Janus Kinase 2-dependent Activation of p38 Mitogen-activated Protein Kinase by Growth Hormone

We demonstrate here that p38 mitogen-activated protein (MAP) kinase is activated in response to cellular stimulation by human GH (hGH) in Chinese hamster ovary cells stably transfected with GH receptor cDNA. This activation requires the proline-rich box 1 region of the GH receptor required for JAK2 association and is prevented by pretreatment of cells with the JAK2-specific inhibitor AG490. ATF-2 is both phosphorylated and transcriptionally activated by hGH, and its transcriptional activation also requires the proline-rich box 1 region of the GH receptor. Expression of wild type JAK2 can further enhance hGH-induced ATF-2-, CHOP-, and Elk-1-mediated transcriptional activation, whereas pretreatment with AG490 is inhibitory. Use of either specific pharmacological inhibitors or transient transfection of cells with p38α MAP kinase cDNA or a dominant negative variant demonstrated that hGH-stimulated transcriptional activation of ATF-2 and CHOP, but not Elk-1, is regulated by p38 MAP kinase. Both the p38 MAP kinase and p44/42 MAP kinase are critical for hGH-stimulated mitogenesis, whereas only p38 MAP kinase is required for hGH-induced actin cytoskeletal re-organization. p38 MAP kinase is therefore an important regulator in coordinating the pleiotropic effects of GH.

We demonstrate here that p38 mitogen-activated protein (MAP) kinase is activated in response to cellular stimulation by human GH (hGH) in Chinese hamster ovary cells stably transfected with GH receptor cDNA. This activation requires the proline-rich box 1 region of the GH receptor required for JAK2 association and is prevented by pretreatment of cells with the JAK2-specific inhibitor AG490. ATF-2 is both phosphorylated and transcriptionally activated by hGH, and its transcriptional activation also requires the proline-rich box 1 region of the GH receptor. Expression of wild type JAK2 can further enhance hGH-induced ATF-2-, CHOP-, and Elk-1-mediated transcriptional activation, whereas pretreatment with AG490 is inhibitory. Use of either specific pharmacological inhibitors or transient transfection of cells with p38␣ MAP kinase cDNA or a dominant negative variant demonstrated that hGH-stimulated transcriptional activation of ATF-2 and CHOP, but not Elk-1, is regulated by p38 MAP kinase. Both the p38 MAP kinase and p44/42 MAP kinase are critical for hGH-stimulated mitogenesis, whereas only p38 MAP kinase is required for hGH-induced actin cytoskeletal re-organization. p38 MAP kinase is therefore an important regulator in coordinating the pleiotropic effects of GH.
Growth hormone (GH), 1 as the major regulator of postnatal body growth, possesses diverse and pleiotropic effects on the growth, differentiation, and metabolism of cells (1,2). The GH receptor is a single membrane-spanning glycoprotein in the cytokine receptor superfamily (3) that consists of a ligandbinding external domain and a 350-amino acid residue cytoplasmic domain required for intracellular signal transduction (1,2). GH receptor signaling involves ligand-induced receptor homodimerization and activation of the tyrosine kinase JAK2 by association with the GH receptor (1,2,4). The proline-rich Box1 motif of the receptor has been identified as the site of association with JAK2. Multiple effector molecules downstream of JAK2-mediating GH signal transduction have now been identified (5)(6)(7)(8)(9)(10)(11). We have also recently demonstrated that GH stimulates the formation of a multiprotein signaling complex centered around p125 FAK (12) and p130 Cas -CrkII (13) leading to JNK/SAPK activation.
We and others (54 -56) have previously demonstrated GH activation of both p44/42 MAP kinase and c-Jun amino-terminal kinase (JNK/SAPK1) (13). We demonstrate here that hGH transiently phosphorylates and activates p38 MAP kinase in CHO cell lines stably transfected with rat GH receptor cDNA and that the activation of the p38 MAP kinase pathway by hGH is JAK2-dependent. Furthermore, both ATF-2 and CHOP (GADD153) are transcriptionally activated by hGH in a JAK2dependent manner, and p38 MAP kinase is required for both cytoskeletal re-organization and cell proliferation stimulated by hGH. The p38 MAP kinase pathway is therefore an important regulator of the pleiotropic effects of GH. polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were blocked with 5% non-fat dry milk in phosphate-buffered saline with 0.1% Tween 20 (PBST) for 1 h at 22°C. The blots were then treated with the primary antibody in PBST containing 1% non-fat dry milk at 4°C overnight. After three washes with PBST, immunolabeling was detected by ECL according to the manufacturer's instructions. Membranes were then stripped by incubation for 30 min at 50°C in a solution containing 62.5 mM Tris-HCl, pH 6.7, 2% SDS, and 0.7% mercaptoethanol. Blots were then washed for 30 min with several changes of PBST at room temperature. Efficacy of stripping was determined by re-exposure of the membranes to ECL. Thereafter, blots were reblocked and immunolabeled as described above.
Kinase Assays-p38 MAP kinase assays were performed using New England Biolabs assay kit specific for p38␣ and p38␤ MAP kinase according to the manufacturer's instructions. In brief, CHO-GHR 1-638 cells were serum-deprived for 16 h, treated with 50 nM hGH, and the cells lysed at 4°C in 1 ml of lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM glycerol phosphate, 1 mM Na 3 VO 4 , 0.1% phenylmethylsulfonyl fluoride, 1 mg/ml leupeptin) per sample. The lysates were centrifuged at 15,000 ϫ g for 15 min at 4°C. The supernatant containing 300 g of protein per sample was incubated overnight at 4°C with 20 l of immobilized phospho-p38 MAP kinase (Thr-180/Tyr-182) monoclonal antibody in a final volume of 500 l in 1ϫ lysis buffer. The bead slurries were washed twice with 500 l of lysis buffer containing 0.1% phenylmethylsulfonyl fluoride and twice with 500 l of kinase buffer (25 mM Tris-HCl, pH 7.5, 5 mM glycerol phosphate, 2 mM dithiothreitol, 0.1 mM Na 3 VO 4 , 10 mM MgCl 2 ). 3 g of recombinant glutathione S-transferase-ATF2 protein containing 1-109 amino-terminal amino acids of ATF-2 was added to each sample. The kinase reactions were carried out in the presence of 100 mM ATP at 30°C for 30 min. Phosphorylation of ATF-2 at Thr-71 was determined by Western blot using phospho-ATF-2 (Thr-71) antibody (1:1000). p44/42 MAP kinase assays were also performed using New England Biolabs assay kit according to the manufacturer's instructions. In brief, cells were serumdeprived for 16 h, treated with 50 nM hGH, and lysed at 4°C in 1 ml of lysis buffer (as described above) per sample. The lysates were centrifuged at 15,000 ϫ g for 15 min at 4°C. The supernatant containing 200 g of protein per sample was incubated overnight at 4°C with 15 l of immobilized phospho-specific p44/42 MAP kinase (Thr-202/Tyr-204) monoclonal antibody in a final volume of 500 l in 1ϫ lysis buffer. The pellets were washed twice with 500 l of lysis buffer containing 0.1% phenylmethylsulfonyl fluoride and washed twice with 500 l of kinase buffer (25 mM Tris-HCl, pH 7.5, 5 mM glycerol phosphate, 2 mM dithiothreitol, 0.1 mM Na 3 VO 4 , 10 mM MgCl 2 ). The kinase reactions were performed in the presence of 2 g of Elk-1 fusion protein and 200 M ATP at 30°C for 30 min. Elk-1 phosphorylation was selectively detected by western immunoblotting using a chemiluminescent detection system and a specific phospho-Elk-1 (Ser-383) antibody (1:1000).
Transfection Procedure and Luciferase Assays-CHO-GHR 1-638 cells were cultured to 60 -80% confluence for transfection experiments in 6-well plates. 1 g of reporter plasmid pFR-Luc was transfected together with 20 ng of the respective fusion trans-activator plasmid (pFA-ATF-2, pFA-CHOP, pFA-Elk-1, or pFC2-dbd). For transient transfection of MAP kinases, either 1.5 g of pCDNA3-p38␣ MAPK, dominant negative pCDNA3-p38␣ MAPK, or pCDNA3-p42 MAPK were transfected in each well. Control empty vector was used to normalize the amount of plasmids in each well. For each well, 4 g of DOTAP for each g of DNA was used as per the manufacturer's instructions. DNA and the DOTAP reagents were diluted separately in 100 l of serumfree medium mixed and incubated at room temperature for 30 min. DNA-lipid complex was diluted to a final volume of 6 ml (for triplicate samples) with serum-free medium. Cells in each well were rinsed once with 2 ml of serum-free medium, and 2 ml of diluted DNA-lipid complex was overlaid in each well and incubated for 6 h. After incubation, complete Ham's F-12 medium containing 2% FBS was added to each well so as to incubate the cells in 0.5% serum for 12-16 h. Cells were pretreated with 10 M SB203580 or 30 M PD98059 or with the equivalent amount of solvent (Me 2 SO) as a control for 45 min prior to stimulation with hGH. 50 nM hGH was added for an additional 5-7 h. The cells were washed in PBS and lysed with 300 l of 1ϫ lysis buffer (25 mM Tris phosphate, pH 7.8, 2 mM EDTA, 2 mM dithiothreitol, 10% glycerol, 1% Triton X-100) by a freeze-thaw cycle, and lysate was collected by centrifugation at 14,000 rpm for 15 min. The supernatant was used for the assay of luciferase and ␤-galactosidase activity. The luciferase activities were normalized on the basis of protein content as well as on the ␤-galactosidase activity of pCMV␤ vector. The ␤-galactosidase assay was performed with 20 l of precleared cell lysate according to a standard protocol (62). Mean and standard deviations of at least three independent experiments are shown in the figures.
Visualization of Filamentous Actin and Confocal Laser Scanning Microscopy-Cells were fixed in ice-cold 4% paraformaldehyde, washed with PBS, permeabilized for 5 min with 0.1% Triton X-100, blocked in 2% bovine serum albumin, and incubated with phalloidin-TRITC (0.2 mg/ml) after the indicated treatments. Excess phalloidin-TRITC was removed by extensive washing with PBS. Labeled cells were visualized with a Carl Zeiss Axioplan microscope equipped with epifluorescence optics and a Bio-Rad MRC1024 confocal laser system. Images were converted to the tagged information file format and processed with the MacIntosh Photoshop program.
Cell Proliferation Assays-Cell proliferation was assayed by measuring incorporation of [ 3 H]thymidine during DNA synthesis (63). Subconfluent CHO-GHR 1-638 cell monolayers in 24-well plates were grown to quiescence in serum-free Ham's F-12 medium at 37°C for 16 h. Cells were then incubated for 24 h in Ham's F-12 medium with hGH to a final concentration as indicated in the figure legends. Each cell line was plated in triplicate for each treatment. In case of pretreatment of cells with chemical inhibitors, cells were pretreated with 10 M SB203580 or 30 M PD98059 or with the equivalent amount of solvent (Me 2 SO) as a control for 45 min. [ 3 H]Thymidine (1 Ci per well, 1 Ci ϭ 37 GBq, Amersham Pharmacia Biotech) was added, and the cells were incubated at 37°C for a further 8 h. Cells were rinsed twice with ice-cold PBS incubated with 1 ml of ice-cold 5% trichloroacetic acid for 30 min at 4°C and 0.5 ml of 0.5 N NaOH, 0.5% SDS was added subsequently at room temperature. Solubilized samples were subjected to liquid scintillation counting in a scintillation counter.
Statistics and Presentation of Data-All experiments were repeated at least three times. Figures presented for Western blot analyses are representative of multiple experiments. The text under "Results" summarizes the results from multiple Western blot analyses. Consequently, the text under "Results" (e.g. description of the time course of phosphorylation and dephosphorylation) may not exactly correspond to the actual figure presented. All numerical data are expressed as mean Ϯ S.D. Data were analyzed using the two-tailed t test or analysis of variance.

Time-and Dose-dependent Activation of p38 MAP Kinase by
Cellular Stimulation with hGH-We and others (13, 54 -56) have previously shown that GH activates both p42/44 MAP kinase and JNK. In order to determine whether cellular stimulation with hGH also results in the activation of p38 MAP kinase, we first examined whether hGH stimulation of cells resulted in the phosphorylation of p38 MAP kinase. Dual phosphorylation of p38 MAP kinase on Thr-180 and Tyr-182 is required for its activation (14,25,64). We therefore treated serum-deprived CHO-GHR 1-638 cells for 0 -60 min with 50 nM hGH and prepared cell extracts for Western blot analysis. Phosphorylated p38 MAP kinase was detected by use of a phospho-p38 MAP kinase (Thr-180/Tyr-182) antibody that detects p38 MAP kinase after dual phosphorylation at Thr-180 and Tyr-182. hGH stimulation of CHO-GHR 1-638 cells resulted in increased p38 MAP kinase phosphorylation detectable 5 min after stimulation. Maximal hGH-stimulated phosphorylation of p38 MAP kinase was observed at 15 min and followed by a subsequent decline in phosphorylation state to 60 min (Fig.  1A). Equal loading of samples was verified by stripping of the membrane and subsequent reblotting with a p38 MAPK antibody (Fig. 1B). We also examined whether the hGH-stimulated phosphorylation of p38 resulted in enhanced p38 MAP kinase activity using ATF-2 as an in vitro substrate (26,41). We therefore treated serum-deprived CHO-GHR 1-638 cells for 0 -60 min with 50 nM hGH and immunoprecipitated dualphosphorylated p38 MAP kinase from cell extracts with immobilized phospho-p38 MAP kinase (Thr-180/Tyr-182) monoclonal antibody and measured p38 MAP kinase activity by using ATF-2 as the substrate. Phosphorylation of ATF-2 by the immunoprecipitated p38 MAP kinase was detected by immunoblotting with an antibody that detects ATF-2 when phosphorylated on threonine residue 71 (Thr-71). Thus the appearance of phosphorylation of p38 MAP kinase observed in hGH-stimulated cell extracts corresponds temporally to the increase in p38 MAP kinase activity in hGH-stimulated cell extracts. We also examined the dose dependence of the hGH stimulation of p38 MAP kinase phosphorylation and activity. We therefore stimulated CHO-GHR 1-638 cells with 0, 0.005, 0.05, 0.5, 5, and 50 nM hGH for 15 min and prepared cell extracts. Determination of p38 MAP kinase phosphorylation and p38 MAP kinase activity was performed as described above. The maximal phosphorylation ( Fig. 1D) and activation (Fig. 1F) of p38 MAP kinase was achieved at a concentration of 50 nM hGH in CHO-GHR 1-638 cells. Again, equal loading of samples was verified by stripping of the membrane and subsequent reblotting with a p38 MAPK antibody (Fig. 1E). Therefore, 50 nM hGH was used for all subsequent experiments.
Activation of p38 MAP Kinase by hGH Is JAK2-dependent-The activation of JAK2 upon cellular stimulation with GH is thought to be required for subsequent signal transduction events (4,5,12,65). To examine the potential role of JAK2 in the activation of p38 MAP kinase by hGH, we utilized well characterized CHO cell clones stably expressing the wild type receptor (CHO-GHR 1-638 ), a receptor mutation in which the proline-rich box 1 region had been deleted (CHO-GHR 1-638 ⌬- Kinase assays (C and F) were performed by immunoprecipitating activated p38 MAP kinase with immobilized phospho-p38 MAP kinase (Thr-180/Tyr-182) monoclonal antibody, and p38 MAP kinase activity was determined as described under "Experimental Procedures." The data presented are representative of at least three separate experiments. 297-311), and a receptor mutation in which the individual proline residues of box 1 had been converted to alanine (CHO-GHR 1-638 P300,301,303,305A) ( Fig. 2A). The proline rich box 1 region of the GH receptor is required for the association of JAK2 with the receptor and its subsequent activation after receptor dimerization (59). We therefore stimulated the respective serum-deprived cell line with 50 nM hGH for 15 min and prepared cell extracts for determination of p38 MAP kinase activity. hGH stimulation of CHO-GHR 1-638 cells resulted in activation of p38 MAP kinase, whereas no hGH-dependent activation of p38 MAP kinase activity was observed in either the CHO-GHR 1-638 ⌬297-311 or CHO-GHR 1-638 P300,301,303, 305A cell lines (Fig. 2B). For control purposes we also treated CHO-GHR 1-638 , CHO-GHR 1-638 ⌬297-311, and CHO-GHR 1-638 -P300,301,303,305A cell lines with sorbitol (41, 66) and demonstrated equipotent activation of p38 MAP kinase in all three cell lines. Thus it is apparent that the hGH activation of p38 MAP kinase requires the proline-rich box 1 region of the GH receptor indicative that p38 MAP kinase is activated in a JAK2-dependent manner.
Tyrphostin AG490, a reportedly specific inhibitor of JAK2 tyrosine kinase activity (67)(68)(69), was also utilized to verify that hGH stimulation of p38 MAP kinase activity was JAK2-dependent. CHO-GHR 1-638 cells were therefore pretreated with 100 M AG490 or vehicle (Me 2 SO) for 16 h, followed by stimulation with 50 nM hGH for 15 min and determination of p38 MAP kinase activity. The hGH-stimulated p38 MAP kinase activation was abolished by the pretreatment of the cells with AG490 (Fig.  2C). In contrast, pretreatment of the CHO-GHR 1-638 cells with AG490 as described above did not alter the ability of sorbitol to stimulate p38 MAP kinase activity (Fig. 2D). Thus, hGH stimulation of p38 MAP kinase activity by hGH is JAK2-dependent.
Time-and Dose-dependent Phosphorylation of Cellular ATF-2 by Cellular Stimulation with hGH-Since ATF-2 has been demonstrated to be a substrate of p38 MAP kinase (23,26,42), we wished to determine if hGH stimulation of CHO-GHR 1-638 cells also resulted in phosphorylation of cellular ATF-2. CHO-GHR 1-638 cells were serum-deprived and stimulated with 50 nM hGH for the indicated times. Cell lysates were prepared, and phosphorylation of ATF-2 was detected by immunoblotting with a specific antibody against ATF-2 phosphorylated on Thr-71 (Fig. 3A). Phosphorylation of ATF-2 was first observed 5 min after cellular stimulation with hGH, and maximal hGHdependent phosphorylation of ATF-2 was observed at 30 min with a relative decline in ATF-2 phosphorylation observed at 60 min. Equal loading of samples was verified by stripping of the FIG. 2. JAK2 dependent activation of p38 MAP kinase by hGH in CHO cells stably transfected with wild type GH receptor or Box1-deficient GH receptor cDNAs. A, schematic diagram of the GH receptor and the various GH receptor mutations/deletions used. The wild type receptor has the extracellular, transmembrane, and intracellular regions indicated. CHO-GHR 1-638 ⌬297-311 has the proline-rich region (Box1) deleted, CHO-GHR 1-638 P300,301,303,305A has the individual proline residues in box 1 mutated to alanine. CHO-GHR 1-638 cells and CHO cells expressing the GH receptor box 1 mutations or deletions were stimulated with 50 nM hGH for 15 min, cell extracts prepared, and p38 MAP kinase activity determined (B). CHO-GHR 1-638 cells were preincubated with 100 M tyrphostin AG490 or vehicle Me 2 SO for 16 h at 37°C prior to treatment with 50 nM hGH for 0 or 15 min, cell extracts prepared, and p38 MAP kinase activity determined (C). CHO-GHR 1-638 cell (preincubated Ϯ AG490), CHO-GHR 1-638 ⌬297-311 cells, and CHO-GHR 1-638 P300,301,303,305A cell were treated with sorbitol for 5 min, cell extracts prepared, and p38 MAP kinase activity determined (D). The data presented are representative of at least three separate experiments. membrane and subsequent reblotting with an ATF-2 antibody (Fig. 3B). We also examined the dose dependence of the hGH stimulation of ATF-2 phosphorylation. We therefore stimulated CHO-GHR 1-638 cells with 0, 0.005, 0.05, 0.5, 5, and 50 nM hGH for 15 min and prepared cell extracts. The maximal phosphorylation of ATF-2 was achieved at a concentration of 5-50 nM hGH in CHO-GHR 1-638 cells (Fig. 3C). Equal loading of samples was again verified by stripping of the membrane and subsequent reblotting with an ATF-2 antibody (Fig. 3D).
Phosphorylation of the ATF-2 by hGH Is JAK2dependent-We also examined the requirement of JAK2 for hGH-stimulated phosphorylation of ATF-2. We utilized the same well characterized CHO cell clones stably expressing either the wild type receptor (CHO-GHR 1-638 ), a receptor mutation in which the proline-rich box 1 region had been deleted (CHO-GHR 1-638 ⌬297-311), or a receptor mutation in which the individual proline residues of box 1 had been converted to alanine (CHO-GHR 1-638 P300,301,303,305A) ( Fig. 2A) as described above. Each serum-deprived cell line was stimulated with 50 nM hGH for 30 min, and cell extracts were prepared for Western blot analysis of ATF-2 phosphorylation. Human GHdependent phosphorylation of ATF-2 was only detected in the CHO cell line that expressed the full-length wild type GH receptor (Fig. 4A). Equal loading of total cellular ATF-2 from the three CHO cell clones was verified by reblotting the membrane with an ATF-2 antibody (Fig. 4B). Thus, it is apparent that the GH-induced ATF-2 phosphorylation requires the proline-rich box 1 region of the GH receptor.
To verify further the JAK2 dependence of the hGH-stimu-lated phosphorylation of ATF-2, we also pretreated CHO-GHR 1-638 cells with 100 M AG490 or vehicle followed by cellular stimulation with 50 nM hGH for 30 min. Pretreatment with the JAK2-specific inhibitor prevented hGH stimulation of ATF-2 phosphorylation (Fig. 4C). Again equal loading of total cellular ATF-2 for the various experimental conditions was verified by reblotting the membrane with an ATF-2 antibody (Fig. 4D).
hGH Stimulates the Transcriptional Activation of ATF-2, CHOP, and Elk-1-To determine if hGH-stimulated phosphorylation of ATF-2 resulted in ATF-2-dependent transcription, we utilized a trans-activation reporter assay specific for ATF-2. We therefore transiently transfected CHO-GHR 1-638 cells with the fusion trans-activator plasmid pFA-ATF-2 consisting of the DNA binding domain of GAL4 (residue 1-147) and the transactivation domain of ATF-2 together with the fusion transactivator plasmids pFA-ATF-2 consisting of the DNA binding domain of GAL4 (residue 1-147) and the trans-activation domain of ATF-2, together with luciferase reporter plasmid and pCMV␤ vector, respectively. The luciferase activities were measured and normalized on the basis of protein content as well as on the ␤-galactosidase activity of pCMV␤ vector. We demonstrated an hGH-dependent transcriptional activation of ATF-2 ( Fig. 5A) in CHO-GHR 1-638 cells. hGH failed to stimulate ATF-2-mediated reporter expression in cells transfected with a plasmid encoding the GAL4 DNA binding domain (residue 1-147) lacking an activation domain, indicative that the ATF-2 transcriptional activation domain is required for hGHstimulated reporter expression. Since transcriptional activation of CHOP has also been shown to be p38 MAP kinase-dependent, we utilized a trans-activation reporter assay specific for CHOP to determine if hGH can also stimulate CHOP-dependent transcription. We demonstrated a hGH-dependent transcriptional activation of CHOP (Fig. 5B) in CHO-GHR 1-638 cells. hGH failed to stimulate CHOP-mediated reporter expression in cells transfected with a plasmid encoding the GAL4 DNA binding domain (residue 1-147) lacking an activation domain, indicative that the CHOP transcriptional activation domain is required for hGH-stimulated reporter expression. We have also incorporated an Elk-1 reporter assay for control purposes (Fig. 5C), since it was previously shown that GHstimulated transcriptional activation of Elk-1 was p44/42 MAP kinase-dependent (70,71). We demonstrated a hGH-dependent transcriptional activation of Elk-1 in CHO-GHR 1-638 cells. hGH failed to stimulate Elk-1-mediated reporter expression in cells transfected with a plasmid encoding the GAL4 DNA binding domain (residue 1-147) lacking an activation domain, indicating that the Elk-1 transcriptional activation domain is required for hGH-stimulated reporter expression.
hGH-stimulated ATF-2, CHOP, and Elk-1 Transcriptional Activation Is JAK2-dependent-We have demonstrated above that the activation of p38 MAP kinase by hGH is JAK2-dependent and therefore reasoned that hGH-stimulated transcriptional activation of ATF-2 and CHOP should also be JAK2-dependent. hGH stimulation of Elk-1 has previously been demonstrated to be JAK2-dependent (73). We first examined the hGH-dependent transcriptional activation of ATF-2, CHOP, and Elk-1 in CHO cell clones stably expressing either the wild type receptor (CHO-GHR 1-638 ), a receptor mutation in which the proline-rich box 1 region had been deleted (CHO-GHR 1-638 ⌬297-311), or a receptor mutation in which the individual proline residues of box 1 had been converted to alanine (CHO-GHR 1-638 P300,301,303,305A) as described above. Neither the CHO-GHR 1-638 ⌬297-311 nor CHO-GHR 1-638 P300, 301,303,305A) cell lines responded to hGH with transcriptional activation of ATF-2, CHOP, or Elk-1 in contrast to CHO-GHR 1-638 cells that responded with the expected magnitude of transcriptional

FIG. 4. JAK2-dependent phosphorylation of ATF-2 by hGH in CHO cells stably transfected with wild type GH receptor or
Box1-deficient GH receptor cDNAs. The schematic diagram of the GH receptor and the various GH receptor mutations/deletions is presented in Fig. 2A. CHO-GHR 1-638 , CHO-GHR 1-638 ⌬297-311, and CHO-GHR 1-638 P300,301,303,305A cells were stimulated with 50 nM hGH for 30 min. The samples containing equal amounts of proteins were electrophoresed and immunoblotted with the antibodies specific for ATF-2 phosphorylated at Thr-71 (A) and total ATF-2 (B). CHO-GHR 1-638 cells were preincubated with 100 M tyrphostin AG490 or vehicle Me 2 SO for 16 h at 37°C prior to treatment with 50 nM hGH. The samples containing equal amount of proteins were electrophoresed and immunoblotted with the antibodies specific for ATF-2 phosphorylated at Thr-71 (C) and total ATF-2 (D). The data presented are representative of at least three separate experiments. activation (Fig. 6A). We also demonstrated that transient transfection of a JAK2 expression plasmid enhanced the hGH-dependent transcriptional activation of ATF-2, CHOP, and Elk-1. Furthermore, pretreatment of the cells with 100 M AG490 prevented both the hGH-dependent transcriptional activation of ATF-2, CHOP, and Elk-1 and the enhancement of hGH-dependent transcriptional activation observed upon transient transfection of the JAK2 ex-  ). B, CHO-GHR 1-638 cells were co-transfected with the following: 1) ␤-galactosidase expression vector pCMV␤; 2) reporter plasmid pFR-Luc; 3) fusion trans-activator plasmid either pFA-ATF-2, pFA-CHOP, or pFA2-Elk-1 as well as pRC-JAK2 as indicated. CHO-GHR 1-638 cells were preincubated with 100 M tyrphostin AG490 or vehicle Me 2 SO for 16 h at 37°C prior to treatment with 50 nM hGH for 6 -8 h, and cell extracts were prepared. The relative luciferase activities presented were normalized by protein concentrations as well as ␤-galactosidase activity (mean Ϯ S.D., n ϭ 3). Luciferase activity in cell lysates is normalized to the activity in untreated control cells (control ϭ 1). The data presented are representative of at least three separate experiments, each measured in triplicate. Bars represent mean Ϯ S.D. *, p Ͻ 0.01. pression plasmid (Fig. 6B).
Lack of Cross-inhibition between the MEK1-specific Inhibitor and the p38 MAP Kinase-specific Inhibitor-PD98059 selectively blocks the activity of MEK1 thereby inhibiting the activation of the p44/42 MAP kinase and the subsequent phosphorylation of p44/42 MAP kinase substrates both in vitro and in vivo (73)(74)(75). SB203580 is a pyridinyl imidazole derivative that is a highly specific inhibitor of p38 MAP kinase activity (26,76,77). We utilized these pharmacologic agents to determine which pathway was utilized by hGH to stimulate the transcriptional activation of ATF-2, CHOP, and Elk-1 and therefore needed to establish the absence of cross-inhibition of the respective pathways. CHO-GHR 1-638 cells were pretreated with either 30 M PD98059 or 10 M SB203580 prior to hGH stimulation, and the activity of both MAP kinases was determined. 30 M PD98059 did not inhibit the hGH-stimulated activation of p38 MAP kinase (Fig. 7A), and 10 M SB203580 did not inhibit the hGH-stimulated activation of p42/44 MAP kinase (Fig. 7B). Both of the inhibitors at the same concentration specified above effected inhibition of the desired respective pathway (Fig. 7, A and B). Thus, PD98059 and SB203580 were suitable for delineation of the MAP kinase pathway required for hGH-dependent activation of ATF-2, CHOP, and Elk-1 in the CHO-GHR 1-638 cell line.
Inhibition of p38 MAP Kinase Attenuates hGH-stimulated ATF-2 and CHOP but Not Elk-1-mediated Transcriptional Activation-To investigate the possible role of p44/42 MAP kinase and p38 MAP kinase in hGH-induced ATF-2-and CHOP-mediated transcriptional activation, CHO-GHR 1-638 cells transfected with the fusion transactivator plasmid pFA-ATF-2 or pFA-CHOP, and the luciferase reporter plasmids were treated with PD98059, SB203580, or Me 2 SO prior to hGH stimulation. Human GH-stimulated ATF-2-mediated transcriptional activation in CHO-GHR 1-638 cells was significantly abrogated by the pretreatment with SB203580 but was unaffected by pre-treatmentwithPD98059,indicatingthathGH-stimulatedATF-2dependent gene expression requires p38 MAP kinase activation but not MEK1 and subsequent p44/42 MAP kinase activation  (Fig. 7C). Similar results were demonstrated in the pFA-CHOP-transfected CHO-GHR 1-638 cells (Fig. 7D). These data indicate that hGH-induced ATF-2-and CHOP-mediated transcriptional activation are downstream of p38 MAP kinase and are independent of p44/42 MAP kinase activation. In contrast, both p44/42 MAP kinase and p38 MAP kinase have shown to be upstream of Elk-1 (78). To determine whether inhibition of the p44/42 MAP kinase or the p38 MAP kinase activity affected the ability of Elk-1 to mediate transcriptional activation in response to hGH, CHO-GHR 1-638 cells transfected with pFA-Elk-1 fusion transactivator plasmid and luciferase reporter plasmid were treated with 30 M PD98059, 10 M SB203580, or Me 2 SO prior to hGH stimulation. Human GH-stimulated Elk-1-mediated transcriptional activation in CHO-GHR 1-638 cells was largely abrogated by pretreatment with PD98059 but unaffected by pretreatment with SB203580, indicating that hGHstimulated Elk-1-dependent gene expression requires MEK1 activation but not p38 MAP kinase activation (Fig. 7E).
Expression of p38␣ MAP Kinase Enhances hGH-induced ATF-2-and CHOP-mediated but Not Elk-1-mediated Transcriptional Activation-To verify further that hGH-stimulated ATF-2 and CHOP-mediated transcriptional activation were p38 MAP kinase-dependent, we transiently transfected either wild type p38␣ MAP kinase, dominant negative p38␣ MAP kinase, or p42 MAP kinase expression plasmids with the respective fusion trans-activator plasmids (pFA-ATF-2, pFA-CHOP, and pFA2-Elk-1) and the luciferase reporter plasmid and examined hGH-stimulated ATF-2-, CHOP-, and Elk-1-mediated transcription. As observed in Fig. 8, transfection of wild type p38␣ MAP kinase enhanced both hGH-induced ATF-2and CHOP-but not Elk-1-mediated transcriptional activation. Transfection of the dominant negative p38␣ MAP kinase abolished both hGH-induced ATF-2-and CHOP-mediated but not Elk-1-mediated transcription. Transfection of p42 MAP kinase did not alter hGH-induced ATF-2-or CHOP-mediated transcriptional activation, although hGH-induced Elk-1-mediated transcriptional activation was remarkably enhanced upon cotransfection of p42 MAP kinase. These data suggest that activation of p38 MAP kinase is required for hGH-dependent ATF-2-and CHOP-mediated but not Elk-1-mediated transcriptional activation in CHO-GHR 1-638 cells.
hGH-induced Actin Cytoskeletal Re-organization Is p38 MAP Kinase-dependent-hGH has previously been demonstrated to stimulate re-organization of the actin cytoskeleton including stress fiber breakdown and the formation of membrane ruffles (79). Since p38 MAP kinase has been implicated to be the required for actin cytoskeletal reorganization stimulated by a variety of stimuli (18,80,81), we therefore examined whether pretreatment of cells with SB203580 affected hGH-induced actin cytoskeletal re-organization. CHO-GHR 1-638 cells were grown on coverslips to 50% confluence and pretreated with PD98059 or SB203580 or the equivalent volume of Me 2 SO for 45 min followed by stimulation with 50 nM hGH for 5 min. In vehicle-treated cells, hGH initially stimulated the depolymerization of actin stress fibers with maximal depolymerization of stress fibers observed 5 min after addition of hGH (Fig. 9, A and  B). Pretreatment of cells with SB203580 prevented the ability of hGH to stimulate stress fiber breakdown in CHO-GHR 1-638 cells (Fig. 9D). In contrast, CHO-GHR 1-638 cells pretreated with PD98059 still exhibited stress fiber breakdown upon hGH stimulation similar to that of the vehicle-treated cells (Fig. 9C). These results suggest that the stress fiber breakdown stimulated by hGH is mediated specifically through the p38 MAP kinase pathway.
STAT5-mediated Transcriptional Activation Is Independent of p44/42 MAP Kinase and p38 MAP Kinase-One major mechanism by which hGH affects cellular function is by use of the JAK-STAT pathway, especially JAK2 and STAT5 (1,2). We have demonstrated here that the p38 MAP kinase pathway is JAK2-dependent, and therefore, we wished to determine if p38 MAP kinase was therefore also required for STAT5 activation. CHO-GHR 1-638 cells were transfected with the SPI-GLE1-LUC plasmid and pCMV␤ vector and were pretreated with 30 M Roles of p44/42 MAP Kinase and p38 MAP Kinase Pathways in hGH-induced Mitogenesis-To characterize the roles of p38 MAP kinase and p44/42 MAP kinase pathways in hGH-induced cell mitogenesis, we examined the effect of the specific chemical inhibitors of the p44/42 MAP kinase and p38 MAP kinase pathways on hGH-induced cell proliferation in CHO-GHR 1-638 cells. Cell proliferation was estimated by the [ 3 H]thymidine incorporation assay. The effects of these inhibitors on DNA synthesis were first examined over a range of inhibitor concentrations with or without the presence of 50 nM hGH as indicated. Thus we observed that both PD98059 and SB203580 inhibit hGH-stimulated [ 3 H]thymidine incorporation in a dosedependent manner (Fig. 10, A and B). We next examined [ 3 H]thymidine incorporation over a range of hGH concentrations when the concentration of inhibitors was held constant (Fig. 10C). Inhibition of MEK1 by PD98059 attenuated hGHinduced DNA synthesis by 40 -45% in CHO-GHR 1-638 cells. A similar pattern of attenuation of hGH-induced CHO-GHR 1-638 cell proliferation was observed upon inhibition of p38 MAP kinase by SB203580. A combination of both inhibitors resulted in an additive effect on the inhibition of cell proliferation, reducing hGH-induced DNA synthesis by 60 -70% compared with cells treated with Me 2 SO alone. DISCUSSION In the present study we have demonstrated that hGH induces the rapid phosphorylation and activation of p38 MAP kinase. We (13,54) and others (55,56) have previously demonstrated that other members of the MAP kinase family, namely p44/42 MAP kinase and c-Jun amino-terminal kinase/ stress-activated protein kinase (JNK/SAPK), are also activated by GH. In support of our demonstration of GH-dependent activation of p38 MAP kinase, the groups of Schwartz and Carter-Su (73) have recently reported a minimal (0.6-fold) increase in p38 MAP kinase activity after GH stimulation of 3T3-F442A cells. The role of activation of the MAP kinases in the cellular effects of GH requires further delineation. It has been demonstrated that GH activation of p44/42 MAP kinase is required for Elk-1-mediated transcription and also for expression of c-fos, egr-1, and junB (73). We have also previously demonstrated that p44/42 MAP is required for the hyperproliferation of mammary carcinoma cells induced by autocrine production of GH (82). We have demonstrated here that p38 MAP kinase is required for hGH stimulation of ATF-2-and CHOP-mediated transcriptional activation, for hGH-stimulated re-organization of the actin cytoskeleton, and also for hGH-stimulated mitogenesis. Thus it is apparent that GH activation of p38 MAP kinase is pivotal in mediation of the pleiotropic cellular effects of GH.
It is apparent that hGH stimulation of p38 MAP kinase activity is JAK2-dependent. JAK2 has been proposed to be the initial event in GH signal transduction (1,4,83), and no GHdependent tyrosine phosphorylation of proteins is observed in cells transfected with cDNAs encoding GH receptor mutations lacking the ability to activate JAK2 (58). JAK2 has also been demonstrated to be required for GH-dependent activation of p44/42 MAP kinase and a variety of other signaling molecules including p125FAK (12), IRS-1, and IRS-2 (6). In the case of p44/42 MAP kinase, JAK2 is required for the phosphorylation of SHC and Grb-2 (8) and therefore subsequent coupling of Ras-Raf upstream of MEK1 (5,58,83). Interestingly though, JAK2 is apparently not required for GH stimulation of calcium ion influx through L-type calcium channels (84). In any case this is the first demonstration that p38 MAP kinase is activated in a JAK-dependent manner despite the fact that p38 MAP kinase has been reported to be activated by other members of the cytokine receptor superfamily which also utilize JAK2 such as erythropoietin (85) and granulocyte colony-stimulating factor (19). It is most probable that JAK2 is upstream of a previously described pathway for the activation of p38 MAP kinase (3). We (12) and others (86) have placed the focal adhesion protein-tyrosine kinase family (including FAK and Pyk2) as downstream components of JAK activation (FAK downstream of JAK2 and Pyk2 downstream of JAK3). It has recently been reported that Pyk2 is critical for the JAK-mediated p44/42 MAP kinase (87) and p38 MAP kinase (39) activation, and therefore a JAK2-FAK coupling may be one mechanism for GH activation of the p38 MAP kinase pathway. In this regard it is interesting that GH utilizes JAK2-dependent phosphorylation of the EGF receptor to activate p44/42 MAP kinase (89).
ATF-2 is one member of the ATF/CREB family of transcription factors and binds to the cAMP response element as a homodimer or heterodimer with c-Jun (90). Interestingly, GH has previously been demonstrated to increase the cellular level of c-Jun resulting in increased binding to AP-1 sites (91). The amino-terminal region of ATF-2 contains the transcriptional activation domain, which is phosphorylated on Thr-69, Thr-71, and Ser-90 by stress-activated kinases (including p38 MAP kinase) (3,23,42) leading to increased ATF-2 trans-activation (42,90). We have demonstrated here that hGH stimulation of cells results in phosphorylation of ATF-2 and subsequent ATF-2-mediated transcriptional activation. As could be expected from the JAK2-dependent activation of p38 MAP kinase, we also demonstrated that ATF-dependent. ATF-2 trans-activation stimulated by GH may also provide a mechanism for the increase in c-Jun transcription noted after cellular exposure to GH since ATF-2 is found in a multiprotein complex with p300 binding to the Jun2 element of the c-jun promoter (92). ATF-2 trans-activation may also provide one mechanism for the p38 MAP kinase-dependent portion of GH-stimulated cell proliferation since ATF-2 has been demonstrated to cooperate with v-Jun to promote cell proliferation (93).
We have also demonstrated here that hGH stimulation of cells results in the transcriptional activation of CHOP. CHOP, also known as GADD153 (growth arrest and DNA damage), is a mammalian gene that encodes for a small nuclear protein related to the CCAAT/enhancer-binding protein (C/EBP) family of transcription factors (44). We have demonstrated here that the hGH-stimulated transcriptional activation of CHOP is p38 MAP kinase-dependent. This is concordant with the demonstration that CHOP undergoes stress-induced phosphorylation on two adjacent serine residues (78 and 81) by p38 MAP kinase in vivo with a resultant enhancement of CHOP transactivation (44). CHOP was initially proposed to be involved in cell cycle arrest and apoptosis (94 -96). It is therefore interesting that we observed hGH-stimulated proliferation, both here in CHO-GHR 1-638 cells and in hGH-producing mammary carcinoma cells (82), to be p38 MAP kinase-dependent. Autocrine production of hGH in mammary carcinoma cells results in the transcriptional up-regulation of CHOP and increased CHOPmediated transcription. 2 Overexpression of CHOP by stable transfection of CHOP cDNA in mammary carcinoma cells results in enhanced cell proliferation in response to hGH, 2 suggesting that CHOP is also a positive regulator of cell number.
Recent evidence derived from in vivo models of CHOP deficiency indicates that CHOP may possess a dual function by regulation of both apoptosis and cellular regeneration (97). It has been observed that CHOP protein influences gene expression as both a dominant negative regulator of C/EBP binding to one class of DNA targets and positively by directing CHOP-C/ EBP heterodimers to other sequences (98), and this observation may constitute the mechanism of its apparent dichotomous function. hGH may therefore utilize CHOP as a positive regulator of gene transcription with resultant mitogenesis in the cell systems described. It is also interesting that GH itself has also been reported to be both mitogenic and growth-inhibitory depending on the cellular context (1,2).
We have demonstrated here that both p38 MAP kinase and p44/42 MAP kinase cooperate during hGH-stimulated mitogenesis in the CHO-GHR 1-638 cell line. Such cooperation between p38 MAP kinase and P44/42 MAP kinase has also been observed for granulocyte colony-stimulating factor stimulated proliferation of hemopoietic cells (19). Granulocyte colony-stimulating factor is another member of the cytokine receptor superfamily (1-3). Interestingly, however, we did not observe such cooperativity between p38 MAP kinase and p44/42 MAP kinase in mammary carcinoma cells with autocrine hGH production, as the autocrine hGH-stimulated cell proliferation was completely inhibited with either the inhibitor for MEK1 (PD98058) or the inhibitor for p38 MAP kinase (SB203580) (82). The target molecules activated by either p38 MAP kinase or P44/42 MAP kinase responsible for the cooperative effect on proliferation in the CHO-GHR 1-638 cell line are not defined. Presumably, however, they must represent a class of molecules, which are exclusively activated by GH stimulation of either p38 MAP kinase or p44/42 MAP kinase. Such an example may be provided in this paper whereby the GH-stimulated activation of ATF-2 and CHOP were p38 MAP kinase-dependent, whereas GH-stimulated activation of Elk-1 was p44/42 MAP kinase-dependent (73). GH has been reported to utilize Elk-1 to mediate GH-induced transcription of egr-1 (99) which may provide a mechanism for the p44/42 MAP kinase-dependent component of GH-stimulated proliferation. GH is a comparatively weak mitogen, and many effects of GH are exerted on differentiated cell function. GH may therefore also utilize p38 MAP kinase for regulation of differentiation or differentiated cell function. For example, p38 MAP kinase has been reported to be required for erythroid differentiation (100), chondrogenesis (101), and myotube formation (102).
We have demonstrated here that inhibition of p38 MAP kinase prevents hGH-stimulated stress fiber breakdown in the CHO-GHR 1-638 cell line, whereas inhibition of MEK1 was without effect. Previous reports (80,103,104) have demonstrated that modulation of actin dynamics by p38 MAP kinase requires the phosphorylation of HSP27 downstream of MAP-KAP-2 and MAPKAP-3 (80,(103)(104)(105). MAPKAP-2 has previously been reported to be activated by GH (73). It is therefore likely that GH also modulates the phosphorylation state of HSP27, although this has not yet been demonstrated. GHstimulated re-organization of the actin cytoskeleton has been demonstrated to require phosphatidylinositol 3-kinase activity (79), suggesting that phosphatidylinositol 3-kinase may also be upstream of the GH-dependent increase in p38 MAP kinase activity. A previous study has reported that p38 MAP kinase activated by the chemotactic peptide N-formyl-Met-Leu-Phe requires phosphatidylinositol 3-kinase activity (106). Such is the case for hGH stimulation of both p44/42 MAP kinase (73,110) and JNK/SAPK. 3 In summary we have demonstrated here that hGH transiently phosphorylates and activates p38 MAP kinase in CHO cell lines stably transfected with rat GH receptor cDNA and that the activation of the p38 MAP kinase pathway by hGH is JAK2-dependent. Furthermore, both ATF2 and CHOP are transcriptionally activated by hGH in a JAK2-dependent manner, and p38 MAP kinase is required for both cytoskeletal re-organization and cell proliferation stimulated by hGH. Since p38 MAP kinase has also been reported to phosphorylate and activate several other protein kinases, including MNK1, MNK2, MAPKAPK2, MAPKAPK3, MSK1, and PRAK (22, 52, 108 -112), it is likely to be central to the pleiotropic cellular effects of GH.