Activation of p38 mitogen-activated protein kinase by norepinephrine in T-lineage cells

Abstract

The catecholamine norepinephrine (NE) stimulates T lymphocytes through a beta-adrenergic receptor (βAR)/adenylyl cyclase (AC)/cyclic AMP (cAMP)/protein kinase A (PKA) pathway, leading to altered cell responsiveness and apoptosis. p38 Mitogen-activated protein kinase (MAPK), a major intracellular signalling mediator for cellular and environmental stressors, is involved in the production of immune modulators and in the regulation of T-cell development, survival and death. In these studies we investigated the relationship among NE signalling, p38 MAPK activity and T-cell death. We showed that NE stimulation of BALB/c mouse thymocytes and S49 thymoma cells selectively increases the dual phosphorylation and activity of p38α MAPK. p38 MAPK activation involves the βAR, Gs protein, AC, cAMP and PKA, as determined through the use of a βAR antagonist, activators of AC and cAMP, and S49 clonal mutants deficient in Gs and PKA. Dual phosphorylation of p38 MAPK is also dependent on its own catalytic activity. Inhibition of p38 MAPK activity revealed its involvement in cAMP-mediated activating transcription factor-2 (ATF-2) phosphorylation, Fas ligand messenger RNA (mRNA) up-regulation, and cell death. These results identify a mechanism through which NE stimulation of the βAR/Gs/PKA pathway activates p38 MAPK, which can be potentiated by autophosphorylation, and leads to changes in T-cell dynamics, in part through the regulation of Fas ligand mRNA expression.

Introduction

Psychological stress can initiate a series of events culminating in the release of steroids, catecholamines (CAs) and other neurotransmitters previously associated with altered immune functions.^1,2 CAs can act as neurotransmitters in the nervous system, particularly the sympathetic nervous system, and play an additional role in the integration of signalling between the nervous system and the immune system. CAs released from sympathetic nerve terminals innervating organs can interact directly with cells by binding to the cell-surface G protein-coupled receptor family of adrenergic receptors (ARs).^1,3 Classically, CA stimulation of ARs initiates a signalling cascade whereby G protein activation leads to the generation of cyclic AMP (cAMP), through adenylyl cyclase (AC), and the subsequent activation of protein kinase A (PKA). This signalling mechanism results in a plethora of cellular changes, including cell proliferation, differentiation and apoptosis.^1,4–6 In the immune system, CAs and cAMP can play an inhibitory role in the modulation of T lymphocytes and cellular immune responses.^2,3,7,8

Previously, we reported that the major stress CA, norepinephrine (NE) affects T lymphocytes by inhibiting their activation and by regulating the expression of the cell-surface protein, Thy-1 (CD90), in both BALB/c mouse thymocytes and in BALB/c-derived S49 thymoma cells through the βAR/AC/cAMP/PKA pathway.^9–11 Other observations in S49 cells are consistent with an important role for cAMP in the regulation of genes that control cell cycle progression and apoptosis.^6,12–14 In this study, we investigated the effect of NE on p38 MAPK, an important stress-activated enzyme involved in regulating transcriptional and post-transcriptional events for many genes including cytokines, cell cycle regulators and apoptotic factors.

Mitogen-activated protein kinases (MAPKs) are a family of serine/threonine protein kinases involved in the transduction of signals from the cell membrane to the nucleus upon stimulation with growth factors, cytokines and stress mediators.¹⁵ p38 MAPK, a major transducer of stress stimuli, is activated in response to cellular and environmental stresses, including DNA damage, inflammatory cytokines and oxidative stress, and influences cellular events including cell proliferation, cell cycle progression and apoptosis.^16–18 The cellular response to p38 MAPK activation is difficult to predict because it is dependent on many factors, including cell type, stage of growth and experimental condition.^18,19 In immune cells, p38 MAPK participates in regulating the production of cytokines, such as tumour necrosis factor alpha (TNF-α) and interleukin-10, pro-inflammatory mediators and angiogenic factors.^20–22 p38 MAPK activity is also important in the development and maturation of T cells in the thymus; however, persistent p38 MAPK activity has been correlated with negative selection and reduced production of mature T cells.^23–29 The biological activity of p38 MAPK is modulated through dual phosphorylation of Thr180 and Tyr182 and its downstream effector molecular targets, including phosphatases, transcription factors [such as activating transcription factor-2 (ATF-2) and cAMP response element-binding (CREB)] and kinases [such as MAPK-activated protein kinase 2 (MK2)].^15,30–34

Although the effects of CA on p38 MAPK in B cells have been reported, there is a paucity of information about p38 MAPK in T-lineage cells.³⁵ Previous work has shown that p38 activation initiated by βAR stimulation using isoproterenol, a synthetic analogue of NE, can be inhibitory in human peripheral blood lymphocytes, or stimulatory in Jurkat cells, but the mechanisms involved have yet to be determined.^36,37 To further delineate the p38 kinase function in T-lineage cells, our studies employed the physiologically natural βAR agonist, NE, to investigate the mechanism of p38 MAPK regulation in a less-differentiated thymocyte population, namely thymocytes isolated from thymuses of 4- to 6-week-old BALB/c mice, and a BALB/c-derived S49 thymoma cell line characterized as double-positive (DP) T cells.⁶ The S49 thymoma cell was chosen as the model system for this study because, in addition to the wild-type (WT) cell type, there are a number of defined somatic cell mutants, including mutants deficient in Gs protein (cyc−) or in PKA function (kin−), for contrast with the WT cell. Historically, S49 cells have been the cell of choice for studying signalling mechanisms, as in the elucidation of the mechanisms for G-protein-coupled receptor signalling.

In this study we demonstrate that NE activates p38 MAPK in both mouse thymocytes and S49 WT cells, but not in S49 mutant cells lacking Gs protein or PKA activity. The mechanism of p38 MAPK activation involves the βAR/Gs/PKA pathway and is dependent on p38 kinase activity. Downstream physiological effects associated with p38 MAPK activity include ATF-2 phosphorylation, Fas ligand (FasL) messenger RNA (mRNA) regulation and cell death. To our knowledge, these studies demonstrate, for the first time, CA-mediated p38 MAPK activation through a βAR/Gs/PKA pathway.

Materials and methods

Cell culture

Mouse thymocytes were isolated from 4- to 6-week-old pathogen-free male BALB/c mice. The mice were killed by CO₂ inhalation followed by cervical dislocation, thymuses were surgically removed and minced, and thymocytes were isolated by passing the minced thymuses through nylon mesh. Thymocytes were cultured in RPMI-1640 (Hyclone, Logan, UT) supplemented with 0·3% gentamycin (Gibco Invitrogen, Carlsbad, CA), 0·4% fungizone (Gibco), 0·1% penicillin/streptomycin (Gibco) and 0·1% Tylosine (Sigma-Aldrich, St. Louis, MO) in a humidified atmosphere containing 5% CO₂ at 37°. These procedures comply with Federal and Institutional guidelines.

BALB/c mouse-derived S49 T-thymoma WT cells, and Gs-deficient and PKA-deficient mutant S49 cell clones (cyc− and kin−, respectively), described previously,¹¹ were maintained at a density of 1 × 10⁵ to 2 × 10⁶ cells/ml in suspension in RPMI-1640 (Hyclone) supplemented with 22% Dulbecco’s modified Eagle’s minimal essential medium (DMEM) (CellGro Mediatech Inc., Manassas, VA), 10% heat-inactivated fetal bovine serum (FBS) (CellGro), 0·3% gentamycin (Gibco), 0·4% fungizone (Gibco), 0·1% penicillin/streptomycin (Gibco) and 0·1% Tylosine (Sigma-Aldrich) in a humidified atmosphere containing 5% CO₂ at 37°.

Cells were exposed to concanavalin A (Con A) (0·5 μg/ml; Sigma-Aldrich), NE (150 μm, freshly prepared in 10 mm HCl; Sigma-Aldrich), 8-bromoadenosine 3′,5′-cyclic monophosphate (8-bromo cAMP) (600 μm, freshly prepared in RPMI-1640; Sigma-Aldrich), forskolin [25 μm, prepared in dimethyl sulphoxide (DMSO); Sigma-Aldrich], pertussis toxin (20 ng/ml; Sigma-Aldrich), cholera toxin (100 ng/ml; Sigma-Aldrich), SB 203580 (1 μm, prepared in DMSO; Sigma-Aldrich), nadolol (20 μm in RPMI-1640; Sigma-Aldrich), or somatostatin (1 μm; Sigma-Aldrich), as indicated, starting at a cell density of 5 × 10⁶ cells/ml for thymocytes or 1 × 10⁶ cells/ml for S49 cells. Cell viabilities, as determined by Trypan Blue exclusion, were at least 95% at the start of experiments.

Cell viability study

S49 WT cells, at a starting density of 2·5 × 10⁵ cells/ml, were cultured, as described above, using the following treatments: no agent, 8-bromo cAMP + DMSO (vehicle for SB 203580), or 8-bromo cAMP + SB 203580 added simultaneously. Cell viabilities, determined by Trypan Blue exclusion, were completed in duplicate at 24, 48 and 72 hr after exposure to the agents.

Preparation of protein extracts

Total cellular protein was prepared using Cell Extraction Buffer (#FNN0011; Biosource Invitrogen, Carlsbad, CA) supplemented with 1 mm phenylmethylsulphonyl fluoride (PMSF) (Sigma-Aldrich) and 0·5 μl of Protease Inhibitor Cocktail (Sigma-Aldrich) per ml of buffer. Cells were isolated by centrifugation (200 g, 7 min, RT), washed in ice-cold phosphate-buffered saline (PBS) and lysed on ice for 30 min at a ratio of 10 μl of lysis buffer to 1 × 10⁶ cells. Insoluble fractions (DNA) were separated by centrifugation (12 000 g, 10 min, 4°) and discarded. The supernatant was aliquoted and stored at −80° for use in western blotting or enzyme-linked immunosorbent assays (ELISAs).

Nuclear protein was prepared by isolating the nuclear fraction followed by lysis in cell-extraction buffer. Then, the cells were collected by centrifugation (200 g, 7 min, RT), washed in ice-cold PBS and lysed in HEPES-based lysis buffer [10 mm HEPES, 3 mm MgCl₂, 40 mm KCl, 5% glycerol, 0·2% Nonidet P-40 (NP-40)] supplemented with 1 mm PMSF (Sigma-Aldrich), 0·5 μl of Protease Inhibitor Cocktail (Sigma-Aldrich) per ml of buffer and 10 μl of Phosphatase Inhibitor Cocktails 1 and 2 (Sigma-Aldrich) per ml of buffer, at a ratio of 10 μl buffer to 1 × 10⁶ cells, for 20 min on ice. The lysate was centrifuged (300 g, 5 min, 4°), the supernatant was discarded and the nuclear pellet was lysed in cell extraction buffer, as described above.

Western blotting

Total cellular protein (20 μg) [unless specified as nuclear protein (5 μg) in the figure legend] was resolved by electrophoresis on 10% sodium dodecyl sulphate (SDS) polyacrylamide gels and then transferred to nitrocellulose membrane. Membranes were blocked for 1 hr at room temperature in TBST (10 mm Tris/base, pH 7·6, 0·1 m NaCl, 0·1% Tween-20) containing 5% membrane-blocking agent (GE Healthcare, Buckinghamshire, UK) and then probed with the following primary antibodies: rabbit anti-(p38 MAPK) (1/1000 dilution; Cell Signaling, Danvers, MA), rabbit anti-(phospho-p38 MAPK) (pT180/pY182) (1/1000 dilution; Cell Signaling), mouse anti-(phospho-CREB) (1/1000 dilution; Cell Signaling), rabbit anti-CREB (1/1000 dilution; Cell Signaling), rabbit anti-(phospho-ATF-2) (pT69/pT71) (1/1000 dilution; Cell Signaling), rabbit anti-(ATF-2) (1/1000 dilution; Santa Cruz Biotechnology, Santa Cruz, CA), goat anti-actin (1/4000 dilution; Santa Cruz Biotechnology), rabbit anti-(phospho-MK2) (pT222) (1/1000 dilution; Cell Signaling), or rabbit anti-MK2 (1/1000 dilution; Cell Signaling), overnight at 4°. After washing with TBST, membranes were probed with the corresponding horseradish peroxidase-labelled secondary antibody for 1 hr at room temperature and bound antibody was detected using the ECL Plus Western Blotting detection reagents (GE Healthcare). Luminescence was visualized on a Kodak Biomax MS film, and the signal intensity was measured by densitometry using Scion image software.

ELISA

Total cellular protein (6 μg) was analyzed in quadruplicate for phospho-p38 MAPK (pTpY180/182) using the BioSource Immunoassay kit, according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA).

Phospho-MAPK protein array

Following exposure to NE (150 μm, 15 min), total cellular protein was obtained using the reagents and protocol provided with the Human Phospho-MAPK Array kit (R&D Systems, Minneapolis, MN). Arrays were incubated with 100 μg of protein and the phospho-MAPK levels were detected by chemiluminescence, according to the manufacturer’s instructions.

In vitro kinase activity assay

p38 MAPK catalytic activity was determined using the p38 MAPK Activity Assay kit (Sigma-Aldrich) according to the manufacturer’s instructions. In brief, following cell exposure to agents as indicated in figure legends, 1 × 10⁷ cells were lysed in CelLytic M Cell Lysis Reagent (Sigma-Aldrich) supplemented with 0·5 μl of Protease Inhibitor Cocktail (Sigma-Aldrich), 1 mm PMSF (Sigma-Aldrich) and 10 μl of Phosphatase Inhibitor Cocktails 1 and 2 (Sigma-Aldrich) per ml of buffer, for 30 min on ice. The lysate was incubated with 2 μl of anti-p38 MAPK and 30 μl of Protein A affinity gel beads (provided with the kit) overnight at 4°, and the kinase assay was performed using the immunoprecipitate, as instructed. The reaction was stopped by the addition of 12 μl 4 × SDS-loading dye, and 5 μl of the sample was resolved by SDS–polyacrylamide gel electrophoresis. The activity of p38 MAPK was determined by its ability to phosphorylate the substrate ATF-2, as detected by western blotting [using anti-(phospho-ATF-2), 1/2000 dilution, as the probe].

RNA extraction and quantitative real-time polymerase chain reaction

Following cell exposure to agents as indicated in figure legend exposures, cells were collected by centrifugation (200 g, 7 min, RT) and lysed in TRIzol LS Reagent (Invitrogen), at a ratio of 1 ml of reagent to 1 × 10⁷ cells. Total cellular RNA was extracted using the RNEasy Mini Prep kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions. Complementary DNA (cDNA) was prepared from polyadenylated mRNA by reverse transcription using Superscript III (Invitrogen) and Oligo(DT)12–18 primers (Invitrogen), according to the manufacturer’s instructions. The resulting cDNA was used in quantitative polymerase chain reaction (qPCR) amplifications containing gene-specific primers for FasL, Bim or the housekeeping gene, hypoxanthine phosphoribosyltransferase 1 (HPRT1) and SYBR Green master mix (Applied Biosystems, Carlsbad, CA). Samples were run in triplicate in the 7900HT Applied Biosciences real-time PCR thermocycler using default settings. The cDNA expression levels were analyzed using the qGene software (http://www.biotechniques.com/softlib/qgene.html).

Primer sequences, selected from a primerbank, (http://pga.mgh.harvard.edu/primerbank/) were as follows: FasL forward, 5′-TCC GTG AGT TCA CCA ACC AAA-3′, and FasL reverse, 5′-GGG GGT TCC CTG TTA AAT GGG- 3′; HPRT1 forward, 5′-CTG GTG AAA AGG ACC TCT CG-3′, and HPRT1 reverse, 5′-TGA AGT ACT CAT TAT AGT CAA GGG CA-3′; and Bim forward, 5′-CCC TGG CCC TTT TGC TAC CC-3′, and Bim reverse, 5′-ACT TGT CAC AAC TCA TGG GTG-3′. All primers were synthesized and purchased from Sigma Genosys (Sigma-Aldrich).

Statistical analysis

Data are presented as the mean ± standard deviation (SD). Statistical differences between the mean values of control and experimental groups were determined using two-tailed, unpaired Student’s t-tests.

Results

NE activates p38 MAPK in mouse thymocytes and in S49 T-thymoma cells

To investigate the regulation of p38 MAPK by the CA, NE, thymocytes isolated from BALB/c mice were stimulated with 150 μm NE for 15 min under resting or Con A-activated conditions. ELISA analysis for phosphorylated p38 MAPK (pT180/pY182) revealed that NE treatment of both resting and Con A-activated thymocytes resulted in increased p38 MAPK phosphorylation, of 1·5- and 1·3-fold, respectively (Fig. 1a). Extension of these studies to include BALB/c-derived S49 thymoma cells revealed that NE-mediated increase in phosphorylation of p38 MAPK occurred at a similar level, 1·6-fold, as that observed in thymocytes (Fig. 1a). Immunoblot analysis confirmed the ELISA results, such that NE treatment resulted in an increase in phospho-p38 MAPK, while total p38 MAPK remained unchanged (Fig. 1b). To determine whether the observed increase in phosphorylation of p38 MAPK corresponded with an increase in kinase activity, its ability to phosphorylate the substrate, ATF-2, was assayed using an in vitro kinase assay, and it was demonstrated that the NE-mediated increase in p38 MAPK phosphorylation paralleled an increase in its catalytic activity (Fig. 1c).

Figure 1.

p38 Mitogen-activated protein kinase (MAPK) is activated in mouse thymocytes and in S49 wild-type (WT) T-thymoma cells. (a) Thymocytes – resting (Con A−) or activated with concanavalin A (ConA+) and S49 WT cells – were stimulated with norepinephrine (NE) (15 min). The phosphorylation of p38 MAPK (p-p38) was detected by enzyme-linked immunosorbent assay (ELISA). Results represent the fold change with respect to no agent and are given as the mean ± standard deviation (SD) of three to six independent experiments. (b) NE-induced phosphorylation of p38 MAPK in S49 WT cells at 15 min was detected by immunoblotting. A representative immunoblot of phosphorylated p38 MAPK (p-p38), stripped and reprobed for total p38 MAPK (p38), is shown; n = 3 independent experiments. (c) NE-induced p38 MAPK activity in S49 WT cells at 15 min was determined by in vitro kinase analysis, as described in the Materials and Methods. A representative immunoblot for phosphorylated activating transcription factor-2 (p-ATF-2) is shown; n = 3 independent experiments. (d) NE-induced phosphorylation of p38 MAPK in S49 cells occurs for the p38α isoform, as detected by a phospho-MAPK protein array incubated with extract from NE-treated cells (15 min). Shown is a representative array region from two independent trials. (e) S49 WT cells were stimulated with NE at the indicated times and p38 MAPK phosphorylation was detected by ELISA. Results represent the fold change with respect to no agent and are given as mean ± SD for n ≥ 5 independent experiments. *P < 0·05, **P < 0·01, ***P < 0·005.

The p38 MAPK family consists of four different isoforms – p38α, p38β, p38δ and p38γ– and the p38α isoform is the most predominant in T cells.²³ To identify which p38 isoform is phosphorylated by NE in S49 cells, a phospho-MAPK protein array was performed using total cellular protein from NE-treated S49 WT cells (150 μm, 15 min); only the phospho-p38α isoform was detected (Fig. 1d).

To characterize the temporality of the NE-mediated p38 MAPK phosphorylation, a time course from 5–60 min was investigated using ELISA. Stimulation of S49 cells with NE (150 μm) increased the phosphorylation of p38 MAPK within 5 min of treatment by 1·7-fold and was sustained for at least 60 min (Fig. 1e). Collectively, these results reveal that NE stimulation of mouse thymocytes activates p38 MAPK rapidly by phosphorylating its α isoform.

βAR stimulation and signalling through Gs is required for the NE-induced activation of p38 MAPK

NE stimulation of cells can occur through AR-dependent, as well as non-AR-dependent, mechanisms.¹⁰ To elucidate the role of βAR (i.e. the AR expressed on T lymphocytes) in NE-mediated p38 MAPK activation, the βAR antagonist, nadodol (20 μm), was used to block βAR before stimulation with NE, and the activity of p38 MAPK was analyzed. The NE-mediated increase in phosphorylated p38 MAPK, as detected by ELISA, was reversed by pretreatment with nadolol (Fig. 2a). Moreover, nadolol treatment also antagonized the kinase activity of p38 MAPK, as seen by decreased phosphorylation of ATF-2 in the kinase assay (Fig. 2b). Thus, signalling through the βAR is involved in the activation of p38 MAPK.

Figure 2.

Norepinephrine (NE)-mediated p38 mitogen-activated protein kinase (MAPK) activation occurs through the beta-adrenergic receptor (βAR)/Gs protein/adenylyl cyclase (AC)/cyclic AMP (cAMP) pathway. (a,b) S49 wild-type (WT) cells were untreated, stimulated with NE (15 min), or stimulated with NE after a 20 min pretreatment with nadolol (NE nadolol). (a) Phosphorylated p38 MAPK (p-p38) was detected using an enzyme-linked immunosorbent assay (ELISA). The results represent the fold change and are given as the mean ± standard deviation (SD) of three independent experiments. (b) p38 MAPK activity was assessed using an in vitro kinase assay, as described in the Materials and methods. A representative immunoblot of phosphorylated activating transcription factor-2 (p-ATF-2) is shown; n = 3 independent experiments. (Images shown are from the same exposure; however, unrelated lanes have been removed.) (c) S49 WT cells were stimulated with forskolin (FSK) or cholera toxin (CTX) for 120 min. The results shown represent an immunoblot for phosphorylated p38 MAPK (p-p38), stripped and reprobed for total p38 MAPK (p38). The signal intensity, analyzed by densitometry, is mean fold change ± SD; n ≥ 3 independent experiments. (d) S49 WT cells were untreated (PTX−) or pretreated with pertussis toxin (PTX+) for 18 hr followed by stimulation with NE (15 min). Phosphorylated and total p38 MAPK was detected by immunoblotting. A representative immunoblot is shown and the signal intensity, analyzed by densitometry, is depicted as mean fold change ± SD for three independent trials. NS, not significant. (e) S49 WT cells were treated with somatostatin (SOM) for 15 min, and phosphorylated and total p38 MAPK were detected by immunoblotting. The images are representative of three independent experiments. Densitometry shows mean fold change ± SD for three independent experiments. (f) S49 WT cells were stimulated with 8-bromo cAMP at the indicated times, and p38 MAPK phosphorylation was detected by ELISA. The results represent the fold change with respect to no agent and are mean ± SD for n ≥ 5 independent experiments. (g) S49 cyc− cells were treated with NE, FSK, or 8-bromo cAMP for 15 min. The results shown represent immunoblots for p-p38 MAPK and p38 MAPK from one independent trial. The signal intensity, analyzed by densitometry, is depicted as the mean fold change ± SD for four independent trials. *P < 0·05, **P < 0·01, ***P < 0·005.

The most common signalling mechanism initiated by βAR stimulation is the βAR/Gs/AC/cAMP/PKA pathway. To determine if NE/βAR-mediated p38 MAPK activation in T-lineage cells results from this pathway, the involvement of these signalling molecules was investigated through genomic and pharmacological approaches.

The Gα family includes two subtypes, Gαs (Gs) and Gαi (Gi), which associate with the βAR and are involved in the activation of p38 MAPK.^38,39 To investigate the role of Gs, cholera toxin, an agent that catalyzes the ADP-ribosylation of Gs, was used. As there is a 2 hr lag in the elevation of cAMP levels by cholera toxin treatment of S49 WT cells, we chose to use a 2 hr treatment time in these experiments.⁴⁰ Treatment of S49 WT cells with cholera toxin (100 ng/ml, 2 hr) resulted in a 1·5-fold increase of p38 MAPK phosphorylation, as revealed by immunoblotting (Fig. 2c), suggesting a role for Gs in NE-mediated p38 MAPK activation. βAR also signals through Gi, and thus the role of Gi in p38 MAPK phosphorylation was investigated. Pretreatment of S49 cells with the Gi inhibitor, pertussis toxin (20 ng/ml), for 18–24 hr, a treatment time shown to be inhibitory in S49 cells,⁴¹ had no effect on NE-mediated p38 MAPK phosphorylation, as detected by immunoblotting (Fig. 2d). Furthermore, treatment with the Gi activator somatostatin (1 μm) for 15 min did not increase p38 MAPK phosphorylation (Fig. 2e), showing that Gi signalling does not activate p38 MAPK in S49 cells. These results suggest a role for Gs signalling, but not for Gi signalling, in NE-mediated activation of p38 MAPK.

To determine if Gs protein is required for the NE-mediated activation of p38 MAPK, we utilized the S49 cyclase mutant cell, S49 cyc−, which lacks Gs protein. Treatment of S49 cyc− cells with NE (150 μm, 15 min) had no effect on the levels of phosphorylated p38 MAPK (Fig. 2g), whereas activation of Gs downstream mediators with 8-bromo cAMP (600 μm, 15 min) or forskolin (25 μm, 15 min), an agent that activates the Gs target, AC, increased the levels of phosphorylated p38 MAPK by 2-fold and 1·9-fold, respectively (Fig. 2g). Thus, NE signalling through Gs is required for p38 MAPK activation.

As AC is the major target for Gs, its involvement in p38 MAPK phosphorylation was investigated. S49 WT cells were treated with forskolin (25 μm) and the dual phosphorylation of p38 MAPK was measured. Utilizing ELISA, p38 MAPK phosphorylation was increased by 1·9-fold after 15 min (data not shown) and by 3-fold after 2 hr of forskolin treatment (Fig. 2c). βAR signalling through Gs activates AC and thereby results in the accumulation of cAMP. To determine if cAMP is the second messenger involved in the NE-mediated activation of p38 MAPK, S49 cells were treated with the cAMP analogue, 8-bromo cAMP (600 μm), and the levels of phospho-p38 MAPK were detected using ELISA (Fig. 2f). p38 MAPK phosphorylation increased within 5 min of cAMP treatment and remained elevated, reaching a fourfold increase at 60 min. Collectively, these results suggest that NE activation of p38 MAPK occurs through a mechanism involving βAR/Gs/AC/cAMP.

PKA is necessary for NE-mediated p38 MAPK phosphorylation in S49 cells

PKA, the cAMP-dependent protein kinase, is a major effector for cAMP mediation, but recently alternative pathways for cAMP effects have been demonstrated.¹⁴ To determine if PKA is involved in the NE-mediated activation of p38 MAPK, an S49 clonal mutant, deficient in PKA activity (kin−), was used. To verify the S49 kin− PKA deficiency, phosphorylation of the common PKA substrate, CREB, was analyzed by immunoblotting following NE treatment (150 μm, 15 min). The level of phospho-CREB increased in S49 WT cells, but not in the kinase mutant (Fig. 3a). Although kin− cells express a basal level of phospho-CREB, which could result from compensatory kinase activity, the lack of an increase in response to NE, as observed in the WT cells, verifies that PKA is not functional in kin− cells. Immunoblots of phospho- and total p38 MAPK protein levels from S49 kin− cells in the presence and absence of 8-bromo cAMP (600 μm, 15 min) revealed no change in the phosphorylation of p38 MAPK (Fig. 3b). Furthermore, the in vitro kinase assay revealed a basal level of active p38 MAPK in the kinase mutant, as in WT cells, yet there was no enhancement of p38 MAPK activity in the kin− cell in the presence of 8-bromo cAMP (Fig. 3c). Thus, PKA activity is required for NE-mediated p38 MAPK activation.

Figure 3.

Norepinephrine (NE)-mediated p38 mitogen-activated protein kinase (MAPK) activation is protein kinase A (PKA) dependent. (a) Immunoblot detection of phosphorylated CREB (p-CREB), stripped and reprobed for total CREB, from S49 wild-type (WT) and S49 kin− cells stimulated with NE for 15 min. Images are representative of three independent experiments. (b) Immunoblot detection of phosphorylated p38 MAPK (p-p38), stripped and reprobed for total p38 MAPK (p38) in S49 kin− cells treated with 8-bromo cyclic AMP (cAMP) (15 min). Images are representative of three independent experiments. (c) p38 MAPK activity in S49 kin− cells stimulated with 8-bromo cAMP (15 min) was assessed using the in vitro kinase assay, as described in the Materials and methods. A representative immunoblot for phosphorylated activating transcription factor-2 (p-ATF-2) is shown; n = 3 independent experiments. (d) S49 kin− cells were stimulated with NE or 8-bromo cAMP for the indicated times and p38 MAPK phosphorylation was detected by enzyme-linked immunosorbent assay (ELISA). The results represent the fold change with respect to no agent and are given as mean ± standard deviation (SD) for n ≥ 3 independent experiments.

It has been reported that βAR signalling can lead to activation of p38 MAPK by both PKA-dependent and non-PKA-dependent mechanisms in a time-dependent manner.⁴² To determine if PKA is responsible for the sustained increase in phosphorylation of p38 MAPK, as observed in WT cells (Fig. 1e), we assayed the levels of phospho-p38 MAPK in kin− cells at varying time-points from 5 to 60 min (Fig. 3d). In contrast to the WT cells, 8-bromo cAMP (600 μm) or NE (150 μm) did not increase the levels of phosphorylated p38 MAPK at the time-points tested (5, 15, 30 and 60 min). These results unambiguously demonstrate that NE/cAMP-mediated p38 MAPK activation in S49 cells is PKA dependent.

NE-mediated p38 MAPK activation involves autophosphorylation

The activation of p38 MAPK most commonly results from a MAPK cascade in which the MAPK kinases (MAP2Ks), MKK6/MKK3, dually phosphorylate p38 at T180/Y182.⁴³ However, in T cells, an alternative mechanism, involving p38 MAPK autophosphorylation, has been identified.⁴⁴ The p38 MAPK inhibitor, SB 203580, binds specifically to the ATP-binding site and blocks its catalytic activity; however, it does not interfere with the ability of p38 MAPK to be phosphorylated at T180/Y182 by MAP2Ks.^45,46 Thus, we reasoned that pretreatment of cells with SB 203580 would only block autophosphorylation because this mechanism requires the p38 MAPK catalytic function. Pretreatment of S49 cells with SB 203580 (1 μg/ml, as determined in ref. 47) inhibited NE-mediated p38 MAPK catalytic activity, as detected by the in vitro kinase assay (Fig. 4a), and decreased NE-mediated p38 MAPK phosphorylation, as detected by ELISA (Fig. 4b). Because inhibiting the catalytic activity of p38 MAPK led to a decrease in the phosphorylation of T180/Y182 residues, these data imply that autophosphorylation is involved in the p38 MAPK activation through NE/βAR/PKA signalling.

Figure 4.

Norepinephrine (NE)-mediated p38 mitogen-activated protein kinase (MAPK) autophosphorylation in S49 cells. S49 wild-type (WT) cells were untreated, treated with NE (15 min), or pretreated with SB 203580 for 20 min then treated with NE for 15 min (NE SB). (a) The activity of p38 MAPK was measured using an in vitro kinase assay, as described in the Materials and Methods. The image shows a representative immunoblot for phosphorylated activating transcription factor-2 (p-ATF-2); n = 3 independent experiments. (b) Phosphorylated p38 MAPK levels as detected by enzyme-linked immunosorbent assay (ELISA). Data shown represent the mean fold change with respect to no agent ± standard deviation (SD) of three independent experiments. ***P < 0·005.

p38 MAPK is involved in cAMP-mediated apoptosis in S49 WT thymoma cells

As βAR stimulation of S49 cells results in cell death by apoptosis,^6,12 and p38 MAPK activity is involved in the regulation of apoptosis for many cell types,¹⁷ we investigated the role of p38 MAPK in βAR-mediated apoptosis in S49 cells. A consequence of treatment with CAs in cell culture is the potential for oxidation of the media.⁴⁸ To obviate this possibility, we used 8-bromo cAMP in these studies to mimic NE-mediated elevation of intracellular cAMP. S49 WT cells were treated with 8-bromo cAMP in the presence or absence of the p38 MAPK inhibitor, SB 203580, and cell viability was monitored over the course of 72 hr (Fig. 5a). Treatment with 8-bromo cAMP resulted in a 60% decrease in cell viability at 72 hr, which was partially reversed by p38 MAPK inhibition with SB 203580. As controls, treatment with SB 203580 or DMSO (SB 203580 vehicle) for 72 hr did not affect cell viability as compared with the no-agent condition (data not shown). Microscopic examination of the dead cells following treatment with 8-bromo cAMP revealed cell shrinkage, a morphological characteristic indicative of apoptotic cell death. These data suggest that p38 MAPK participates in the regulation of cell death in S49 T lymphocytes.

Figure 5.

Inhibition of p38 mitogen-activated protein kinase (MAPK) activity partially reverses cyclic AMP (cAMP)-mediated apoptosis, up-regulation of Fas ligand (FasL) messenger RNA (mRNA) expression and activating transcription factor-2 (ATF-2) phosphorylation in S49 wild-type (WT) cells. (a) S49 WT cells were treated with 8-bromo cAMP or with 8-bromo cAMP and SB 203580 simultaneously (cAMP SB) for up to 72 hr. Cell viability was measured by Trypan Blue stain exclusion and expressed as percentage of live cells with respect to the total count. A representative experiment of five independent trials is shown. (b) S49 WT cells were untreated or treated for 2 hr with 8-bromo cAMP or with 8-bromo cAMP pretreated with SB 203580 for 20 min (cAMP SB). Steady-state levels of FasL and Bim mRNA were detected by quantitative polymerase chain reaction (qPCR) [normalized to hypoxanthine phosphoribosyltransferase 1 (HPRT1]. Data shown represent the mean fold change with respect to no agent ± standard deviation (SD); n = 4 (FasL), or n = 2 (Bim) independent experiments were performed. *P < 0·05. (c) Nuclear protein from S49 WT cells treated for 2 hr with 8-bromo cAMP or with 8-bromo cAMP pretreated with SB 203580 (20 min). Phosphorylated ATF-2 (p-ATF-2) and total ATF-2 levels were detected by immunoblotting. The images shown represent one of four independent experiments. Densitometry of p-ATF-2 normalized to total ATF-2 is presented as the fold change with respect to no agent. (d) Total cellular protein from S49 WT cells treated with 8-bromo cAMP or norepinephrine (NE) for 15 min. Phosphorylated MK2 (p-MK2) and total MK2 were detected by immunoblotting. The data shown represent one of three independent trials.

Cell death by apoptosis can occur by two different mechanisms: an intrinsic mechanism involving mitochondrial release of apoptotic factors, or an extrinsic mechanism involving ligation of death receptors.^14,49 cAMP-mediated apoptosis of S49 cells has been reported to involve an intrinsic mechanism; however, we found that cAMP up-regulates FasL mRNA, an apoptotic factor important in the extrinsic mechanism¹⁴ (D. A. Chambers and D. H. Davis, Manuscript in preparation). FasL can be regulated by p38 MAPK in T cells, contributing to T-cell-receptor-mediated activation-induced cell death.^50,51 To determine if the involvement of p38 MAPK in cAMP-mediated apoptosis of S49 WT cells is associated with FasL regulation, we analyzed the steady-state levels of FasL mRNA after 2 hr of treatment with 8-bromo cAMP in the presence or absence of SB 203580 (Fig. 5b). FasL mRNA levels increased 9·8-fold after treatment with 8-bromo cAMP; however; in the presence of SB 203580, the up-regulation of FasL mRNA decreased by over 50%. The reversal of cAMP-mediated elevation of FasL mRNA expression by p38 MAPK inhibition suggests its role in FasL mRNA regulation. In addition to FasL, another apoptotic factor, Bim, has been reported to be up-regulated in response to elevated cAMP in S49 cells.⁵² Accordingly, we also examined the role of p38 MAPK on Bim mRNA expression. Surprisingly, in contrast to the findings with FasL, inhibition of p38 MAPK did not alter the cAMP-mediated increase of Bim mRNA (Fig. 5b).

Because the increase of FasL mRNA in the presence of cAMP is only partially dependent on p38 MAPK activity, we questioned if p38 MAPK regulates transcription factors involved in regulating the FasL promoter. Proteins comprising the transcription factor activator protein-1, including ATF-2, bind to a region within the FasL promoter and are involved in its transcriptional regulation.⁵³ As cAMP-mediated p38 MAPK activation increases ATF-2 phosphorylation in vitro, we analyzed its effect in vivo and determined that treatment of S49 WT cells with cAMP for 120 min resulted in an increase in ATF-2 phosphorylation, which was reversed in the presence of SB 203580 (Fig. 5c). We also analyzed MK2, another common substrate of p38 MAPK, to see if it also was phosphorylated in response to NE or cAMP. Unlike ATF-2, MK2 was not phosphorylated after 15 (Fig. 5d) or 60 min of treatment (data not shown). Thus, cAMP-mediated p38 MAPK activation results in the specific activation of the p38 MAPK substrate, ATF-2. These results argue that p38 MAPK participates in the regulation of S49 cell death, possibly as a consequence of its regulation of FasL mRNA expression and ATF-2 phosphorylation.

Discussion

The major focus of this research was to gain a better understanding of how mediators of psychogenic stress, namely CAs, affect the immune response. We have shown previously that the CA, NE, can inhibit T-lymphocyte activation and proliferation.^9,10 Using the S49 thymoma cell as our model system, we found that NE stimulation, through a βAR/cAMP/PKA signalling pathway, alters gene expression both transcriptionally and post-transcriptionally¹¹ (D. A. Chambers and D. H. Davis, Manuscript in preparation). To enhance our understanding of how NE mediates T-lymphocyte physiology, we investigated the response of the major stress-activated kinase, p38 MAPK, to NE stimulation. In this report we demonstrated that (i) p38 MAPK is activated by NE in mouse thymocytes and in S49 thymoma cells, (ii) NE signalling through a βAR/Gs/PKA pathway activates p38 MAPK through a mechanism dependent on p38 kinase activity and (iii) the physiological effects of p38 MAPK in S49 thymoma cells include ATF-2 phosphorylation, FasL mRNA regulation and cell death.

The effects of CA on p38 MAPK activity have been studied in many cell types, including cardiomyocytes, B cells, PC-12 cells, spinal microglia and haematopoietic progenitor cells.^39,54–57 In the T-lineage cells in this study, we showed that NE treatment results in the dual phosphorylation (pT180/pY182) of p38α MAPK, which corresponds to an increase in its catalytic activity (Fig. 1). Others have reported that regulation of p38 MAPK in the presence of NE requires a costimulant.^54,57 In B cells or PC-12 cells, costimulation with NE and CD40, or NE and nerve growth factor, respectively, activates p38 MAPK, suggesting that cross-talk among multiple pathways is necessary for p38 regulation.^54,57 We found that NE stimulation of thymocytes, alone, or when costimulated with Con A, activates p38 MAPK in T-lineage cells, and suggest that this NE effect involves only a βAR-induced signalling cascade.

In addition to classical βAR signalling through cAMP/PKA, alternate pathways and effector molecules have been reported to be involved in NE-mediated p38 MAPK regulation, including a β-arrestin, non-PKA-dependent mechanism, a cAMP/EPAC (exchange factor activated by cAMP) pathway, and signalling events dependent on Gi or α-AR.^{19,38,42,58,59} Using molecular and genetic approaches, in this study we highlight the importance of the classical βAR/Gs/cAMP/PKA signalling pathway in NE-mediated activation of p38 MAPK in T-lineage cells. Evidence for this mechanism, and not for the alternate pathways, is based on (i) the dependence on βAR, as shown by the complete reversal of NE-mediated p38 MAPK activation upon pretreatment with the βAR-selective antagonist, nadolol (Fig. 2a,b); (ii) the requirement for Gs, as shown by the lack of NE-induced p38 MAPK phosphorylation in cyc− cells (Fig. 2g); and (iii) the requirement for PKA, as determined by the lack of NE-induced p38 MAPK activity in S49 kin− cells (Fig. 3), which are deficient in PKA activity, but express EPAC and β-arrestin.^13,60 While a dependency for AC was not directly investigated through loss-of-function experiments, the finding that cholera toxin and forskolin, two activators of AC which work through different mechanisms, lead to p38 MAPK phosphorylation (Fig. 2c), strongly suggests the involvement of AC. Additionally, the product of AC, cAMP, also increased p38 MAPK phosphorylation (Fig. 2f). Thus, we propose that the classical βAR/Gs/AC/PKA pathway activates p38 MAPK (Fig. 6).

Figure 6.

Proposed mechanism of norepinephrine (NE)/p38 mitogen-activated protein kinase (MAPK) signalling in T-lineage cells. NE stimulation of beta-adrenergic receptors (βARs) on T-lineage cells induces signalling events activating the classical Gs protein/adenylyl cyclase (AC)/cyclic AMP (cAMP)/protein kinase A (PKA) pathway and leading to the dually phosphorylated, active form of p38 MAPK. Although the events relating PKA and p38 remain unidentified, the process of activation involves p38 kinase activity. Downstream events of p38 MAPK activation include phosphorylation of activating transcription factor-2 (ATF-2), upregulation of FasL mRNA and cell death, which may be directly related. These events might contribute to the decreased cellularity and thymocyte number, and increased apoptosis, observed in the thymus of stressed mice.^3,67 Dashed lines represent as-yet-undefined steps in the mechanism.

Given that PKA is necessary for p38 MAPK activation, the question arises as to how a serine/threonine-specific kinase leads to phosphorylation of the threonine and tyrosine residues in p38. An in vitro kinase assay has revealed that PKA does not directly phosphorylate p38 MAPK,⁶¹ suggesting the involvement of an intermediary substrate. PKA has been shown to phosphorylate and activate kinases, such as transforming growth factor β-activated kinase 1 (TAK1), and to inhibit phosphatases, such as haematopoietic protein tyrosine phosphatase (HePTP), both of which lead to p38 activation.^54,62,63 It is possible that either of these proteins may be candidate effector molecules responsible for the NE-mediated p38 phosphorylation observed in this study. Preliminary findings have revealed that inhibiting tyrosine phosphatase activity by treating S49 cells with sodium orthovanadate alone is not sufficient to increase p38 phosphorylation (D. A. Chambers and M. D. LaJevic, unpublished data). Therefore, although PKA-mediated inhibition of a phosphatase may occur in response to NE stimulation of S49 thymoma cells, another mechanism, possibly involving kinase activity, may account for the increase in p38 MAPK phosphorylation observed.

p38 MAPK activation typically occurs via a MAPK signalling cascade, wherein MAPK kinase kinases (MAP3Ks) activate MAP2Ks, which activate p38 MAPK through dual phosphorylation (pT180/p182).⁴³ However, alternative mechanisms involving p38 MAPK autophosphorylation have been reported.^44,64 One such mechanism, specific to T cells, involves T-cell-receptor-mediated, non-MAP2K-dependent phosphorylation of p38 MAPK (pY323), followed by p38 MAPK-dependent autophosphorylation (pT180/pY182).^43,44 In our studies, we showed that NE-mediated p38 dual phosphorylation requires p38 MAPK catalytic activity (Fig. 4), suggesting that autophosphorylation is involved. Although we have not functionally ruled out MAPK signalling, if MAP2Ks were active, we would expect to detect dually phosphorylated p38 protein in the presence of the inhibitor, SB 203580, which was not the case. This conclusion is supported by reports that in cells lacking the T-cell-receptor-mediated autophosphorylation mechanism (i.e. non-T cells), inhibition of p38 activity did not interfere with MAP2K activity or p38 dual phosphorylation.^43,44 Additionally, it has been observed that the mechanism of p38 MAPK activation can dictate the substrates targeted by p38 MAPK.^64,65 For example, while classical MAPK activation of p38 typically results in the activation of MK2 and the expression of pro-inflammatory genes, p38 activation by an autophosphorylation mechanism can have the opposite effect.⁶⁴ In our experiments, cAMP treatment resulted in p38-dependent ATF-2 phosphorylation (Fig. 5c); however, the phosphorylation of MK2 in response to NE or cAMP was unchanged (Fig. 5d). These data suggest that activation of p38 results from an NE-mediated autophosphorylation mechanism and leads to signal-specific physiological responses in the cell.

In T cells, one physiological effect of p38 MAPK activity is the regulation of cell growth and cell death, especially important in the thymus during T-cell development.^16,26,27 Dysregulation of p38 can result in negative selection-induced cell death and the subsequent absence of T-cell populations in the peripheral immune system.^26,27 Activation of p38 MAPK in T cells has been associated with the up-regulation of FasL, which can trigger apoptosis through interacting with the Fas receptor.^14,49,66 In S49 cells, βAR/cAMP stimulation is associated with cell death, and with the up-regulation of apoptotic factors, including Bim, Fas and FasL mRNAs (D. A. Chambers and D. H. Davis, unpublished data).¹⁴ Our data reveal that p38 MAPK is involved in the cAMP-mediated regulation of FasL mRNA (Fig. 5b) and cell death (Fig. 5a). As p38 inhibition does not completely antagonize activation of the cell death programme, it is likely that cAMP activates additional apoptotic mechanisms. cAMP-mediated apoptosis of S49 cells has been reported to occur through a mitochondrial mechanism initiated by multiple pathways.¹⁴ One such mechanism involves the cAMP-mediated up-regulation of Bim mRNA, which, unlike FasL, is not affected by p38 MAPK inhibition (Fig. 5b). A Gs, non-PKA-dependent mechanism has also been proposed for βAR-mediated apoptosis of S49 cells.¹² As βAR stimulation of S49 cells triggers apoptosis, it is likely that a variety of mechanisms are activated to ensure cellular elimination during unfavourable conditions, including psychogenic stress; our results suggest that in T-lineage cells such a mechanism may include p38 MAPK.

In these studies we identified a relationship between NE and p38 MAPK in thymocytes through which NE stimulation of the βAR/Gs/PKA signalling pathway activates p38 MAPK by a mechanism involving autophosphorylation. This autophosphorylation mechanism may specify the targeting of p38 MAPK to substrates such that ATF-2, but not MK2, is phosphorylated. One physiological consequence of p38 MAPK activity is cell death, which can result from the involvement of p38 MAPK in the up-regulation of FasL mRNA. These results are consistent with a proposed mechanism (Fig. 6) in which p38 MAPK contributes to the inhibitory effects of stress on immune function, and more specifically in the thymus, where chronic stress leads to reduced thymus mass, reduced thymocyte numbers (affecting mainly DP thymocytes) and an increased number of apoptotic cells.^3,67,68 Although there are a multitude of stress agents and cellular events that can contribute to these physiological responses, the NE/p38 MAPK relationship we have identified, coupled with previous findings relating NE to selective gene regulation of apoptotic factors and cellular signalling molecules by transcriptional and post-transcriptional events, could be participating factors leading to altered T-cell populations and consequent immune dysfunction.¹¹ Determining the specifics of how cellular decisions are accomplished within this broad array of choices represents the next challenge for understanding the signal transduction of psychogenic stress in immune cells.

Disclosures

The authors have no financial conflict of interest.