Insulin-like growth factor I-stimulated melanoma cell migration requires phosphoinositide 3-kinase but not extracellular-regulated kinase activation

Available online at www.sciencedirect.com R

Experimental Cell Research 286 (2003) 128 –137

www.elsevier.com/locate/yexcr

Insulin-like growth factor I-stimulated melanoma cell migration requires phosphoinositide 3-kinase but not extracellular-regulated kinase activation Cheryl L. Neudauer and James B. McCarthy* Department of Laboratory Medicine and Pathology, University of Minnesota, MMC 609, 420 Delaware Street SE, Minneapolis, MN 55455, USA Received 1 August 2002, revised version received 9 December 2002

Abstract Dysregulated signaling contributes to altered cellular growth, motility, and survival during cancer progression. We have evaluated the ability of several factors to stimulate migration in WM1341D, a cell line derived from an invasive human vertical growth phase melanoma. Basic fibroblast growth factor, hepatocyte growth factor, interleukin-8, and CCL27 each slightly increased migration. Insulin-like growth factor I (IGF-I), however, stimulated a 15-fold increase in migration. This response required the IGF-I receptor, which activates phosphoinositide 3-kinase (PI3K) and mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK) pathways. Both pathways have been implicated in migration in a variety of cell types, but the signaling required for IGF-I-induced melanoma cell migration is not well defined. IGF-I-stimulated activation of MAPK/ERK signaling in WM1341D cells was inhibited by U0126, but a 33-fold higher dose of U0126 was needed to inhibit IGF-I-stimulated cellular migration. In contrast, similar concentrations of either wortmannin or LY294002 were required to inhibit both IGF-I-induced PI3K activation and migration. These results indicate that IGF-I-stimulated migration of WM1341D cells requires PI3K activation but is independent of MAPK/ERK signaling. Determining the contributions of IGF-I signaling pathways to migration will help us to understand melanoma progression and may lead to new therapeutic targets of this highly metastatic cancer. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Melanoma; Migration; Insulin-like growth factor I; Signaling; Phosphoinoinositide 3-kinase; Mitogen-activated protein kinase/extracellularregulated kinase

Introduction Despite many recent advances, the mechanisms regulating melanoma progression are not completely understood. In certain patient subgroups, primary melanomas are thought to progress from dysplastic nevi to radial growth phase (RGP)1 to vertical growth phase (VGP) tumors. Fur* Corresponding author. Fax: ⫹612-625-1121. E-mail address: [email protected] (J.B. McCarthy). 1 Abbreviations used: bFGF, basic fibroblast growth factor; BSA, bovine serum albumin; ddH2O, distilled deionized water; DMSO, dimethyl sulfoxide; ERK, extracellular signal-regulated kinase; HGF, hepatocyte growth factor; IC50, inhibitory concentration 50%; IGF, insulin-like growth factor; IGF-IR, IGF-I receptor; IL-8, interleukin-8, MAPK, mitogen-activated protein kinase; MEK, MAPK kinase; PBS, phosphate-buffered saline; PI3K, phosphoinositide 3-kinase; RGP, radial growth phase; SDS–

ther progression of VGP cells leads to metastases, which are refractory to current therapeutic strategies. Early RGP primary tumors are confined mainly to the epidermis with microinvasion into the dermis. RGP tumors are poorly tumorigenic and are thought to be nonmetastatic. Advanced VGP primary tumor cells have increased motility, invade into the dermis and subcutis, and migrate to the lymphatics and capillaries. VGP tumors are tumorigenic, metastatic, and can be lethal [1– 4]. The transition from a thin, potentially curable melanoma to a thick invasive melanoma with poor prognosis is thought to correlate with the transition from RGP to VGP and a more migratory and invasive PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; SIFM, serum- and insulin-free medium; TBST, Tris-buffered saline with Tween 20; VGP, vertical growth phase.

0014-4827/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0014-4827(03)00049-1

C.L. Neudauer, J.B. McCarthy / Experimental Cell Research 286 (2003) 128 –137

phenotype. Therefore, increased invasive potential of melanoma cells represents an important biological and clinical step and underlies the importance of identifying factors that lead to the migration and invasion of melanoma cells. In melanoma, growth factors and chemokines are important for progression and pathogenesis [5,6]. Dysregulated growth factor and chemokine signaling contributes to transformation, tumorigenesis, and progression in cancer [7]. Many of the growth factors and chemokines upregulated in tumors stimulate motility [5]. In the tumor stroma, fibroblasts produce a variety of factors including basic fibroblast growth factor (bFGF), hepatocyte growth factor (HGF), interleukin-8 (IL-8), and insulin-like growth factor I (IGF-I) [5]. In melanoma, bFGF stimulates migration, growth, and tumorigenicity [8]. In other cancers, bFGF has been reported to promote growth, metastasis, and angiogenesis [7]. HGF can also stimulate melanoma cell migration [9]. In other cell types, HGF has been shown to disrupt cell– cell associations, help cells survive anoikis [10], and enhance mitogenesis, migration, and invasion [7]. IL-8 was originally characterized as a neutrophil chemoattractant, and it induces many proinflammatory effects. IL-8 is also important in melanoma metatasis and progression. It can stimulate angiogensis and proliferation, tumorigenesis, and migration of melanoma cells [6,11,12]. IGF-I can stimulate mitogenesis and migration, protect cells from apoptosis, regulate cellular adhesion, induce and maintain a transformed phenotype, and induce terminal differentiation in a variety of cell types [13]. In melanoma, IGF-I can stimulate mitogenesis [14,15] and migration [16 –18]. CCL27 is a skin-associated chemokine that induces chemotaxis of skin-homing T cells and is important for T cell-mediated inflammation of the skin [19]. CCR10, the receptor for CCL27, is expressed by many melanoma cell lines [20], but the function of CCL27 in melanoma is unknown. Here, we screened these growth factors and chemokines for their ability to stimulate the migration of WM1341D, a cell line derived from a human, primary VGP melanoma. Although all ligands stimulated a slight increase in migration, IGF-I stimulated, on average, a 15-fold increase in migration, which was dependent on the IGF-I receptor (IGF-IR). IGF-IR is a receptor tyrosine kinase that is activated by IGF-I and IGF-II and, to a lesser extent, insulin. Extracellular binding of ligand to the receptor leads to activation of two well-defined pathways: phosphoinositide 3-kinase (PI3K) and mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK). PI3K is a protein and lipid kinase that activates many downstream signaling pathways. PI3K activation can result in gene transcription, protein synthesis, mitogenesis, and inhibition of apoptosis. In the MAPK/ERK pathway, the MAPK kinases MEK1 and MEK2 are activated downstream of Ras by MEK kinases. ERK1 and ERK2 are then specifically phosphorylated by MEK1 and MEK2. The MAPK/ERK pathway can regulate gene transcription, mitogenesis, and cell survival [21]. Both the PI3K and MAPK/ERK signaling pathways have been

129

implicated in cell migration. PI3K helps define the leading edge of the cell and induces polarized migration through amplification of signaling asymmetry as localized stimulation of PI3K leads to localized activation of its downstream effectors. PI3K also regulates F-actin assembly during chemotaxis [22]. Integrins and growth factors coordinately stimulate the MAPK/ERK pathway. MAPK/ERK activation results in myosin light chain phosphorylation, which promotes polymerization of actin cables and generates the contractile forces to stimulate cell migration [23]. Inhibitors of PI3K and MEK1/2 have been used to implicate one [24 –32] or both [33–36] of these pathways in IGF-I-stimulated migration. These results suggest that involvement of a particular IGF-I signaling pathway may depend on cell type. Because of this, one needs to carefully define the system in which he or she is interested. In addition, these inhibitors are not completely specific for their target enzymes [37]. Systematic analyses of inhibitors, which compare inhibition of the cellular response with inhibition of the target enzyme, are needed to establish the contributions of each pathway to the cellular response. An understanding of the signaling pathways that mediate cell migration would help to delineate the mechanisms of tumor invasion and provide strategies for the rational design of novel therapeutics. Here, we used a pharmacological approach to study the contributions of signaling pathways to the IGF-I-stimulated migration of a primary, invasive melanoma cell line (WM1341D). We first examined the signaling pathways activated downstream of IGF-I receptor ligand binding in WM1341D cells. We then used the MEK1/2 inhibitor U0126 and the PI3K inhibitors wortmannin and LY294002 to study the requirement for these two pathways in the IGF-I-induced migration of these invasive, primary melanoma cells. All inhibitors decreased migration at high concentrations; however, dose–response experiments revealed that IGF-I stimulation of the MAPK/ERK pathway was completely inhibited at concentrations that were unable to inhibit IGF-I-induced migration. Conversely, the doses of PI3K inhibitors needed to inhibit PI3K activation were similar to the concentrations required to inhibit migration in response to IGF-I. These results suggest that PI3K, but not MAPK/ERK, signaling is necessary for IGF-I-stimulated WM1341D cell migration.

Materials and methods Materials WM1341D cells were generously provided by Dr. Meenhard Herlyn (The Wistar Institute, Philadelphia, PA). MCDB 153 medium and insulin were purchased from Sigma Chemical Company (St. Louis, MO, USA). Leibovitz’s L-15 medium was purchased from Invitrogen (Carlsbad, CA, USA). Fetal bovine serum was purchased from

130

C.L. Neudauer, J.B. McCarthy / Experimental Cell Research 286 (2003) 128 –137

Mediatech, Inc. (Herndon, VA, USA). Recombinant human bFGF, CCL27, HGF, IGF-I, IL-8, and goat polyclonal IGF-I antibodies were purchased from R & D Systems, Inc. (Minneapolis, MN, USA). Monoclonal ␣IR3 IGF-I receptor antibodies were purchased from Oncogene Research Products (Boston, MA, USA). Wortmannin, LY294002, and U0124 were purchased from Calbiochem–Novabiochem Corporation (San Diego, CA, USA). U0126 was purchased from Promega Corporation (Madison, WI, USA). Polyclonal p44/42 MAPK, Akt, and phospho-Akt (Ser473) antibodies and monoclonal phospho-p44/42 MAPK (Thr202/ Tyr204) antibodies were purchased from Cell Signaling Technology, Inc. (Beverly, MA, USA). Cell culture WM1341D melanoma cells were maintained in MCDB153:L15 (4:1) medium supplemented with 2% fetal bovine serum and 5 ␮g/ml insulin. Cells were maintained at 37°C in a humidified atmosphere with 5% CO2. Migration assays Migration assays were performed in a modified Boyden chamber (NeuroProbe, Inc, Gaithersburg, MD, USA). Cells were starved overnight in serum-free and insulin-free medium (SIFM), released in 3 mM EDTA in phosphate-buffered saline (PBS), and washed in SIFM containing 20 mM HEPES, pH 7.1 (SIFM/HEPES). Ligands were diluted in SIFM/HEPES and added (33 ␮l) to the lower wells; SIFM/ HEPES was placed in control wells. Cells (2.2 ⫻ 104 in 55 ␮l) were placed in the upper wells and separated from the lower wells by an 8-␮m-pore-size polycarbonate filter (Osmonics, Inc., Minnetonka, MN, USA). Filters were coated on both sides (30 min each side at 37°C) in a solution of 2.5 ␮g/ml rat tail collagen type I (BD Biosciences, Bedford, MA, USA) in PBS, rinsed in ddH2O, and dried at room temperature. For antibody inhibition studies, cells were pretreated with antibody dilutions (in SIFM/HEPES) for 15 min at 37°C with rotation prior to placement in the migration chamber; antibodies were also present in the lower wells. Isotype-matched antibodies, at the highest concentration of inhibitory antibody, were added to the controls. For inhibitor studies, cells were pretreated with inhibitor dilutions (in DMSO; final concentration of DMSO was 0.5%) for 30 min at 37°C with rotation prior to placement in the migration chambers. Controls were pretreated with DMSO or U0124 in DMSO for U0126 inhibition. After pretreatment with wortmannin, cells were pelleted at 300g for 5 min and resuspended in SIFM/HEPES. Migration assays were incubated at 37°C for 4 h. Filters were fixed and stained with Diff-Quik solutions (Dade Behring, Inc., Newark, DE, USA), and unmigrated cells were wiped from the filters. Filters were mounted on glass slides in immersion oil, and migrated cells in five random fields were counted at 400⫻ magnification.

Fig. 1. IGF-I stimulates migration of WM1341D melanoma cells. Modified Boyden chambers were used to test the migration of starved WM1341D cells to the indicated concentrations of bFGF (open circle), CCL27 (closed circle), HGF (open square), IGF-I (closed triangle), and IL-8 (closed square). Each concentration was assayed in triplicate, and the number of cells in five random fields was counted for each well. Data are presented as means (⫾) SEM.

Immunoblotting Activation of kinases was detected by immunoblotting whole-cell lysates with phospho-specific antibodies. Cells were plated in 35-mm tissue culture plates (3 ⫻ 105 cells per plate) in 2 ml of growth medium, incubated for 24 h, and then starved overnight in SIFM. For kinase activation studies, cells were stimulated with 10 ng/ml IGF-I in SIFM for various amounts of time at 37°C. For inhibitor studies, cells were pretreated with SIFM containing inhibitor dilutions (in DMSO; final concentration of DMSO was 0.5%) for 30 min at 37°C. Control cells were treated with DMSO or U0124 in DMSO for U0126 studies. Inhibitor solutions were then replaced with SIFM containing 10 ng/ml IGF-I and inhibitors; wortmannin samples were replaced with SIFM containing 10 ng/ml IGF-I. Cells were incubated at 37°C for 1 h for MAPK/ERK or 10 min for PI3K inhibition experiments. Cells were then lysed in 1.5⫻ Laemmli sample buffer heated to 95°C, and DNA was sheared by passage through a 27-gauge needle. Proteins were fractionated by SDS– PAGE and transferred to Immobilon-P membranes (Millipore Corp. Bedford, MA, USA). Membranes were blocked with 5% (w/v) milk in Tris-buffered saline with Tween 20 (TBST: 20 mM Tris, pH 7.5, 150 mM NaCl, 0.1% Tween 20). Blots were probed with a 1:2000 dilution of antiphospho-p42/p44 (Thr202/Tyr204) or 1:1000 dilution of anti-phospho-Akt (Ser473) in 3% (w/v) BSA in TBST. Proteins were detected with 1:25,000 dilutions of horseradish peroxidase-conjugated anti-mouse or anti-rabbit (Jackson ImmunoResearch Laboratories, Inc, West Grove, PA, USA) antibody in 3% BSA/TBST followed by enhanced chemiluminescence (Amersham Pharmacia Biotech, Piscataway, NJ, USA). Equal loading of the MAPK/ERK blots was determined by reprobing the blot with a 1:1000 dilution of anti-p44/42 MAPK. Phospho-Akt blots were stripped and reprobed with a 1:1000 dilution of anti-Akt to confirm equal loading.

C.L. Neudauer, J.B. McCarthy / Experimental Cell Research 286 (2003) 128 –137

131

Fig. 2. IGF-I-induced migration of WM1341D cells is inhibited by IGF-I ligand or receptor antibodies. Indicated concentrations of either anti-IGF-I (A) or anti-IGF-I receptor (B) antibodies were placed in the lower wells in modified Boyden chambers. Isotype-matched antibody controls were performed with and without IGF-I (10 ng/ml) in the lower wells. Starved WM1341D cells, pretreated with the same concentrations of antibodies, were placed into the upper wells. Each antibody concentration was assayed in triplicate and presented as the mean (⫾ SEM) of five random fields per well.

Results IGF-I induces migration of WM1341D cells Growth factors and chemokines are important for the progression of melanoma [5,6]. Tumor stromal fibroblasts produce bFGF, HGF, IGF-I, and IL-8 [5], and these factors have been shown to induce migration of melanoma cells [8,9,11,16 –18]. Numerous melanoma cells lines express CCR10, the receptor for skin-associated CCL27 [20], but the function of this chemokine in melanoma is unknown. We evaluated the effects bFGF, HGF, IGF-I, IL-8, and CCL27 on the migration of WM1341D, a cell line derived from a human primary VGP melanoma. All ligands increased migration of WM1341D cells (Fig. 1). bFGF, CCL27, HGF, and IL-8 stimulated a 2- to 4-fold increase in migration. However, IGF-I-induced a 15-fold increase in WM1341D migration, on average. We observed a similar stimulation of migration by IGF-I in WM115 cells, another cell line derived from a VGP tumor (data not shown). A concentration of 10 ng/ml IGF-I reproducibly induced migration in the linear range of the dose–response curve and was used in all subsequent assays. IGF-I-stimulated migration of WM1341D cells requires IGF-IR WM1341D cells express IGF-I receptor ([14] and our unpublished data). To established the requirement for IGF-I and its receptor in the migration of WM1341D cells, we assayed IGF-I-stimulated migration in the presence of increasing concentrations of IGF-I or IGF-IR antibodies. Migration was stimulated with 10 ng/ml of IGF-I (negative control containing no IGF-I is indicated) in the lower wells of modified Boyden chambers. Starved cells were pretreated with various concentrations of antibody for 15 min, and antibodies were present in the upper and lower wells during the migration assays. Control wells (⫾IGF-I) contained

isotype-matched antibodies at the highest concentration of inhibitory antibody. The IGF-I-stimulated migration of starved WM1341D cells was inhibited in a concentrationdependent manner with IGF-I antibodies (Fig. 2A) or IGF-IR antibodies (Fig. 2B). These results demonstrate that IGF-I specifically stimulated WM1341D cell migration primarily through the IGF-IR. IGF-I-induced migration of WM1341D cells is independent of MAPK/ERK signaling The MAPK/ERK signaling pathway is stimulated after IGF-IR activation. In this pathway, the MAPK kinases MEK1/2 phosphorylate and activate ERK1/2 [21]. To determine if the MAPK/ERK pathway was activated by incubating starved WM1341D cells with 10 ng/ml IGF-I, total cell lysates were prepared at the indicated times and analyzed by immunoblotting. IGF-I induced a slight, timedependent increase in ERK1 and ERK2 phosphorylation, as detected with phospho-specific antibodies (Fig. 3A). Maximal activation occurred at 60 min (Fig. 3A) and was sustained for 2 h (data not shown). Several studies have used MEK1/2 inhibitors to implicate the MAPK/ERK pathway in IGF-I-induced cell migration [25,26,33–36]. To determine whether the endogenous ERK1/2 signaling pathway is directly related to IGF-Iinduced migration in WM1341D cells, we treated cells with U0126, a MEK1/2 inhibitor [37,38]. Starved cells were pretreated with the indicated doses of inhibitor or control (50 ␮M U0124, a negative control compound for U0126) for 30 min. Cells were then stimulated with 10 ng/ml IGF-I for 60 min. Inhibitors were present for the duration of the assay. IGF-I-stimulated phosphorylation of endogenous ERK1/2 was inhibited in a dose-dependent manner in the presence of increasing concentrations of MEK inhibitor (Fig. 3B). The inhibitory concentration 50% (IC50) for ERK1/2 activation was 0.78 ␮M. Having demonstrated the direct effect of U0126 on

132

C.L. Neudauer, J.B. McCarthy / Experimental Cell Research 286 (2003) 128 –137

MEK1/2 kinase activity, we next assessed whether IGF-Istimulated migration was dependent on activated MEK1/2. Starved cells were treated with the indicated concentrations of U0126 (controls were treated with 50 ␮M U0124) for 30 min prior to the migration assay, and the inhibitor was present during the migration assay. Migration was stimulated with 10 ng/ml IGF-I in the lower wells. U0126 decreased migration only at the highest doses tested (16.67 and 50 ␮M) (Fig. 3C). The IC50 of U0126 for migration was 26 ␮M. Since inhibition of 50% of migration required U0126 at concentrations 33-fold higher than the IC50 for MEK1/2 activation, we conclude that IGF-I-induced migration of WM1341D cells is not dependent on MEK1/2 activation. IGF-I-induced migration of WM1341D cells is dependent on PI3K activation PI3K is also stimulated after IGF-I receptor activation and activates many downstream signaling pathways. In one of these pathways, the kinase Akt is phosphorylated and activated downstream of PI3K [21]. The phosphorylation of Ser473 of Akt is often used to assay in vivo PI3K activation [39]. To determine if IGF-I stimulated the activation of PI3K in starved WM1341D cells, we performed immunoblot analysis with PI3K activation detected by increases in phosphorylation of Akt. Total cell lysates were prepared after treatment of cells with 10 ng/ml IGF-I for the indicated amounts of time. IGF-I stimulated a time-dependent activation of PI3K, detected by phospho-specific Akt Ser473 antibodies (Fig. 4A). Maximal activation occurred at 10 min and diminished after 15 min. We also detected IGF-I stimulation of Akt activity, as assayed with an in vitro kinase assay (data not shown). Several studies have also used PI3K inhibitors to implicate this pathway in IGF-I-induced cell migration [24,27– 36]. To determine whether endogenous PI3K signaling is directly related to IGF-I-induced migration of WM1341D cells, we treated starved cells with the indicated concentrations of wortmannin, an irreversible PI3K inhibitor [40], for 30 min prior to addition of 10 ng/ml IGF-I for 10 min. IGF-I-induced phosphorylation of endogenous Akt was inhibited in a dose-dependent manner in the presence of increasing concentrations of wortmannin (Fig. 4B). Having characterized the direct effect of wortmannin on PI3K activity, we then evaluated the requirement for activated PI3K in IGF-I-stimulated migration of WM1341D cells. Starved cells were treated with indicated concentrations of wortmannin for 30 min prior to addition to the migration assay. Wortmannin treatment resulted in a dosedependent decrease in IGF-I-stimulated migration (Fig. 4C). The levels of inhibition of IGF-I-enhanced PI3K activity and migration at each wortmannin concentration were very similar (compare Figs. 4B and C). These data clearly link IGF-stimulated migration to PI3K activation in WM1341D cells.

To confirm these results, we used the competitive PI3K inhibitor LY294002 [40] in the PI3K activation and migration assays as described above for wortmannin. LY294002 also decreased IGF-I-enhanced phosphorylation of cellular Akt in a dose-dependent manner (Fig. 4D). As observed with wortmannin, the LY294002 concentrations required for inhibition of both IGF-I-induced PI3K signaling and migration of WM1341D cells were very comparable (Figs. 4D and E). Our data show dose-dependent decreases in IGF-I-stimulated PI3K activation and migration by two PI3K inhibitors. Importantly, the inhibition of IGF-I-induced cellular migration by the PI3K inhibitors occurred at the same concentrations that prevented the phosphorylation of a physiological PI3K effector (Akt) in the same cell type. These observations indicate that PI3K activity directly mediates IGF-I-induced migration of WM1341D cells.

Discussion Invasive melanoma cells exhibit altered phenotypic properties when compared with noninvasive cells, including enhanced responses to various motility stimuli. Our laboratory is interested in the biology of tumor progression, and we use WM1341D cells, which were derived from a primary VGP tumor, to study the events in melanoma invasion. Melanoma cells have been shown to migrate in response to bFGF, HGF, IL-8, and IGF-I [8,9,11,16 –18], and these factors are expressed by stromal fibroblasts [5]. CCL27 is a skin-associated chemokine, and its receptor, CCR10, is expressed by many melanoma cell lines [20]; a function for CCL27 in melanoma has not been reported. To study migration in a cell line derived from a primary tumor, we tested the migration of WM1341D cells in response to these factors. Each of these factors induced slight, 2- to 4-fold increases in migration (Fig. 1). However, IGF-I induced a striking response, with an average 15-fold increase in migration. Our data also indicate that the IGF-I-induced migration of WM1341D cells was dependent on IGF-I ligand binding to the IGF-IR (Fig. 2). This is consistent with previous studies where the same receptor blocking antibody was used to inhibit the 5- to 6-fold IGF-I chemotactic stimulation of A2058 cells, a highly metastatic melanoma cell line [17,18]. Recently, Satyamoorthy et al. [16] used IGF-I antibodies to block the 3-fold increase in migration in response to IGF-I in SBcl2 cells, a cell line derived from a RGP melanoma. Our study represents the first report of the stimulation of migration of a cell line derived from a VGP melanoma and illustrates a response to IGF-I much greater than in other melanoma cell lines. IGF-I has been shown to stimulate migration in a variety of other cell types including other tumor cell lines such as prostate cancer [41], colon carcinoma [26,34], breast cancer [28,42,43], and neuroblastoma [30,35] cells. Studies using PI3K and MEK inhibitors to examine the role of IGF-I

C.L. Neudauer, J.B. McCarthy / Experimental Cell Research 286 (2003) 128 –137

133

Fig. 3. IGF-I induced migration of WM1341D cells is independent of MAPK/ERK signaling. (A) Starved WM1341D cells were stimulated with 10 ng/ml IGF-I for the indicated amounts of time and then lysed with Laemmli sample buffer. Immunoblots of total cell lysates were probed with anti-phospho-p44/42 MAPK (Thr202/Tyr204) monoclonal antibodies to detect activation (top: upper band is p44 ERK1; lower band is p42 ERK2) after probing with anti-p44/42 MAPK polyclonal antibodies to demonstrate equal protein loading (bottom: only p42 ERK2 was detected in these cells under these conditions). (B) Starved WM1341D cells were incubated with 10 ng/ml for 60 min at 37°C in the presence of the indicated concentrations of U0126, a MEK1/2 inhibitor, or 50 ␮M U0124 as a control. Cells were pretreated with inhibitor or control for 30 min at 37°C prior to the addition of IGF-I. ERK1/2 were detected as in (A). Top: phosphorylated ERK1/2. Bottom: ERK1/2 loading control. (C) Modified Boyden chambers were used to assay the migration of starved WM1341D cells to 10 ng/ml IGF-I. Cells were pretreated with the indicated concentrations of U0126 or 50 ␮M U0124 control, for 30 min at 37°C. Each concentration was assayed in triplicate, and the number of cells in five random fields was counted for each well. Data represent means ⫾ SEM

signaling pathways in the migratory response of cells have generated variable results. For example, in skeletal myoblasts, neither LY294002 nor U0126 decreases IGF-I-induced migration [44]. In neuroblastoma [35], hepatic stellate [36], vascular smooth muscle [33], and colon carcinoma [34] cells, both PI3K inhibitors (wortmannin and/or LY294002) and a MEK1/2 inhibitor (PD98059) diminish IGF-I-stimulated migration. Imai and Clemmons [24] reported that PI3K inhibitors decrease IGF-I-induced migration at doses that inhibit PI3K activation, but a dose of MEK inhibitor that completely inhibits IGF-I-stimulated proliferation is able to inhibit migration by only 34%. The authors conclude that the contributions of MAPK signaling to IGFI-stimulated migration are less important than stimulation of PI3K signaling. Pukac et al. [32] reached similar conclusions after observing a 34% inhibition of IGF-I-induced migration by the PI3K inhibitor wortmannin and 19% inhibition with PD98059 in vascular smooth muscle cells where IGF-I did not activate ERK1/2. In Schwann [27], breast cancer [28], and vascular endothelial [29] cells, the PI3K inhibitor LY294002 decreases IGF-I-enhanced migration and MEK1/2 inhibitors do not. IGF-I-stimulated migration of vascular smooth muscle [25] and colon carcinoma [26] cells is diminished with the MEK1/2 inhibitor PD98059, but PI3K inhibitors were not tested. In neuroblastoma [30] and vascular smooth muscle [31] cells, IGF-I-induced migration

is decreased with PI3K inhibitors, but MEK1/2 inhibitors were not tested in these studies. Collectively, these data suggest that either PI3K or MAPK/ERK signaling pathways or both may be involved in IGF-I-induced cell migration and that the involvement of each pathway may need to be confirmed for each cell type. In these studies, dose dependence for inhibition of endogenous enzyme activation and migration was not compared in the same cell type, making it difficult to establish the contributions of each pathway to IGF-induced cell migration in a particular cell type. To determine if IGF-I-stimulated migration of WM1341D cells was dependent on the MAPK/ERK and PI3K pathways, we performed dose–response studies for endogenous kinase activation and migration with inhibitors of these two pathways. Dose–response studies of MEK inhibitor illustrated that the concentration required to inhibit 50% of enzyme activation was 33-fold lower than that needed to inhibit 50% of migration. At two of the MEK inhibitor concentrations tested, there was no decrease in migration despite complete inhibition of ERK1/2 phosphorylation. These results suggest that IGF-I-induced migration of these cells is independent of MAPK/ERK activation. Inhibition of migration, observed only at the highest doses (16.67 and 50 ␮M), was probably the result of nonspecific enzyme inhibition in these cells. For example, in this concentration range, U0126 has been shown to inhibit MEK5

134

C.L. Neudauer, J.B. McCarthy / Experimental Cell Research 286 (2003) 128 –137

activity [37,38]. Even though this inhibitor may not be completely specific for MEK1/2, it is useful for excluding this signaling pathway [37]. If we had used just one high dose of MEK inhibitor (e.g., 50 ␮M) for our migration studies, we would have observed greater than 50% inhibition and could have concluded that MAPK/ERK signaling is important for IGF-I-induced migration of these cells. This dose is within the range (5–50 ␮M) recommended for this inhibitor [38]. We would have then found that this same concentration inhibited enzyme activation, further supporting such a claim. We used a similar approach to determine if PI3K signaling was involved in the IGF-I-stimulated migration of WM1341D cells. IGF-I stimulation of PI3K signaling in RGP, early VGP, and metastatic melanoma cell lines has been demonstrated [14]. The PI3K inhibitors LY294002 and wortmannin each decreased cellular PI3K activation and migration with comparable dose responses (compare Figs. 4B with C and 4D with E). These concentrations are within the range (10 –20 ␮M for LY294002 and 20 –50 nM for wortmannin) recommended for inhibition of cellular PI3K [40], and similar concentrations were needed to inhibit IGF-I-stimulated Akt phosphorylation in skeletal myoblasts cells [39]. These inhibitors are not completely specific for their target enzymes, but the use of both of them diminishes the likelihood that the observed decrease in migration is the result of nonspecific inhibition of another signaling pathway [37]. Also, a comparison of the concentration needed to inhibit the cellular migration to that needed to inhibit endogenous PI3K activation supports our conclusions. Imai and Clemmons [24] also used PI3K and MEK inhibitors to demonstrate that PI3K signaling is the primary pathway involved in IGF-I stimulation of vascular smooth muscle cell migration. Since PI3K has been shown to stimulate the MAPK/ERK pathway [45,46], Imai and Clemmons [24] specifically addressed this cross-talk in their system. They demonstrated that a concentration of the PI3K inhibitor LY294002 that inhibits 58% of IGF-I-stimulated migration does not inhibit IGF-I-induced ERK phosphorylation. In our system, we observed a high basal level of ERK1/2 phosphorylation and a slight increase in ERK1/2 phosphorylation in response to IGF-I, suggesting that this pathway is not a major contributor to IGF-I-induced signaling and migration in these cells. Satyamoorthy et al. [14] have shown that MAPK/ERK can be activated in SBcl2 RGP cells but detected constitutive phosphorylation of ERK1/2 in a late VGP cell line and four metastatic melanoma cell lines that could not be further increased by IGF-I. BRAF, which is upstream of ERK1/2, has recently been shown to be mutated in a majority of melanoma cell lines [47], and this may explain the high level of endogenous ERK1/2 phosphorylation in these melanoma cell lines. In the WM1341D cells, we did not observe any decrease in migration with the MEK inhibitor U0126 at two doses that completely inhibited ERK1/2 phosphorylation. Since cells could still migrate in response to IGF-I even when there was

no detectable ERK1/2 phosphorylation, our results suggest that the IGF-I-induced migration in these cells was not the result of PI3K stimulation of the MAPK/ERK pathway. In the RGP cell line SBcl2, IGF-I-induced migration is decreased by IL-8 antibodies, suggesting that IGF-I-mediated migration is dependent on autocrine IL-8 stimulation in these cells. IGF-I-stimulated IL-8 production in SBcl2 cells is blocked by the MEK1/2 inhibitor PD98059 but not by the PI3K inhibitor wortmannin [16]. We observed a much greater stimulation of WM1341D cell migration with IGF-I than with IL-8 (Fig. 1), suggesting that IGF-I-induced migration of these VGP cells is not dependent on IL-8. Our IGF-I-induced migration was inhibited by PI3K inhibitors but not by a MEK1/2 inhibitor. In addition, IGF-I-stimulated migration of A2058 melanoma cells is not inhibited by pertussis toxin [17,18], which blocks signaling through the IL-8 receptors CXCR1 and CXCR2 [48], demonstrating that IL-8 signaling is not required for IGF-I-induced migration of this metastatic cell line. Collectively, these data suggest that signaling pathways involved in IGF-I-stimulated cell migration may differ among cell lines and may change as a function of progression. In the skin, the majority of IGF-I is produced by fibroblasts that reside in the dermis just beneath the basement membrane [49]. We propose that IGF-I produced in the dermis acts as a chemoattractant for VGP melanoma cells in the epidermis and stimulates these cells to migrate into the dermis. IGF-I may then also stimulate proliferation of the cells and protect against apoptosis. IGF-I can stimulate the mitogenesis of WM1341D and other early VGP and RGP cell lines in culture, and SBcl2 cells expressing IGF-I have increased anchorage-independent survival [14]. IGF-I also stimulates melanoma cell production of VEGF [16], which may then act as an autocrine growth factor and/or stimulate angiogenesis [7]. These effects would promote growth and survival of the tumor in the dermis. In addition, blocking the IGF-IR receptor with antibodies [50] or downregulation of IGF-IR with antisense oligonucleotides [51] inhibits melanoma growth in mouse tumor models. The dysregulation of IGF-I signaling may result from changes in receptor expression, alterations in interactions with or activities of other receptors, or modifications in integrin affinity for extracellular ligands. Interestingly, WM1341D VGP cells express more IGF-I receptor than RGP cells ([14] and our unpublished results), suggesting that changes in IGF-I receptor levels may play a role in melanoma progression. IGF-I has been shown to increase ␣V␤3 ligand affinity in vascular smooth muscle cells; blocking ␣V␤3 ligand binding reduces IGF-I-stimulated migration of these cells [52] through decreased IGF-I-induced receptor phosphorylation [53,54]. PI3K inhibitors also decrease IGF-I-stimulated migration of these cells, but PI3K or MEK inhibitors do not alter integrin affinity [24]. In our system, blocking antibodies to ␤1 integrin completely inhibited IGF-I-induced migration (data not shown), suggesting that the combined effects of integrin engagement

C.L. Neudauer, J.B. McCarthy / Experimental Cell Research 286 (2003) 128 –137

135

Fig. 4. IGF-I-induced migration of WM1341D cells requires PI3K activation. (A) Starved WM1341D cells were stimulated with 10 ng/ml IGF-I for the indicated times and lysed with Laemmli sample buffer. Immunoblots were probed with anti-phospho-Akt polyclonal antibodies to detect activation (top), stripped, and reprobed with anti-Akt polyclonal antibodies to confirm equal loading (bottom). (B) Starved WM1341D cells were pretreated with the indicated concentrations of the PI3K inhibitor wortmannin or DMSO (controls) for 30 min at 37°C prior to addition of 10 ng/ml IGF-I for 10 min at 37°C. Akt was detected as in (A). Top: phosphorylated Akt. Bottom: Akt loading control. (C) Modified Boyden chambers were used to test the migration of starved WM1341D cells to 10 ng/ml IGF-I in the absence or presence of the indicated concentrations of wortmannin for 30 min at 37°C. Each inhibitor concentration was assayed in triplicate and presented as the mean (⫾ SEM) number of cells in five random fields per well. (D) Starved WM1341D cells, pretreated for 30 min with the indicated concentrations of the PI3K inhibitor LY294002, were incubated with 10 ng/ml IGF-I for 10 min at 37°C. Akt was detected as in (A). Top: phosphorylated Akt. Bottom: Akt loading control. (E) The migration of WM1341D cells, in the presence of the indicated concentrations of LY294002, was assayed as in (C).

and IGF-I receptor activation are necessary for migration to occur. Such combined effects may include IGF-I-stimulated changes in the ligand affinity of the ␤1 integrin or ␤1 integrin-mediated alterations in IGF-I receptor signaling. Doerr and Jones [42] observed that IGF-I-induced migration of breast cancer cells on collagen type IV-coated membranes is partially inhibited by ␣2 blocking antibodies and completely inhibited by ␤1 blocking antibodies. Signaling through ␣5␤1 also potentiates insulin receptor signaling including insulin receptor substrate and PI3K complexes, and insulin activates ␣5␤1 adhesion [55,56]. Additionally, ␣5␤1 co-immunoprecipitates with insulin receptor substrate-1. An understanding of the cooperation between

IGF-I and integrins is needed to fully understand IGF-Iinduced cellular migration. Determining the contributions of the IGF-I signaling pathways in each of the cellular events of progression would increase our understanding of the molecular mechanisms of melanoma progression and may lead to the rational design of treatments for this highly malignant tumor.

Acknowledgments We thank Dr. Meenhard Herlyn at The Wistar Institute for the WM1341D human melanoma cells. We also thank

136

C.L. Neudauer, J.B. McCarthy / Experimental Cell Research 286 (2003) 128 –137

Dr. Yun Qiu and Dr. Julie Vrana for helpful discussions and Dr. Melanie Simpson and Dr. Kelli Bullard for careful review of the data and the manuscript. This work was supported by Grant CA82295 from the National Institutes of Health (to J.B.M.) References [1] M. Nesbit, M. Herlyn, Adhesion receptors in human melanoma progression, Invasion Metast. 14 (1994) 131–146. [2] T. Saida, Recent advances in melanoma research, J. Dermatol. Sci. 26 (2001) 1–13. [3] F. Meier, K. Satyamoorthy, M. Nesbit, M.Y. Hsu, B. Schittek, C. Garbe, M. Herlyn, Molecular events in melanoma development and progression, Front. Biosci. 3 (1998) D1005–D1010. [4] L. Chin, G. Merlino, R.A. DePinho, Malignant melanoma: modern black plague and genetic black box, Genes Dev. 12 (1998) 3467– 3481. [5] D. Ruiter, T. Bogenrieder, D. Elder, M. Herlyn, Melanoma–stroma interactions: structural and functional aspects, Lancet Oncol. 3 (2002) 35– 43. [6] E. Lazar-Molnar, H. Hegyesi, S. Toth, A. Falus, Autocrine and paracrine regulation by cytokines and growth factors in melanoma, Cytokine 12 (2000) 547–554. [7] R.E. Favoni, A. d Cupis, The role of polypeptide growth factors in human carcinomas: new targets for a novel pharmacological approach, Pharmacol. Rev. 52 (2000) 179 –206. [8] F. Meier, M. Nesbit, M.Y. Hsu, B. Martin, P. Va Belle, D.E. Elder, G. Schaumburg-Lever, C. Garbe, T.M. Waltz, P. Donatien, T.M. Crombleholme, M. Herlyn, Human melanoma progression in skin reconstructs: biological significance of bFGF, Am. J. Pathol. 156 (2000) 193–200. [9] M. Komada, N. Kitamura, The cell dissociation and motility triggered by scatter factor/hepatocyte growth factor are mediated through the cytoplasmic domain of the c-Met receptor, Oncogene 8 (1993) 2381– 2390. [10] P.M. Comoglio, C. Boccaccio, Scatter factors and invasive growth, Semin. Cancer Biol. 11 (2001) 153–165. [11] J.M. Wang, G. Taraboletti, K. Matsushima, J. Va Damme, A. Mantovani, Induction of haptotactic migration of melanoma cells by neutrophil activating protein/interleukin-8, Biochem. Biophys. Res. Commun. 169 (1990) 165–170. [12] M. Bar-Eli, Role of interleukin-8 in tumor growth and metastasis of human melanoma, Pathobiology 67 (1999) 12–18. [13] B. Valentinis, R. Baserga, IGF-I receptor signalling in transformation and differentiation, Mol. Pathol. 54 (2001) 133–137. [14] K. Satyamoorthy, G. Li, B. Vaidya, D. Patel, M. Herlyn, Insulin-like growth factor-1 induces survival and growth of biologically early melanoma cells through both the mitogen-activated protein kinase and beta-catenin pathways, Cancer Res. 61 (2001) 7318 –7324. [15] U. Rodeck, M. Herlyn, H.D. Menssen, R.W. Furlanetto, H. Koprowsk, Metastatic but not primary melanoma cell lines grow in vitro independently of exogenous growth factors, Int. J. Cancer 40 (1987) 687– 690. [16] K. Satyamoorthy, G. Li, B. Vaidya, J. Kalabis, M. Herlyn, Insulinlike growth factor-I-induced migration of melanoma cells is mediated by interleukin-8 induction, Cell Growth Differ. 13 (2002) 87–93. [17] M.L. Stracke, J.D. Engel, L.W. Wilson, M.M. Rechler, L.A. Liotta, E. Schiffmann, The type I insulin-like growth factor receptor is a motility receptor in human melanoma cells, J. Biol. Chem. 264 (1989) 21544 –21549. [18] L.M. Stracke, E.C. Kohn, S.A. Aznavoorian, L.L. Wilson, D. Salomon, H.C. Krutzsch, L.A. Liotta, E. Schiffmann, Insulin-like growth factors stimulate chemotaxis in human melanoma cells, Biochem. Biophys. Res. Commun. 153 (1988) 1076 –1083.

[19] B. Homey, H. Alenius, A. Muller, H. Soto, E.P. Bowman, W. Yuan, L. McEvoy, A.I. Lauerma, T. Assmann, E. Bunemann, M. Lehto, H. Wolff, D. Yen, H.H. Marxhausen, W. To, J. Sedgwick, T. Ruzicka, P. Lehmann, A. Zlotnik, CCL27–CCR10 interactions regulate T cellmediated skin inflammation, Nat. Med. 8 (2002) 157–165. [20] A. Muller, B. Homey, H. Soto, N. Ge, D. Catron, M.E. Buchanan, T. McClanahan, E. Murphy, W. Yuan, S.N. Wagner, J.L. Barrera, A. Mohar, E. Verastegui, A. Zlotnik, Involvement of chemokine receptors in breast cancer metastasis, Nature 410 (2001) 50 –56. [21] K. Siddle, B. Urso, C.A. Niesler, D.L. Cope, L. Molina, K.H. Surinya, M.A. Soos, Specificity in ligand binding and intracellular signalling by insulin and insulin-like growth factor receptors, Biochem. Soc. Trans. 29 (2001) 513–525. [22] C.Y. Chung, S. Funamoto, R.A. Firtel, Signaling pathways controlling cell polarity and chemotaxis, Trends Biochem. Sci. 26 (2001) 557–566. [23] S.P. Holly, M.K. Larson, L.V. Parise, Multiple roles of integrins in cell motility, Exp. Cell Res. 261 (2000) 69 –74. [24] Y. Imai, D.R. Clemmons, Roles of phosphatidylinositol 3-kinase and mitogen-activated protein kinase pathways in stimulation of vascular smooth muscle cell migration and deoxyriboncleic acid synthesis by insulin-like growth factor-I, Endocrinology 140 (1999) 4228 – 4235. [25] S. Goetze, X.P. Xi, H. Kawano, T. Gotlibowski, E. Fleck, W.A. Hsueh, P.E. Law, PPAR gamma-ligands inhibit migration mediated by multiple chemoattractants in vascular smooth muscle cells, J. Cardiovasc. Pharmacol. 33 (1999) 798 – 806. [26] V. Rigot, M. Lehmann, F. Andre, N. Daemi, J. Marvaldi, J. Luis, Integrin ligation and PKC activation are required for migration of colon carcinoma cells, J. Cell Sci. 111 (Pt. 20) (1998) 3119 –3127. [27] H.L. Cheng, M.L. Steinway, J.W. Russell, E.L. Feldman, GTPases and phosphatidylinositol 3-kinase are critical for insulin-like growth factor-I-mediated Schwann cell motility, J. Biol. Chem. 275 (2000) 27197–27204. [28] M. Bartucci, C. Morelli, L. Mauro, S. Ando, E. Surmacz, Differential insulin-like growth factor I receptor signaling and function in estrogen receptor (ER)-positive MCF-7 and ER-negative MDA-MB-231 breast cancer cells, Cancer Res. 61 (2001) 6747– 6754. [29] W. Liu, Y. Liu, W.L. Lower Jr., The role of phosphatidylinositol 3-kinase and the mitogen-activated protein kinases in insulin-like growth factor-I-mediated effects in vascular endothelial cells, Endocrinology 142 (2001) 1710 –1719. [30] A. Puglianiello, D. Germani, P. Rossi, S. Cianfarani, IGF-I stimulates chemotaxis of human neuroblasts: involvement of type 1 IGF receptor, IGF binding proteins, phosphatidylinositol-3 kinase pathway and plasmin system, J. Endocrinol. 165 (2000) 123–131. [31] C. Duan, J.R. Bauchat, T. Hsieh, Phosphatidylinositol 3-kinase is required for insulin-like growth factor-I-induced vascular smooth muscle cell proliferation and migration, Circ. Res. 86 (2000) 15–23. [32] L. Pukac, J. Huangpu, M.J. Karnovsky, Platelet-derived growth factor-BB, insulin-like growth factor-I, and phorbol ester activate different signaling pathways for stimulation of vascular smooth muscle cell migration, Exp. Cell Res. 242 (1998) 548 –560. [33] R. Cospedal, H. Abedi, I. Zachary, Platelet-derived growth factor-BB (PDGF-BB) regulation of migration and focal adhesion kinase phosphorylation in rabbit aortic vascular smooth muscle cells: roles of phosphatidylinositol 3-kinase and mitogen-activated protein kinases, Cardiovasc. Res. 41 (1999) 708 –721. [34] F. Andre, V. Rigot, M. Remacle-Bonnet, J. Luis, G. Pommier, J. Marvaldi, Protein kinases C-gamma and -delta are involved in insulin-like growth factor I-induced migration of colonic epithelial cells, Gastroenterology 116 (1999) 64 –77. [35] G.E. Meyer, E. Shelden, B. Kim, E.L. Feldman, Insulin-like growth factor I stimulates motility in human neuroblastoma cells, Oncogene 20 (2001) 7542–7550. [36] A. Gentilini, F. Marra, P. Gentilini, M. Pinzani, Phosphatidylinositol-3 kinase and extracellular signal-regulated kinase mediate the

C.L. Neudauer, J.B. McCarthy / Experimental Cell Research 286 (2003) 128 –137

[37]

[38] [39]

[40] [41]

[42]

[43]

[44]

[45]

[46]

[47]

chemotactic and mitogenic effects of insulin-like growth factor-I in human hepatic stellate cells, J. Hepatol. 32 (2000) 227–234. S.P. Davies, H. Reddy, M. Caivano, P. Cohen, Specificity and mechanism of action of some commonly used protein kinase inhibitors, Biochem. J. 351 (2000) 95–105. J.M. English, M.H. Cobb, Pharmacological inhibitors of MAPK pathways , Trends Pharmacol. Sci. 23 (2002) 40 – 45. S. Adi, N.Y. Wu, S.M. Rosenthal, Growth factor-stimulated phosphorylation of Akt and p70(S6K) is differentially inhibited by LY294002 and wortmannin, Endocrinology 142 (2001) 498 –501. R.C. Stein, Prospects for phosphoinositide 3-kinase inhibition as a cancer treatment, Endocr. Relat. Cancer 8 (2001) 237–248. K. Reiss, J.Y. Wang, G. Romano, X. Tu, F. Peruzzi, R. Baserga, Mechanisms of regulation of cell adhesion and motility by insulin receptor substrate-1 in prostate cancer cells, Oncogene 20 (2001) 490 –500. M.E. Doerr, J.I. Jones, The roles of integrins and extracellular matrix proteins in the insulin-like growth factor I-stimulated chemotaxis of human breast cancer cells, J. Biol. Chem. 271 (1996) 2443–2447. M.A. Guvakova, J.C. Adams, D. Boettiger, Functional role of alphaactinin, PI 3-kinase and MEK1/2 in insulin-like growth factor I receptor kinase regulated motility of human breast carcinoma cells, J. Cell Sci. 115 (2002) 4149 – 4165. J. Suzuki, Y. Yamazaki, G. Li, Y. Kaziro, H. Koide, Involvement of Ras and Ral in chemotactic migration of skeletal myoblasts, Mol. Cell Biol. 20 (2000) 4658 – 4665. M. Lopez-Ilasaca, P. Crespo, P.G. Pellici, J.S. Gutkind, R. Wetzker, Linkage of G protein-coupled receptors to the MAPK signaling pathway through PI 3-kinase gamma, Science 275 (1997) 394 –397. T. Bondeva, L. Pirola, G. Bulgarelli-Leva, I. Rubio, R. Wetzker, M.P. Wymann, Bifurcation of lipid and protein kinase signals of PI3Kgamma to the protein kinases PKB and MAPK, Science 282 (1998) 293–296. H. Davies, G.R. Bignell, C. Cox, P. Stephens, S. Edkins, S. Clegg, J. Teague, H. Woffendin, M.J. Garnett, W. Bottomley, N. Davis, E. Dicks, R. Ewing, Y. Floyd, K. Gray, S. Hall, R. Hawes, J. Hughes, V. Kosmidou, A. Menzies, C. Mould, A. Parker, C. Stevens, S. Watt, S. Hooper, R. Wilson, H. Jayatilake, B.A. Gusterson, C. Cooper, J.

[48] [49]

[50]

[51]

[52]

[53]

[54]

[55]

[56]

137

Shipley, D. Hargrave, K. Pritchard-Jones, N. Maitland, G. ChenevixTrench, G.J. Riggins, D.D. Bigner, G. Palmieri, A. Cossu, A. Flanagan, A. Nicholson, J.W. Ho, S.Y. Leung, S.T. Yuen, B.L. Weber, H.F. Seigler, T.L. Darrow, H. Paterson, R. Marais, C.J. Marshall, R. Wooster, M.R. Stratton, P.A. Futreal, Mutations of the BRAF gene in human cancer, Nature 417 (2002) 949 –954. M. Thelen, Dancing to the tune of chemokines, Nat. Immunol. 2 (2001) 129 –134. S.M. Rudman, M.P. Philpott, G.A. Thomas, T. Kealey, The role of IGF-I in human skin and its appendages: morphogen as well as mitogen, J. Invest. Dermatol. 109 (1997) 770 –777. R.W. Furlanetto, S.E. Harwell, R.B. Baggs, Effects of insulin-like growth factor receptor inhibition on human melanomas in culture and in athymic mice, Cancer Res. 53 (1993) 2522–2526. M. Resnicoff, D. Coppola, C. Sell, R. Rubin, S. Ferrone, R. Baserga, Growth inhibition of human melanoma cells in nude mice by antisense strategies to the type 1 insulin-like growth factor receptor, Cancer Res 54 (1994) 4848 – 4850. J.I. Jones, T. Prevette, A. Gockerman, D.R. Clemmons, Ligand occupancy of the alpha-V-beta3 integrin is necessary for smooth muscle cells to migrate in response to insulin-like growth factor, Proc. Natl. Acad. Sci. USA 93 (1996) 2482–2487. B. Zheng, D.R. Clemmons, Blocking ligand occupancy of the alphaVbeta3 integrin inhibits insulin-like growth factor I signaling in vascular smooth muscle cells, Proc. Natl. Acad. Sci. USA 95 (1998) 11217–11222. L.A. Maile, D.R. Clemmons, The alphaVbeta3 integrin regulates insulin-like growth factor I (IGF-I) receptor phosphorylation by altering the rate of recruitment of the Src-homology 2-containing phosphotyrosine phosphatase-2 to the activated IGF-I receptor, Endocrinology 143 (2002) 4259 – 4264. A. Guilherme, M.P. Czech, Stimulation of IRS-1-associated phosphatidylinositol 3-kinase and Akt/protein kinase B but not glucose transport by beta1-integrin signaling in rat adipocytes, J. Biol. Chem. 273 (1998) 33119 –33122. A. Guilherme, K. Torres, M.P. Czech, Cross-talk between insulin receptor and integrin alpha5 beta1 signaling pathways, J. Biol. Chem. 273 (1998) 22899 –22903.