Interleukin 1A (IL-1A) Induced Activation of P38 Mitogen-Activated

Home , Interleukin

Molecular Psychiatry (2004) 9,65–75 & 2004 Nature Publishing Group All rights reserved 1359-4184/04 $25.00 www.nature.com/mp ORIGINAL RESEARCH ARTICLE Interleukin 1a (IL-1a) induced activation of p38 mitogen-activated protein kinase inhibits glucocorticoid receptor function X Wang, H Wu and AH Miller Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, Atlanta, GA, USA

Previous studies have demonstrated that interleukina (IL-1a) inhibits glucocorticoid receptor (GR) nuclear translocation and dexamethasone (Dex)-induced gene transcription. Given that IL-1a is a potent activator of the p38 mitogen-activated protein kinase (MAPK) signal transduction pathway and p38 MAPK has been associated with reduced GR function, we examined the role of p38 MAPK in IL-1a-mediated inhibition of GR function in mouse fibroblast cells stably transfected with a GR-mediated reporter gene construct (LMCAT cells). Treatment of LMCAT cells with IL-1a (1000 U/ml) for 24 h inhibited Dex (50 nM)-induced GRE-CAT activity by B35%. When cells were cotreated for 24 h with IL-1a plus SB-203580 (0.5–1 lM), a selective p38 inhibitor, IL-1a’s inhibitory effect on GR function as determined by Dex-induced GRE-CAT activity was reversed. Using gel mobility shift assay, SB-203580 was also found to reverse IL-1a inhibition of GR-GRE binding. Further confirming the role of p38 pathways, pretreatment of LMCAT cells with antisense oligonucleotides targeted to p38 MAPK completely abrogated IL-1a inhibition of Dex-induced GRE-CAT activity. Taken together, these results demonstrate that activation of p38 MAPK pathways are involved in IL-1a-mediated inhibition of GR function. In addition, these findings extend the intracellular targets of p38 to include the GR and indicate that p38 inhibitors may have special utility in immunologic and/or neuropsychiatric disorders associated with impaired GR-mediated feedback inhibition. Molecular Psychiatry (2004) 9, 65–75. doi:10.1038/sj.mp.4001339 Keywords: cytokines; HPA axis; glucocorticoid receptor; depression; p38 MAPK

Substantial data provide evidence that glucocorticoid may include hypersecretion of CRH as well as over- receptor (GR) function is reduced in patients with activation of the immune response, both of which may major depression.1,2 These data are derived primarily contribute to the behavioral and physiological altera- from in vivo and in vitro challenge tests that have tions of the disorder.2,3 Although the mechanism(s) of examined the ability of glucocorticoids (presumably GR alterations in major depression are unknown, through their receptors) to feedback negatively on consideration has been given to the possibility that relevant neuroendocrine and immune responses that proinflammatory cytokines may be involved.2,3 are under glucocorticoid control. In vivo challenge Emerging evidence has shown that major depres- tests using the synthetic glucocorticoid, dexametha- sion is associated with inflammation/immune activa- sone (Dex), have reliably demonstrated nonsuppres- tion with associated increases of proinflammatory sion of cortisol in depressed patients after oral cytokines (including IL-1 and IL-6) in blood and administration of Dex in the context of both the Dex cerebrospinal fluid.4 Moreover, elevations in blood suppression test and the Dex/corticotrophin releasing IL-1 have been correlated with post-Dex cortisol hormone (CRH) test. In vitro studies have also concentrations, suggesting a direct relation between demonstrated reduced responsiveness to glucocorti- proinflammatory cytokine concentrations and im- coids in peripheral blood mononuclear cells from paired glucocorticoid-mediated feedback inhibi- depressed patients, obviating concerns that in vivo tion.3,5 Proinflammatory cytokines are also potent tests are solely related to alterations in Dex bioavail- inducers of a syndrome of sickness behavior that has ability and/or increased hypothalamic drive. Conse- many features in common with major depression, quences of altered GR function in major depression including anhedonia, anorexia, fatigue, sleep distur- bance, and impaired cognition, and therefore may contribute to the behavioral features of the disorder.3,6 Correspondence: Dr AH Miller, MD, Department of Psychiatry Nevertheless, of most relevance to the GR are studies and Behavioral Sciences, Emory University School of Medicine, demonstrating that proinflammatory cytokines can 1639 Pierce Dr, Suite 4000, Atlanta, GA 30322, USA. directly inhibit GR function.3 Exposure of cells to E-mail: [email protected] Received 6 January 2003; revised 19 February 2003; accepted 25 proinflammatory cytokines has been shown to de- February 2003 crease cell sensitivity to the effects of glucocorticoids p38 MAPK inhibition of GR function X Wang et al 66 on various functional end points7–9 as well as Sciences Inc. (Boston, MA, USA), RNeasy Mini kit from decrease GR affinity for ligand.7,10–12 Additionally, Qiagen Inc. (Valencia, CA, USA), and BCA Protein studies performed on peripheral cells and tissues of Assay Reagent from Pierce Endogen (Rockford, IL, USA). patients with inflammatory diseases, such as asthma, ulcerative colitis, acquired immunodeficiency syn- Cell culture drome, and rheumatoid arthritis, especially those Mouse fibroblast LMCAT cells stably transfected with exhibiting resistance to the therapeutic effects of a construct containing MMTV and CAT reporter gene glucocorticoids, have demonstrated reductions in were generously provided by Dr ER Sanchez (Medical GR function and affinity.10,13–15 Of note, these inflam- College of Ohio, Toledo, OH, USA). Cells were grown matory disorders also have been found to exhibit a in 175-cm2 flasks in DMEM medium (Invitrogen high rate of comorbid depression.16–20 Corporation, Carlsbad, CA, USA) with 10% stripped In exploring the mechanism of the effects of newborn calf serum (NCS) (charcoal/dextran-ex-

cytokines on GR, work in our laboratory has shown tracted, heat-inactivated) at 371C with 5% CO2 in that the proinflammatory cytokine, interleukin 1a (IL- ambient air. The stably transfected GRE-CAT reporter 1a), inhibits the translocation of the GR from cytosol construct was maintained in the presence of 0.2 mg/ to nucleus and reduces Dex-induced, GR-mediated ml G418 sulfate (geneticin) antibiotic. For drug gene transcription.21 In addition, IL-1a administration treatment experiments, cells were subcultured in was associated with evidence of increased cellular six-well plates and grown for 24 h to 95% confluence. concentrations of GR protein, as manifested by Appropriate vehicle controls were used in all the increased GR binding and increased GR immuno- experiments, and each experimental condition in- staining (especially in the cytoplasm). The signal cluded at least three replicates. transduction pathways that are involved in these effects of IL-1 have yet to be elucidated. One CAT assay possibility includes IL-1 activation of mitogen-acti- CAT concentrations induced by Dex were measured vated protein kinases (MAPK).22,23 Through binding using the commercial CAT ELISA kit from Roche to the IL-1/toll-like receptor (TLR), IL-1 is well known Diagnostics (Indianapolis, IN, USA). Before assay, the to activate MAPK signaling pathways including the cells were washed three times with cold 1 Â PBS and p38 MAPK pathway, as evidenced by increased then lysed with the provided buffer after incubation phosphorylation of ATF224 and heat shock protein at room temperature for 30 min. Cell extracts were 27.25 Of note, in a recent report, activation of p38 then spun at 23 000 Â g at 41C for 10 min to remove MAPK pathways was found to result in GR phosphor- cellular debris. The supernatant was collected and ylation that in turn was associated with reduced stored at À801C until assayed. CAT assay was sensitivity to glucocorticoids.26 performed following manufacturer’s instructions. To further investigate the molecular mechanism(s) Briefly, antibodies to CAT were precoated on the by which IL-1a inhibits GR function, we set out to surface of the microtiter plate modules. Cell extracts examine the role of IL-1a-induced activation of p38 were then added to the wells to allow CAT in the cell MAPK pathways. Experiments were conducted in a extracts to bind specifically to the precoated anti-CAT mouse fibroblast cell line (LMCAT cells) stably antibodies. A digoxigenin conjugated to peroxidase transfected with the mouse mammary tumor virus- was added to bind to digoxigenin on anti-CAT chloramphenicol acetyltransferase (MMTV-CAT) re- antibodies. After incubation for 1 h, CAT concentra- porter gene. Cells were treated with IL-1a for 24 h in tions were determined by colorimetry (absorption at the presence or absence of the synthetic steroid, Dex 405 nm against the reference wavelength of 492 nm) and/or SB-203580, a selective p38 inhibitor. Anti- using a VERSAmax tunable microtiter plate reader sense oligonucleotides targeted to p38 MAPK were from Molecular Devices (Sunnyvale, CA, USA). also used to examine the involvement of p38 pathways. GR function was measured by means of GR- Intracellular [3H]Dex concentration determination mediated gene transcription in LMCAT cells. In LMCAT cells were grown in 12-well culture dishes addition, GR-GRE binding from different treatments until 95% confluent. Cells were then treated with SB- was determined using gel mobility shift assay, and GR 203580 at various concentrations for 24 h. After mRNA levels were quantitated using real-time PCR. incubation with [3H]Dex (5 mCi/well) for 1.5 h, cells were harvested by washing with cold PBS twice and Materials and methods trypsinization. Cells were then counted using a LKB Chemicals Liquid Scintillation Counter. Sources of reagents for indicated experiments were as follows: SB-203580 from BIOMOL Research Labora- Nuclear extraction tories Inc. (Plymouth Meeting, PA, USA), Dex and Cells were cultured in 150-mm dishes until confluent. corticosterone (Cort) from Sigma-Aldrich CO. (St Louis, Cells were treated/cotreated with various concentra- MO, USA), G418 (Geneticin), Trizol reagent and tions of Dex and other agents for different time Lipofectin from Invitrogen corporation (Carlsbad, CA, intervals. Cells were then harvested by trypsinization USA), IL-1a from R&D Systems Inc. (Minneapolis, MN, and washed with 10 ml of ice-cold PBS for 5 min at USA), [3H]Dex (37 MBq) from Perkin-Elmer Life 1200 rpm (Beckman J6 at 41C). The cell pellets were

Molecular Psychiatry p38 MAPK inhibition of GR function X Wang et al 67 resuspended in 1 ml of PBS, transferred to a 1.5 ml was added, and samples were vortexed vigorously for Eppendorf tube and centrifuged for 30 s at 10 000 Â g 1 min. RNA samples were precipitated and washed (41C). The cell pellets were resuspended in 50 mlof following the manufacturer’s instructions. After brief cold, low-salt buffer (containing 20 mM HEPES, pH air-drying at RT, RNA samples were dissolved in

7.9, 10 mM KCl, 0.1 mM Na3VO4, 1 mM EDTA, 1 mM RNase-free water, and their concentrations and A260/ EGTA, 0.2% Nonidet P-40, 10% glycerol, 0.5 mM 280 ratio were determined using the MBA 2000 PMSF, and 1 mM DTT) and mixed by gentle pipetting. System (Perkin-Elmer, Shelton, CT, USA). After 15 min incubation on ice, cell pellets were centrifuged at 13 000 g for 2 min (41C), and the Real-time PCR analysis supernatants were discarded. The nuclei pellets were Changes in GR gene transcription were analyzed suspended in 50 ml ice-cold high salt buffer (contain- using real-time PCR technique, GeneAmp 7000 ing 20 mM HEPES, pH 7.9, 420 mM NaCl, 10 mM KCl, Sequence Detection System from Applied Biosystems

0.1 mM Na3VO4, 1 mM EDTA, 1 mM EGTA, 20% (Norwalk, CT, USA). Briefly, 2 mg total RNA isolated glycerol, 0.5 mM PMSF, and 1 mM DTT) and vigor- from different treatments of cultured cells was reverse ously rocked at 41C for 30 min. The nuclear extracts transcribed using SuperScriptt First-strand Synthesis were collected after centrifugation at 13 000 g for System for RT-PCR from Invitrogen (Carlsband, CA, 30 min (41C) and stored at À801C for later analysis. USA) using random primers following the manufacturer’s instructions. Primers for mouse GR and 18S Protein concentration determination rRNA gene were designed using Primer Express

Protein concentrations were determined using a software with Tm at 591C and G/C content in the 20– commercial bicinchoninic acid (BCA) kit from Pierce 80% range. The primers used for the amplification of (Rockford, IL, USA) following the manufacturer’s the mouse GR gene were 50-ACC TGG ATG ACC AAA instructions. Briefly, protein samples were alkalized TGA CC (forward) and 50-TTG GCT CTT CAG ACC with the supplied reagent and incubated with the TTC CT (reverse), corresponding to nucleotides 1799– reaction reagent containing Cu2 þ to produce purple 1818 and 2066–2085 of the mouse GR cDNA. PCR color for colorimetric detection against BSA protein reaction was carried out using SYBR Green PCR standard. Master Mix from Applied Biosystems after optimiza- tion of primer concentrations, template concentra- Gel shift assay tions and PCR cycles. In brief, 20 ng RT product was Synthetic oligonucleotides containing GRE (50-AAG mixed with 5 ml SYBR Green PCR Master Mix and ATT CAG GTC ATG ACC TGA GGA GA) sequences specific primers (300 nM) in a total volume of 10 ml. were obtained from Invitrogen (Frederick, MD, USA). The reactions in 96-well plate were held at 501C for The oligonucleotides were annealed and then labeled 2 min and 951C for 10 min, followed by thermal at the 50end using T4-polynucleotide kinase and cycling for 40 cycles at 951C for 15 s and 601C for [g-32P]ATP according to the manufacturer’s instruc- 1 min. The quantitation of the GR gene was analyzed tions. A measure of 5–10 mg of cell nuclear extract using the included software against the standard and from different treatments were incubated with 1 mg normalized against 18S ribosomal RNA. poly d(I-C) for 15 min at room temperature (RT) to bind nonspecific DNA-binding proteins. A volume of p38 antisense oligonucleotide treatment 1 ml[32P]-labeled DNA (final concentration, 1 nM) was Phosphorothiolate-modified antisense oligonucleo- added and incubated for an additional 15 min at RT. tides specific for p38-MAPK (50-GTC TTG TTC AGC The 100-fold excess unlabeled DNA was added 5 min TCC TGC-30) as well as sense (50-GCA GGA GCT GAA before additional [32P]-labeled DNA to compete with CAA GAC-30) oligonucleotides 27 were synthesized. the specific DNA–protein binding. Reaction mixtures Cells were incubated with 1 mM of each oligonucleo- were loaded onto a 5% nondenaturing polyacryla- tide in serum-free RPMI 1640 for 6 h before addition mide gel (acrylamide : bisacrylamide, 30 : 0.8) and run of 10% charcoal stripped/heat-inactivated NCS. After at constant 150 V in 1 Â TBE buffer (0.09 M Tris, 24 h of culture, cells were treated with either vehicle 0.09 M borate, and 2 mM EDTA, pH 8.3) until the control, or Dex in the presence or absence of IL-1a, bromophenol blue reached the bottom. Gels were and cells were prepared for CAT assay as described dried and protein–DNA binding was visualized by above. autoradiography or analyzed using PhosphorImager. Western blot analysis of p38 protein RNA isolation Antibodies against p38-MAPK or phospho-p38- Total RNA was isolated from cells grown in 60-mm MAPK were purchased from Cell Signaling Technol- culture dishes after different treatments for 24 h as ogy, Inc. (Beverly, MA, USA). After different described. Cells were washed with 1 Â ice-cold PBS treatments, cells were directly lysed in the culture once, and 1 ml of Trizol reagent (Invitrogen Corpora- dishes using 500 ml of lysing buffer (containing 20 mM tion, Carlsbad, CA, USA) was directly added into the Tris, pH 7.5; 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, dish to lyse cells. Cell lysates were incubated for 1% Triton, 2.5 mM sodium pyrophosphate, 1 mM

5 min at RT to permit complete dissociation of b-glycerophosphate, 1 mM Na3VO4,1mg/ml leupep- nucleoprotein complexes. Then 0.2 ml of chloroform tin, 1 mM PMSF). Cell lysates were sonicated and

Molecular Psychiatry p38 MAPK inhibition of GR function X Wang et al 68 1 centrifuged at 15 000 rpm for 10 min at 4 C. Protein a IL-1 (min) (50 mg) from each group was mixed with SDS sample C 5 15 30 60 buffer and separated by a 12% SDS-PAGE, transferred to nitrocellulose membrane, and probed with the first p-ATF-2 antibody (1/1000 dilution) at room temperature for IL-1 (min) 3 h. Bound primary Ab was detected by horseradish b IL-1-10 min+ peroxidase-conjugated secondary antibody (1 : 2000 C105 SB-1µM SB-1µM dilution), protein bands were visualized using a commercially available chemoluminescence kit from p-ATF-2 Amersham Biosciences Corp. (Piscataway, NJ, USA). c Pos C C IL-1 (15 min) IL-1+ IL-1+ SB-10µM p38-MAPK activity measurement SB-1µM SB-10µM p38 MAP Kinase Assay kit was purchased from Cell p-MAPKAPK-2

Signaling Technology, Inc. Cells were directly lysed MAPKAPK-2 in the culture dishes. Cell lysates were sonicated and centrifuged at 15 000 rpm for 10 min at 41C as above. Figure 1 Effect of IL-1a and SB-203580 on p38 MAPK The supernatant containing equivalent amounts of activity. LMCAT cells were grown in 60-mm culture dishes protein (200 mg) was incubated by gentle rocking with until 80% confluency. Cells were then treated with IL-1a 20 ml of immobilized phospho-p38-MAPK monoclo- (1000 U/ml) for various lengths of time as indicated. Cells nal antibody for 16 h at 41C. The immunoprecipitates were harvested using lysing buffer and 50 mg protein from were washed twice with the lysing buffer and pelleted each group was loaded and separated on a 12% SDS-PAGE by centrifugation. The p38-MAPK assay was carried and transferred to a nitrocellulose membrane. p38 MAP out using ATF-2 fusion protein (2 mg) as a substrate in kinase assay kit from Cell Signaling using ATF-2 as the substrate was used to determine p38 activity as described in the presence of 200 mM ATP and 1 Â kinase buffer the Materials and methods. (a) shows time-dependent following the manufacturer’s recommendations. Sam- induction of p38 activity by IL-1a treatment. (b) shows the ples were resolved on a 12% SDS-PAGE gel and effect of 1 h pretreatment of SB-203580 (1 mM), a selective visualized by autoradiography. p38 inhibitor, on IL-1a (1000 U/ml) induction of p38 activity measured using ATF-2 as the substrate. (c) shows the effect Data analysis and statistics of 1 h pretreatment of SB-203580 (1 and 10 mM) on IL-1a Descriptive statistics (including the mean and stan- (1000 U/ml) induction of p38 activity using antibody to the phospho-MAPKAPK-2, an endogenous p38 substrate. The dard deviation) were used to characterize the depen- lower panel in (c) depicts a loading control using an dent measures in all experiments. One- or two-way antibody to the nonphosphorylated MAPKAPK-2. Results analysis of variance (ANOVA) was used to assess for shown are from a representative experiment of three main effects and interactions of treatment conditions separate experiments. CFvehicle control, Pos CFpositive and/or dose and/or time. For all ANOVAs, both the phospho-MAPKAPK-2 control. Student–Newman–Keuls method and the Student’s t-test were used for post hoc tests of differences only endogenous MAPKAPK-2 phosphorylation was between specific means in order to include both a inhibited by SB-203580 using in vitro assay.28 conservative (Student–Newman–Keuls) and a power- ful (Student’s t-test) assessment of statistical significance. The level of significance was set at Po0.05, Selective p38 inhibitor, SB-203580, reverses IL-1a and all tests of significance were two-tailed. inhibition of GRE-CAT activity and GR-GRE binding As shown in Figure 2, there was a significant main Results effect of treatment condition on GRE-CAT activity (F[6, 14] ¼ 120.8, Po0.0001). Treatment with IL-1a IL-1a induces p38 activity as measured by (1000 U/ml) for 24 h led to significant inhibition of phosphorylation of ATF2 and MAPKAPK-2 kinase Dex-induced GRE-CAT activity by B35%. When cells As shown in Figure 1, IL-1a significantly induced p38 were cotreated with IL-1a plus SB-203580 (1 mM) for activity in mouse fibroblasts as evidenced by in- 24 h, IL-1a-mediated inhibition of GRE-CAT activity creased expression of phosphorylated ATF2 (Panels A was significantly attenuated. SB-203580 alone (1 mM) and B) as well as increased expression of phosphory- had no effects on Dex-induced GRE-CAT activity. To lated MAPKAPK-2 kinase (Panel C), both of which are examine the effects of IL-1a and SB-203580 on GR substrates of activated p38 MAPK. Interestingly, DNA binding, nuclear extracts were obtained from however, when cells were pretreated with SB- LMCAT cells after treatment with Dex alone or in 203580 (1 mM), a selective p38 inhibitor, for 1 h, combination with IL-1a and/or SB-203580. Results IL-1a-induced p38 activity as reflected by phosphory- from gel mobility shift assay (see Figure 3) using lated ATF2 was unchanged (Panel B), whereas IL-1a- synthetic [32P]-labeled GRE oligos showed that Dex induced MAPKAPK-2 phosphorylation was signifi- (50 nM, 1.5 h) induced marked GR-GRE binding (lane cantly blocked by SB-203580 at both 1 and 10 mM. 2) that was inhibited by IL-1a (lane 3). SB-203580 Similar results have been reported by others in which reversed IL-1a-mediated inhibition of Dex-induced

Molecular Psychiatry p38 MAPK inhibition of GR function X Wang et al 69 4 GR-GRE binding (lane 4). These results were repli- cated in at least three independent experiments. a 3 Effect of IL-1a and SB-203580 on GR mRNA expression To determine whether IL-1a inhibition of GR- * mediated gene transcription and SB-203580 attenua- 2 tion of IL-1a inhibitory effects were associated with changes in GR mRNA levels, total RNA was isolated from LMCAT cells treated with IL-1a in the presence Fold Induction or absence of Dex and/or SB-203580. Real-time PCR 1 was used with specific primers to mouse GR cDNA, and 18S rRNA was used as the internal loading standard. As depicted in Figure 4, there was a 0 α α α significant main effect of treatment condition SB Dex IL-1 (F[7, 16] ¼ 63.9, Po0.0001) on GR mRNA levels. Control SB + Dex Cells treated with IL-1a for 24 h exhibited a slight Dex + IL-1 but significant increase in GR mRNA. Of note, the SB + Dex + IL-1 addition of SB-203580 (1 mM) had no effect on IL-1a- induced changes. SB-203580 alone (1 mM) signifi- Figure 2 Effect of IL-1a and SB-203580 on Dex-induced cantly increased GR mRNA. Dex plus IL-1a reversed MMTV-CAT activity. LMCAT cells were seeded in six-well the effects of IL-1a alone. plates at a density of 900 Â 103 cells/well overnight. Cells were treated with IL-1a (1000 U/ml) or cotreated with SB- 203580 (1 mM), a selective p38 inhibitor for 24 h. Cells were Effect of SB-203580 on GR function and the then harvested for CAT assay after an additional 1.5-h p-glycoprotein pump treatment with Dex (50 nM). *IL-1a treatment caused 35% Based on gel shift assay, SB-203580 plus IL-1a and (Po0.001) inhibition of Dex-induced GRE-CAT activity Dex exhibited evidence of some enhancement of GR compared to Dex alone. aFCotreatment with 1 mM SB- 203580 and IL-1a significantly attenuated IL-1a inhibition of ** ** Dex-induced GRE-CAT activity (Po0.01). SB-203580 alone 1.5 had no effects on Dex-induced GRE-CAT activity (P40.05). **

1.0 1 2 3 4 5 6 7 8

GR-GRE

Fold Induction 0.5

0.0

Dex IL-1 SB Control Dex+SB Dex+IL-1 SB+IL-1 Dex+IL-1+SB Figure 4 Effect of IL-1a and SB-203580 on GR mRNA levels. Total RNA was isolated from LMCAT cells after 24 h treatment with IL-1a (1000 U/ml) with or without cotreat- Figure 3 Effect of IL-1a and SB-203580 on GR-DNA ment with SB-203580 (1 mM) or 1.5 h treatment with Dex binding. Nuclear extracts were obtained from LMCAT cells (50 nM). RNA was reverse-transcribed and 20 ng RT pro- after treatment with DMSO (lane 1), Dex (50 nM, 1.5 h, lane ducts were used from each treatment. PCR reaction was 2), Dex þ IL-1a (lane 3), Dex þ IL-1a þ SB (lane 4), IL-1a carried out using SYBR Green PCR Master Mix. The (lane 5), SB alone (lane 6), Dex þ SB (lane 7) and IL-1a þ SB quantitation of the GR gene was analyzed using the (lane 8). SB-203580 concentration in all treatments was included software against the standard and normalized 1 mM. Cells were pretreated for 30 min with IL-1a and/or SB- against 18S ribosomal RNA. IL-1a caused 25% (Po0.01) 203580 before adding Dex. Gel mobility shift assay was increase of GR mRNA compared with the control, and performed, and the results are shown. Dex induced cotreatment with SB-203580 had no additional effects on IL- significant GR-GRE binding, IL-1a inhibited Dex-induced 1a’s induction. SB-203580 alone also caused significant GR-GRE binding, whereas the p38 inhibitor, SB-203580 increase of GR mRNA levels by 60% (Po0.001) and the reversed IL-1a-mediated inhibition. Data are from a repre- combination of SB-203580 and IL-1a had no additional sentative experiment of three separate experiments. effects.

Molecular Psychiatry p38 MAPK inhibition of GR function X Wang et al 70 15.0 10000 * ** 12.5 7500

10.0 **

5000 7.5

Fold Induction 5.0

H]Dex DPM/mg protein 2500 3 [

2.5

0 0.0 µM µM µM µM m m m Dex µm µ µ µ Dex Control Control Dex+SB1 Dex+SB0.1 Dex+SB0.5 Dex+SB10 Dex + SB-1 Dex + SB-10 Dex + SB-0.1Dex + SB-0.5 Figure 6 Dose-dependent effect of SB-203580 on intracel- 3 Figure 5 Dose-dependent effect of SB-203580 on Dex- lular [ H]Dex concentrations. LMCAT cells were grown in induced MMTV-CAT activity. LMCAT cells were seeded in 12-well culture dishes until 80% confluent. Cell were then six-well plates at the density of 900 Â 103 cells/well over- treated with SB-203580 at various concentrations for 24 h. night. Cells were then treated with SB-203580 at various Cells were harvested by washing with cold PBS twice and 3 m concentrations for 24 h. Cells were harvested for CAT trypsinization after incubation with [ H]Dex (5 Ci/well) for assay after an additional 1.5 h treatment with Dex (50 nM). 1.5 h. Cells were then counted using a LKB Liquid m *SB-203580 at 10 mM caused significant enhancement of Scintillation Counter. SB-203580 at 0.5 and 1 M concentrations caused a slight increase of intracellular [3H]Dex Dex-induced MMTV-CAT activity (Po0.0001), whereas SB-203580 at other concentrations had no effect on Dex- concentrations by 14–16% (Po0.05). When used at high m induced MMTV-CAT activity (P40.05). concentration (10 M), SB-203580 significantly increased intracellular [3H]Dex concentration by 60% (Po0.001).

DNA binding over Dex alone. Moreover, as noted centration accounted for SB-203580-enhancement of above, SB-203580 alone significantly increased GR GR activity, we examined the effects of SB-203580 on mRNA. These data suggested some independent GRE-CAT activity in LMCAT cells treated with Cort, effect of SB-203580 on the GR. To further investigate which is not a substrate for pump.30 Although less independent effects of SB-203580 on GR activity, pronounced than the effects on Dex, SB-203580 also LMCAT cells were treated with various concentra- significantly enhanced Cort-induced GRE-CAT activ- tions of SB-203580 (0.1, 0.5, 1, and 10 mM) for 24 h, ity by 16 and 32%, respectively, at concentrations of 5

and Dex (50 nM) was added 1.5 h before harvest for and 10 mM, but not at 1 mM (Figure 7) (F[4, 10] ¼ 68.7, CAT assay. There was a significant main effect of Po0.0001). treatment condition (F[5, 12] ¼ 636.3, Po0.0001). Inter- estingly, SB-203580 significantly enhanced Dex-in- p38 antisense oligonucleotide treatment reduces p38 duced CAT activity at 10 mM but not at lower protein levels and reverses IL-1a inhibition of GR concentrations (Figure 5). function and IL-1a-induced p38 phosphorylation SB-203580 at higher dose (30 mM) has been reported Given the direct effects of SB-203580 on the GR, the to increase significantly intracellular accumulation of role of p38 pathways in IL-1a-mediated GR inhibition [3H]-vincristine in multidrug resistant L1210/VCR was further explored using antisense oligonucleotides mouse leukemic cell line by inhibition of the p- targeting p38 MAPK. Cells were pretransfected with glycoprotein pump,29 the effects of SB-203580 on 1 mM antisense oligonucleotide for 24 h, and p38 and intracellular dexamethasone (also a substrate for the phospho-p38 were measured by Western blotting. As p-glycoprotein pump) in LMCAT cells in the presence seen in Figure 8a and b, both p38 and phospho-p38 of various concentrations of SB-203580 was exam- levels were significantly reduced in the antisense ined. As shown in Figure 6, there was a significant treatment groups vs the sense or nonoligo control main effect of treatment condition on intracellular groups. When cells were pretransfected with 1 mM 3 [ H]Dex concentrations (F[5, 12] ¼ 277.3, Po0.0001). antisense oligonucleotide for 24 h, IL-1a-mediated GR Treatment of LMCAT cells with SB-203580 for 24 h inhibition was completely reversed (Po0.01). The increased intracellular [3H]Dex concentrations by 14, p38 sense oligonucleotide had no significant effects 16, and 60%, at the dosages of 0.5, 1, and 10 mM, on IL-1a-mediated GR inhibition (P40.05) (see Figure respectively. SB-203580 at 0.1 mM had no effect on 9). Of note, p38 antisense oligonucleotide treatment intracellular [3H]Dex concentrations. To determine alone noticeably enhanced basal GRE-CAT activity whether this increase in intracellular [3H]Dex con- (Po0.05).

Molecular Psychiatry p38 MAPK inhibition of GR function X Wang et al 71 4 a ** Antisense Sense * 3

Con IL-1-5 min IL-1-10 minCon IL-1-5 min IL-1-10 minCon IL-1-5 minIL-1-10 min

2 p38 b Fold Induction

1 Antisense Sense

in in

0 Con IL-1-5 min IL-1-10 m Con IL-1-5 min IL-1-10 m M µ µM µM Cort p-p38 Control

Cort + SB 1 Cort + SB 5 Figure 8 The effect of p38 antisense oligonucleotide on Cort + SB 10 p38 and phospho-p38. LMCAT cells were grown in 60-mm Figure 7 Dose-dependent effect of SB-203580 on Cort- culture dishes until 80% confluent. Cells were then induced MMTV-CAT activity. LMCAT cells were seeded in tranfected with p38 antisense or sense oligonucleotide six-well plates at the density of 900 Â 103 cells/well over- (1 mM) using lipofectin from Amersham for 24 h. Cells were night. Cells were then treated with SB-203580 at different treated with vehicle control or with IL-1a (1000 U/ml) for concentrations (1–10 mM) for 24 h. Cells were harvested for various times as indicated. Cells were harvested using CAT assay after an additional 1.5 h treatment with Cort lysing buffer, and 50 mg protein from each group was loaded (50 nM). *, **SB-203580 at 5 mM concentration caused 16% and separated on a 12% SDS-PAGE and transferred to a (Po0.05) and at 10 mM caused 32% (Po0.01) further nitrocellulose membrane. p38 (a) and phosphorylated p38 enhancement of Cort-induced MMTV-CAT activity. SB- (b) levels were determined using monoclonal antibodies 203580 at 1 mM had no effect on Cort-induced MMTV-CAT from Cell Signaling as described in the Materials and activity (P40.05). methods. (a) Antisense transfection reduced p38 levels in both vehicle and IL-1a treatment groups vs sense or nonoligonucleotide transfection control groups. (b) Anti- sense transfection reduced phosphorylated p38 levels in To determine the role of p38 in the effect of higher both vehicle and IL-1a treatment groups vs the sense group. doses of SB-203580 (X10 mM) on GR function, A representative experiment of three replicates is shown. LMCAT cells were treated with SB-203580 in the presence and absence of p38 antisense oligonucleotides. As shown in Figure 10, antisense oligonucleo- proliferation34 and apoptosis caused by osmotic shock tides had no effect on SB-203580 (10 mM) or ultraviolet irradiation.35 The signaling pathway by enhancement of Dex-induced GRE-CAT activity, which IL-1 activates p38 pathways is not yet fully further confirming an effect of SB-203580 on GR understood, although it has been proposed that the function independent of its activity on p38. early events of signal transduction are similar to IL-1 activation of NF-kB and JNK. Briefly, IL-1 binds to IL- Discussion 1 type I receptor on the cell membrane and forms a complex with the IL-1 receptor accessory protein (IL- Consistent with our previous studies, the results of 1RAcP), triggering the association of an adaptor these experiments demonstrate that IL-1a signifi- protein myeloid differentiation factor 88 (MyD88). cantly inhibits GR-mediated gene transcription and MyD88 then interacts with IL-1 receptor-associated GR-GRE binding. Relevant to the mechanism of these kinase (IRAK), subsequently causing IRAK phosphor- effects, IL-1a inhibitory effects on GR function were ylation. Phosphorylated IRAK forms a complex with significantly attenuated by inhibition of p38 activity tumor necrosis factor receptor-associated factor 6 using SB-203580, a selective p38 MAPK inhibitor, (TRAF-6), resulting in the phosphorylation of trans- and were completely reversed by antisense oligonu- forming growth factor-b-activated kinase-1 (TAK1).36 cleotides targeted to p38, indicating that p38 MAPK TAK1 is reported to be an upstream kinase of MKK6 pathways are involved in IL-1a-mediated GR func- and MKK3,37 both of which are upstream kinases of tional impairment. p38 MAPK. Proinflammatory cytokines such as IL-1 bind to Although the data suggest that p38 pathways are their membrane receptors and activate MAPK path- involved in IL-1a-induced GR impairment, the exact ways including p38, ERK, and JNK as well as the NF- molecular mechanism has yet to be established. A kB pathway.31 Activation of p38 MAPK has been recent study demonstrated that activation of p38 by reported to play a pivotal role in cytokine production combined treatment of human peripheral blood (IL-1, TNFa, and IL-2),32,33 cytokine-induced cell mononuclear cells with IL-2 and IL-4 was associated

Molecular Psychiatry p38 MAPK inhibition of GR function X Wang et al 72 8 * a 8 7 *

6 6 * 5 * 4 4

Fold Induction 3

2 Fold Induction 2 1

0 α 0 + AS + SS + AS + SS Control α α α α (1000 U) Dex + ASDex + SS α IL-1 IL-1 ControlControl + AS + SS Dex (50Dex nM) + IL-1 IL-1 Dex+SS Dex+AS Dex + DexIL-1 + IL-1 Control+SS Dex+SB+SS Dex+SB+AS Figure 9 The effect of p38 antisense oligonucleotide on IL- 1a-mediated inhibition of MMTV-CAT activity. LMCAT cells Figure 10 The combined effect of p38 antisense oligonu- were treated with 1 mM phosphorothiolate-modified anti- cleotide and SB-203580 on Dex-induced MMTV-CAT sense oligonucleotides specific for p38-MAPK or sense activity. LMCAT cells were treated with 1 mM phosphor- oligonucleotides in serum-free RPMI 1640 for 6 h before othiolate-modified antisense oligonucleotides specific for addition of 10% charcoal stripped/heat-inactivated NCS for p38-MAPK or sense oligonucleotides in serum-free RPMI additional 18 h. Cells were then treated with either vehicle 1640 for 6 h before addition of 5% charcoal stripped/heat- control, or Dex in the presence or absence of IL-1a as inactivated FCS for additional 18 h. Cells were then treated described. After treatment, cells were harvested for CAT with either vehicle control, or Dex (50 mM, 1.5 h) in the presence or absence of pretreatment with SB-203580 assay as described. *IL-1a treatment caused 35% (Po0.001) inhibition of Dex-induced GRE-CAT activity compared to (10 mM). After treatment, cells were harvested for CAT assay Dex alone. aFPretreatment of cells with antisense p38 oligo as described. *SB treatment in the presence of the p38 sense significantly reversed IL-1a inhibition of GR function or antisense oligonucleotide transfection caused significant further enhancement of Dex-induced GRE-CAT activity (Po0.001). Pretreatment with antisense or sense oligo had no effects on Dex-induced CAT activity (P40.05). Pre- (Po0.001), 260 and 290%, respectively. treatment with sense oligo had no significant additional effects on IL-1a inhibition of GR function (P40.05).

with phosphorylation of the GR and reduced GR targeting p38 pathways, while leaving NF-kB signal function.26 Although it could not be determined transduction ostensibly intact. Thus, in some cell whether GR phosphorylation was a direct effect of types (eg microglia), where IL-1 activates p38 phos- p38 MAPK, phosphorylation at hGR serine 226 was phorylation without effects on NF-kB pathways, IL-1 reversed using SB-203580. It should be noted that may predominately influence GR function through many p38 target proteins may be involved in activation of p38 signaling pathways.48 disrupted GR function including the immediate Interestingly, p38 MAPK has been reported to downstream kinase, MAPK-activated protein kinase interact with other members of the steroid receptor 2 and 3 (MAPKAPK-2, -3),38,39 CREB,40 hsp-27,25,38 superfamily including the estrogen receptor49 and ATF2,24 Elk1,41 CHOP,42 and Max.43 For example, retinoid acid receptor g50 as well as other nuclear overexpression of CREB has been found to inhibit GR receptors such as peroxisome proliferator-activated transactivation in placenta cells,44 and CREB activa- receptor a51 and transcription coactivator PGC-1.52 tion by treatment with high concentrations of b2- These results suggest that there may be shared adrenoceptor agonist inhibits GR-GRE binding in rat mechanisms among these effects on steroid hormone lung.45 Given the multiple steps in both GR and p38 receptors. signaling pathways, it is also possible that other GR In further evaluation of the effects of SB-203580 on co-factors/repressors or p38 target proteins such as p38 activity as well as the GR, it is apparent that this nuclear transcription factors may participate in IL-1a- compound does not block all p38 activities under all induced GR impairment. Of note, previous studies experimental conditions and has effects on the GR have implicated NF-kB in disruption of GR function, that are independent of its antagonism of p38. SB- primarily through protein–protein interactions.46,47 203580 treatment failed to inhibit phosphorylation of Nevertheless, data from this study suggest that p38 exogenous ATF-2 by immunoprecipitated p38. A pathways are just as relevant to GR regulation, in that similar finding was reported by Baldassare et al.,28 the effects of IL-1, which activates NF-kB in concert who hypothesized that ‘the intracellular interaction with p38, were completely reversed by specifically of the inhibitor with the catalytic domain of p38 may

Molecular Psychiatry p38 MAPK inhibition of GR function X Wang et al 73 render the kinase more accessible as a target substrate matory cytokines on the GR, thereby restoring in stimulated cells.’ In a related fashion, in a recent glucocorticoid-mediated negative feedback on inflam- review of p38 inhibitors, it was noted that SB-203580 matory responses. Moreover, through their activity on inhibits sorbitol-activated phosphorylation of Hsp27 NF-kB signaling pathways, glucocorticoids are some by p38-MAPKAPK-2 in HeLa cells, but had no of the most potent anti-inflammatory factors in the inhibitory effect on p38 kinase activity in the same body.66 samples using an in vitro kinase assay.53 Furthermore, Of note, p38 has become a target in the develop- although SB-203580 exhibits high selective inhibition ment of novel treatments for inflammatory disorders. of p38a (IC50 ¼ 48 nM) and p38b (IC50 ¼ 50 nM), it For example, one recent report showed that oral also inhibits other kinases including JNK2b1 (IC50 ¼ administration of a p38 MAPK inhibitor (BIRB 796 53 280 nM) and c-raf (IC50 ¼ 360 nM) as well as the p- BS) significantly inhibited LPS-induced cytokine glycoprotein mdr pump.29 Taken together, the data production including TNFa, IL-6, and IL-10 in healthy indicate that SB-203580 has broad biological effects volunteers.67 Moreover, several widely used nonster- beyond p38 inhibition. According to the data reported oidal anti-inflammatory drugs (aspirin, ibuprofen) herein, this broad activity extends to effects on the GR have been found to suppress T-cell activation through as manifested by (1) enhancement of Cort-induced the inhibition of p38 MAPK pathways.68 Finally, GR-mediated gene transcription (an effect that is mdr activation of the phosphatidyl serine receptor has pump independent because Cort is not a substrate for been shown to reduce IL-1 activation of p38 MAPK, the pump) and (2) enhancement of Dex-induced GR- suggesting another interesting therapeutic target for mediated gene transcription in cells treated with p38 abrogating IL-1 effects on GR function.48 Taken antisense oligonucleotides (thus eliminating any together, the results provide evidence that pharmaco- contribution of p38). Therefore, we cannot conclude logical agents that have p38 inhibitory properties may that the effects of SB-203580 in reversing IL-1- not only be useful for the treatment of inflammatory mediated inhibition of GR function occur solely immunologic disorders but also may have utility in through blockade of p38 activity, complicating the neuropsychiatric disorders associated with inflamma- interpretation of the experimental and/or clinical tion and impaired GR-mediated feedback inhibition, activity of this compound. including depression in the medically ill. It should be noted that all of the experiments conducted as part of this study utilized fibroblasts. Fibroblasts have been widely used as models to study Acknowledgements a number of biological activities, including the This work was supported in part by grants from the interaction of the GR with a variety of signaling National Institute of Mental Health (MH00680 and pathways.54–59 In addition, human fibroblasts have MH47674-08). Xiaohong Wang, MD, PhD is a recipi- been reported to contain receptors, second messen- ent of NIMH Minority Investigator Research Supple- gers, and protein kinase systems, which are similar to ment, Janssen Faculty Career Development Award those found in other tissues including the brain.60–62 and APA/Wyeth-Ayerst, MD/PhD Psychiatric Indeed, reduced PKA activity in fibroblasts from Research Fellowship. depressed patients63 has also been reported in brain tissue from depressed individuals.64 As previously discussed, activation of p38 activity by IL-2 plus IL-4 References leading to impaired GR function was first described in 1 Holsboer F. The corticosteroid receptor hypothesis of depression. peripheral blood mononuclear cells, further suggest- Neuropsychopharmacology 2000; 23: 477–501. ing that the pathway from p38 to disrupted GR 2 Pariante CM, Miller AH. Glucocorticoid receptors in major function generalizes to other cell types including depression: relevance to pathophysiology and treatment. Biol immune cells.26 Nevertheless, further studies must be Psychiatry 2001; 49: 391–404. 3 Miller AH, Pariante CM, Pearce BD. Effects of cytokines on conducted to determine the generalizability of the glucocorticoid receptor expression and function. Glucocorticoid findings to other cell types including neurons. resistance and relevance to depression. Adv Exp Med Biol 1999; Resistance to the effects of glucocorticoids not only 461: 107–116. exists in depression, but also appears in subgroups of 4 Maes M, Meltzer HY, Scharpe S, Bosmans E, Suy E, De Meester I et al. Relationships between lower plasma L-tryptophan levels and patients with acute or chronic inflammatory diseases immune-inflammatory variables in depression. Psychiatry Res such as sepsis, asthma, ulcerative colitis, acquired 1993; 49: 151–165. immunodeficiency syndrome, rheumatoid arthritis, 5 Weidenfeld J, Yirmiya R. Effects of bacterial endotoxin on the and allogenic organ transplantation who also exhibit glucocorticoid feedback regulation of adrenocortical response to high rates of depression.16–20,65 Although the molecu- stress. Neuroimmunomodulation 1996; 3: 352–357. 6 Kent S, Bluthe RM, Kelley KW, Dantzer R. Sickness behavior as a lar mechanism of glucocorticoid resistance is yet to be new target for drug development. Trends Pharmacol Sci 1992; 13: fully elucidated, its presence in disorders character- 24–28. ized by inflammation supports the potential role of 7 Kam J, Szefler S, Surs W, Sher E, Leung D. Combination IL-2 and inflammatory cytokines, including IL-1, and the IL-4 reduces glucocorticoid receptor-binding affinity and T cell response to glucocorticoids. J Immunol 1993; 151: 3460–3466. activation of p38 MAPK pathways. Data from this 8 Hill M, Stith R, McCallum R. Interleukin 1: a regulatory role in study provide evidence that inhibitors of p38 MAPK glucocorticoid-regulated hepatic metabolism. J Immunol 1986; signaling pathways may reverse the effects of inflam- 137: 858–862.

Molecular Psychiatry p38 MAPK inhibition of GR function X Wang et al 74 9 Hill M, Stith R, McCallum R. Human recombinant IL-1 alt- 30 Sharom FJ. The P-glycoprotein efflux pump: how does it transport ers glucocorticoid receptor function in Reuber hepatoma cells. drugs? J Membr Biol 1997; 160: 161–175. J Immunol 1988; 141: 1522–1528. 31 Stylianou E, Saklatvala J. Interleukin-1. Int J Biochem Cell Biol 10 Spahn J, Szefler S, Surs W, Doherty D, Nimmagadda S, Leung D. A 1998; 30: 1075–1079. novel action of IL-13: induction of diminished monocyte gluco- 32 Lee JC, Laydon JT, McDonnell PC, Gallagher TF, Kumar S, Green D corticoid receptor-binding affinity. J Immunol 1996; 157: 2654– et al. A protein kinase involved in the regulation of inflammatory 2659. cytokine biosynthesis. Nature 1994; 372: 739–746. 11 Falus A, Biro J, Rakasz E. Cytokine networks and corticosteroid 33 Ward SG, Parry RV, Matthews J, O’Neill L. A p38 MAP kinase receptors. Ann NY Acad Sci 1995; 762: 71–77; discussion 77–78. inhibitor SB203580 inhibits CD28-dependent T cell proliferation 12 Liu LY, Sun B, Tian Y, Lu BZ, Wang J. Changes of pulmonary and IL-2 production. Biochem Soc Trans 1997; 25: 304S. glucocorticoid receptor and phospholipase A2 in sheep with acute 34 Crawley JB, Rawlinson L, Lali FV, Page TH, Saklatvala J, Foxwell lung injury after high dose endotoxin infusion. Am Rev Respir Dis BM. T cell proliferation in response to interleukins 2 and 7 1993; 148: 878–881. requires p38MAP kinase activation. J Biol Chem 1997; 272: 13 Corrigan CJ, Brown PH, Barnes NC, Szefler SJ, Tsai JJ, Frew AJ et 15023–15027. al. Glucocorticoid resistance in chronic asthma. Glucocorticoid 35 Whitmarsh AJ, Davis RJ. Transcription factor AP-1 regulation by pharmacokinetics, glucocorticoid receptor characteristics, and mitogen-activated protein kinase signal transduction pathways. inhibition of peripheral blood T cell proliferation by glucocorti- J Mol Med 1996; 74: 589–607. coids in vitro. Am Rev Respir Dis 1991; 144: 1016–1025. 36 McDermott EP, O’Neill LA. Ras participates in the activation of 14 Sher ER, Leung DY, Surs W, Kam JC, Zieg G, Kamada AK et al. p38 MAPK by interleukin-1 by associating with IRAK, IRAK2, Steroid-resistant asthma. Cellular mechanisms contributing to TRAF6, and TAK-1. J Biol Chem 2002; 277: 7808–7815. inadequate response to glucocorticoid therapy. J Clin Invest 1994; 37 Moriguchi T, Kuroyanagi N, Yamaguchi K, Gotoh Y, Irie K, Kano T 93: 33–39. et al. A novel kinase cascade mediated by mitogen-activated 15 Norbiato G, Bevilacqua M, Vago T, Clerici M. Glucocorticoids and protein kinase kinase 6 and MKK3. J Biol Chem 1996; 271: interferon-alpha in the acquired immunodeficiency syndrome. 13675–13679. J Clin Endocrinol Metab 1996; 81: 2601–2606. 38 Cuenda A, Rouse J, Doza YN, Meier R, Cohen P, Gallagher TF et al. 16 Evans DL, Staab JP, Petitto JM, Morrison MF, Szuba MP, Ward HE SB 203580 is a specific inhibitor of a MAP kinase homologue et al. Depression in the medical setting: biopsychological interac- which is stimulated by cellular stresses and interleukin-1. FEBS tions and treatment considerations. J Clin Psychiatry 1999; 60: Lett 1995; 364: 229–233. 40–55; discussion 56. 39 McLaughlin MM, Kumar S, McDonnell PC, Van Horn S, Lee JC, 17 Dantzer R. Cytokine-induced sickness behavior: mechanisms and Livi GP et al. Identification of mitogen-activated protein (MAP) implications. Ann NY Acad Sci 2001; 933: 222–234. kinase-activated protein kinase-3, a novel substrate of CSBP p38 18 Leonard BE. The immune system, depression and the action of MAP kinase. J Biol Chem 1996; 271: 8488–8492. antidepressants. Prog Neuropsychopharmacol Biol Psychiatry 40 Tan Y, Rouse J, Zhang A, Cariati S, Cohen P, Comb MJ. FGF and 2001; 25: 767–780. stress regulate CREB and ATF-1 via a pathway involving p38 MAP 19 Yirmiya R, Pollak Y, Morag M, Reichenberg A, Barak O, Avitsur R kinase and MAPKAP kinase-2. EMBO J 1996; 15: 4629–4642. et al. Illness, cytokines, and depression. Ann NY Acad Sci 2000; 41 Raingeaud J, Whitmarsh AJ, Barrett T, Derijard B, Davis RJ. MKK3- 917: 478–487. and MKK6-regulated gene expression is mediated by the p38 20 Yirmiya R, Weidenfeld J, Pollak Y, Morag M, Morag A, Avitsur R et mitogen-activated protein kinase signal transduction pathway. al. Cytokines, ‘‘depression due to a general medical condition,’’ Mol Cell Biol 1996; 16: 1247–1255. and antidepressant drugs. Adv Exp Med Biol 1999; 461: 283–316. 42 Wang XZ, Ron D. Stress-induced phosphorylation and activation 21 Pariante CM, Pearce BD, Pisell TL, Sanchez CI, Po C, Su C et al. of the transcription factor CHOP (GADD153) by p38 MAP Kinase. The proinflammatory cytokine, interleukin-1alpha, reduces glu- Science 1996; 272: 1347–1349. cocorticoid receptor translocation and function. Endocrinology 43 Zervos AS, Faccio L, Gatto JP, Kyriakis JM, Brent R. Mxi2, a 1999; 140: 4359–4366. mitogen-activated protein kinase that recognizes and phosphor- 22 Krstic MD, Rogatsky I, Yamamoto KR, Garabedian MJ. Mitogen- ylates Max protein. Proc Natl Acad Sci USA 1995; 92: activated and cyclin-dependent protein kinases selectively and 10531–10534. differentially modulate transcriptional enhancement by the glu- 44 Stauber C, Altschmied J, Akerblom IE, Marron JL, Mellon PL. cocorticoid receptor. Mol Cell Biol 1997; 17: 3947–3954. Mutual cross-interference between glucocorticoid receptor and 23 Saklatvala J, Dean J, Finch A. Protein kinase cascades in CREB inhibits transactivation in placental cells. New Biologist intracellular signalling by interleukin-I and tumour necrosis 1992; 4: 527–540. factor. Biochem Soc Symp 1999; 64: 63–77. 45 Peters MJ, Adcock IM, Brown CR, Barnes PJ. Beta-adrenoceptor 24 Raingeaud J, Gupta S, Rogers JS, Dickens M, Han J, Ulevitch RJ et agonists interfere with glucocorticoid receptor DNA binding in rat al. Pro-inflammatory cytokines and environmental stress cause lung. Eur J Pharm 1995; 289: 275–281. p38 mitogen-activated protein kinase activation by dual phos- 46 Ray A, Prefontaine KE. Physical association and functional phorylation on tyrosine and threonine. J Biol Chem 1995; 270: antagonism between the p65 subunit of transcription factor NF- 7420–7426. kappa B and the glucocorticoid receptor. Proc Natl Acad Sci USA 25 Rouse J, Cohen P, Trigon S, Morange M, Alonso-Llamazares A, 1994; 91: 752–756. Zamanillo D et al. A novel kinase cascade triggered by stress and 47 McKay LI, Cidlowski JA. Cross-talk between nuclear factor-kappa heat shock that stimulates MAPKAP kinase-2 and phosphoryla- B and the steroid hormone receptors: mechanisms of mutual tion of the small heat shock proteins. Cell 1994; 78: 1027–1037. antagonism. Mol Endocrinol 1998; 12: 45–56. 26 Irusen E, Matthews JG, Takahashi A, Barnes PJ, Chung KF, Adcock 48 Ajmone-Cat MA, De Simone R, Nicolini A, Minghetti L, Yalcin A, IM. p38 Mitogen-activated protein kinase-induced glucocorticoid Koulich E et al. Effects of phosphatidylserine on p38 mitogen receptor phosphorylation reduces its activity: role in steroid- activated protein kinase, cyclic AMP responding element binding insensitive asthma. J Allergy Clin Immunol 2002; 109: 649–657. protein and nuclear factor-kappaB activation in resting and 27 Aoshiba K, Yasui S, Hayashi M, Tamaoki J, Nagai A. Role of p38- activated microglial cells. J Neurochem 2003; 84: 413–416. mitogen-activated protein kinase in spontaneous apoptosis of 49 Lee H, Bai W. Regulation of estrogen receptor nuclear export by human neutrophils. J Immunol 1999; 162: 1692–1700. ligand-induced and p38-mediated receptor phosphorylation. Mol 28 Baldassare JJ, Bi Y, Bellone CJ. The role of p38 mitogen-activated Cell Biol 2002; 22: 5835–5845. protein kinase in IL-1 beta transcription. J Immunol 1999; 162: 50 Gianni M, Bauer A, Garattini E, Chambon P, Rochette-Egly C. 5367–5373. Phosphorylation by p38MAPK and recruitment of SUG-1 are 29 Barancik M, Bohacova V, Kvackajova J, Hudecova S, Krizanova O, required for RA-induced RAR gamma degradation and transacti- Breier A. SB203580, a specific inhibitor of p38-MAPK pathway, is vation. EMBO J 2002; 21: 3760–3769. a new reversal agent of P-glycoprotein-mediated multidrug 51 Barger PM, Browning AC, Garner AN, Kelly DP. p38 mitogen- resistance. Eur J Pharm Sci 2001; 14: 29–36. activated protein kinase activates peroxisome proliferator-acti-

Molecular Psychiatry p38 MAPK inhibition of GR function X Wang et al 75 vated receptor alpha: a potential role in the cardiac metabolic 5-hydroxytryptamine1C and 5-hydroxytryptamine2 receptors. Mol stress response. J Biol Chem 2001; 276: 44495–44501. Pharmacol 1992; 42: 328–335. 52 Knutti D, Kressler D, Kralli A. Regulation of the transcriptional 61 Julius D, Huang KN, Livelli TJ, Axel R, Jessell TM. The 5HT2 coactivator PGC-1 via MAPK-sensitive interaction with a repres- receptor defines a family of structurally distinct but functionally sor. Proc Natl Acad Sci USA 2001; 98: 9713–9718. conserved serotonin receptors. Proc Natl Acad Sci USA 1990; 87: 53 Lee JC, Kassis S, Kumar S, Badger A, Adams JL. p38 mitogen- 928–932. activated protein kinase inhibitorsFmechanisms and therapeutic 62 Manier DH, Eiring A, Shelton RC, Sulser F. Beta-adrenoceptor- potentials. Pharmacol Ther 1999; 82: 389–397. linked protein kinase A (PKA) activity in human fibroblasts from 54 Wion D, Neveu I, Jehan F, Brachet P. Pertussis toxin provides normal subjects and from patients with major depression. evidence for two independent signalling pathways leading to the Neuropsychopharmacology 1996; 15: 555–561. activation of the nerve growth factor gene. J Neurosci Res 1992; 31: 63 Shelton RC, Mainer DH, Sulser F. cAMP-dependent protein kinase 294–300. activity in major depression. Am J Psychiatry 1996; 153: 1037– 55 Wadekar SA, Li D, Periyasamy S, Sanchez ER. Inhibition of heat 1042. shock transcription factor by GR. Mol Endocrinol 2001; 15: 64 Dwivedi Y, Conley RR, Roberts RC, Tamminga CA, Pandey GN. 1396–1410. [(3)H]cAMP binding sites and protein kinase a activity in the 56 Kim AH, Khursigara G, Sun X, Franke TF, Chao MV. Akt prefrontal cortex of suicide victims. Am J Psychiatry 2002; 159: phosphorylates and negatively regulates apoptosis signal-regulat- 66–73. ing kinase 1. Mol Cell Biol 2001; 21: 893–901. 65 Lamberts SW. The glucocorticoid insensitivity syndrome. Horm 57 Lee JY, Hannun YA, Obeid LM. Functional dichotomy of protein Res 1996; 45: 2–4. kinase C (PKC) in tumor necrosis factor-alpha (TNF-alpha) signal 66 McKay LI, Cidlowski JA. Molecular control of immune/inflamma- transduction in L929 cells. Translocation and inactivation of PKC tory responses: interactions between nuclear factor-kappa B and by TNF-alpha. J Biol Chem 2000; 275: 29290–29298. steroid receptor-signaling pathways. Endocr Rev 1999; 20: 435– 58 Miller AH, Vogt GJ, Pearce BD. The phosphodiesterase type 4 459. inhibitor, rolipram, enhances glucocorticoid receptor function. 67 Branger J, van den Blink B, Weijer S, Madwed J, Bos CL, Gupta A Neuropsychopharmacology 2002; 27: 939–948. et al. Anti-inflammatory effects of a p38 mitogen-activated protein 59 Pariante CM, Pearce BD, Pisell TL, Owens MJ, Miller AH. Steroid- kinase inhibitor during human endotoxemia. J Immunol 2002; 168: independent translocation of the glucocorticoid receptor 4070–4077. by the antidepressant desipramine. Mol Pharmacol 1997; 52: 68 Paccani SR, Boncristiano M, Ulivieri C, D’Elios MM, Del Prete G, 571–581. Baldari CT. Nonsteroidal anti-inflammatory drugs suppress T-cell 60 Leonhardt S, Gorospe E, Hoffman BJ, Teitler M. Molecular activation by inhibiting p38 MAPK induction. J Biol Chem 2002; pharmacological differences in the interaction of serotonin with 277: 1509–1513.

Molecular Psychiatry