c-Jun is a downstream target for ceramide-activated protein phosphatase in A431 cells.

Stimulation of [3H]serine-labeled A431 cells with tumor necrosis factor-alpha (TNFalpha) or bacterial sphingomyelinase (SMase) resulted in a rapid decrease (approximately 50% by 15 min) in cellular [3H]sphingomyelin content and generation of the lipid moiety [3H]ceramide, which remained elevated 60 min later. Sphingomyelin hydrolysis in response to TNFalpha or bacterial SMase resulted in a time-dependent decrease in the phosphorylation state of c-Jun protein, an effect that was also observed in cells treated with the membrane-permeable ceramide analogue N-hexanoylsphingosine (C6-ceramide). The rapid dephosphorylation of the c-Jun gene product in response to TNFalpha, SMase, or C6-ceramide was not observed in A431 cells treated with the serine-threonine phosphatase inhibitor okadaic acid. After the initial steps of previously described methods for the purification of a ceramide-activated protein phosphatase termed CAPP (Dobrowsky, R. T., Kamibayashi, C., Mumby, M. C., and Hannun, Y. A. (1993) J. Biol. Chem. 268, 15523-15530), we obtained a cytosolic fraction from A431 cells that specifically dephosphorylated 32Pi-labeled c-Jun protein used as substrate in an immunocomplex phosphatase assay. Phosphatase activity in vitro was apparent only in the presence of ceramide (5 micro) and was specifically abrogated when okadaic acid (1 n) was included in the immunocomplex phosphatase assay. These results provide strong evidence for c-Jun as a downstream target for CAPP activated in response to post-TNF signaling in A431 cells.


Stimulation of [ 3 H]serine-labeled A431 cells with tumor necrosis factor-␣ (TNF␣) or bacterial sphingomyelinase (SMase) resulted in a rapid decrease (ϳ50% by 15 min) in cellular [ 3 H]sphingomyelin content and generation of the lipid moiety [ 3 H]ceramide
, which remained elevated 60 min later. Sphingomyelin hydrolysis in response to TNF␣ or bacterial SMase resulted in a timedependent decrease in the phosphorylation state of c-Jun protein, an effect that was also observed in cells treated with the membrane-permeable ceramide analogue N-hexanoylsphingosine (C 6 -ceramide). The rapid dephosphorylation of the c-Jun gene product in response to TNF␣, SMase, or C 6 -ceramide was not observed in A431 cells treated with the serine-threonine phosphatase inhibitor okadaic acid. After the initial steps of previously described methods for the purification of a ceramide-activated protein phosphatase termed CAPP (Dobrowsky, R. T., Kamibayashi, C., Mumby, M. C., and Hannun, Y. A. (1993) J. Biol. Chem. 268, 15523-15530), we obtained a cytosolic fraction from A431 cells that specifically dephosphorylated 32 P i -labeled c-Jun protein used as substrate in an immunocomplex phosphatase assay. Phosphatase activity in vitro was apparent only in the presence of ceramide (5 M) and was specifically abrogated when okadaic acid (1 nM) was included in the immunocomplex phosphatase assay. These results provide strong evidence for c-Jun as a downstream target for CAPP activated in response to post-TNF signaling in A431 cells.
Efforts aimed at delineating a sphingolipid-dependent pathway for signal transduction led to the characterization of different protein kinases that include a ceramide-activated and proline-directed protein kinase first identified by its ability to phosphorylate Thr-669 of the epidermal growth factor receptor in A431 cells (herein referred to as ceramide-activated protein kinase (12,13,17)), components of the mitogen-activated protein kinase cascade (16,18), and a non-phorbol ester-stimulable and diacylglycerol-independent protein kinase C isotype (19,20). In addition, a cytosolic Ser/Thr protein phosphatase of the 2A-type specifically activated by ceramide, and thus termed CAPP, has also been reported (21)(22)(23)(24)(25).
Although regulation of transcription factors by posttranslational modification via protein phosphorylation plays a paramount role in cellular regulation (reviewed in Refs. [25][26][27][28][29], only a reduced number of observations support a role for a ceramide-activated multicomponent system involving kinases and phosphatases as a mechanism whereby signals emanating from the plasma membrane are transduced to the nucleus to affect the transcriptional machinery of the target cell (3 -10). These evidences include CAPP-mediated down-regulation of c-myc gene expression (11), protein kinase C-induced translocation of preformed nuclear factor-B from the cytosol to the nucleus (13,14,19,20), and more recently, ceramide-mediated activation of JNK (c-Jun N-terminal kinase, also referred to as stress-activated kinase) and the subsequent augmented expression of c-jun gene (30 -32) and AP-1 activity (31,33).
Members of the c-Jun gene family encode components that can either homodimerize or heterodimerize with proteins of the Fos family to form AP-1 (25)(26)(27)(28)(29), a sequence-specific transcription factor that is strictly dependent for its DNA binding and transcriptional activity of the site-specific phosphorylation status of the preexisting and/or newly synthesized c-Jun component (29,34,35).
Although JNK/SAPK-induced phosphorylation of sites at the N-terminal transactivation domain of c-Jun is important for functional AP-1 activation (36 -40), in unstimulated fibroblastic and epithelial cells c-Jun is constitutively phosphorylated near the DNA binding region, and removal of these phosphates is also important for transcriptional activity in quiescent cells (41)(42)(43)(44)(45)(46).
Because ceramide activates JNK/SAPK (30 -32), which in turn results in enhanced transcription of c-Jun and DNA binding activity of AP-1 (31,33), we investigated whether activation of sphingomyelin metabolism may also affect c-Jun phosphorylation in resting epithelial A431 cells. Results presented herein strongly suggest that ceramide generated in response to TNF␣-induced sphingomyelin hydrolysis stimulates a cytosolic OA-sensitive phosphatase activity in A431 cells that specifically dephosphorylates c-Jun in intact cells or in an immunocomplex phosphatase assay.

MATERIALS AND METHODS
Reagents and Hormones-The human recombinant TNF␣ was purchased from Calbiochem. Bacterial SMase from Bacillus cereus, phospholipase A 2 from Crotalus adamanteus, phospholipase C from Clostridium perfringens, phospholipase D from Streptomyces chromofuscus, OA, sphingomyelin, phenylmethylsulfonyl fluoride, and other protease inhibitors (aprotinin, leupeptin, and pepstatin A) were purchased from Sigma. The membrane-permeable analogue N-hexanoylsphingosine (C 6 -cer) was obtained from Biomol Research Laboratories (Plymouth Meeting, PA). Dulbecco's modified Eagle's medium, Ham's F-12 medium, fetal calf serum, and other tissue culture reagents were purchased from Life Technologies, Inc.
Cell Culture-The A431 human epithelial tumor cells (kindly provided by Dr. F. Casanueva, University of Santiago de Compostela, Santiago de Compostela, Spain) were grown in a mixture of Dulbecco's modified Eagle's medium and Ham's F-12 medium (1:1) supplemented with 10% fetal calf serum, 2 mM L-glutamine, and antibiotics (100 units/ml penicillin and 100 g/ml streptomycin). Cultures were maintained at 37°C under a water-saturated atmosphere of 5% CO 2 and 95% air, and cells were passaged biweekly by trypsinization before reaching confluence. For experiments, cells were harvested by trypsinization and were inoculated (1-3 ϫ 10 5 cells/well) into multicluster (6 ϫ 34-mm/well or 24 ϫ 16-mm/well) Falcon plastic culture plates (Becton & Dickinson, Oxnard, CA) and were grown in the same medium until 60% confluence.
Quantitation of Sphingomyelin Metabolites in A431 Cells-Serumstarved subconfluent cultures of A431 cells (4 -5 ϫ 10 6 cells/well) were labeled for 24 h in serine-free medium containing 2 Ci/ml of [6-3 H]serine (DuPont NEN), the time period necessary to achieve isotopic steady state in these cells (results not shown). Labeled cells were rinsed twice with fresh medium supplemented with 25 mM serine, were allowed to equilibrate for 30 min in this medium, and thereafter were treated with TNF␣ (10 ng/ml), SMase (0.3 unit/ml), or an equivalent volume (50 l) of diluent. At the indicated time periods, the culture plates were placed on ice, and the medium was rapidly removed; then cells were scrapped with 1 ml of ice-cold methanol and transferred to clean glass tubes containing 2 ml of chloroform (47). The organic phases were washed three times by vigorous mixing with 3 ml of chloroform:0.1 M KCl (1:1, v/v.), were separated from the aqueous phases by centrifugation, and were dried under nitrogen. The organic phases were redissolved in 100 l of chloroform and were spotted on activated Silica Gel 60 G chromatography plates (Merck), and lipids were separated by sequential one-dimensional chromatography in the solvent system chloroform:benzene:ethanol (80:40:75, v/v/v) followed by another run in the basic mobile phase chloroform:methanol:28% ammonium hydroxide (65:25:5, v/v/v), as described (48). Lipids were visualized with molybdenum blue spray or iodine vapors, and fractions that comigrated with the same retention factors as the unlabeled standards were scrapped into scintillation vials; the associated radioactivity was determined by liquid scintillation counting as described (49,50).
[ 32 P i ]Orthophosphate Labeling of A431 Cells-Subconfluent cultures (ϳ4 ϫ 10 6 cells/well) were rinsed twice with phosphate-free Dulbecco's modified Eagle's medium and were labeled for 4 h with 500 Ci/ml carrier-free [ 32 P i ]orthophosphate (DuPont NEN). All experimental agents were freshly diluted in sterile culture medium and were added in 50-l aliquots with control incubations, receiving the same volume of medium and a similar final concentration of diluent (less than 0.1% diluent of the appropriate stock solutions). After stimulation, the culture plates were placed on ice, the labeling medium was discarded, and cells were rapidly rinsed two times with ice-cold phosphate-buffered saline supplemented with 400 M sodium orthovanadate; thereafter, cells were incubated for 30 min at 4°C in lysis buffer (10 mM phosphate, pH 7.4, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, and 0.1% SDS) supplemented with 5 mM EDTA, 1 mM NaF, 1 mM sodium orthovanadate, 30 mM sodium pyrophosphate, and protease inhibitors (5 g/ml aprotinin, 10 g/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, and 15 mM iodoacetamide).
The lysates were precleared by centrifugation in a refrigerated Beckman table-top centrifuge (14,000 ϫ g for 20 min at 4°C), and suitable aliquots of the supernatants were used to determine trichloroacetic acid-precipitable radioactivity or total protein content by a commercial BCA method (Sigma).
Immunoprecipitation of c-Jun-Equal amounts of cell proteins were immunoprecipitated at 4°C by an overnight incubation with 1 g of a purified rabbit polyclonal antiserum raised against a peptide spanning the 15 C-terminal residues of the human c-Jun protein (Oncogene Science, Cambridge, MA) that was preadsorbed to protein A-Sepharose CL4B (Pharmacia Biotech Inc.). In some experiments, the immunospecificity of the antibody was assessed by preadsorbing the antiserum (2 h at 4°C) before addition to the samples with different concentrations (1-5 g) of the synthetic peptide used to generate anti-Jun antibody.
Labeled immunocomplexes were collected by centrifugation (14,000 ϫ g for 2 min at 4°C) and were washed four times with 1 ml of lysis buffer, and protein complexes were eluted from the beads by boiling in 2 ϫ concentrated sample buffer (10% [v/v] glycerol, 1 mM dithiothreitol, 1% (w/v) SDS, 50 mM Tris-HCl, pH 6.8, and 0.002% bromphenol blue). After 7.5% SDS-polyacrylamide gel electrophoresis (51), the gels were dried under vacuum and were recorded as autoradiographs by photostimulable storage imaging (52), followed by heliumneon laser scanning in a Molecular Dynamics 400A PhosphorImager (Molecular Dynamics, Sunnydale, CA). Volume integrations with subtraction of appropriate backgrounds were performed with software provided by the manufacturers.
Extraction of a CAPP-like Activity from the Cytosol of A431 Cells and in Vitro c-Jun Dephosphorylation-An enriched preparation of CAPP was obtained from A431 cells essentially by the methods as described by Okazaki et al. (23) and Wolf et al. (24). Cells were harvested by trypsinization and were resuspended (2 ϫ 10 7 cells/ml) in homogenization buffer (20 mM Tris-HCl, pH 7.4) supplemented with 1 mM EDTA, 0.1 mM EGTA, 0.5 mM dithiothreitol, and protease inhibitors (5 g/ml aprotinin, 2.5 g/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, and 2 g/ml pepstatin A). Cell suspensions were sonicated (at 4°C for 10 s at 50% setting) in a Braun-Labsonic 2000 (Barcelona, Spain) ultrasonicator, and complete cell lysis was assessed by light microscopy. Cell homogenates were centrifuged (100,000 ϫ g for 60 min at 4°C), and the cytosol was passed through a column of Sephadex G-50 (Pharmacia) preequilibrated with the same buffer. The clarified cytosol was mixed for 60 min at 4°C with DEAE-Sephacel preequilibrated with buffer A (20 mM Tris-HCl, pH 7.4, 1 mM EDTA, 0.1 mM EGTA, 1 mM benzamidine, 0.5 mM dithiothreitol, and 10% glycerol), and the slurry was poured into 4-ml plastic columns and was washed with 100 ml of the same buffer. The columns were washed with buffer A supplemented with 150 mM NaCl until A 280 returned to base line; thereafter, columns were eluted at a flow rate of 0.5 ml/min with a linear (200 -400 mM) sodium chloride gradient in the same buffer. Fractions (1 ml) were collected, snap-frozen with liquid nitrogen, and stored frozen (Ϫ80°C). Phosphatase activity was assayed essentially as described by Okazaki et al. (23) and Wolff et al. (24), except that pooled c-Jun immunoprecipitates (obtained from three to four 32 P i -labeled 150-mm dishes of A431 cells) were used as a substrate.
Briefly, proteins (1-5 g) of each fraction were diluted to 75 l with assay buffer (50 mM Tris-HCl, pH 7.4, and 1 mM EDTA) and were incubated in parallel (5 min at 30°C) with C 6 -cer (5 M), OA (1 nM), or a combination thereof. Lipids were added in 5-l aliquots freshly diluted by sonication from the appropriate stock solutions, with control incubations receiving an equivalent amount of diluent (final concentration in the assay of less than 0.5% ethanol). Dephosphorylation reactions were initiated by adding equal amounts of the pooled 32 P-immunoprecipitated substrate (40 l of a 50% slurry in assay buffer), and dephosphorylation reactions were typically conducted for 15 min at 37°C.
To terminate reactions, each incubation was quenched with 4 volumes of ice-cold assay buffer centrifuged at 14,000 rpm for 5 min at 4°C, and the immunocomplexes were washed three times, boiled in SDS-sample buffer, and resolved and analyzed as described above.  (Fig. 1).

TNF␣ and Bacterial
Time-dependent Effect of TNF␣ Bacterial SMase or C 6 -cer on Cellular Phosphorylation of c-Jun-In unstimulated cultures of 32 P i -labeled A431 cells, the immunoprecipitated c-jun gene product migrated as a single band, as determined by SDSpolyacrylamide gel electrophoresis and autoradiography (Fig.  2). The c-Jun protein was not precipitated by control preimmune serum (results not shown), and immunoprecipitation of 32 P i -labeled c-Jun was blocked by pretreating the antibody with unlabeled cognate antigen (Fig. 2). It is of interest to note that treatment of [ 32 P]orthophosphate-labeled A431 cells with TNF␣, SMase, or the cell-permeable analog induced a time-dependent decrease in 32 P i -labeled c-Jun protein levels (Fig. 2) that closely correlated with TNF␣ or bacterial SMase-induced ceramide generation (Fig. 1). In contrast, stimulation with heat-inactivated TNF␣, SMase, or other phospholipases (phospholipase C, phospholipase A 2 , or phospholipase D) did not affect cellular sphingomyelin levels or the phosphorylation status of c-Jun in resting A431 cells (results not shown).
Effect of OA on c-Jun Phosphorylation in [ 32 P i ]Orthophos-phate-labeled Cells Treated with TNF␣, Bacterial SMase, or C 6 -cer-In an effort to determine the mechanism(s) associated with a ceramide-induced decrease in preexisting 32 P i -labeled Jun levels, 32 P i -labeled cells were treated for 30 min with TNF␣, SMase, or C 6 -cer in the presence or absence of 10 nM OA (Fig. 3). Whereas OA alone slightly reduced c-Jun phosphorylation levels, the cell-permeant phosphatase inhibitor effectively blocked a TNF␣-induced decrease in 32 P i -labeled c-Jun levels in A431 cells. Although OA has been reported to induce down-modulation and shedding of the p55-type TNF␣ receptor in A431 and other cell types (53), it seems unlikely that the antagonistic effect of OA on TNF␣-induced c-Jun dephosphorylation may simply reflect a nonspecific consequence of receptor transmodulation because a similar, albeit less pronounced, effect of the phosphatase inhibitor on 32 P i -labeled c-Jun levels was observed in cells treated with exogenous SMase or the ceramide analogue.

Effect of Ceramide and OA on c-Jun Dephosphorylation by Partially Purified Cytosolic Extracts from A431
Cells-A phosphatase activity was partially purified by the two-step procedure described under "Materials and Methods" from cytosolic extracts of A431 cells, and equal amounts of c-Jun obtained from 32 P i -labeled A431 epithelial cells served as substrate in the immunocomplex phosphatase assays. As shown in Fig. 4, no differences were observed in c-Jun phosphorylation after In parallel experiments, no effect on sphingomyelin hydrolysis was observed when cells were treated with equivalent amounts of other phospholipases (phospholipase C, phospholipase A 2 , or phospholipase D), heat-inactivated SMase, or TNF␣ (results not shown).

FIG. 2. Time course of c-Jun dephosphorylation in A431 cells treated with TNF␣, bacterial
SMase, or membrane-permeable C 6 -cer. A431 cells (ϳ4 ϫ 10 6 cells/well) were labeled for 4 h in phosphate-free medium in the presence of [ 32 P i ]orthophosphate (500 Ci/ml) and were treated for the time periods indicated with 10 ng/ml of TNF␣ (E), 0.3 unit/ml bacterial SMase (Ⅺ), or 5 M of cell-permeant C 6 -cer (Ç). All experimental agents were added in 50-l aliquots freshly diluted in sterile culture medium, with control incubations receiving the same volume of medium and a similar final concentration (less than 0.1%) of diluent. To terminate experiments, the medium was aspirated, and cell lysates were obtained and immunoprecipitated overnight with 1 g of c-Jun polyclonal antiserum, as described under "Materials and Methods." After separation of proteins by SDS-polyacrylamide gel electrophoresis, the dried gels were analyzed on a Molecular Dynamics 400A PhosphorImager (lower panel), as described under "Materials and Methods." The immunospecificity in TNF␣-treated cells was determined by preadsorbing the antiserum (2 h at 4°C) with the synthetic peptide used to generate anti-c-Jun antibody, and similar results were obtained with cells treated with SMase or C 6 -cer (results not shown).
incubation of immunoprecipitated beads with heat-inactivated cytosolic extracts or cytosol preincubated with diluent alone. In contrast, C 6 -cer (5 M) clearly reduced 32 P i -labeled c-Jun levels, and incubation with 1 nM OA specifically reversed this effect. No effect on 32 P i -labeled c-Jun was observed when cell extracts were assayed in the presence of other lipids tested, including sphingosine, phosphatidylcholine, or dioleine (results not shown).
Dephosphorylation of c-Jun by a Phosphatase Activity Extracted from A431 Cells Treated with TNF␣, Bacterial SMase, or C 6 -cer-Because preincubation with ceramide induced c-Jun dephosphorylation by a partially purified cytosolic phosphatase (Fig. 4), we next tested whether in intact A431 cells stimulation of the sphingomyelin pathway may also activate a similar OA-sensitive phosphatase activity under the same immuno-complex assay conditions described under "Materials and Methods." As shown in Fig. 5, cultured A431 cells were treated (30 min) with and without OA (10 nM) in the presence or absence of TNF␣ (10 ng/ml), SMase (0.3 units/ml), or C 6 -cer (5 M), and the cytosolic phosphatase was obtained and assayed using 32 P i -labeled c-Jun as substrate. Partially purified cytosolic extracts from TNF␣-, SMase-, or ceramide-treated cells induced the rapid dephosphorylation of 32 P i -labeled c-Jun, an effect that was specifically abrogated when A431 cells were simultaneously treated with OA. DISCUSSION The multitude of biological responses induced by TNF␣ relies on the activation of two dissimilar receptors of 55 kDa (p55TNFR) and 75 kDa (p75TNFR) that are coexpressed on the membranes of virtually all cell types (reviewed in Refs. 54 -56). Although the mechanism of signal transduction has not been fully elucidated, early biochemical events detectable after TNF␣-stimulation include GTP-binding proteins, activation of phospholipases (phospholipase A 2 and phosphatidylcholinespecific phospholipase C), and different sphingomyelin phosphodiesterases (SMases) characterized by its differential pH optimum and cellular localization (54,55). Although most cell types coexpress both types of TNF␣ receptors, the 55-kDa TNFR is responsible for the majority of cytokine actions, being the contribution of the larger receptor species explained, at least in part, by the so-called ligand passing model in which the pp75TNFR presents TNF␣ to the signal-transducing pp55TNFR molecule (56). In the present study, the demonstration that TNF␣ induces sphingomyelin hydrolysis to phosphocholine and a ceramide moiety in human A431 cells complements earlier observations placing the positive coupling of the activated p55TNFR to a neutral and/or an acidic SMase as an early transmembrane event in other cell types (3)(4)(5)(6)(7)(8)(9)(10). Although a complex pattern of target proteins, which include ceramideactivated protein kinase (12,13,17), mitogen-activated protein kinases (16,18), protein kinase C (19,20), and protein phosphatase 2A-type CAPP (21)(22)(23)(24), may account, at least in part, for enhanced phosphorylation of cellular proteins observed after TNF␣ stimulation (54,55), only few data support a role for ceramide-mediated activation of gene transcription (3)(4)(5)(6)(7)(8)(9)(10).
Ligand-stimulated sphingomyelin hydrolysis has been recently reported to activate JNK/SAPK (30 -32), a proline-directed kinase that phosphorylates the c-Jun component of AP-1 at its N-terminal transactivation domain (36 -40). Because AP-1 plays a major role as a convergence point coupling extracellular signals from the membrane receptor to the nucleus (25)(26)(27)(28)(29), the above-mentioned results provide an important link for the sphingomyelin pathway in TNF␣-induced phenotypic responses.
Ceramide-activated JNK contributes to cellular responses via phosphorylation of c-Jun in hepG2 (30), HL-60 cells (31), or bovine aortic endothelial cells (32), and in the present study, we present strong evidence supporting the notion that TNF␣-induced generation of ceramide activates an OA-sensitive phosphatase that specifically dephosphorylates the c-Jun component of AP-1 in resting A431 epithelial cells (Figs. 2 and 3). Moreover, after the initial steps described by Okazaki et al. (23) and Wolff et al. (24) for CAPP purification from rat brain and cultured T9 rat glioma cells, we show (Fig. 4) that cytosolic extracts of A431 cells incubated with ceramide (5 M) dephosphorylated 32 P i -labeled c-Jun used as substrate in an immunocomplex phosphatase assay, an effect that was specifically abrogated by concentrations of OA (1 nM) compatible with protein phosphatase 2A inhibition in cell-free systems (reviewed in Ref. 57). As shown in Fig. 5, a similar OA-sensitive effect on 32 P ilabeled c-Jun dephosphorylation was also observed after incubation of cytosolic extracts obtained from TNF␣, SMase, or short-chain ceramide-treated A431 cells. Because the cytosolic heterotrimeric CAPP shares several biochemical properties with protein phosphatase 2A, including its sensitivity to OA in a range (1-10 nM) similar to concentrations used in this study (21)(22)(23)(24), it seems reasonable to conclude that c-Jun dephosphorylating activity in A431 cells is mediated by CAPP.
Functional activation of AP-1 requires two events that are frequently, but not always, coordinately induced in resting cells: phosphorylation by JNK/SAPK of serines proximal to the N-terminal domain and removal of phosphates from inhibitory sites at the C-terminal region that are also important for DNA association and enhanced transcription of c-jun-responsive genes (25-29, 34, 35). In addition to JNK/SAPK (36 -40), c-Jun is targeted by members of the glycogen synthase kinase-3 family (41)(42)(43)(44)(45) and the nuclear casein kinase CKII (46), which under a variety of experimental conditions have been shown to phosphorylate similar sites at the DNA binding domain of c-Jun, resulting in reduced AP-1 association to DNA (41)(42)(43)(44)(45)(46).
Although a physiological role for these kinases in AP-1 regulation derives from the observation that overexpression of glycogen synthase kinase-3 (45) or microinjection of CKII enhances c-Jun phosphorylation in resting fibroblastic and epithelial cells (46), the mechanisms whereby the activity of these kinases is regulated in vivo are not fully understood (25-29, 34, 35). Both kinases are constitutively active, tonically phosphorylating important cellular proteins that are normally activated by dephosphorylation reactions involving protein phosphatases 1 and 2A; thus far, in vivo studies have failed to detect measurable changes in CKII or glycogen synthase kinase-3 in response to extracellular agents (25)(26)(27)(28)58). In addition, although JNK/SAPK is activated by phosphorylation and is inactivated by a mitogen-activated protein kinase phosphatase-1 (59), it is presently unknown whether dephosphorylation of c-Jun at its C-terminal domain in intact cells is mediated by protein kinase C-induced phosphorylation and inactivation of glycogen synthase kinase-3 (42,44) and/or direct activation of an as yet unidentified Jun-phosphatase (29).
Results showing the ability of ceramide-treated cytosolic ex-tracts to dephosphorylate c-Jun can be a faithful measure of the enzymatic activity of CAPP and support the notion that c-Jun may be a downstream target for CAPP under in vivo conditions. Although this possibility is reinforced by the observation that extracts obtained from TNF␣, SMase, or ceramidetreated cells are also endowed with a similar phosphatase activity (Fig. 5), the cascade of events that may ultimately result in c-Jun dephosphorylation in intact cells is complex; therefore, the present findings provide a framework for future studies, which are currently under way. Nevertheless, the results presented here are important by themselves, and together with previous reports showing a role for CAPP in c-myc transmodulation (11), constitute a significant advance toward understanding the role of CAPP as a downstream target of ceramide-induced nuclear events.