appropriate stimulus (26). To test the hypothesis that PKA may oppose the ability of calcineurin to induce NF-ATc nuclear accumulation, we studied the subcellular localization of exogenous NF-ATc1, co-transfected with PKA or calcineurin (either alone or in combination) in HEK 293T cells. Transfection of NF-ATc1 alone resulted in its complete cytoplasmic accumulation in these cells (Fig. 2A). In contrast, NF-ATc1 co-transfected with a constitutively active calcineurin construct (cCaN) (47) resulted in exclusive nuclear accumulation of NF-ATc1 in ~90% of the transfected cells (Fig. 2A). This was similar to results obtained in other adherent cell models, thus supporting the use of this system to study NF-ATc subcellular localization (4, 31). We then tested for the ability of PKA to oppose calcineurin-mediated NF-ATc1 nuclear accumulation. Cotransfection of PKAc opposed the cCaN-mediated nuclear accumulation of NF-ATc1 by ~64% (Fig. 2A). We also tested an NF-ATc1-GFP construct that was previously shown to behave in an identical manner to the full-length protein in a subcellular localization assay (31). NF-ATc1-GFP behaved similarly to full-length NF-ATc1 in response to either cCaN or cCaN plus PKAc, although the GFP fusion construct appeared to be moderately less responsive to PKAc than full-length NF-ATc1 (Fig. 2A).

View larger version (37K): [in this window] [in a new window]   Fig. 2.   PKA opposes calcineurin-mediated nuclear accumulation and dephosphorylation of NF-ATc1. A, calcineurin-mediated nuclear accumulation of both full-length NF-ATc1 and NF-ATc1-GFP are opposed by PKA. HEK 293T cells were co-transfected with pSH102 (full-length HA-tagged NF-ATc1) and the following test constructs from upper left, clockwise: pSR, cCaN, cCaN + PKAc at 1:14 ratio, and cCaN + cCaMKII at 1:14 ratio. 48 h later, cells were immunostained with monoclonal anti-HA antibody (12CA5) and rhodamine-conjugated goat anti-mouse IgG. Original magnification, ×40. Graph shows mean percentage of nuclear ± S.D. of HEK 293T cells transfected with 0.5 µg of either pSH160 (full-length FLAG-tagged NF-ATc1) or pSH160E-J (FLAG-tagged NF-ATc1-GFP) and either pSR empty control vector (no stim), 0.1 µg of cCaN, or 0.1 µg of cCaN + 0.05 µg of PKAc. After 24 h, cells were fixed. The full-length NF-ATc1 transfectants were further permeabilized and stained with anti-FLAG antibody as described under "Experimental Procedures." 100 cells were counted in each of three independent experiments and were scored as mostly nuclear, mostly cytoplasmic, or both. B, HEK 293T cells were transfected with 0.5 µg of each of the indicated constructs and pSR control vector to a final 2.0 µg of total DNA. Cell lysates were immunoblotted with a rabbit polyclonal anti-NF-ATc1 antibody.

As a control for nonspecific kinase effects, we tested another serine/threonine protein kinase calcium/calmodulin-dependent protein kinase II because both PKAc and cCaMKII (48) are known to inhibit the expression of the IL-2 and NF-ATc/AP-1 composite DNA element reporter gene constructs (55-57). However, unlike PKAc, cCaMKII had no effect on cCaN-mediated NF-ATc1 nuclear accumulation (Fig. 2A), suggesting that cCaMKII may inhibit these promoters through an alternate mechanism.

Next, we examined whether PKA could also oppose calcineurin-mediated dephosphorylation of NF-ATc1, as the localization of NF-ATc1 within a cell has been correlated with its phosphorylation state; phosphorylated NF-ATc1 accumulates in the cytoplasm, whereas dephosphorylated NF-ATc1 accumulates in the nucleus (1). Cotransfection with PKAc opposed cCaN-mediated NF-ATc1 dephosphorylation, as shown by its decrease in mobility on an SDS-polyacrylamide gel (Fig. 2B, compare lanes 3 and 10). In contrast, cCaMKII and another serine-threonine protein kinase in its constitutively active form, protein kinase C (cPKC) (46), had no effect on cCaN-mediated dephosphorylation of NF-ATc1 (Fig. 2B, compare lane 3 with lanes 5, 9, and 11), which is consistent with the roles of protein kinase C and calcineurin as co-activators of IL-2 gene transcription (58, 59) and of the inability of cCaMKII to oppose cCaN-mediated NF-ATc1 nuclear accumulation (Fig. 2A). Together these results strongly support our hypothesis that PKA can down-regulate NF-AT transcriptional activity by opposing the calcineurin-mediated dephosphorylation and nuclear accumulation of NF-ATc1. Interestingly, PKAc appeared to cause the hyperphosphorylation of NF-ATc1 compared with control levels (Fig. 2B, compare lanes 6 and 2) and was able to phosphorylate both immunoprecipitated cytoplasmic and nuclear forms of full-length NF-ATc1 in vitro (data not shown), suggesting that PKAc induced the phosphorylation of additional sites beyond those phosphorylated in the resting basal state. In contrast, both cPKC and cCaMKII did not cause hyperphosphorylation of NF-ATc1 (Fig. 2B, lanes 4 and 7), nor did either kinase phosphorylate NF-ATc1 in vitro (data not shown).

PKA Opposes Calcineurin-mediated NF-ATc1 Nuclear Accumulation by Direct Phosphorylation of Three Serines Found in the Ser-Pro Repeat Domain and NH₂-terminal NLS-- We next assessed whether PKA opposes calcineurin-mediated dephosphorylation and nuclear accumulation via direct phosphorylation of NF-ATc1. We have previously reported that PKA phosphorylates NF-ATc1 in vitro at serines 245 and 269, which are found in the NHR and localized to the carboxyl terminus of the SP2 repeat motif and the N-NLS, respectively (Fig. 3A; Ref. 35). A third site has now been identified on Ser-294 and is localized to the carboxyl terminus of the SP3 repeat motif in a homologous position to that of Ser-245 in the SP2 repeat (Fig. 3A). Single serine-to-alanine mutations of each site in full-length human NF-ATc1 demonstrated the absence of its respective in vitro PKA-phosphorylated phosphopeptide on phosphotryptic peptide maps (Fig. 3B). Mutation of all three serines to alanines in the NF-ATc1-GST fusion protein, m3 NF-ATc1-GST, resulted in the complete absence of PKA phosphorylation in vitro (Fig. 3C).

View larger version (60K): [in this window] [in a new window]   Fig. 3.   PKA phosphorylates NF-ATc1 at Ser-245, Ser-269, and Ser-294. A, diagram of NF-ATc1 and the amino acid sequence detail surrounding the PKA phosphorylation sites. Asterisk, NLS; number sign, NES. The serines phosphorylated by PKA are circled. The circled aa, filled triangles, and carets indicate serine to alanine mutations in the m3, mSRR, and mGSKSP23 mutant NF-ATc1 constructs, respectively. The underlined KTT is a homologous PKA consensus site, not phosphorylated by PKA on NF-ATc1, as explained in the text. The multiple SP repeats are shown in solid line boxes. The basic amino acids KRK of the N-NLS are shown in a dashed box. B, in vitro PKA-phosphorylated wt NF-ATc1-GST (wt, upper left panel) substrate was analyzed by phosphotryptic peptide mapping. Analysis of the duplicated spots at each site found them to be derived from the same phosphopeptide, and they are thought to be a result of alkylation during the mapping process. The resulting three phosphopeptides were isolated from the cellulose plates and sent for Edman degradation sequencing. Site 1 = Ser-269, site 2 = Ser-245, site 3 = Ser-294. Single serine to alanine full-length NF-ATc1 mutants S269A-pSH102 (S269A, upper right panel), S245A-pSH102 (S245A, lower left panel), and S294A-pSH102 (S294A, lower right panel) were expressed in HEK 293T cells, immunoprecipitated with anti-HA antibody, phosphorylated in vitro with PKA, and analyzed by phosphotryptic peptide mapping. Dashed circles indicate absence of respective phosphopeptides. Note that the peptide containing Ser-294 in wt NF-ATc1-GST is 12 aa, whereas in pSH102 it is the full-length 42 aa. This longer peptide does not elute as well as the shorter peptide from PVDF, resulting in a less pronounced appearance. C, GST, wt NF-ATc1-GST, and the triple S245A/S269A/S294A serine to alanine mutant m3 NF-ATc1-GST were assayed for in vitro PKA phosphorylation. The upper panel is the Coomassie stain, and the bottom panel is the autoradiograph of the same gel. Several degradation products of the NF-ATc1-GST fusion proteins can be seen after Coomassie stain; the full-length product is indicated. D, PKA increases overall phosphorylation of wt NF-ATc1, not just at PKA sites. HEK 293T cells were transfected with either 0.5 µg of pSH102 (wt NF-ATc1) or 0.5 µg of m3 pSH102 (m3 NF-ATc1) and either pSR control vector (Control), 0.5 µg of PKAc, or 0.5 µg of cCaN + 0.5 µg of PKAc. NF-ATc1 was immunoprecipitated with anti-NF-ATc1 (7A6) ascites from [32P]orthophosphate-labeled cells, gel-purified, and assayed by phosphotryptic peptide mapping. Dashed circles indicate the relative positions of the phosphopeptides that are present on wt NF-ATc1, but missing on m3 NF-ATc1. The solid line circle on the left indicates the relative position of the Ser-294 phosphopeptide. Another phosphopeptide on the in vivo labeled protein also runs in a similar position on the electrophoresis neutral axis. The cCaN/PKAc-treated m3 NF-ATc1 protein could not be visualized because of substantial dephosphorylation. The arrows indicate the presence of an additional phosphopeptide that appears in the presence of PKAc.

In addition, other sites with similarities to the PKA consensus sequence (RRX(S/T)Y, where Y is a hydrophobic residue) were also mutated to evaluate other possible PKA phosphorylation sites. Single serine or threonine to alanine mutations at T26A, S153A, S169A, S261A, T339A, T558A, or S686A did not prevent PKA from fully phosphorylating full-length human NF-ATc1 as analyzed by phosphotryptic peptide mapping (data not shown). Interestingly, mutation at T215A, which is found in a homologous position in the SP1 repeat as the Ser-245 and Ser-294 PKA phosphorylation sites (Fig. 3A, underlined), also did not prevent PKA phosphorylation (data not shown).

The ability of PKA to phosphorylate these sites within cells was then determined (Fig. 3D). Comparison of the in vivo-labeled phosphotryptic peptide maps of wild-type (wt) and triple mutant S245A/S269A/S294A (m3) NF-ATc1 immunoprecipitated from HEK 293T cells cotransfected with PKAc confirmed that PKA phosphorylated NF-ATc1 at these sites in situ (Fig. 3D, PKAc). Two of these phosphopeptides (Fig. 3D, dashed line circles) that were absent in m3 NF-ATc1 were highly phosphorylated in the wild-type protein, whereas the third phosphopeptide (Fig. 3D, solid line circle) on the far left in the neutral axis of the thin-layer electrophoresis dimension appeared to have another peptide unrelated to PKA phosphorylation running in the same position. Although it cannot be stated with certainty that the S294A bearing peptide is not phosphorylated in this assay, it does correspond to the relative position of the phosphopeptide that disappeared upon mutation of S294A in the in vitro PKA phosphorylation assay (Fig. 3B), and the intensity of that spot is decreased in m3 NF-ATc1 compared with wild-type. In addition, the control maps of unstimulated NF-ATc1 also showed the presence of these three phosphopeptides in the wild-type protein, although at significantly decreased levels when compared with wild type cotransfected with PKAc (Fig. 3D; compare wt Control and PKAc). Transfected PKAc also increased the phosphorylation of an additional NF-ATc1 peptide that did not show up in control maps, nor did it correspond to the PKA phosphorylation sites mapped in vitro (Fig. 3D, arrows). This suggests that PKAc may also induce the phosphorylation of an alternate regulatory site not found in the basal resting state through activation of another kinase. Interestingly, the phosphotryptic peptide map of wt NF-ATc1 in the presence of both PKAc and cCaN not only showed the expected PKA phosphopeptides, but also showed that other phosphopeptides on wt NF-ATc1 remained in a highly phosphorylated state. In contrast, the PKA site triple mutant (m3 NF-ATc1) was highly dephosphorylated under these same conditions (Fig. 3D, cCaN+PKAc). These results demonstrate that PKA phosphorylates NF-ATc1 in situ at the sites mapped for PKA phosphorylation in vitro and that PKA phosphorylation of serines 245, 269, and 294 on NF-ATc1 is required for other NF-ATc1 kinases to compete effectively with calcineurin and oppose the substantial dephosphorylation of the NF-ATc1 protein.

We then determined whether the triple mutation S245A/S269A/S294A (m3) in NF-ATc1 would prevent the effect of PKA on calcineurin-mediated NF-ATc1 nuclear accumulation and dephosphorylation in situ. In the subcellular localization assay, both wt and m3 NF-ATc1-GFP were cytoplasmic in the resting (pSR) state and nuclear in the cCaN-activated (cCaN) state (Fig. 4A). However, when cotransfected with both cCaN and PKAc, a majority of cells exhibited cytoplasmic wt NF-ATc1-GFP and nuclear m3 NF-ATc1-GFP accumulation (Fig. 4A). Western blot analysis of the wt and m3 full-length NF-ATc1 proteins under these various conditions confirmed that the resting and calcineurin-mediated mobility shifts were similar for both wt and m3 NF-ATc1 (Fig. 4B), indicating that mutation of these serines did not dramatically affect the resting state phosphorylation levels or the ability of cCaN to dephosphorylate it. In contrast, PKAc was able to hyperphosphorylate wt, but not m3, NF-ATc1 in these cells (Fig. 4B). Furthermore, the serine-to-alanine mutations of S245A/S269A/S294A now enabled cCaN to dephosphorylate m3 NF-ATc1 in the presence of PKAc, whereas wt NF-ATc1 was not dephosphorylated under these same conditions (Fig. 4B). These results strongly suggest that PKA opposes calcineurin-mediated dephosphorylation and nuclear accumulation of NF-ATc1 by direct phosphorylation at Ser-245, Ser-269, and Ser-294.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   PKA phosphorylation sites on NF-ATc1 are required for PKA to block calcineurin-mediated nuclear accumulation and dephosphorylation. A, HEK 293T cells were transfected with 0.5 µg of either pSH160E-J (wt) or the triple S245A/S269A/S294A mutant m3 pSH160E-J (m3) and either pSR control vector, 0.1 µg of cCaN, or 0.1 µg of cCaN + 0.1 µg of PKAc as indicated. Cells were observed with fluorescence microscopy (magnification, ×40). B, HEK 293T cells were transfected with 0.5 µg of either wt pSH160E-J (wt) or m3 pSH160E-J (m3) and either pSR, 0.5 µg of PKAc, 0.5 µg of cCaN, or 0.5 µg of cCaN + 0.5 µg of PKAc. Cell lysates were immunoblotted with monoclonal anti-NF-ATc1 7A6 antibody. C, mutation of S245A/S269A/S294A prevents PKA from blocking ionomycin-induced mobility shift of NF-ATc1 in TAg Jurkat T cells. TAg Jurkat T cells were transfected with 20 µg of either pSH160c (wt NF-ATc1) or m3 pSH160c (m3 NF-ATc1) and stimulated as described under "Experimental Procedures." NF-ATc1 was both immunoprecipitated from whole cell extracts and immunoblotted with anti-FLAG antibody. Lines indicate divisions between hyperphosphorylated, dephosphorylated, and phosphorylated NF-ATc1 mobility shifts based on control wt NF-ATc1 lane.

To determine whether endogenous PKA is also able to phosphorylate NF-ATc1 at the sites phosphorylated by PKA in vitro and by exogenous PKAc in situ, the cell-permeable cAMP analog Bt₂cAMP and the phosphodiesterase inhibitor IBMX were used as activators of endogenous PKA in TAg Jurkat T cells. The cells were transfected with either wt NF-ATc1 or the triple mutant m3 NF-ATc1 and stimulated with either Me₂SO control, CsA, ionomycin, Bt₂cAMP, and IBMX, or the combination of ionomycin, Bt₂cAMP, and IBMX. In contrast to the mobility shift seen in control (pSR) HEK 293T cells (Fig. 4B), the m3 NF-ATc1 mobility shift in control TAg Jurkat T cells was slightly lower than wt NF-ATc1 (Fig. 4C). In addition, CsA, used to ensure the inhibition of any basal calcineurin activity, did not change the mobility of the triple mutant compared with wt, nor did Bt₂cAMP and IBMX alone result in hyperphosphorylation of wt NF-ATc1 (Fig. 4C), suggesting that Ser-245, Ser-269, and Ser-294 were phosphorylated more efficiently in resting TAg Jurkat T cells than in HEK 293T cells. Furthermore, although the ionomycin-induced mobility shift caused by dephosphorylation was similar in both wt and m3 proteins, the endogenous PKA-activating agents Bt₂cAMP and IBMX opposed this ionomycin-induced downward mobility shift of wt NF-ATc1, but were unable to oppose this shift with m3 NF-ATc1 (Fig. 4C). This suggests that endogenous PKA is able to phosphorylate NF-ATc1 on Ser-245, Ser-269, and Ser-294, in TAg Jurkat T cells, as was demonstrated by exogenous PKA both in vitro and in situ with HEK 293T cells.

PKA Requires an Additional Limiting Endogenous Factor for Complete Block of Ionomycin-induced Nuclear Accumulation of NF-ATc1-- Previous reports have described opposition of nuclear targeting of proteins by phosphorylation adjacent to the NLS (60, 61). To determine whether this mechanism could explain PKA-mediated opposition of NF-ATc1 nuclear targeting, we examined PKA phosphorylation of Ser-269 adjacent to the N-NLS at aa 265-267 (Fig. 3A, dashed line box) as an obvious prime candidate for this mechanism. However, single mutation of S269A did not prevent the ability of PKA to oppose NF-ATc1 nuclear accumulation in a localization assay (data not shown). In a complementary assay, we tested the ability of PKA to oppose the nuclear accumulation of SOS-265, a constitutively nuclear fusion protein of the normally cytoplasmic SOS exchange factor and the NF-ATc1 N-NLS (aa 261-274) (31). This construct was used to identify the NF-ATc1 N-NLS and contains within the NF-ATc1 fragment the PKA consensus site RKYS²⁶⁹. However, cotransfection of PKAc was unable to oppose the nuclear accumulation of SOS-265 (data not shown). These results demonstrate that PKA-mediated opposition of NF-ATc1 is not simply the result of phosphorylation of the NF-ATc1 N-NLS at Ser-269.

Another hypothesis is that PKA requires an additional limiting endogenous kinase to complete its effects on NF-ATc1 opposition. As discussed earlier in HEK 293T cells, PKAc opposed cCaN-mediated nuclear accumulation of NF-ATc1, but did not result in its exclusive cytoplasmic accumulation, causing, instead, a large number of cells to exhibit both nuclear and cytoplasmic NF-ATc1 (Fig. 2A). However, increasing the amount of transfected PKAc 5-10-fold was unable to further increase cytoplasmic accumulation of NF-ATc1 in opposition to either ionomycin or cCaN (Fig. 5 and data not shown). We then used the human EBV-transformed TAB B cell line as an in vivo system to test whether activation of endogenous PKA alone was able to oppose ionomycin-induced nuclear accumulation of endogenous NF-ATc1. TAB cells express all three NF-ATc1 isoforms, NF-ATc1. , , and , that result from differential RNA splicing of the NH₂- and COOH-terminal ends while leaving the regulatory NHR identical in each (1, 62), and have 4 times greater concentration of NF-ATc1 protein than TAg Jurkat T cells (data not shown), making them amenable for use in in vivo labeling and confocal fluorescence microscopy experiments to determine PKA effects on NF-ATc1 phosphorylation levels and subcellular localization. After in vivo labeling, we found that NF-ATc1 in unstimulated TAB cells was already substantially dephosphorylated by calcineurin, as the addition of CsA increased the amount of phosphate on NF-ATc1 by greater than 3-fold and significantly decreased its mobility as seen by Western blot analysis (Fig. 6A, compare lanes 1 and 2). The addition of ionomycin to the cells caused a further reduction in phosphate levels, although it did not appreciably increase the mobility of NF-ATc1 relative to unstimulated levels (Fig. 6A, compare lanes 2 and 3). Notably, another family member, NF-ATc2, exhibited a characteristic increase in mobility upon ionomycin stimulation in cells from this same experiment (data not shown), thus serving as a positive control for ionomycin stimulation. In contrast, the addition of Bt₂cAMP and IBMX, used as activators of endogenous PKA, overcame the ionomycin-induced dephosphorylation of NF-ATc1 by increasing the amount of phosphate to unstimulated levels and significantly decreasing NF-ATc1 mobility to an even greater degree than the unstimulated control (Fig. 6A, compare lane 4 with lanes 3 and 2), although these agents did not bring the phosphate levels to that seen in cells treated with CsA (Fig. 6A, compare lanes 4 and 1).

View larger version (45K): [in this window] [in a new window]   Fig. 5.   Increasing amounts of PKAc are unable to induce complete cytoplasmic accumulation of NF-ATc1 during ionomycin stimulation. HEK 293T cells were transfected with 1.0 µg of pSH160E-J (wt NF-ATc1-GFP) ± indicated amounts of PKAc plasmid DNA and pSR control vector to normalize total DNA per transfection. Increasing the amount of PKAc to 0.5 µg and beyond caused the cells to be too rounded for accurate distinction between nucleus and cytoplasm. Cells were stimulated with 2 µM ionomycin + 10 mM CaCl2 in normal media at 37 °C for 1 h and fixed. Original magnification, ×40. 200 cells in each of three independent experiments were counted as "cytoplasmic only," "both," or "nuclear only." "Cytoplasmic only" cells had no visually detectable punctate nuclear NF-ATc1-GFP, and "nuclear only" had no visually detectable cytoplasmic NF-ATc1-GFP. "Both" cells were counted as having visually detectable NF-ATc1-GFP in both compartments regardless of amount in each. Pie charts graph the mean percentage of cells showing "cytoplasmic only," "both," or "nuclear only" wt NF-ATc1-GFP.

View larger version (31K): [in this window] [in a new window]   Fig. 6.   PKA phosphorylates NF-ATc1 in vivo but requires an additional limiting endogenous factor for complete block of ionomycin-induced nuclear accumulation. A, endogenous PKA is able to phosphorylate endogenous NF-ATc1 in human TAB B cells. TAB cells were [32P]orthophosphate-labeled and stimulated as described under "Experimental Procedures." Anti-NF-ATc1 (7A6) antibody immunoprecipitates (this antibody recognizes all three NF-ATc1 family members , , ) were separated by SDS-PAGE and transferred to nitrocellulose. This membrane was then analyzed by phosphorimaging and anti-NF-ATc1 (7A6) Western blotting as explained under "Experimental Procedures." Data shown are representative of three independent experiments. The levels for NF-ATc1. are similar to NF-ATc1., but a radioactive co-precipitating band with NF-ATc1. prevented efficient analysis of this isoform. B, endogenous PKA does not oppose ionomycin-induced NF-ATc1 nuclear accumulation in TAB cells. TAB cells were stimulated as in A, immunostained, and then viewed on a confocal fluorescence microscope as described under "Experimental Procedures." Anti-NF-ATc1 (7A6) is shown in green, and the nuclear stain TO-PRO-3 is shown in red.

Surprisingly, the ability of endogenous PKA to oppose full dephosphorylation by calcineurin in these TAB cells did not correlate with a subsequent block of nuclear accumulation. Using confocal fluorescence microscopy, control cells showed NF-ATc1 to be prevalent in both cytoplasmic and nuclear compartments (Fig. 6B), which correlated well with the intermediate phosphorylation state of their NF-ATc1 protein (Fig. 6A, lane 2). Stimulation with CsA resulted in complete cytoplasmic accumulation, whereas stimulation with ionomycin resulted in complete nuclear accumulation of NF-ATc1, as expected (Fig. 6B). However, stimulation with Bt₂cAMP, IBMX, and ionomycin did not oppose NF-ATc1 nuclear accumulation. Rather, NF-ATc1 demonstrated largely nuclear accumulation with only diffuse perinuclear staining (Fig. 6B). Nuclear and cytoplasmic extracts of TAg Jurkat T cells stimulated with Bt₂cAMP, IBMX, and ionomycin also showed a majority of partially phosphorylated NF-ATc1 localized to the nucleus (data not shown). These results indicate that endogenous PKA is able to oppose full calcineurin-mediated dephosphorylation of NF-ATc1 in vivo, but still gives rise to an intermediate phosphorylation state that is primarily nuclear. This raises the possibility that an additional kinase, which may be limiting in these cell types, is necessary for full PKA-mediated opposition of calcineurin-mediated dephosphorylation and nuclear accumulation of NF-ATc1.

PKA and GSK-3 Synergize to Completely Block Ionomycin-induced Nuclear Accumulation of NF-ATc1-- Two of the PKA phosphorylation sites, Ser-245 and Ser-294, map to the COOH terminus of the SP2 and SP3 repeat motifs, respectively (Fig. 3A), and phosphorylation of these sites creates ideal consensus sequences for GSK-3. GSK-3 is found in both nuclear and cytoplasmic compartments and consists of the two isoforms  and  (16, 63). It is constitutively active in unstimulated cells, but can be inhibited by serine phosphorylation, and is also effectively regulated by its dependence on a priming kinase for the creation of a pre-phosphorylated substrate (64-67). This "hierarchical phosphorylation" mechanism (68) requires an initial priming kinase for the pre-phosphorylation of the substrate, creating a (S/T)XXXpS GSK-3 consensus sequence, where X denotes any amino acid and pS denotes a pre-phosphorylated serine. We have previously shown that PKA can supply this requisite priming kinase activity for GSK-3 phosphorylation of NF-ATc1 in vitro, and that overexpression of GSK-3 can oppose ionomycin-induced nuclear accumulation of NF-ATc1 and increase NF-ATc1 export rate after termination of calcium/calcineurin signaling (35, 69). Thus, we hypothesize that the mechanism of PKA opposition of NF-ATc1 is to act as a priming kinase for subsequent rephosphorylation by GSK-3, and that GSK-3 is the limiting endogenous kinase required for PKA opposition of NF-ATc1 nuclear accumulation.

To test this hypothesis we assayed for exclusive cytoplasmic versus exclusive nuclear accumulation of wt NF-ATc1-GFP in the presence of increasing amounts of PKAc and GSK-3, alone or in combination, with a constant ionomycin stimulus. Transfection of either kinase alone partially opposed nuclear accumulation of wt NF-ATc1-GFP during ionomycin stimulation, yet neither could efficiently cause exclusive cytoplasmic accumulation by itself (Fig. 7, A and B, wt NF-ATc1-GFP). However, together PKAc and GSK-3 synergized to cause dramatic and virtually exclusive cytoplasmic accumulation of NF-ATc1-GFP during ionomycin stimulation (Fig. 7, A and B, wt NF-ATc1-GFP). The highest concentrations of transfected kinases (0.25 µg of PKAc + 3.0 µg of GSK-3) caused 66% of the cells to exhibit exclusive cytoplasmic accumulation of NF-ATc1-GFP, compared with a maximum of 1% with PKAc alone or 7% with GSK-3 alone. On the other hand, exclusive nuclear accumulation of NF-ATc1-GFP was reduced from 87% in control cells to 53% with GSK-3 alone, 33% with PKAc alone, and to a mere 3% with PKAc and GSK-3 together (Fig. 7B, wt NF-ATc1-GFP). Thus, both PKA and GSK-3 increased the cytoplasmic accumulation of NF-ATc1 by themselves, but only to a partial degree. Only together could the two kinases dramatically and completely oppose the ionomycin-induced nuclear accumulation of NF-ATc1. In addition, the presence of PKAc allowed limiting amounts of GSK-3 to more efficiently oppose the ionomycin-induced nuclear accumulation of NF-ATc1 (Fig. 7B, wt).

View larger version (70K): [in this window] [in a new window]   Fig. 7.   PKA and GSK-3 synergize to oppose ionomycin-induced nuclear accumulation of NF-ATc1. A, HEK 293T cells were co-transfected with either 1.0 µg of pSH160E-J (wt NF-ATc1-GFP), 1.0 µg of m3 pSH160E-J (m3 NF-ATc1-GFP), or 1.0 µg of mGSKSP23 pSH160E-J (mGSKSP23 NF-ATc1-GFP), and combinations of increasing amounts of PKAc and GSK-3 plasmids as indicated, including pSR control vector to normalize total DNA concentrations. 24 h later, cells were stimulated with 2 µM ionomycin + 10 mM CaCl2 in normal media at 37 °C for 1 h and fixed. Original magnification, ×40. Pictures are representative of three independent experiments. B, graphs of results from A. Top panels: closed symbols, wt NF-ATc1-GFP; open symbols, m3 NFATc1-GFP. Bottom panels: closed symbols, mGSKSP23 NF-ATc1-GFP; open symbols, m3 NFATc1-GFP for comparison. 200 cells in each of three independent experiments were counted as "cytoplasmic only," "both," or "nuclear only." "Cytoplasmic only" cells had no visually detectable punctate nuclear NF-ATc1-GFP, and "nuclear only" had no visually detectable cytoplasmic NF-ATc1-GFP. "Both" cells were counted as having visually detectable NF-ATc1-GFP in both compartments regardless of amount in each. Graphs are mean ± S.E. C, as in B, except that it is a graph of PKA data only, showing that the ability of PKA to oppose nuclear accumulation of NF-ATc1 requires both the PKA phosphorylation sites on NF-ATc1 and the consensus GSK-3 phosphorylation sites directly upstream. Graphs are mean ± S.E.

If PKA functions through this priming kinase mechanism, then mutation of either the PKA phosphorylation sites or the upstream GSK-3 phosphorylation sites should prevent the ability of PKA to affect NF-ATc1 subcellular localization. Therefore, we compared the effects of PKAc and GSK-3 on wt, the m3 NF-ATc1-GFP mutant (which, as already described, has the PKA sites Ser-245, Ser-269, and Ser-294 mutated to alanine to prevent PKA phosphorylation; Fig. 3A, circles), and the mGSKSP23 NF-ATc1-GFP mutant (which has serine to alanine mutations in the SP2 and SP3 repeat motifs at serines 233, 237, 241, 278, 282, 286, and 290 upstream of the intact PKA phosphorylation sites to prevent subsequent phosphorylation by GSK-3 after PKA priming; Fig. 3A, caret symbols). Although m3 behaved like wt in both resting and ionomycin-stimulated cells (Fig. 4A, pSR), mGSKSP23 was predominantly, but not exclusively, cytoplasmic in resting cells (data not shown), and localized exclusively to the nucleus after ionomycin stimulation (Fig. 7C, white columns). In this localization assay, increasing amounts of PKAc caused a steady reduction in the percentage of cells expressing exclusive nuclear accumulation of wt NF-ATc1-GFP (Fig. 7C). In contrast, although a slight effect on the nuclear accumulation of the m3 and mGSKSP23 mutants was evident, PKAc was clearly unable to cause a similar reduction of the nuclear accumulation of either of these two mutants compared with wt (Fig. 7C). This suggests that additional phosphorylation of the SP2 and SP3 serines (presumably by endogenous GSK-3) is required for PKA to block the nuclear targeting of NF-ATc1. Furthermore, either of these mutations prevents the ability of GSK-3 to synergize with PKAc (Fig. 7B), thereby blocking the dramatic cytoplasmic accumulation seen with wt NF-ATc1-GFP, providing evidence that phosphorylation of these two repeats is required for this effect. All of these results provide strong support for the hypothesis that PKA functions by directly priming NF-ATc1 for further phosphorylation by GSK-3.

Interestingly, overexpression of either PKAc or GSK-3 alone was able to partially oppose the nuclear accumulation of wt NF-ATc1 (Fig. 7B, Nuclear). It is likely that PKAc was able to do so by synergizing with the limited endogenous GSK-3 activity available in the HEK 293T cells, because mutation of the GSK-3 phosphorylation sites in the SP2 and SP3 repeats prevented this PKAc effect (Fig. 7C). GSK-3 alone was able to decrease the exclusive nuclear accumulation of both mutants to a similar extent as seen with wt NF-ATc1-GFP, although significantly less than the effect of combined PKAc and GSK-3 on wt (Fig. 7B, Nuclear). We also found that mutation of the SRR (Fig. 3A, solid triangles) in NF-ATc1 prevents the complete cytoplasmic accumulation induced by the combination of PKAc and GSK-3 (data not shown). A possible explanation is that GSK-3 may also act through additional serine-containing domains such as the SRR and SP1 repeat with the aid of an additional endogenous priming kinase activity.

To assess whether GSK-3 was in fact the limiting endogenous kinase responsible for completing the PKA effect, we incubated ionomycin-stimulated HEK 293T cells co-transfected with wt NF-ATc1-GFP and PKAc with one of three different GSK-3 inhibitors: lithium chloride (Ref. 70; Fig. 8A), indirubin-3-monoxime, and alsterpaullone (Ref. 71; Fig. 8B). Our results show that these three GSK-3 inhibitors blocked the ability of PKA to oppose ionomycin-induced nuclear accumulation of NF-ATc1. In addition, these inhibitors were able to surpass the level of nuclear accumulation seen with ionomycin alone, which may be the result of inhibition of all opposing GSK-3 kinase activity. Previous reports have found that, although lithium has no effect on PKA activity (72), indirubin-3-monoxime has a 300-fold greater effect on GSK-3 than PKA (73), and alsterpaullone has an undetermined effect on PKA. Interestingly, however, both indirubin-3-monoxime and alsterpaullone are also able to inhibit cyclin-dependent kinases (73, 74). Therefore, as a negative control we also tested the cyclin-dependent kinase inhibitor roscovitine, which has no effect on PKA activity and has a 1000-fold greater effect on cyclin-dependent kinases than GSK-3 (71, 75), and found that it was unable to significantly block this PKA effect. Together these results suggest that GSK-3 is the limiting endogenous factor that is required for PKA to oppose the ionomycin-induced nuclear accumulation of NF-ATc1.

View larger version (17K): [in this window] [in a new window]   Fig. 8.   GSK-3 inhibitors block the ability of PKA to oppose ionomycin-induced nuclear accumulation of NF-ATc1. HEK 293T cells were transfected with 1.0 µg of pSH160E-J (wt NF-ATc1-GFP) ± 0.05 µg of PKAc plasmid DNA and pSR control vector to normalize total DNA per transfection. Indicated cells were pre-incubated with 5, 10, or 20 mM LiCl (A) or either Me2SO vehicle control (), the GSK-3 inhibitors indirubin-3-monoxime or alsterpaullone, or the control cyclin-dependent kinase inhibitor roscovitine for 30 min (B) and then stimulated with 2 µM ionomycin + 10 mM CaCl2 ± inhibitor as indicated in normal medium at 37 °C for 1 h, then fixed. 200 cells in each of three independent experiments were counted and scored as "cytoplasmic only," "both," or "nuclear only." Graph shows the mean percentage ± S.D. of cells showing "nuclear only" wt NF-ATc1-GFP.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The large number of phosphorylation sites on the NF-ATc proteins (33) may allow them to integrate a myriad of impinging signaling pathways into graded levels of nuclear accumulation and subsequent transcriptional activity. In this study we describe a mechanism for the inhibition of NF-ATc1 activity by the cAMP/PKA pathway and show that PKA counteracts calcium/calcineurin signaling through direct phosphorylation of the NF-ATc1 transcription factor, thus "priming" for the subsequent phosphorylation by GSK-3. Together, these kinases are capable of negatively modulating the calcineurin-mediated dephosphorylation and nuclear accumulation of NF-ATc1 (Fig. 9). Similar cooperative regulation by PKA and GSK-3 has been observed in cAMP-responsive element-binding protein regulation (76, 77), the Alzheimer-like phosphorylation of tau (78), and Hedgehog signaling in Drosophila via Cubitus interruptus (Ci), a transcription factor that can be primed by PKA for further GSK-3 phosphorylation, thereby opposing Hedgehog-mediated dephosphorylation of Ci (79).

View larger version (23K): [in this window] [in a new window]   Fig. 9.   PKA/GSK-3 priming kinase mechanism opposes calcineurin to determine subcellular distribution of NF-ATc1. Subcellular localization of NF-ATc1 is determined by the opposing actions of dephosphorylation by calcineurin, which leads to nuclear accumulation, and of phosphorylation by kinases, which leads to cytoplasmic accumulation of NF-ATc1. In this model, phosphorylation by PKA on serines 245, 269 and 294 "primes" NF-ATc1 for further phosphorylation on the SP2 and SP3 repeat motifs by GSK-3 (and possibly other kinases). This hierarchical phosphorylation causes translocation of NF-ATc1 to the cytoplasm and thereby lowers the amount of NF-ATc1 available for transcription. Dephosphorylation of these sites by calcineurin reverses this process and leads to increased nuclear accumulation and transcriptional activation of NF-ATc1.

Much evidence supports the role of PKA as an in vivo regulatory NF-ATc1 kinase. Consistent with earlier reports (5-7, 9, 80, 81), we demonstrate that PKA inhibits NF-ATc/AP-1- but not AP-1-mediated transcription, suggesting that PKA inhibition is likely to function through the NF-ATc protein. In addition, exogenous PKA is also able to oppose calcineurin-mediated dephosphorylation and nuclear accumulation of NF-ATc1, suggesting that PKA may inhibit NF-ATc/AP-1 transcriptional activity by competing directly with calcineurin. Furthermore, we provide evidence that PKA phosphorylates NF-ATc1 directly at serines 245, 294, and 269, located at the carboxyl termini of the SP2 and SP3 repeat domains, and adjacent to the ²⁶⁵KRK²⁶⁷ N-NLS, respectively. Mutation of these three serines to alanines not only prevents PKA phosphorylation, but also the ability of PKA to oppose calcineurin-mediated dephosphorylation and nuclear accumulation of NF-ATc1.

Many studies of IL-2 gene regulation have indicated inhibition of T cell receptor-activated NF-ATc by cAMP elevation (5-7, 9). Dependent on T cell source, both physiological and pharmacological cAMP-elevating agents and PKA have been found to inhibit NF-AT DNA element transcriptional activity and NF-AT DNA binding (5-7, 9), whereas overexpression of the NF-ATc activator calcineurin (8), or NF-ATc itself (9), has been found to overcome IL-2 transcriptional sensitivity to cAMP inhibition. In addition, the catalytic subunit of PKA is found to rapidly diffuse throughout the cytoplasm and nucleus after cAMP elevation (82), suggesting that PKA may have access to NF-ATc1 in both compartments. Furthermore, an increase in cAMP degradation by the phosphodiesterase PDE7 is required for T cell receptor-stimulated cell proliferation and IL-2 gene transcription to occur (83), suggesting that PKA activation kinetics inversely complement those of calcineurin and NF-ATc during T cell activation.

NF-ATc1 is a very highly phosphorylated protein, containing an estimated 20 phosphorylated serines in the resting cell state, so, not surprisingly, we found that PKA phosphorylation of only three of those serines, although necessary, is not sufficient to oppose calcineurin-mediated activation of the NF-ATc1 protein. Evidence from this and our previous study (35) suggests that subsequent phosphorylation by GSK-3 is necessary for inhibition of NF-ATc1 nuclear targeting by PKA. Several experiments reveal that PKA requires a second limiting endogenous kinase for its effect. First, increasing the amount of PKA increases the frequency of cells exhibiting both cytoplasmic and nuclear NF-ATc1, but not cells exhibiting exclusive cytoplasmic NF-ATc1. Second, in the presence of PKA and calcineurin, in vivo labeling together with phosphotryptic peptide mapping demonstrate that phosphorylation of the PKA sites on wt NF-ATc1 leads to an increase of phosphorylation on other sites, whereas mutation of the PKA sites prevents this increased phosphorylation and results in substantial dephosphorylation of the protein. Finally, in vivo activation of endogenous PKA in both ionomycin-treated TAB and TAg Jurkat cells increases the apparent molecular weight and phosphate level of NF-ATc1 compared with that from cells treated with ionomycin alone, but, in TAB cells, NF-ATc1 carries 3-fold less phosphate than NF-ATc1 from cells treated with CsA and, in both cell lines, PKA activation alone does not result in substantial opposition of NF-ATc1 nuclear accumulation.

We found that this endogenous kinase activity can be supplied by the NF-ATc1 kinase GSK-3. We have previously reported that PKA can prime NF-ATc1 for subsequent phosphorylation by GSK-3 in vitro, and that exogenous GSK-3 is able to oppose ionomycin-induced nuclear accumulation and increase the nuclear export rate of NF-ATc1 (35). GSK-3 (84) is found in both cytoplasmic and nuclear cellular compartments, and a dominant negative version of GSK-3 is able to decrease the nuclear export rate of both NF-ATc1 and -c4 (16, 69). GSK-3 is constitutively active in normal unstimulated T cells and is inhibited upon T cell activation (85), thereby inversely complementing the activation kinetics of both calcineurin and NF-ATc. In addition, immunodepletion of GSK-3 from whole brain extracts removes the NF-ATc1 and -c4 kinase activity that is able to phosphorylate a pre-PKA-phosphorylated NF-ATc substrate (35, 69). Here we found that increasing amounts of PKA and GSK-3 together synergize to promote cytoplasmic accumulation of NF-ATc1 in ionomycin-treated cells. Furthermore, not only does mutation of the PKA phosphorylation sites block the ability of PKA to oppose NF-ATc1 nuclear accumulation, but mutation of the GSK-3 phosphorylation sites in the SP2 and SP3 repeats (leaving the PKA phosphorylation sites intact), or the inhibition of GSK-3 kinase activity in vivo by pharmacological agents, also blocks the effect of PKA. This shows that PKA can function as a priming kinase for GSK-3 to enhance its opposition of calcineurin activation of NF-ATc1. In addition, the ability of GSK-3 to phosphorylate its substrates is highly dependent upon the activity level of its priming kinase. Priming has been shown to increase the GSK-3 phosphorylation rate by 50-1000-fold (66, 86), making it likely that PKA priming of NF-ATc1 enhances the ability of limited GSK-3 activity levels to compete more effectively with activated calcineurin.

The complexity of NF-ATc1 regulation by phosphorylation may allow for very different activation properties in different cell types. The resting and PKA/calcineurin-activated NF-ATc1 phosphorylation states and subcellular localization differed substantially between the HEK 293T, TAg Jurkat T, and TAB B cell lines used in this study. The PKA phosphorylation sites are largely phosphorylated in resting TAg Jurkat T cells, but not in resting HEK 293T cells. In all cell types, PKA is able to oppose calcineurin-mediated dephosphorylation of NF-ATc1, but in TAg Jurkat T and TAB B cells is unable to oppose nuclear accumulation as it does in HEK 293T cells. Because PKA-mediated opposition is dependent upon another kinase, these findings likely relate to differences in the kinases within these cell types. For example, GSK-3 activity may be more severely limited in both TAg Jurkat T and EBV-transformed TAB cells than in HEK 293T cells by the fact that Jurkat cells carry a mutation in their D3 phosphoinositide phosphatase PTEN gene and TAB cells carry the EBV-expressed LMP2A protein, both of which lead to hyperactivation of Akt/protein kinase B (87, 88), a kinase known to phosphorylate and inactivate GSK-3 (89). This may explain the apparent lack of regulation of both transfected and endogenous NF-ATc1 observed in both resting Jurkat (20, 90) and TAB cells (Fig. 6). In addition, resting HEK 293T cells may utilize an alternate priming kinase that phosphorylates the SP repeats upstream of the PKA priming kinase sites, but during PKA activation are able to utilize PKA and GSK-3 as an alternate regulatory pathway. Such kinases as p38, ERK, JNK, and CK2, which have been variably found to phosphorylate the SRR and SP repeat serines and oppose ionomycin-induced NF-ATc1 nuclear accumulation, are possible candidates (37, 38, 91). Furthermore, other endogenous kinase activities, such as a putative SRR and/or SP1 priming kinase, may be limiting in any of these cell types. The complexity of the kinase/phosphatase-mediated regulation of NF-ATc localization highlights the need for a more detailed analysis of this regulatory network, including single-cell measurements of concurrent NF-ATc nuclear concentration and transcriptional activation, which may help us to better understand such intricate modulation of NF-ATc nuclear presence.

We have shown that PKA and GSK-3 synergize to phosphorylate the SP2 and SP3 repeats of NF-ATc1, and that phosphorylation of these regions is necessary for PKA to oppose calcineurin-mediated nuclear accumulation of NF-ATc1. However, we have observed that GSK-3 also has significant effects on additional domains in the NHR and that mutation of the SRR prevents PKA and GSK-3 from inducing exclusive cytoplasmic accumulation of NF-ATc1. How then does phosphorylation of these two domains and Ser-269 contribute to the regulation of the masking and unmasking of the NLS and NES? Several reports have described that nuclear accumulation of NF-ATc in unstimulated cells can be produced through mutation of one or more of the SRR or SP repeat domains (31, 33, 39); the SP2 and SP3 repeats appear to play a supporting role in this respect. One report found that mutation of the SP2 and SP3 serines in NF-ATc2 does not result in constitutive nuclear localization of the protein, consistent with our findings for NF-ATc1, but this mutation does complement an SRR mutation, resulting in more complete nuclear accumulation (33). On the other hand, phosphorylation of NF-ATc has been reported to occur prior to cytoplasmic re-accumulation (4), and the nuclear exportin Crm1 has been found to bind more efficiently to phosphorylated NF-ATc than dephosphorylated NF-ATc, suggesting that phosphorylation also unmasks an NES (33). Thus, the balance between activation of the NLS through dephosphorylation and activation of the NES through rephosphorylation may determine the steady-state subcellular localization of the NF-ATc protein. One may consider then that partial phosphorylation of the NHR at the SP2, SP3, and N-NLS serines by PKA and GSK-3 may partially activate the NES, resulting, therefore, in partial nuclear and cytoplasmic accumulation of NF-ATc1. Because a threshold concentration of nuclear NF-ATc is required for NF-AT-mediated transcription (34), it is possible that PKA activation alone, which only partially decreases NF-ATc nuclear accumulation, may have large consequences on NF-AT-mediated transcription. Additionally, NF-ATc activation thresholds may differ greatly between various promoters; this may explain, in part, the ability of cAMP elevation to inhibit the transcription of the IL-2 gene and not the IL-4 gene (5, 7, 55, 92, 93). GSK-3 also appears to effect other NHR domains in NF-ATc1, although this function of GSK-3 remains to be fully defined.

PKA/GSK-3 phosphorylation of Ser-269 and the SP2 and SP3 motifs may also have additional effects besides regulating NF-ATc1 subcellular localization. GSK-3 has been shown to decrease NF-ATc1 DNA binding affinity by a molecular rheostat mechanism, where binding affinity varies inversely with the number of GSK-3-phosphorylated SP repeats (90, 94), which may further result in increased nuclear export (95). PKA has been shown to increase the association of several NF-ATc family members with 14-3-3 proteins (96), which are postulated to act as phosphoserine-dependent signaling modulators (97). Interestingly, the PKA phosphorylation sites on NF-ATc1 are variably conserved throughout the NF-ATc family, which may result in differential regulation of the NF-ATc family members by PKA. These results have implications for processes such as cardiac hypertrophy (98), vascular development (99), Th1/Th2 development (100), and neuronal synaptic plasticity (69), where cAMP-elevating agents together with GSK-3 may regulate the NF-ATc activity in these systems.

	ACKNOWLEDGEMENTS

We thank K. Stankunas for critically reading the manuscript. C. M. S. also thanks the laboratory of Dr. T. Meyer at Stanford University for support and helpful discussion.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant 5PO1AI36535 (to P. G.).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.

^§ Present address: Inst. for Systems Biology, Seattle, WA 98103-8904.

Present address: Cardiology Division, Massachusetts General Hospital, Boston, MA 02114.

Present address: Pfizer, New London, CT 06333.

^§§ To whom correspondence should be addressed. Tel.: 650-498-4826; Fax: 650-724-7778; E-mail: pgardner@stanford.edu.

Published, JBC Papers in Press, September 25, 2002, DOI 10.1074/jbc.M207029200

² NF-ATc1-c4, HUGO Nomenclature Committee Consensus, 1999.

	ABBREVIATIONS

The abbreviations used are: NF-AT, nuclear factor of activated T cells; PKA, cAMP-dependent protein kinase A; IL, interleukin; CsA, cyclosporin A; NHR, NF-AT homology region; NLS, nuclear localization sequence; NES, nuclear export sequence; SRR, serine-rich region; SP, serine-proline; Bt₂cAMP, dibutyryl cAMP; IBMX, 3-isobutyl-1-methylxanthine; PMA, phorbol 12-myristate 13-acetate; AP-1, activator protein-1; GSK-3, glycogen synthase kinase-3; cCaN, constitutively active truncated form of calcineurin; cCaMKII, constitutively active form of calcium/calmodulin-dependent protein kinase II; cPKC, constitutively active form of protein kinase C, GST, glutathione S-transferase; GFP, green fluorescent protein; SOS, Son of Sevenless GDP/GTP exchange factor; EBV, Epstein-Barr virus; HA, hemagglutinin; Ci, Cubitus interruptus; PKAc, catalytic subunit of protein kinase A; PVDF, polyvinylidene difluoride; m3, triple mutant; wt, wild type; aa, amino acid(s); PBS, phosphate-buffered saline; DMEM, Dulbecco's modified Eagle's medium; BSA, bovine serum albumin; FITC, fluorescein isothiocyanate; N-NLS, NH₂-terminal nuclear localization sequence; PIPES, 1,4-piperazinediethanesulfonic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				A. C. Rosenkranz, B. H. Rauch, K. Freidel, and K. Schror Regulation of protease-activated receptor-1 by vasodilatory prostaglandins via NFAT Cardiovasc Res, June 11, 2009; (2009) cvp163v2. [Abstract] [Full Text] [PDF]

				S. Klein-Hessling, T. Bopp, M. K. Jha, A. Schmidt, S. Miyatake, E. Schmitt, and E. Serfling Cyclic AMP-induced Chromatin Changes Support the NFATc-mediated Recruitment of GATA-3 to the Interleukin 5 Promoter J. Biol. Chem., November 7, 2008; 283(45): 31030 - 31037. [Abstract] [Full Text] [PDF]

				R. I. Jabr, A. J. Wilson, M. H. Riddervold, A. H. Jenkins, B. A. Perrino, and L. H. Clapp Nuclear translocation of calcineurin Abeta but not calcineurin A{alpha} by platelet-derived growth factor in rat aortic smooth muscle Am J Physiol Cell Physiol, June 1, 2007; 292(6): C2213 - C2225. [Abstract] [Full Text] [PDF]

				Y.-Y. Cho, K. Yao, A. M. Bode, H. R. Bergen III, B. J. Madden, S.-M. Oh, S. Ermakova, B. S. Kang, H. S. Choi, J.-H. Shim, et al. RSK2 Mediates Muscle Cell Differentiation through Regulation of NFAT3 J. Biol. Chem., March 16, 2007; 282(11): 8380 - 8392. [Abstract] [Full Text] [PDF]

				W. Zhou, T. S. Blackwell, K. Goleniewska, J. F. O'Neal, G. A. FitzGerald, M. Lucitt, R. M. Breyer, and R. S. Peebles Jr. Prostaglandin I2 analogs inhibit Th1 and Th2 effector cytokine production by CD4 T cells J. Leukoc. Biol., March 1, 2007; 81(3): 809 - 817. [Abstract] [Full Text] [PDF]

				T. Shen, Z. Cseresnyes, Y. Liu, W. R. Randall, and M. F. Schneider Regulation of the nuclear export of the transcription factor NFATc1 by protein kinases after slow fibre type electrical stimulation of adult mouse skeletal muscle fibres J. Physiol., March 1, 2007; 579(2): 535 - 551. [Abstract] [Full Text] [PDF]

				J. L. Belfield, C. Whittaker, M. Z. Cader, and S. Chawla Differential Effects of Ca2+ and cAMP on Transcription Mediated by MEF2D and cAMP-response Element-binding Protein in Hippocampal Neurons J. Biol. Chem., September 22, 2006; 281(38): 27724 - 27732. [Abstract] [Full Text] [PDF]

				M. Hallhuber, N. Burkard, R. Wu, M. H. Buch, S. Engelhardt, L. Hein, L. Neyses, K. Schuh, and O. Ritter Inhibition of Nuclear Import of Calcineurin Prevents Myocardial Hypertrophy Circ. Res., September 15, 2006; 99(6): 626 - 635. [Abstract] [Full Text] [PDF]

				S. Ryzhov, A. E. Goldstein, I. Biaggioni, and I. Feoktistov Cross-Talk between Gs- and Gq-Coupled Pathways in Regulation of Interleukin-4 by A2B Adenosine Receptors in Human Mast Cells Mol. Pharmacol., August 1, 2006; 70(2): 727 - 735. [Abstract] [Full Text] [PDF]

				N. Hermann-Kleiter, N. Thuille, C. Pfeifhofer, T. Gruber, M. Schafer, C. Zitt, A. Hatzelmann, C. Schudt, M. Leitges, and G. Baier PKC{theta} and PKA are antagonistic partners in the NF-AT transactivation pathway of primary mouse CD3+ T lymphocytes Blood, June 15, 2006; 107(12): 4841 - 4848. [Abstract] [Full Text] [PDF]

				T. Shen, Y. Liu, Z. Cseresnyes, A. Hawkins, W. R. Randall, and M. F. Schneider Activity- and Calcineurin-independent Nuclear Shuttling of NFATc1, but Not NFATc3, in Adult Skeletal Muscle Fibers Mol. Biol. Cell, April 1, 2006; 17(4): 1570 - 1582. [Abstract] [Full Text] [PDF]

				T. Oka, Y.-S. Dai, and J. D. Molkentin Regulation of Calcineurin through Transcriptional Induction of the calcineurin A{beta} Promoter In Vitro and In Vivo Mol. Cell. Biol., August 1, 2005; 25(15): 6649 - 6659. [Abstract] [Full Text] [PDF]

				I. Ortega-Perez, E. Cano, F. Were, M. Villar, J. Vazquez, and J. M. Redondo c-Jun N-terminal Kinase (JNK) Positively Regulates NFATc2 Transactivation through Phosphorylation within the N-terminal Regulatory Domain J. Biol. Chem., May 27, 2005; 280(21): 20867 - 20878. [Abstract] [Full Text] [PDF]

				B. Sanna, O. F. Bueno, Y.-S. Dai, B. J. Wilkins, and J. D. Molkentin Direct and Indirect Interactions between Calcineurin-NFAT and MEK1-Extracellular Signal-Regulated Kinase 1/2 Signaling Pathways Regulate Cardiac Gene Expression and Cellular Growth Mol. Cell. Biol., February 1, 2005; 25(3): 865 - 878. [Abstract] [Full Text] [PDF]

				X. K. Liu, X. Lin, and S. L. Gaffen Crucial Role for Nuclear Factor of Activated T Cells in T Cell Receptor-mediated Regulation of Human Interleukin-17 J. Biol. Chem., December 10, 2004; 279(50): 52762 - 52771. [Abstract] [Full Text] [PDF]

				K. A. Kafadar and M. S. Cyert Integration of Stress Responses: Modulation of Calcineurin Signaling in Saccharomyces cerevisiae by Protein Kinase A Eukaryot. Cell, October 1, 2004; 3(5): 1147 - 1153. [Abstract] [Full Text] [PDF]

				H. Okamura, C. Garcia-Rodriguez, H. Martinson, J. Qin, D. M. Virshup, and A. Rao A Conserved Docking Motif for CK1 Binding Controls the Nuclear Localization of NFAT1 Mol. Cell. Biol., May 15, 2004; 24(10): 4184 - 4195. [Abstract] [Full Text] [PDF]

				F. Demarchi, C. Bertoli, P. Sandy, and C. Schneider Glycogen Synthase Kinase-3{beta} Regulates NF-{kappa}B1/p105 Stability J. Biol. Chem., October 10, 2003; 278(41): 39583 - 39590. [Abstract] [Full Text] [PDF]

				P. G. Hogan, L. Chen, J. Nardone, and A. Rao Transcriptional regulation by calcium, calcineurin, and NFAT Genes & Dev., September 15, 2003; 17(18): 2205 - 2232. [Full Text] [PDF]

				O. Marin, V. H. Bustos, L. Cesaro, F. Meggio, M. A. Pagano, M. Antonelli, C. C. Allende, L. A. Pinna, and J. E. Allende Inaugural Article: A noncanonical sequence phosphorylated by casein kinase 1 in {beta}-catenin may play a role in casein kinase 1 targeting of important signaling proteins PNAS, September 2, 2003; 100(18): 10193 - 10200. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/50/48664    most recent M207029200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sheridan, C. M. Articles by Gardner, P. Search for Related Content PubMed PubMed Citation Articles by Sheridan, C. M. Articles by Gardner, P. Social Bookmarking                      What's this?