CrkII Participation in the Cellular Effects of Growth Hormone and Insulin-like Growth Factor-1

We have examined the role of CrkII in the cellular response to both human growth hormone (hGH) and human insulin-like growth factor-1 (hIGF-1). We have demonstrated that overexpression of the adaptor molecule enhances both basal phosphatidylinositol 3-kinase (PI 3-kinase) activity and also dramatically enhances the ability of both hormones to stimulate PI 3-kinase activity in the cell. Many of the effects of CrkII overexpression on hGH- and hIGF-1-stimulated cellular function can then be attributed to CrkII enhancement of PI 3-kinase stimulation by these hormones. Thus, CrkII-enhanced PI 3-kinase activity is used to enhance actin filament reorganization in response to both hGH and hIGF-1, to enhance stress activated protein kinase (SAPK) activity in response to hGH, and to diminish STAT5-mediated transcription in response to hGH. It is apparent, however, that CrkII also regulates cellular function independent of its ability to stimulate PI 3-kinase activity. This is evidenced by the ability of CrkII, in a PI 3-kinase-independent manner, to diminish the activation of p44/42 mitogen-activated protein kinase in response to both hGH and hIGF-1 and to inhibit the activation of SAPK by hIGF-1. Therefore, despite the common use of CrkII to activate PI 3-kinase, CrkII also allows hGH or hIGF-1 to selectively switch the activation of SAPK. Thus, common utilization of CrkII by hGH and hIGF-1 allows the execution of common cellular effects of these hormones, concomitant with the retention of hormonal specificity.

We have examined the role of CrkII in the cellular response to both human growth hormone (hGH) and human insulin-like growth factor-1 (hIGF-1). We have demonstrated that overexpression of the adaptor molecule enhances both basal phosphatidylinositol 3-kinase (PI 3-kinase) activity and also dramatically enhances the ability of both hormones to stimulate PI 3-kinase activity in the cell. Many of the effects of CrkII overexpression on hGH-and hIGF-1-stimulated cellular function can then be attributed to CrkII enhancement of PI 3-kinase stimulation by these hormones. Thus, CrkIIenhanced PI 3-kinase activity is used to enhance actin filament reorganization in response to both hGH and hIGF-1, to enhance stress activated protein kinase (SAPK) activity in response to hGH, and to diminish STAT5-mediated transcription in response to hGH. It is apparent, however, that CrkII also regulates cellular function independent of its ability to stimulate PI 3-kinase activity. This is evidenced by the ability of CrkII, in a PI 3-kinase-independent manner, to diminish the activation of p44/42 mitogen-activated protein kinase in response to both hGH and hIGF-1 and to inhibit the activation of SAPK by hIGF-1. Therefore, despite the common use of CrkII to activate PI 3-kinase, CrkII also allows hGH or hIGF-1 to selectively switch the activation of SAPK. Thus, common utilization of CrkII by hGH and hIGF-1 allows the execution of common cellular effects of these hormones, concomitant with the retention of hormonal specificity.
Growth hormone (GH) 1 is the major regulator of postnatal somatic growth (1). The anabolic and growth-promoting effects of GH are largely thought to be mediated via stimulation of insulin-like growth factor-1 expression in hepatic and extrahepatic tissues (2). Although it is now apparent that GH may act independently of insulin-like growth factor-1 (IGF-1), the two hormones possess multiple common effects on cellular function.
The GH receptor is a member of the cytokine receptor superfamily and therefore mediates tyrosine phosphorylation of cellular proteins by its association with the JAK family of nonreceptor tyrosine kinases (3). In contrast, the IGF-1 receptor is a receptor tyrosine kinase, and ligand stimulation results in direct phosphorylation of signaling molecules by the receptor itself. Despite the difference in which the hormones initiate signal transduction, there exists a common utilization of multiple signal-transducing molecules such as FAK, CrkII, p130 cas , IRS-1, IRS-2, p44/42 MAP kinase, and PI 3-kinase, which may provide the basis for the shared cellular function (4 -10).
We have previously demonstrated that human growth hormone (hGH) stimulates the formation of a large multiprotein signaling complex centered around CrkII and p130 cas . CrkII is a member of a family of adaptor proteins predominantly composed of Src homology 2 and 3 domains whose role in signaling pathways is presently unclear, although it has been implicated in various cellular functions such as cytoskeletal reorganization and mitogenesis (11,12). Similarly, the structure of p130 cas suggests that it is a docking protein with a role in multiprotein signaling complexes (13). Other components of the hGH-stimulated complex include c-Src, c-Fyn, c-Cbl, Nck, FAK, paxillin, IRS-1, C3G, SHC, Grb-2, and Sos-1 (6). Here we have examined the role of CrkII in the cellular responses to both GH and IGF-1 and demonstrate that the adaptor molecule acts as both a point of convergence and divergence in their cellular effects. It appears that many of the effects of CrkII on the cellular functions of hGH and hIGF-1 lie with its ability to dramatically enhance PI 3-kinase activity stimulated by these hormones. Despite the common use of CrkII to activate PI 3-kinase, CrkII also allows these hormones to selectively switch the activation of SAPK. Thus, common utilization of CrkII by hGH and hIGF-1 allows the execution of the common cellular effects of these hormones, concomitant with the retention of hormonal specificity.

EXPERIMENTAL PROCEDURES
Materials-Recombinant hGH was a generous gift of Novo-Nordisk (Singapore). Sodium orthovanadate, PMSF, phalloidin-TRITC, wortmannin, and common chemicals and reagents were purchased from Sigma. Polyclonal antisera against the p85 regulatory subunit of PI 3-kinase was purchased from Upstate Biotechnology, Inc. (Lake Placid, NY). CrkII monoclonal antibody and anti-phosphotyrosine monoclonal antibody (PY20) were from Transduction Laboratories (Lexington, KY). Anti-mouse IgG conjugated to TRITC and anti-rabbit IgG conjugated to FITC were purchased from Roche Molecular Biochemicals. Secondary anti-IgG antibodies and the enhanced chemiluminescence (ECL) kit were from Amersham Pharmacia Biotech.
Treatment of Cells with hGH and hIGF-1-NIH3T3 cells were grown in medium containing 10% fetal calf serum for 48 h before changing to serum-free medium (serum deprivation) for 12-15h. Serum-deprived cells were treated with 50 nM hGH or 100 nM hIGF-1 for the indicated time periods.
SDS-Polyacrylamide Gel Electrophoresis, and Western Blot Analysis-Cells were grown as described above and lysed in immunoprecipitation buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris, pH 7.4, 1 mM EDTA, 1 mM EGTA, 1% Nonidet P-40, 0.2 mM Na 3 VO 4 , 1 g/ml protease inhibitor mixture, and 0.1 mM PMSF). SDS-polyacrylamide gel electrophoresis sample buffer (50 mM Tris (pH 6.8), 2% SDS, 2% ␤-mercaptoethanol, and bromphenol blue) was added to each sample, and the samples were boiled for 5 min. Samples were subjected to discontinuous SDS-polyacrylamide gel electrophoresis with a 10% resolving gel and transferred to nitrocellulose membranes (Hybond C-extra) using standard electroblotting procedures. Membranes were blocked with 2% bovine serum albumin overnight at 4°C and immunolabeled with monoclonal antibody against STAT5b (Upstate Biotechnology) and phosphospecific STAT5a/b antibody (New England Biolabs (Beverly, MA)) or CokII monoclonal antibody (Transduction Laboratories) for 1 h at room temperature. Immunolabeling was detected by the enhanced chemiluminescence kit according to the manufacturer's instructions.
Confocal Laser-scanning Microscopy-At the end of the respective treatment period, cells were rinsed with ice-cold phosphate-buffered saline (PBS), fixed in ice-cold 4% paraformaldehyde, permeabilized for 10 min with 0.1% Triton X-100, blocked in 2% bovine serum albumin, and incubated with a polyclonal antibody against PI 3-kinase or CrkII followed by anti-rabbit IgG or anti-mouse IgG conjugated to FITC at room temperature and phalloidin-TRITC (0.2 mg/ml). 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 Adobe Photoshop program.
Phosphotidylinositol 3-Kinase Assay-NIH3T3 cells were grown to about 80% confluence, incubated for 15 h in serum-free medium, washed once in serum-free medium, and incubated with 50 nM hGH or 100 nM hIGF-1 for the indicated time periods. Cells were lysed with lysis buffer (10 mM Tris-HCl, pH 7.4, 0.25 M sucrose, 1 mM EDTA, 0.1 mM sodium orthovanadate (Na 3 VO 4 ), 0.1 mM diisopropyl fluorophos-FIG. 1. CrkII overexpression in NIH3T3 cells. A, confocal laserscanning photomicrographs of CrkII expression in NIH3T3 cells stably transfected with either pCXN2-CrkII (NIH3T3-CrkII) or pCNX2 (NIH3T3-vector). NIH3T3 cells were grown on coverslips, serum-deprived for 15 h, fixed with 4% paraformaldehyde, and processed for confocal laser-scanning microscopy using monoclonal antibody against CrkII followed by anti-mouse antibody conjugated with FITC (see "Experimental Procedures"). Bar, 10 m. The photomicrographs presented are representative of at least three independently performed experiments. B, Western blot analysis of NIH3T3-vector (stably transfected with empty pCXN2 vector) and NIH3T3-CrkII (stably transfected with pCXN2-CrkII) cells. Cells were grown to confluence and serum-deprived, and total cell extracts were prepared and processed as described under "Experimental Procedures." Immunolabeling was detected by the enhanced chemiluminescence method. The blot presented is representative of at least three independently performed experiments.

FIG. 2. Effect of CrkII overexpression on hGH-and hIGF-1stimulated PI 3-kinase activity.
A, autoradiograph demonstrating formation of PIP by PI 3-kinase activity upon hGH stimulation of NIH3T3 cells stably transfected with either CrkII cDNA (NIH3T3-CrkII) or the empty vector (NIH3T3-vector). The autoradiograph is representative of at least three independent experiments. B, quantitative analysis by densitometry of the level of PI 3-kinase activity upon hGH stimulation of NIH3T3 cells stably transfected with either CrkII cDNA (NIH3T3-CrkII) or the empty vector (NIH3T3-vector). The graph is representative of at least three experiments. The level of PI 3-kinase activity observed in unstimulated NIH3T3-vector cells has been arbitrarily assigned a value of 1, and the other values are plotted as the -fold increase over the unstimulated value resulting in the bar chart. C, autoradiograph demonstrating formation of PIP by PI 3-kinase upon hIGF-1 stimulation of NIH3T3 cells stably transfected with either CrkII cDNA (NIH3T3-CrkII) or the empty vector (NIH3T3-vector). The autoradiograph is representative of at least three independent experiments. D, quantitative analysis by densitometry of the level of PI 3-kinase activity upon hIGF-1 stimulation of NIH3T3 cells stably transfected with either CrkII cDNA (NIH3T3-CrkII) or the empty vector (NIH3T3-vector). The graph is representative of at least three experiments. The level of PI 3-kinase activity observed in unstimulated NIH3T3-vector cells has been arbitrarily assigned a value of 1, and the other values are plotted as the fold increase over the unstimulated value resulting in the bar chart.
phate, 1% Triton X-100, 10 g/ml protease inhibitor mixture, and 0.1 mM PMSF) for 1 h at 4°C, scraped, and spun for 10 min at 4°C. The supernatants were collected, and the proteins in the whole cell extracts were assayed using a Bio-Rad protein assay kit with bovine serum albumin as standard. 800 g of each sample was immunoprecipitated overnight with 3 g of polyclonal antibody against phosphotidylinositol 3-kinase. 25 l of protein G plus/protein A-agarose (Calbiochem) was added for 2 h, and the agarose beads were washed twice with lysis buffer, twice with PBS, and twice with assay buffer (40 mM Tris-HCl, pH 7.4, 5 mM MgCl 2 , and 0.5 mM EGTA). A 50-l assay mix (0.01 mg/ml phosphatidylserine, 0.2 mg/ml phosphatidylinositol, and 0. p44/42 MAP Kinase and SAPK Assays-p44/42 MAP kinase assays were performed using the New England Biolabs assay kit according to the manufacturer's instructions. In brief, cells were serum-deprived for 16 h, treated with 50 nM hGH or 100 nM hIGF-1, and lysed at 4°C in 1 ml of lysis buffer (20 mM Tris, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM glycerolphosphate, 1 mM Na 3 VO 4 , 0.1% PMSF, 1 g/ml leupeptin) 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 a phosphospecific p44/42 MAP kinase (Thr 202 / Tyr 204 ) monoclonal antibody (1:100 dilution) in a final volume of 500 l in 1ϫ lysis buffer. Protein G plus/protein A-agarose beads were added, and the mixture was incubated with gentle rocking for 13 h at 4°C. The pellets were washed twice with 500 l of lysis buffer containing 0.1% PMSF and twice with 500 l of kinase buffer (25 mM Tris, pH 7.5, 5 mM glycerolphosphate, 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 Elk1 fusion protein and 200 M ATP at 30°C for 30 min. Elk1 phosphorylation was selectively detected by Western immunoblotting using a chemiluminescence detection system and a specific phospho-Elk1 (Ser 383 ) antibody (1:1000 dilution). c-Jun N-terminal kinase/SAPK assays were performed using the New England Biolabs assay kit according to the manufacturer's instructions. In brief, cells were serum-deprived for 16 h, treated with 50 nM hGH or 100 nM hIGF-1, and lysed at 4°C in 1 ml of lysis buffer (20 mM Tris, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM glycerolphosphate, 1 mM Na 3 VO 4 , 0.1% PMSF, 1 g/ml leupeptin) 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 2 g of N-terminal c-Jun-(1-89) fusion protein bound to glutathione-Sepharose beads in a final volume of 500 l in 1ϫ lysis buffer. The beads were washed twice with 500 l of lysis buffer containing 0.1% PMSF and twice with 500 l of kinase buffer (25 mM Tris, pH 7.5, 5 mM glycerolphosphate, 2 mM dithiothreitol, 0.1 mM Na 3 VO 4 , 10 mM MgCl 2 ). The kinase reactions were carried out in the presence of 100 M ATP at 30°C for 30 min. c-Jun phosphorylation was selectively detected by Western immunoblotting using a chemiluminescence detection system and specific c-Jun antibodies (1:1000 dilution) recognizing c-Jun phosphorylated at serine 63.
Densitometric Analysis of Band Intensities-The intensity of the respective bands on the blots were quantitated using a Bio-Rad GS-700 imaging densitometer and analyzed with the Multi-Analyst (version 1.0.1) program (Bio-Rad).
Transient Transfection and Reporter Assay-NIH3T3 cells were cultured to 50% confluence in six-well plates. Transient transfection was performed in serum-free Dulbecco's modified Eagle's medium with N-[1-(2, 3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methylsulfate according to the manufacturer's instructions (Roche Molecular Biochemicals). 1.0 g of reporter plasmid (SPI-GLE1-CAT/SIE c-fos-CAT/TK-CAT) were transfected per well in serum-free Dulbecco's mod-ified Eagle's medium. Cells were incubated with N-[1-(2, 3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methylsulfate/DNA for 12 h before the medium was changed to fresh serum free Dulbecco's modified Eagle's medium with or without 50 nM hGH. After a further 24 h, cells were washed in PBS and scraped into lysis buffer (250 mM Tris-HCl, pH 8.0, 1 mM DTT). The protein content of the samples was normalized, and chloramphenicol acetyltransferase assays were performed. Results were normalized to the level of luciferase activity to control for transfection efficiency and calculated as the -fold stimulation of unstimulated (non-hormone-treated) cells.
Statistical Analysis and Presentation of Data-All experiments were performed at least three times. Numerical data are expressed as mean Ϯ S.D. Data were analyzed using the two-tailed t test or analysis of variance.

Characterization of CrkII-overexpressing NIH3T3 Cells-We
have previously demonstrated that GH stimulates the tyrosine phosphorylation of CrkII and p130 cas and their association with multiple known signaling molecules (4). To determine the role of Crk-II in the cellular effects of GH, we utilized NIH3T3 cells stably transfected with either pCXN2-CrkII (NIH3T3-CrkII) or with pCXN2 vector alone (NIH3T3-vector). NIH3T3-CrkII cells exhibited overexpression of CrkII as demonstrated by both Western blot analysis and immunofluorescence with CrkII antibody (Fig. 1). The level of Crk-II expression in NIH3T3-CrkII cells is approximately 10 times that of vectortransfected control cells (NIH3T3-vector) as determined by densitometric analysis of the Western blot for CrkII. Confocal laser-scanning microscopic analysis of CrkII expression in NIH3T3-CrkII and NIH3T3-vector cells also demonstrated overexpression of CrkII in NIH3T3-CrkII cells compared with vector-transfected control cells. CrkII is localized to both the cytoplasm and nucleus in both vector-transfected and CrkIIoverexpressing NIH3T3 cells. The two cell lines possess comparable levels of both GH and IGF-1 receptor levels at the cell surface (5).
CrkII Overexpression Enhances hGH and hIGF-1 Stimulation of PI 3-Kinase Activity-Several molecules stimulated by GH to associate with the p130 cas -CrkII complex have been reported to directly associate with the p85 regulatory subunit of PI 3-kinase including FAK (14), tensin (15), c-Cbl (16), and IRS-1 (17). We therefore examined the effect of CrkII overexpression on hGH-stimulated PI 3-kinase activity. For comparison, we also examined the cellular responses of these cells to hIGF-1. The IGF-1 receptor kinase has been demonstrated to directly phosphorylate CrkII (5). NIH3T3-CrkII cells exhibited a higher level of PI 3-kinase activity in serum-free conditions compared with vector transfected controls (Fig. 2, A and B). The relative level of hGH-stimulated PI 3-kinase activity was also dramatically increased at 5 min compared with vectortransfected control cells. Similarly for IGF-1, we observed a marked enhancement of hIGF-1-stimulated PI 3-kinase activity in NIH3T3-CrkII cells compared with vector transfected control cells (Fig. 2, C and D). Interestingly, CrkII overexpression appeared to preferentially enhance hIGF-1-stimulated, compared with hGH-stimulated, PI 3-kinase activity. Prior treatment of cells with 50 nM wortmannin or 20 M LY294002 abolished both the hGH and hIGF-1-stimulated increases in PI 3-kinase activity in both NIH3T3-CrkII and NIH3T3-vector cells (data not shown).
CrkII Enhances Actin Filament Reorganization in Response to hGH and hIGF-1-PI 3-kinase is pivotal in mediating many of the pleiotropic effects of growth factors and hormones on cellular function. One such effect is both hGH-and hIGF-1stimulated reorganization of the actin cytoskeleton (18). Treatment of cells with either hGH or hIGF-1 has been reported to result in an initial depolymerization of filamentous actin resulting in stress fiber breakdown, followed by a later repoly-FIG. 5. Localization of CrkII and the p85 regulatory subunit of PI 3-kinase in membrane ruffles after hGH or hIGF-1 stimulation. NIH3T3-CrkII cells were grown on coverslips, serum-deprived for 15 h, and stimulated with 50 nM hGH (A) or 100 nM hIGF-1 (B) for 30 min. Cells were fixed with 4% paraformaldehyde-PBS and processed for confocal laser-scanning microscopy. Phalloidin-TRITC was used for visualization of F-actin (left panel) and a rabbit polyclonal antibody against PI 3-kinase or a monoclonal antibody against CrkII followed by FITC-conjugated anti-rabbit or anti-mouse IgG, respectively. Bar, 10 m (right panels as indicated). merization resulting in the formation of focal filamentous actin-containing complexes (18). We therefore examined the response of the actin cytoskeleton to hGH and hIGF-1 treatment in NIH3T3-CrkII cells and compared this response with that in NIH3T3-vector cells (Fig. 3). In the unstimulated state, NIH3T3-vector cells display an elongated morphology with stress fibers traversing the cell. NIH3T3-CrkII cells possessed similar morphology, although in the unstimulated state these CrkII-overexpressing cells possessed small focal complexes of filamentous actin resembling rudimentary membrane ruffles. This was presumably due to the higher level of PI 3-kinase activity present in these cells. Upon hGH stimulation of NIH3T3-vector cells, there was minimal stress fiber breakdown at 4 min and minimal formation of membrane ruffles at 30 min. In contrast, with hGH stimulation of NIH3T3-CrkII cells there was substantial stress fiber breakdown at 4 min accompanied by the early formation of membrane ruffles and marked formation of membrane ruffles 30 min after stimulation with hGH. hIGF-1 stimulation of either cell line produced results similar to those observed with hGH (Fig. 4). The enhanced stress fiber breakdown and membrane ruffle formation observed in NIH3T3-CrkII cells after both hGH and IGF-1 stimulation were abrogated by prior exposure of the cells to 50 nM wortmannin or 20 M LY294002 (data not shown). Thus, overexpression of CrkII dramatically enhances hGH-and hIGF-1stimulated reorganization of the actin cytoskeleton in a PI 3-kinase-dependent manner. To investigate the possible mechanism whereby CrkII enhances PI 3-kinase-dependent reorganization of the actin cytoskeleton stimulated by hGH and hIGF-1, we examined the subcellular localization of both CrkII and the p85 regulatory subunit of PI 3-kinase. We observed that both CrkII and the p85 regulatory subunit of PI 3-kinase are colocalized with filamentous actin in the membrane ruffles formed by cellular stimulation with either hGH (Fig. 5A) or hIGF-1 (Fig. 5B). Thus, both CrkII and the p85 subunit of PI 3-kinase possess a spatially appropriate localization for involvement in membrane ruffle formation after stimulation with either hGH or hIGF-1.
CrkII Serves as a Switch for the Selective Activation of p44/42 MAP Kinase or SAPK by hGH and hIGF-1-GH stim-ulation of p44/42 MAP kinase activity has also been reported to be PI 3-kinase-dependent (19) and thus may be affected by cellular overexpression of CrkII. We first determined whether hGH-or hIGF-1-stimulated p44/42 MAP kinase activity in NIH3T3 cells was also dependent on PI 3-kinase activity (Fig.  6). Both hGH and hIGF-1 treatment of NIH3T3-vector cells resulted in marked stimulation of p44/42 MAP kinase activity (Fig. 6). Pretreatment of NIH3T3-vector cells with 50 nM wortmannin (or 20 M LY294002) partially inhibited the hGH-and hIGF-1-stimulated increase in p44/42 MAP kinase activity. Surprisingly, both hGH and hIGF-1 stimulation of NIH3T3-CrkII cells resulted in a dramatically lower stimulation of p44/42 MAP kinase activity compared with NIH3T3-vector cells stimulated with hGH or hIGF-1 (Fig. 7). Pretreatment of NIH3T3-CrkII cells with 50 nM wortmannin (or 20 M LY294002) further diminished both hGH-and hIGF-1-stimulated p44/42 MAP kinase activity. Thus, overexpression of CrkII results in inhibition of both hGH-and hIGF-1-stimulated p44/42 MAP kinase activity despite the ability of CrkII to enhance both hGH-and hIGF-1-stimulated PI 3-kinase activity.
We have previously reported that hGH activation of SAPK activity is CrkII-dependent (6). We therefore also examined whether hGH stimulation of SAPK activity was PI 3-kinasedependent in light of the above demonstration that hGH-stimulated PI 3-kinase activity is CrkII-dependent. As shown in Fig. 8, CrkII overexpression in NIH3T3-CrkII cells resulted in larger -fold increases in hGH stimulation of SAPK activity despite an increase in the basal activity of SAPK in these cells. Pretreatment of both NIH3T3-vector and NIH3T3-CrkII cells with 50 nM wortmannin inhibited the hGH-stimulated increase in SAPK activity. We also examined the effect of CrkII overexpression on the ability of hIGF-1 to stimulate SAPK activity. hIGF-1 stimulation of NIH3T3-vector cells resulted in activation of SAPK to a level similar to that observed with 50 nM hGH (Fig. 9). However, hIGF-1 stimulation of NIH3T3-CrkII cells resulted in a decrease in SAPK activity to below that of the basal level. Although 50 nM wortmannin prevented the increase in SAPK activity in NIH3T3-vector cells, it was without effect on the hIGF-1-dependent decrease in SAPK activity ob- CrkII Overexpression Diminishes hGH-stimulated STAT5mediated Transcription in a PI 3-Kinase-dependent Manner-One of the major pathways by which hGH regulates transcriptional activity is by the use of STAT1, -3, or -5 (20). hGH and hIGF-1 treatment of NIH3T3-vector cells do not result in increased transcription through STAT1 or -3. hGH but not hIGF-1 stimulates STAT5-mediated transcription through the ␥-activated sequence like element of the serine protease inhibitor 2.1 gene promoter (SPI-GLE1) in NIH3T3-vector cells (Fig.  10A). To determine the role of PI 3-kinase in hGH-stimulated STAT5-mediated transcription, we pretreated NIH3T3-vector cells with LY294002. Surprisingly, inhibition of PI 3-kinase with LY294002 increased hGH-stimulated STAT5-mediated transcription in a dose-dependent manner (Fig. 10B), indicat-ing that PI 3-kinase possesses a negative regulatory role for hGH-stimulated STAT5-mediated transcription. LY294002 had no effect on the lack of stimulation of STAT5-mediated transcription observed for hIGF-1 or on the lack of hGH-stimulated STAT1/3-mediated transcription (data not shown). CrkII overexpression in NIH3T3-CrkII cells diminished the hGH stimulation of STAT5-mediated transcription compared with NIH3T3-vector cells (Fig. 10C). Similar reduction in hGHstimulated STAT5-mediated transcription was also observed upon transient transfection of CrkII cDNA in NIH3T3-vector cells (data not shown). To determine whether the hGH-stimulated CrkII-enhanced increase in PI 3-kinase activity was responsible for the diminished STAT5-mediated transcription, we pretreated NIH3T3-CrkII cells with LY294002. Pretreatment of the CrkII-overexpressing cells with LY294002 restored hGH-stimulated STAT5-mediated transcription to levels observed in NIH3T3-vector cells. We also examined the effect of CrkII overexpression with and without LY294002 treatment on the tyrosine phosphorylation of STAT5 stimulated by hGH. As is observed in Fig. 10D, neither CrkII overexpression nor treatment with LY294002 altered the ability of hGH to stimulate STAT5 tyrosine phosphorylation. Thus, CrkII overexpression in NIH3T3 cells diminishes hGH-stimulated STAT5-mediated transcription in a PI 3-kinase-dependent manner without alteration of STAT5 tyrosine phosphorylation. DISCUSSION We have demonstrated here that the CrkII adaptor molecule regulates various hGH and hIGF-1 signaling pathways. We have identified CrkII as a powerful positive regulator of PI 3-kinase activity stimulated by both hGH and hIGF-1. The mechanism by which CrkII regulates PI 3-kinase activity is not clear, although we have previously demonstrated that both CrkII and the p85 regulatory subunit of PI 3-kinase co-exist in a multiprotein signaling complex stimulated by hGH (6). Other components of this complex include p130 cas , c-Src, c-Fyn, c-Cbl, Nck, FAK, paxillin, IRS-1, C3G, SHC, Grb-2, and Sos-1. One possibility is that CrkII simply acts as an adaptor molecule to facilitate the interactions between components of the hGHstimulated complex regulating PI 3-kinase activity. In this regard, it is interesting that c-Fyn has been reported to phosphorylate tyrosine 731 of c-Cbl, which is required for the binding of the p85 subunit of PI 3-kinase to c-Cbl (21). Similarly, IRS-1 has been reported to be a substrate for FAK, and FAK overexpression has been reported to increase IRS-1 tyrosine phosphorylation and the association between IRS-1 and the p85 subunit of PI 3-kinase (22). Activation of PI 3-kinase by the related hormone prolactin has been demonstrated to require c-Fyn (23), although c-Fyn and the p85 subunit of PI 3-kinase may associate directly (24). The p85 subunit of PI 3-kinase has also been reported to associate with FAK either directly or indirectly, and FAK has been demonstrated to phosphorylate p85 in vitro (25). In any case, the majority of the effects of CrkII overexpression on hGH-and hIGF-1-stimulated cellular function can be attributed to CrkII enhancement of PI 3-kinase stimulation by these hormones. It is apparent, however, that CrkII also regulates cellular function independent of its ability to stimulate PI 3-kinase activity as evidenced by the ability of CrkII to diminish the activation of p44/42 MAP kinase in response to both hGH and hIGF-1 in a PI 3-kinase-independent manner.
We have previously demonstrated that actin cytoskeletal reorganization stimulated by both hGH and hIGF-1 is PI 3kinase-dependent (18). Thus, CrkII-enhanced stimulation of PI 3-kinase activity by both hormones resulted in increased formation of membrane ruffles in these cells (this study); since the effect of CrkII overexpression on cytoskeletal reorganization could be inhibited with PI 3-kinase inhibitors. These results are in agreement with a recent report in which microinjection of CrkII antibody into cells prevented membrane ruffle formation induced by IGF-1 (12). CrkII overexpression in fibroblasts has also been demonstrated to enhance p130 cas phosphorylation and Rac-dependent cell migration (11), and the formation of a p130 cas -CrkII complex has been demonstrated to be suffi-cient for cell migration. We have also demonstrated that CrkII possessed the appropriate spatial distribution within the cell to mediate membrane ruffle formation, with both CrkII and the p85 regulatory subunit of PI 3-kinase localized to the membrane ruffles. It has also been previously demonstrated that CrkII is localized to the membrane ruffles of cells induced by attachment of the cells to vitronectin matrix or upon insulin treatment (11). The association and functional interactions between PI 3-kinase and the mediators of actin rearrangements, the small GTPases Rho and Rac, is well established (26). Recently, CrkII has been shown to regulate the Rho signaling pathway (11,(27)(28)(29)(30)(31)(32)(33). In addition, DOCK180, a CrkII Src homology 3-binding protein, was shown to participate in the activation of Rac1 in integrin signaling (34).
We have demonstrated here that overexpression of CrkII enhances the ability of hGH to stimulate SAPK activity as previously published (6) and in accord with the demonstration that a CrkII-C3G complex activates SAPK through a pathway involving the mixed lineage kinase family of proteins (35). Interestingly, however, we observe that CrkII overexpression diminishes the hGH stimulation of p44/42 MAP kinase activity. We have previously reported that C3G is present in the hGHstimulated multiprotein signaling complex centered around CrkII, and we have also observed hGH-stimulated tyrosine phosphorylation of C3G. 2 C3G appears to function as a specific guanine nucleotide exchange factor for Rap1 (36), and membrane targeting of Crk enhances the Rap1 guanine nucleotide exchange activity of C3G (37). It has been reported that insulin stimulation of cells results in a rapid disassociation of the CrkII-C3G complex (38) with subsequent inhibition of the Rap1-Raf1 interaction. These authors postulate that uncoupling of the CrkII-C3G complex by insulin de-represses Rap1 function, thereby releasing Raf1 for activation by Ras, with resultant MEK activation (38). In contrast, since CrkII overexpression inhibited hGH stimulation of p44/42 MAP kinase, hGH stimulation of cells would be expected to activate Rap1 in a CrkII-dependent manner, and this is indeed what we find. 3 The effect of hIGF-1 on the CrkII-C3G complex has not been reported, but it should not be expected to be similar to insulin, since IGF-1 and insulin differentially regulate CrkII-associated proteins. One example is that insulin stimulates the dephosphorylation of p130 cas , whereas cellular stimulation with IGF-1 results in increased p130 cas phosphorylation (7). One difference that has been reported for hGH and hIGF-1 is that hGH stimulates the association between CrkII and IRS-1 (6), whereas IGF-1 stimulates the disassociation between CrkII and IRS-1 (39). In any case, although both hGH and IGF-1 stimulate the phosphorylation of CrkII, it is differentially utilized for regulation of SAPK but not p44/42 MAP kinase activities. Further studies are required to identify the mechanisms involved in the selective activation of p44/42 MAP kinase or SAPK by these hormones.
We have demonstrated in this paper that overexpression of CrkII diminishes hGH stimulation of STAT5-mediated tran-2 T. Zhu and P. E. Lobie, unpublished observations. 3 T. Zhu, and P. E. Lobie, manuscript in preparation.
(Spi2.1-GLE1-CAT)-mediated transcription in NIH3T3-vector and NIH3T3-CrkII cells. NIH3T3-vector and NIH3T3-CrkII cells were grown to confluence and transiently transfected with Spi2.1-GLE1-CAT. Cells were exposed to 50 M LY294002 as indicated for 1 h prior to stimulation with 50 nM hGH and processed for CAT assay as described under "Experimental Procedures." The data are presented as the mean Ϯ S.D. of triplicate determinations. The data presented are representative of at least three independently performed experiments. D, effect of hGH on STAT5 tyrosine phosphorylation in NIH3T3-vector and NIH3T3-CrkII cells. NIH3T3-vector and NIH3T3-CrkII cells were grown to confluence. Cells were exposed to 50 M LY294002 as indicated for 1 h prior to stimulation with 50 nM hGH and processed for Western blot analysis as described under "Experimental Procedures." Membranes were first blotted with a monoclonal antibody recognizing phosphorylated STAT5, stripped, and reblotted to demonstrate equal loading of STAT5 protein. The data presented are representative of at least three independently performed experiments. scription in a PI 3-kinase-dependent manner without alteration of hGH-stimulated STAT5 tyrosine phosphorylation. This suggests that CrkII regulation of STAT5 transcription occurs distally. It has been demonstrated that type I interferons induced the formation of a CrkL-STAT5 complex that translocates to the nucleus, and CrkL, in this case, modulates STAT5mediated transcription directly at the level of DNA binding (40). A similar direct association between phosphorylated STAT5 and CrkL has been demonstrated upon stimulation of the EPO receptor (41). Thus, it is likely that CrkII also associates with phosphorylated STAT5 and prevents its binding to DNA and therefore transcriptional activation. It is interesting that c-Cbl, which is also found in the multiprotein signaling complex stimulated by hGH, decreases hGH stimulated STAT5-mediated transcription by inhibition of STAT5 phosphorylation. 4 Thus, two adaptor molecules present in the hGH stimulated multiprotein complex act to inhibit STAT5-mediated transcription at either proximal or distal points in the signal transduction pathway. At the present time, it is not clear why CrkII inhibition of STAT5-mediated transcription is PI 3-kinase-dependent. It is interesting to note that the activation of phospholipase C is PI 3-kinase dependent and pharmacological inhibition of phospholipase C also increases GH-stimulated STAT5-mediated transcription (42) by maintaining the activation (or delaying the deactivation) of STAT5. It remains to be determined how PI 3-kinase negatively regulates STAT5-mediated transcription.
In conclusion, we have examined the role of CrkII in the cellular response to both GH and IGF-1 and demonstrate that the adaptor molecule acts as both a point of convergence and divergence in their cellular effects. It appears that many of the effects of CrkII on the cellular function of hGH and hIGF-1 lie with its ability to dramatically enhance PI 3-kinase activity stimulated by these hormones. Despite the common use of CrkII to activate PI 3-kinase, CrkII also allows these hormones to selectively switch the activation of SAPK. Thus, common utilization of CrkII by hGH and hIGF-1 allows the execution of the common cellular effects of these hormones, concomitant with the retention of hormonal specificity.