Effect of Cholesterol Depletion on Mitogenesis and Survival: The Role of Caveolar and Noncaveolar Domains in Insulin-Like Growth Factor-Mediated Cellu
http://www.100md.com
《内分泌学杂志》
Endocrine Sciences (L.C.M., M.W.), Maternal and Fetal Health Research Centre (M.J.T.)
Smooth Muscle Physiology Group (M.J.T.), Cardiovascular Research, University of Manchester, Manchester M13 9PT, United Kingdom
Abstract
The type 1 IGF receptor (IGF-IR) is thought to localize to a subset of lipid rafts, known as caveolae, but the impact on IGF signaling remains controversial. We investigated this potential regulatory mechanism by assessing IGF function in caveolae-positive (3T3L1 and NWTb3) and -negative (HepG2) cells. Coimmunoprecipitation studies demonstrated that IGF-IR and insulin receptor substrate 1 associated with caveolin, a caveolar marker, in 3T3L1 and NWTb3 cells. Subcellular fractionation showed that methyl-cyclodextrin, which disrupts lipid rafts by sequestration of cholesterol, disrupted the colocalization of caveolin and the IGF-IR at the plasma membrane. Methyl-cyclodextrin did not alter IGF-I-induced 3T3L1 or NWTb3 proliferation but significantly impaired the ability of IGF-I to protect these cells from apoptosis. Immunoblotting revealed that methyl-cyclodextrin had no effect on IGF-I-induced activation of the IGF-IR or insulin receptor substrate 1 but increased and decreased the phosphorylation of MAPK and protein kinase B, respectively. In caveolae-negative HepG2 cells, the effect of methyl-cyclodextrin on IGF signaling and cellular function was similar to that observed in caveolae-positive 3T3L1 and NWTb3 cells. Furthermore, transfecting caveolin into HepG2 cells to give morphologically identifiable caveolae made no difference to IGF action, despite a demonstrable interaction between caveolin and the IGF-IR. This suggests that although IGF-IR localizes to caveolin-rich subcellular fractions and coimmunoprecipitates with caveolin, caveolae may not be obligatory for IGF signaling.
Introduction
CAVEOLAE ARE INVAGINATIONS of the plasma membrane suggested to be important for many aspects of cell function including cellular development, growth, and survival. Caveolae-negative mice are lean (1) and insulin resistant (2, 3) and have reduced lifespan (4) in addition to hyperproliferative (5) and cardiomyopathic (6) phenotypes. Caveolin, the primary protein marker for caveolae, binds to and modulates the functions of a broad range of molecular species including the tyrosine kinase receptors for EGF (7) and platelet-derived growth factor (8), nonreceptor tyrosine kinase c-src (9), the G protein-coupled receptor for endothelin (10), and the membrane estrogen receptor (11).
Evidence suggests that the insulin receptor is also localized to caveolae through an association with caveolin (12). Residence within this specialized subset of lipid rafts is reported to have a complex regulatory action on downstream insulin signaling, where caveolae are proposed to sort between mitogenic and metabolic signaling pathways (13, 14).
Components of the closely related IGF system are important mediators of various aspects of cellular function. With actions mediated primarily through the type 1 IGF receptor (IGF-IR), the IGFs exert effects on the control of cellular growth (15), differentiation (16), migration (17), metabolism (18), and survival (19).
Studies have proposed that the IGF-IR is also localized to caveolin-containing subcellular fractions, through direct binding to caveolin (20). However, the functional relevance of this finding has been questioned because a recent report suggested that caveolae were not required for IGF-I-stimulated differentiation and clonal expansion of 3T3L1 adipocytes (21).
Cholesterol modification, using the lipid-binding drug methyl-cyclodextrin (22), disrupts lipid raft domains, including caveolae (23), and thus alters the interaction between lipid raft-associated proteins. If the IGF-IR is a resident of lipid rafts, cholesterol depletion may have ramifications for IGF-mediated signaling and cellular function. Consequently, we have examined the effect of cholesterol depletion on IGF-induced phosphorylation of the putative signaling molecules and determined how this impacts on two other functions of IGFs, cellular mitogenesis and survival, in three cell systems: one that expresses endogenous caveolin and IGF-IR (3T3L1), one that expresses endogenous caveolin but stably overexpresses the IGF-IR (NWTb3), and one that expresses the IGF-IR with (HepG2cav) and without (HepG2) caveolin/caveolae. Using cells that all possess lipid rafts but differing expression of caveolin in relation to the IGF-IR allowed us to examine the role of caveolae/lipid rafts in the regulation of the IGF system.
Materials and Methods
Unless otherwise stated, all chemical reagents were purchased from Sigma (Dorset, UK).
Primary antibodies
Anti--actin (1:1000) and antivimentin (1:1000) were from Sigma; anticaveolin (1:2000) was from BD Transduction Laboratories, (Oxfordshire, UK); anti-IGF-IR (1:2000), anti-caveolin-1 (1:2000), and anti-caveolin-2 (1:1000) were from Santa Cruz Biotechnology (Santa Cruz, CA); anticlathrin (1:2000) was a kind gift of Dr. Martin Lowe, University of Manchester, Manchester, UK; anti-PY162IRS-1 (no. 44-816; 1:1000) and anti-PY1162/1163-IGF-IR (no. 44-804; 1:1000) were from Biosource International (Camarillo, CA); anti-IRS-1 (1:1000) was from Upstate (Milton Keynes, UK); and anti-PKB (1:1000), anti-PS473-PKB (no. 9217; 1:1000), anti-MAPK (1:1000), and anti-PT202/PY204-MAPK (no. 9101; 1:1000) were from Cell Signaling Technology (Beverly, MA).
Secondary antibodies
Horseradish peroxidase (HRP)-conjugated antimouse (1:5000) and HRP-conjugated antirabbit (1:2000 to 1:5000) were from Amersham (Little Chalfont, UK), and HRP-conjugated antigoat (1:10,000) was from Santa Cruz Biotechnology.
Cell culture
3T3L1 mouse fibroblasts (European Collection of Cell Cultures, Wiltshire, UK) were cultured in DMEM supplemented with 10% fetal calf serum (Pierce, Rockford, IL), 1 mM pyruvate, 4 mM glutamine, 0.01% streptomycin, and 0.0005% gentamicin.
NIH-3T3 mouse fibroblasts that stably overexpress the human IGF-IR [NWTb3 cells; a kind gift from Professor D. LeRoith, National Institutes of Health, Bethesda, MD (24)] were cultured in DMEM supplemented with 10% fetal calf serum, 1 mM pyruvate, 4 mM glutamine, 0.01% streptomycin, 0.0005% gentamicin, and 0.05% geneticin.
The human hepatocyte carcinoma line, HepG2 (European Collection of Cell Cultures) was maintained in MEM containing Earls salts supplemented with 10% fetal calf serum, 1 mM pyruvate, 2 mM glutamine, 0.01% streptomycin, 0.0005% gentamicin, and 1% nonessential amino acids.
The HepG2cav cell line was maintained in the same medium as wild-type cells (above) with the addition of the selective antibiotic neomycin (0.1%).
All cells were cultured in a humidified atmosphere of 5% carbon dioxide at 37 C.
Generation of a stable HepG2cav cell line
HepG2 cells were cotransfected (fugene 6 reagent; Roche Diagnostics, Basel, Switzerland) with pcDNA3.1-human caveolin-1 and pcDNA3.1-human caveolin-2 (a kind gift of Dr. Toyoshi Fujimoto, Nagoya University, Japan). Twenty-four hours after transfection, cells were transferred to antibiotic selective media and clones selected, expanded, and subjected to immunoblot analysis for caveolin-1/2 expression. A single clone cell line was selected and used for subsequent experiments.
Electron microscopy
Cells were centrifuged at 1000 x g for 2 min, resuspended in 1 ml human serum, centrifuged at 1500 x g for 3 min, and then fixed (2.5% glutaraldehyde/0.1 M sodium cacodylate buffer, pH 7.3) for 3 h. The resulting solid cell pellet was cut into small cubes, postfixed (1% osmium tetroxide, 0.05 M sodium cacodylate buffer, pH 7.3) for 1 h at room temperature, dehydrated in a graded alcohol series and then propylene oxide, and embedded in Taab epoxy resin (Taab Laboratories Equipment Ltd., Aldermaston, UK). The 0.5-μm sections were stained (1% toluidine blue, 1% borax) and representative areas selected for ultrathin sectioning. Pale gold sections were mounted on copper grids, contrasted with uranyl acetate and Reynold’s lead citrate, and examined under a Philips 301 electron microscope at an accelerating voltage of 60 kV.
Sample preparation
Lysates.
Cells were serum starved for 24 h, treated with 5 mM methyl-cyclodextrin for 1 h, and/or incubated with 5 nM human recombinant IGF-I for 15 min (determined through a series of preliminary time-course experiments) at 37 C, 5% CO2. Lysates were prepared by scraping cells washed with PBS (Oxoid, Hampshire, UK) into RIPA lysis buffer [50 mM Tris-Cl (pH 7.4), 1% Nonidet P40, 0.25% Na-deoxycholate, 150 mM NaCl, 1 mM EDTA] containing protease (Calbiochem, La Jolla, CA) and phosphatase inhibitors before centrifugation for 30 min at 10,000 x g and 4 C. The supernatants were harvested, assayed for protein content (Bio-Rad Laboratories, Hercules, CA), and diluted in reducing loading buffer [0.125 M Tris-Cl (pH 6.8), 0.1% SDS, 20% glycerol, 0.2% -mercaptoethanol, 0.001% bromphenol blue] before boiling for 5 min.
Immunoprecipitates.
Whole-cell lysates (500 μg protein) were precleared with 50 μl protein A-coated Sepharose beads (Zymed Laboratories, South San Francisco, CA) for 30 min at room temperature. These beads were pelleted by centrifugation at 5000 x g for 20 sec and discarded. In the test samples, supernatant was incubated with 2 μg primary antibody and 50 μl protein A-coated Sepharose beads overnight at 4 C. The control sample supernatants were incubated with 50 μl protein A-coated Sepharose beads alone. After this incubation, test and control samples were centrifuged (5000 x g for 20 sec) to pellet the beads, and the supernatants were collected, precipitated with 10% trichloracetic acid (TCA), and boiled in reducing loading buffer for 5 min. The pellets were washed three times in ice-cold PBS and boiled for 5 min in reducing loading buffer, and the beads were removed before electrophoresis by centrifugation at 5000 x g for 20 sec. The precipitated (pellet) and nonprecipitated (supernatant) fractions harvested from test and control samples were analyzed by immunoblot to demonstrate that in lysates known to contain the protein of interest, precipitation was dependent on the presence of the antibody.
Immunoblotting
Whole-cell lysates (50 μg protein) or immunoprecipitates were electrophoresed (Protean II Xi; Bio-Rad) on SDS-acrylamide (10%) gels and transferred to 0.2 μM nitrocellulose membranes (Bio-Rad). Membranes were blocked for 6 h (0.15 M NaCl, 1% dried milk, 0.1% Tween 20) and incubated with primary antibodies (diluted in blocking buffer) overnight at 4 C. After three 10-min washes [88 mM Tris (pH 7.8), 0.25% dried milk, 0.1% Tween 20], membranes were incubated with a species-specific HRP-conjugated secondary antibody (diluted in wash buffer) for 1 h at room temperature and then washed another three times for 10 min each. Immunoreactive proteins were visualized using enhanced chemiluminescence (Supersignal West Pico, West Femto Maximum Sensitivity Substrate; Pierce), and the protein molecular masses were established using calibrated full-range (7–201 kDa) kaleidoscope standards (Bio-Rad).
Subcellular fractionation
Cells were fractionated using a protocol adapted from the previously reported method of Smart et al. (25). This method takes advantage of the relatively light buoyant density of lipid rafts and caveolae and through a series of density gradient centrifugation steps purifies lipid rafts including caveolar membranes. Briefly, cells were washed with PBS, scraped, and pelleted by centrifugation for 5 min at 1500 x g. The pellet was resuspended (0.25 M sucrose, 1 mM EDTA, 20 mM Tricine, pH 7.8), and after 12 strokes with a loose-fitting dounce homogenizer, the homogenate was centrifuged at 1000 x g for 10 min. This step was repeated, and the postnuclear supernatants were combined and assayed for protein content. Two milligrams of protein were then overlaid onto 25 ml 30% Percoll (diluted in homogenization buffer) and centrifuged at 84,000 x g for 30 min in a 70 Ti fixed angle rotor (Beckman Coulter, Fullerton, CA).
After centrifugation, the plasma membrane fraction, which was visible as an opaque band, was extracted (2 ml) through the sidewall of the tube. An aliquot was saved on ice (positive control) and the rest combined with 0.16 ml homogenization buffer and 1.84 ml 50% OptiPrep (0.25 M sucrose, 6 mM EDTA, 120 mM Tricine, pH 7.8). A linear 20–10% density gradient was achieved by the addition of 4 ml 20% OptiPrep, followed by 4 ml 10% OptiPrep (50% OptiPrep diluted accordingly in homogenization buffer) and then centrifuged (Beckman SW40Ti rotor) at 52,000 x g for 90 min at 4 C.
The light fractions (top 6 ml) were removed, mixed with 4 ml 50% OptiPrep, and overlaid with 2 ml 5% OptiPrep (50% OptiPrep diluted in homogenization buffer). After a final centrifugation (Beckman SW40Ti rotor) for 90 min at 52,000 x g and 4 C, 12 1-ml fractions were collected, and the protein was precipitated with 10% TCA and boiled for 5 min in reducing loading buffer for immunoblotting.
[3H]Thymidine incorporation
Cells were serum starved for 24 h, pretreated for either 1) 1 h with 5 mM methyl-cyclodextrin or 2) 30 min with 100 nM LY294002 or 500 nM PD98059, and then incubated with 5 nM IGF-I. Twenty hours later, methyl-[3H]thymidine was added to a final concentration of 0.25 μCi/ml. After an additional 4 h at 37 C and 5% CO2, cells were washed with PBS and incubated with 10% TCA for 2 h at 4 C. After solubilization with 0.1 M NaOH, thymidine incorporation was assessed using a -counter (Packard Tri-Carb LS counter; Global Medical Instrumentation Inc., Ramsey, MN) and OptiPhase HiSafe liquid scintillant. Each condition was examined in triplicate; means of triplicate counts were calculated and values expressed as fold induction compared with serum-free control (mean ± SEM). The concentrations for methyl-cyclodextrin, LY294002, and PD98059 do not modify basal cell activity as determined by preliminary dose-response experiments.
Apoptosis assays
Cells were serum starved for 24 h and either treated for 1) 1 h with 5 mM methyl-cyclodextrin or 2) 30 min with 100 nM LY294002 or 500 nM PD98059 before incubation with 5 nM IGF-I. Twenty hours later, the medium (containing dead cells) was aspirated and combined with trypsinized (live) cells, which were pelleted at 1400 x g and fixed in 3% paraformaldehyde for 2 h at 4 C. Cells were cytospun onto poly-L-lysine-coated slides (5 min at 1400 x g), nuclear DNA stained with 4',6-diamidino-2-phenylindole (DAPI) (Vectashield; Vector Laboratories, Burlingame, CA) and examined under a fluorescence microscope (Leica Microsystems, Wetzlar, Germany). Chromatin fragmentation was used as an indicator of apoptosis and more than 1000 single cells per replicate were scored for apoptosis. Each condition was assessed in triplicate; means of the triplicate counts were calculated and values expressed as the number of apoptotic cells compared with serum-free control (mean ± SEM).
Statistical analysis
For [3H]thymidine uptake and apoptosis assays, triplicate counts of three or four independent experiments were analyzed by nonparametric Student’s t test (Mann-Whitney, SPSS) where P < 0.05 was considered statistically significant.
Results
IGF-IR and insulin receptor substrate 1 (IRS-1) are associated with caveolin in 3T3L1 and NWTb3 cells
Coimmunoprecipitation studies were used to examine whether caveolin interacted with the IGF-IR and IRS-1. Figure 1 demonstrates that incubation of 3T3L1 (Fig. 1A) and NWTb3 (Fig. 1C) whole-cell lysates with an antibody against either the IGF-IR or caveolin resulted in the precipitation of both molecules, although neither the receptor nor caveolin were precipitated by incubation with protein A-Sepharose alone. Duplicate experiments using an antibody against IRS-1, the immediate downstream substrate of the IGF-IR, suggested that IRS-1 was also directly associated with caveolin (Fig. 1B, 3T3L1 cells; Fig. 1D, NWTb3 cells).
Subcellular fractionation was then used to examine whether this relationship is also evident within the plasma membrane of 3T3 L1 and NWTb3 cells. This method purifies lipid rafts and caveolae through three centrifugation steps such that proteins associated with lipid rafts are found in the 12 fractions obtained from the final round of density gradient centrifugation. Immunoblotting of these 12 fractions demonstrated that the IGF-IR was present in the same fractions as caveolin in NWTb3 cells (Fig. 2A, fraction 3) and 3T3L1 cells (data not shown). Proteins such as clathrin and vimentin, which are thought not to reside within caveolae in situ, localized to fractions (6–12, bulk plasma membrane) distinct from those containing caveolin (Fig. 2A).
Although together these data suggest that the IGF-IR is localized to caveolar domains through an association with caveolin, such findings are not necessarily indicative of a modulatory action of caveolae on IGF signaling. We next investigated whether IGF function was affected after treatment with methyl-cyclodextrin, a cholesterol-binding drug that has been proven to disrupt lipid rafts, including caveolae (23).
Methyl-cyclodextrin attenuates IGF-I mediated survival of 3T3L1 and NWTb3 cells but has no affect on IGF-I stimulated proliferation
Initial studies confirmed that treatment with methyl-cyclodextrin (5 mM, 1 h) is sufficient to disrupt the interaction of IGF-IR with lipid rafts in NWTb3 cells because, when lipid rafts/caveolae were purified, IGF-IR was no longer detectable in the fractions obtained from the final round of density centrifugation (Fig. 2B). Similar results were obtained after methyl-cyclodextrin treatment of 3T3L1 cells (data not shown).
In subsequent studies, serum-starved 3T3L1 and NWTb3 cells were treated with methyl-cyclodextrin and/or IGF-I before assessment of cellular proliferation and survival. Figure 3A demonstrates that treatment with IGF-I caused a 2.6-fold increase (P < 0.01, n = 4) in [3H]thymidine uptake of 3T3L1 cells and that neither IGF-I-stimulated nor basal proliferation was significantly affected after pretreatment with methyl-cyclodextrin (IGF, 2.6 ± 0.1-fold; IGF plus methyl-cyclodextrin, 2.3 ± 0.3-fold; P < 0.01, n = 4).
Assessment of nuclear morphology demonstrated that IGF-I was also able to rescue 3T3L1 cells from apoptosis induced by serum withdrawal (Fig. 3B) (serum starved, 33.2 ± 0.2%; IGF, 4.7 ± 0.1% apoptosis; P < 0.01; n = 4). Although there was no independent effect of methyl-cyclodextrin on cell survival, pretreatment with this cholesterol-binding compound significantly attenuated IGF-I mediated survival (Fig. 3B) (IGF, 4.7 ± 0.1%; IGF plus methyl-cyclodextrin, 12.7 ± 0.2% apoptosis; P < 0.01; n = 4). Comparable effects on proliferation and survival were also obtained after methyl-cyclodextrin treatment of NWTb3 cells (Fig. 3C shows proliferation: IGF-I, 2.1 ± 0.1-fold; IGF plus methyl-cyclodextrin, 2.1 ± 0.2-fold; Fig. 3D shows apoptosis: IGF-I 6.1 ± 0.4%; IGF plus methyl-cyclodextrin, 27.1 ± 1.0% apoptosis; P < 0.01; n = 4).
These data suggest that lipid rafts/caveolae are not essential for IGF-I to have an effect on cellular mitogenesis, yet they are important in the transmission of IGF-I signals mediating cell survival.
Methyl-cyclodextrin alters the activation of IGF signaling molecules
It has been hypothesized that lipid rafts/caveolae sort between signaling pathways by differentially modulating the activity of intracellular signaling molecules (13). We have demonstrated that cholesterol removal has no effect on IGF-mediated cell proliferation but markedly disrupts cell survival. To examine this more closely, we next investigated the effect of methyl-cyclodextrin on the activation status of signaling proteins thought to be important for promoting proliferation and survival in response to IGF.
Serum-starved 3T3L1 and NWTb3 cells were treated with 5 mM methyl-cyclodextrin (1 h) and/or 5 nM IGF-I (15 min) before assessing the activation of IGF-IR, IRS-1, protein kinase B (PKB) and MAPK by immunoblotting with antibodies specific for the phosphorylated isoform of each molecule.
As illustrated in Fig. 4, IGF-I stimulated the phosphorylation of IGF-IR, IRS-1, PKB, and MAPK in 3T3L1 (Fig. 4A) and NWTb3 (Fig. 4B) cells. Although IGF-induced activation of both IGF-IR and IRS-1 remained unaltered, pretreatment with methyl-cyclodextrin decreased the phosphorylation of PKB and increased the phosphorylation of MAPK in response to IGF in both 3T3L1 and NWTb3 cell lines.
Having demonstrated that methyl-cyclodextrin treatment induced changes in IGF-stimulated protein phosphorylation and cell survival but not proliferation, we next used chemical inhibitors acting upstream of MAPK (PD98059) and PKB (LY294002) to assess the contribution of each pathway to cellular survival and mitogenesis.
In 3T3L1 and NWTb3 cells, pretreatment with either LY294002 or PD98059 attenuated IGF-I-stimulated proliferation (P < 0.01; n = 4), which suggests that both phosphatidylinositol-3-kinase (PI-3K)/PKB and MAPK pathways contribute to IGF-I-induced proliferation in these cells (data not shown). However, only PI-3K/PKB is necessary for IGF-I-mediated survival of 3T3L1 and NWTb3 cells because the actions of IGF-I were reduced after pretreatment with LY294002 (P < 0.01; n = 4) but not PD98059 (data not shown). This finding suggests that reduced levels of survival in response to methyl-cyclodextrin treatment may be therefore attributable to the reduced activity of PKB.
These data suggest a potential role for caveolae in mediating IGF-induced protein phosphorylation and cell survival, but a fact often overlooked when using methyl-cyclodextrin is that modulating cellular cholesterol levels disrupts all lipid rafts and not specifically the caveolar subset. Consequently, we sought to validate our results by investigating the effect of cholesterol depletion on IGF signaling and function in a caveolae/caveolin-negative cell line.
Methyl-cyclodextrin impacts on IGF signaling and function in a caveolin/caveolae-deficient HepG2 cell line
Cells commonly used to study IGF signaling were screened for the presence of caveolin. Although characterization by immunoblot and electron microscopy demonstrated that caveolin was expressed by 3T3L1 and NWTb3 fibroblasts (Fig. 5A) and that flask-like structures resembling caveolae were evident at the plasma membrane (Fig. 5, B and C), caveolin could not be detected by immunoblot analysis of the human hepatic carcinoma cell line HepG2 (Fig. 5A). These cells also lacked morphologically identifiable caveolae at the plasma membrane (Fig. 5D), suggesting that they were both caveolin and caveolae deficient. However, HepG2 cells are not reported to be devoid of cholesterol or sphingolipid, the main components of lipid rafts. We were therefore able to examine the impact of caveolae on IGF signaling because HepG2 cells also express the IGF-IR and associated intracellular signaling molecules (Fig. 5E).
After serum starvation and thymidine uptake analysis, data showed that 5 nM IGF-I induced a significant increase in proliferation of HepG2 cells (1.4 ± 0.0-fold; P < 0.01; n = 4) and that this increased rate of proliferation was unaltered after pretreatment with 5 mM methyl-cyclodextrin (Fig. 6A) (IGF-I, 1.4 ± 0.0-fold; IGF-I plus methyl-cyclodextrin, 1.2 ± 0.1-fold; P < 0.01; n = 4). IGF-I also significantly reduced the number of HepG2 cells undergoing apoptosis in response to serum withdrawal (Fig. 6B) (serum starved, 4.8 ± 0.0%; IGF, 1.3 ± 0.2% apoptosis; P < 0.01; n = 4). This effect was partially reversed after treatment with methyl-cyclodextrin (Fig. 6B) (IGF-I, 1.3 ± 0.2%; IGF-I plus methyl-cyclodextrin, 2.6 ± 0.2% apoptosis; P < 0.01; n = 4), even though methyl-cyclodextrin alone had no significant effect on HepG2 cell survival.
Phosphoimmunoblot analysis demonstrated that whereas methyl-cyclodextrin did not alter the IGF-induced phosphorylation of either the IGF-IR or its immediate substrate IRS-1, phosphorylation of PKB was markedly reduced and MAPK phosphorylation enhanced in response to IGF (Fig. 6C).
The fact that methyl-cyclodextrin had a similar affect on IGF function in cells with and without caveolae suggests that caveolae do not impact on IGF-I-mediated proliferation or survival; we then sought to confirm this hypothesis by assessing IGF signaling and function in HepG2 cells engineered to overexpress human caveolin-1 and human caveolin-2.
Caveolin associates with the IGF-IR and IRS-1 in HepG2cav cells
HepG2 cells were transfected with caveolin-1 and caveolin-2 and five stable clones selected for screening. Immunoblot analysis (Fig. 7A) demonstrated that the transfected HepG2 cells express both caveolin-1 and caveolin-2 to a similar degree as the NWTb3 cells (clone 5 was selected for study and clone 4 used as a transfected control), and by electron microscopy, we were able to show that the transfected proteins form morphologically identifiable caveolae at the plasma membrane (Fig. 7C).
Moreover, coimmunoprecipitation studies demonstrated that caveolin associates with IGF-IR (Fig. 7D) and IRS-1 (Fig. 7E) in HepG2cav cells. Taken together, these data suggest that in HepG2cav cells, the exogenous caveolin-1 and -2 can mimic the action of endogenous caveolin in the two fibroblast cell lines.
We next examined whether caveolin expression could impact IGF function by comparing IGF-mediated proliferation and survival in caveolin-negative and -positive HepG2 cells; Fig. 8 shows that functionally, the two cell lines respond in a similar manner and that the presence of caveolin has no significant effect on the phosphorylation of IGF-IR, IRS-1, PKB, or MAPK (Fig. 8C) after treatment with 5 nM IGF-I for 15 min. This reinforces the data presented after methyl-cyclodextrin treatment and suggests that caveolae are not absolute modulators of IGF signaling. Our data therefore indicate that the broader class of lipid rafts and not the caveolar subset are required for proper transmission of IGF signals in these cell types.
Discussion
The IGF-IR is proposed to localize to caveolin-rich domains of the plasma membrane, and as a result, it has also been suggested that the caveolins, a family of small cholesterol-binding proteins found in caveolae (26), are necessary for IGF function (20, 27). In this study, we too have shown that through a direct association with caveolin, the IGF-IR is present within caveolar membrane fractions. However, our results suggest that these membrane domains are not an essential component in the mediation of IGF actions.
The suggestion that caveolae were required for IGF signaling originated from a study in which IGF-I-induced differentiation of preadipocytes was impaired by treating cells with methyl-cyclodextrin to deplete cellular cholesterol (20). The formation of caveolae is critically dependent on surrounding cholesterol, and it has been shown that changes in the level of cellular cholesterol results in disrupted function of caveolin (23). However, the use of cholesterol-binding drugs such as methyl-cyclodextrin also leads to the disruption of other liquid-ordered domains within the plasma membrane, making it difficult to draw conclusions about the relative importance of caveolar vs. noncaveolar lipid rafts. Consequently, in this study, we determined the effect of cholesterol depletion on IGF signaling and function in three cell systems that all have the cholesterol and sphingolipid component of lipid rafts but different caveolin and IGF-IR expression; one expresses endogenous caveolin and IGF-IR (3T3L1), one expresses endogenous caveolin but stably overexpresses the IGF-IR (NWTb3), and one expresses the IGF-IR with (HepG2cav) and without (HepG2) caveolin/caveolae.
Treatment with methyl-cyclodextrin had no affect on the ability of IGF-I to activate the IGF-IR or the immediate downstream signaling molecule IRS-1 in 3T3L1, NWTb3, or HepG2 cells. Yet the phosphorylation of more distal components of IGF-I signaling pathways (which are also used by a variety of other hormones and growth factors) were differentially altered after cholesterol depletion. This finding may suggest that the altered activation of MAPK and PKB may be mediated, at least in part, through input from additional pathways such as those involving shc or src. Indeed, there is evidence to suggest that IGF stimulates protein kinase C translocation to lipid rafts, which may provide an alternative mechanism for activation of PKB (28).
The opposing effect on MAPK and PKB activity might have been anticipated because a differential activation of proteins after methyl-cyclodextrin treatment has been reported for the insulin receptor (29). It is thought that the integration of changes in the activity of signaling molecules may be the means by which lipid rafts/caveolae can sort downstream cell function.
In all three cell types, IGF-I-stimulated phosphorylation of PKB was markedly reduced after exposure to methyl-cyclodextrin; this supports data from other studies that have shown PI-3K is lost from lipid rafts after methyl-cyclodextrin treatment (30). Our inhibitor data indicate that this molecule is critical to the prosurvival role of IGF-I, and therefore it is not surprising that the capacity for IGF-I to rescue cells from apoptosis was also significantly impaired after methyl-cyclodextrin treatment (summarized in Fig. 9). However, cholesterol depletion had no effect on IGF-mediated proliferation even though we have shown through the use of specific inhibitors that the PI-3K/PKB pathway is required for IGF-I to affect this cellular response. This may reflect enhanced IGF-I activation of MAPK after cyclodextrin treatment, because an increase in the activity of one pathway (Figs. 4, A and B, and 6C; increased MAPK phosphorylation) coupled to a decrease in the activity of the second pathway (Figs. 4, A and B, and 6C; reduced PKB phosphorylation) could dampen any change in proliferation that may be otherwise observed (Fig. 9). This is consistent with existing data to suggest that the contribution of PI-3K and MAPK to IGF-mediated cell proliferation depends on the exact physiological setting; for example, although PI-3K is implicated in the survival but not the proliferative actions of IGF in myogenic cells (31), both PI-3K and MAPK have been reported to mediate IGF-I-induced proliferation in osteoblasts (32) and mammary cancer cells (33).
Most importantly, because methyl-cyclodextrin had a similar effect in caveolae-positive and caveolae-negative cells, our results indicate that cholesterol, but not caveolin, is important for transmission of IGF-I signals through the IGF-IR/ PI-3K/PKB pathway. A recent study of EGF-induced survival of prostate cancer cells (34) suggested that this might also be the case for activation of PKB by other tyrosine kinase receptors.
Similarly, the enhanced MAPK phosphorylation observed in 3T3L1, NWTb3, and HepG2 cells after methyl-cyclodextrin treatment indicates that caveolin is not necessary for IGF-I activation of MAPK but instead may suggest a negative regulatory role for lipid rafts. Indeed, other groups have observed enhanced basal and stimulated MAPK phosphorylation in cells treated with methyl-cyclodextrin, leading to the suggestion that after cholesterol depletion, MAPK becomes hypersensitive to activation (35). In contrast, some studies show reduced growth factor activation of MAPK after lipid raft disruption (36, 37, 38). These inconsistencies may be a product of the interaction between PKB and MAPK signaling pathways (39, 40) in some cell types. There is evidence to suggest that PKB can negatively regulate MAPK, and so the hyperactivation observed in response to methyl-cyclodextrin treatment could reflect attenuated inhibition of MAPK as a result of reduced PKB activation.
Initially we were concerned that the similarities between IGF function in caveolae-positive and caveolae-negative cell lines may reflect fundamental differences in the mouse and human cell signaling pathways. However, after stable transfection of caveolin-1 and caveolin-2 to give morphologically identifiable caveolae at the plasma membrane of HepG2 cells, we were able to show that caveolin has no significant effect on the ability of IGF-I to act as a growth or survival factor.
Our data therefore demonstrate that caveolae, and their resident marker protein caveolin, may not be obligatory for facilitating IGF effects on cellular proliferation and survival. Together with a recent study indicating that caveolae are not involved in IGF-induced differentiation of adipocytes (21), this suggests that IGF signaling and function may be dependent on noncaveolar lipid microdomains. The generality of these findings to additional IGF-responsive cell lines and, indeed, other membrane-associated receptor signaling pathways, requires investigation.
The role of the broader class of lipid rafts is often ignored, because experimentally they are difficult to define with no specific marker to distinguish them from caveolae. This study highlights the necessity for caution when interpreting data from experiments using methyl-cyclodextrin as a model of caveolar dysfunction and the potential importance for the broader class of lipid rafts in signal modulation.
Nonetheless, the role of caveolae must not be completely dismissed. We and others (20, 41) have shown a direct association between caveolin and the IGF-IR and also between caveolin and IRS-1; thus it is possible that although caveolae and caveolin are not essential for mediating IGF effects through the IGF-IR, they do have a role in modulating the cellular response to IGF. A study of caveolae function in relation to the structurally similar insulin receptor suggested that the presence of caveolae may facilitate the activation of signaling molecules that favor a metabolic response to insulin (29). The receptors for insulin and IGF have common downstream effectors, and thus caveolae may be involved in directing which signaling molecules are stimulated after binding of IGF to the IGF-IR.
It is speculated that caveolae may function as platforms for cross-talk between different receptor signaling pathways; Ushio-Fukai et al. (42) demonstrated that although transactivation of EGF receptor (EGFR) by angiotensin II was cholesterol dependent, cholesterol was not required for activation of the EGFR by its cognate ligand. IGF-I-mediated activation of IGF-IR is known to cause ligand-independent activation of the EGFR (43, 44), and phosphorylation of unbound IGF-IR has been observed after activation of the angiotensin (45) and EGF (46) receptors. It is possible that caveolin may play a key role in mediating this transactivation.
The structure of caveolae may underlie their ability to organize receptor cross-talk because they could represent an internalization pathway for lipid raft domains. The function of caveolae may therefore be analogous to that proposed for clathrin-coated pits, where intracellular trafficking of endocytic vesicles to endosomes serves to amplify and propagate signals (47). Indeed, it has been shown that stimulation with EGF recruits caveolin to early endosomal compartments (48); the coalescence of multiple internalized caveolae to form so-called caveosomes may thus provide an additional compartment for intracrine signaling.
The function of caveolae may therefore be 2-fold, first to direct cell function by sorting the downstream signals of a particular receptor and second to orchestrate a much larger-scale integration of signals from different receptor species; methyl-cyclodextrin may therefore mediate its effect by uncoupling cross-talk between different signaling molecules.
Our data suggest that caveolae are not essential for cellular mitogenesis and survival in response to IGF-I, but these microdomains may still be important for fine-tuning and integrating IGF signals with those received from other growth factors and cytokines within the cellular environment.
Acknowledgments
We gratefully acknowledge Derek Le Roith, Toyoshi Fujimoto, and Martin Lowe for kind donation of cells, plasmids, and antibodies, respectively, and the Biotechnology and Biological Sciences Research Council for financial support, and we thank Dr. Carolyn J. P. Jones and Michael Mackness for technical assistance.
Footnotes
First Published Online September 15, 2005
Abbreviations: DAPI, 4',6-Diamidino-2-phenylindole; HRP, horseradish peroxidase; IGF-IR, type 1 IGF receptor; IRS-1, insulin receptor substrate 1; PI-3K, phosphatidylinositol-3-kinase; PKB, protein kinase B; TCA, trichloracetic acid.
Accepted for publication September 10, 2005.
References
Razani B, Combs TP, Wang XB, Frank PG, Park DS, Russell RG, Li M, Tang B, Jelicks LA, Scherer PE, Lisanti MP 2002 Caveolin-1-deficient mice are lean, resistant to diet-induced obesity, and show hypertriglyceridemia with adipocyte abnormalities. J Biol Chem 277:8635–8647
Cohen AW, Razani B, Wang XB, Combs TP, Williams TM, Scherer PE, Lisanti MP 2003 Caveolin-1-deficient mice show insulin resistance and defective insulin receptor protein in adipose tissue. Am J Physiol 285:C222–C235
Oshikawa J, Otsu K, Toya Y, Tsunematsu T, Hankins R, Kawabe JI, Minamisawa S, Umemura S, Hagiwara Y, Ishikawa Y 2004 Insulin resistance in skeletal muscles of caveolin-3-null mice. Proc Natl Acad Sci USA 101:12670–12675
Park D, Cohen AW, Frank PG, Razani B, Lee H, Williams TM, Chandra M, Shirani J, De Souza AP, Tang B, Jelicks LA, Factor SM, Weiss LM, Tanowitz HB, Lisanti MP 2003 Caveolin-1 null mice show dramatic reductions in life span. Biochemistry 42:15124–15131
Razani B, Engelman JA, Wang XB, Schubert W, Zhang XL, Marks CB, Macaluso F, Russell RG, Li M, Pestell RG, Di Vizio D, Hou H, Kneitz B, Lagaud G, Christ GJ, Edelmann W, Lisanti MP 2001 Caveolin-1 null mice are viable but show evidence of hyperproliferative and vascular abnormalities. J Biol Chem 276:38121–38138
Park DS, Woodman SE, Schubert AL, Cohen AW, Frank PG, Chandra M, Shirani J, Razani B, Tang B, Jelicks LA, Factor SM, Weiss LM, Tanowitz HB, Lisanti MP 2002 Caveolin-1/3 double-knockout mice are viable, but lack both muscle and non-muscle caveolae, and develop a severe cardiomyopathic phenotype. Am J Pathol 160:2207–2217
Mineo C, James GL, Smart EJ, Anderson RGW 1996 Localization of epidermal growth factor-stimulated Ras/Raf-1 interaction to caveolae membrane. J Biol Chem 271:11930–11935
Yamamoto M, Toya Y, Jensen RA, Ishikawa Y 1999 Caveolin is an inhibitor of platelet-derived growth factor receptor signaling. Exp Cell Res 247:380–388
Li S, Couet J, Lisanti MP 1996 Src tyrosine kinases, G subunits, and H-Ras share a common membrane-anchored scaffolding protein, caveolin. J Biol Chem 271:29182–29190
Chun M, Liyanage U, Lisanti MP, Lodish HF 1994 Signal transduction of a G protein-coupled receptor in caveolae: colocalization of endothelin and its receptor with caveolin. Proc Natl Acad Sci USA 91:11728–11732
Schlegel A, Wang C, Pestell RG, Lisanti MP 2001 Ligand-independent activation of oestrogen receptor by caveolin-1. Biochem J 359:203–210
Gustavsson J, Parpal S, Karlsson M, Ramsing C, Thorn H, Borg M, Lindroth M, Peterson KH, Magnusson KE, Stralfors P 1999 Localization of the insulin receptor in caveolae of adipocyte plasma membrane. FASEB J 13:1961–1971
Parpal S, Karlsson M, Thorn H, Stralfors P 2001 Cholesterol depletion disrupts caveolae and insulin receptor signalling for metabolic control via insulin receptor substrate-1, but not for mitogen-activated protein kinase control. J Biol Chem 276:9670–9678
Yamamoto M, Toya Y, Schwenke C, Lisanti MP, Myers MG, Ishikawa Y 1998 Caveolin is an activator of insulin receptor signaling. J Biol Chem 273:26962–26968
Coppola D, Ferber A, Miura M, Sell C, D’Ambrosio C, Rubin R, Baserga R 1994 A functional insulin-like growth factor I receptor is required for the mitogenic and transforming activities of the epidermal growth factor receptor. Mol Cell Biol 14:4588–4595
Benito M, Valverde AM, Lorenzo M 1996 IGF-I: a mitogen also involved in differentiation processes in mammalian cells. Int J Cell Biol 28:499–510
Leventhal PS, Feldman EL 1997 Insulin-like growth factors as regulators of cell motility. Trends Endocrinol Metab 8:1–6
Baudry A, Lamothe B, Bucchini D, Jami J, Montarras D, Pinset C, Joshi RL 2001 IGF-I receptor as an alternative receptor for metabolic signaling in insulin receptor-deficient muscle cells. FEBS Lett 488:174–178
O’Connor R, Kauffmann-Zeh A, Liu Y, Lehar S, Evan GI, Baserga R, Blattler WA 1997 Identification of domains of the insulin-like growth factor I receptor that are required for protection from apoptosis. Mol Cell Biol 17:427–436
Huo H, Guo X, Hong S, Jiang M, Liu X, Liao K 2003 Lipid rafts/caveolae are essential for insulin-like growth factor-1 receptor signaling during 3T3–L1 preadipocyte differentiation induction. J Biol Chem 278:11561–11569
Hong S, Huo H, Xu J, Liao K 2004 Insulin-like growth factor-1 receptor signalling in 3T3–l1 adipocyte differentiation requires lipid rafts but not caveolae. Cell Death Differ 11:714–723
Christian AE, Haynes MP, Phillips MC, Rothblat GH 1997 Use of cyclodextrins for manipulating cellular cholesterol content. J Lipid Res 38:2264–2272
Hailstones D, Sleer LS, Parton RG, Stanley KK 1998 Regulation of caveolin and caveolae by cholesterol in MDCK cells. J Lipid Res 39:369–379
Blakesely VA, Kato H, Roberts CT, LeRoith D 1995 Mutation of a conserved amino acid residue (tryptophan 1171) in the tyrosine kinase domain of the IGF-I receptor abolishes autophosphorylation but does not eliminate biologic function. J Biol Chem 270:2764–2769
Smart EJ, Ying Y, Mineo C, Anderson RGW 1995 A detergent-free method for purifying caveoolae membrane from tissue culture cells. Proc Natl Acad Sci USA 92:10104–10108
Murata M, Peranen J, Schreiner R, Wieland F, Kurzchalia TV, Simons K 1995 VIP21/caveolin is a cholesterol-binding protein. Proc Natl Acad Sci USA 92:10339–10343
Podar K, Tai YT, Cole CE, Hideshima T, Sattler M, Hamblin A, Mitsiades N, Schlossman RL, Davies FE, Morgan GJ, Munshi NC, Chauhan D, Anderson KC 2003 Essential role of caveolae in IL-6- and IGF-I-triggered Akt-1-mediated survival of multiple myeloma cells. J Biol Chem 278:5794–5801
Partovian C, Simons M 2004 Regulation of protein kinase B/Akt activity and Ser473 phosphorylation by protein kinase C in endothelial cells. Cell Signal 16:951–957
Cohen AW, Combs TP, Scherer PE, Lisanti MP 2003 Role of caveolin and caveolin in insulin signaling and diabetes. Am J Physiol Endocrinol Metab 285:E1151–E1160
Peres C, Yart A, Perret B, Salles JP, Raynal P 2003 Modulation of phosphoinositide 3-kinase activation by cholesterol level suggests a novel positive role for lipid rafts in lysophosphatidic acid signalling. FEBS Lett 534:164–168
Halevy O, Cantley LC 2004 Differential regulation of the phosphoinositide 3-kinase and MAP kinase pathways by hepatocyte growth factor vs. insulin-like growth factor-I in myogenic cells. Exp Cell Res 297:224–234
Radcliff K, Tang TB, Lim J, Zhang Z, Abedin M, Demer, LL, Tintut Y 2005 Insulin-like growth factor-I regulates proliferation and osteoblastic differentiation of calcifying vascular cells via extracellular signal-regulated protein kinase and phosphatidylinositol 3-kinase pathways. Circ Res 96:398–400
Moelling K, Schad K, Bosse M, Zimmermann S, Schweneker M 2002 Regulation of Raf-Akt crosstalk. J Biol Chem 277:31099–31106
Zhuang L, Lin J, Lu ML, Solomon KR, Freeman MR 2002 Cholesterol-rich lipid rafts mediate akt-regulated survival in prostate cancer cells. Cancer Res 62:2227–2231
Furuchi T, Anderson RGW 1998 Cholesterol depletion of caveolae causes hyperactivation of extracellular signal-regulated kinase (ERK). J Biol Chem 273:21099–21104
Hua H, Munk S, Whiteside CI 2003 Endothelin-1 activates mesangial cell ERK1/2 via EGF-receptor transactivation and caveolin-1 interaction. Am J Physiol Renal Physiol 284:F303–F312
Karlsson M, Thorn H, Danielsson A, Stenkula KG, Ost A, Gustavsson J, Nystrom FH, Stralfors P 2004 Colocalisation of insulin receptor and insulin receptor substrate-1 to caveolae in primary human adipocytes. Cholesterol depletion blocks insulin signalling for metabolic and mitogenic control. Eur J Biochem 271:2471–2479
Nystrom F, Chen H, Cong L, Li Y, Quon MJ 1999 Caveolin-1 interacts with the insulin receptor and can differentially modulate insulin signaling in transfected Cos-7 cells and rat adipose cells. Mol Endocrinol 13:2013–2024
Zimmermann S, Moelling K 1999 Phosphorylation and regulation of Raf by Akt (protein kinase B). Science 286:1741–1744
Rommel C, Clarke BA, Zimmermann S, Nunez L, Rossman R, Reid K, Moelling K, Yancopoulos GD, Glass DJ 1999 Differentiation stage-specific inhibition of the Raf-MEK-ERK pathway by Akt. Science 286:1738–1741
Panetta D, Biedi C, Repetto S, Cordera R, Maggi D 2004 IGF-I regulates caveolin 1 and IRS-1 interaction in caveolae. Biochem Biophys Res Commun 316:240–243
Ushio-Fukai M, Hilenski L, Santanam N, Becker PL, Ma X, Griendling KK, Alexander RW 2001 Cholesterol depletion inhibits epidermal growth factor receptor transactivation by angiotensin II in vascular smooth muscle cells. J Biol Chem 276:48269–48275
Pietrzkowski Z, Sell C, Lammers R, Ullrich A, Baserga R 1992 Roles of insulinlike growth factor 1 (IGF-1) and the IGF-1 receptor in epidermal growth factor-stimulated growth of 3T3 cells. Mol Cell Biol 12:3883–3889
Gilmore AP, Valentijn AJ, Wang P, Ranger AM, Bundred N, O’Hare MJ, Wakeling A, Korsmeyer SJ, Streuli CH 2002 Activation of BAD by therapeutic inhibition of epidermal growth factor receptor and transactivation by insulin-like growth factor receptor. J Biol Chem 277:27643–27650
Zahradka P, Litchie B, Storie B, Helwer G 2004 Transactivation of the insulin-like growth factor-I receptor by angiotensin II mediates downstream signaling from the angiotensin II type 1 receptor to phosphatidylinositol 3-kinase. Endocrinology 145:2978–2987
Roudabush FL, Pierce KL, Maudsley S, Khan KD, Luttrell LM 2000 Transactivation of the EGF receptor mediates IGF-I-stimulated Shc phosphorylation and ERK1/2 activation in COS-7 cells. J Biol Chem 275:22583–22589
Di Fiore P, Gill GN 1999 Endocytosis and mitogenic signaling. Curr Opin Cell Biol 11:483–488
Pol A, Lu A, Pons M, Peiro S, Enrich K 2000 Epidermal growth factor-mediated caveolin recruitment to early endosomes and MAPK activation. J Biol Chem 275:30566–30572(Laura C. Matthews, Michael J. Taggart an)
Smooth Muscle Physiology Group (M.J.T.), Cardiovascular Research, University of Manchester, Manchester M13 9PT, United Kingdom
Abstract
The type 1 IGF receptor (IGF-IR) is thought to localize to a subset of lipid rafts, known as caveolae, but the impact on IGF signaling remains controversial. We investigated this potential regulatory mechanism by assessing IGF function in caveolae-positive (3T3L1 and NWTb3) and -negative (HepG2) cells. Coimmunoprecipitation studies demonstrated that IGF-IR and insulin receptor substrate 1 associated with caveolin, a caveolar marker, in 3T3L1 and NWTb3 cells. Subcellular fractionation showed that methyl-cyclodextrin, which disrupts lipid rafts by sequestration of cholesterol, disrupted the colocalization of caveolin and the IGF-IR at the plasma membrane. Methyl-cyclodextrin did not alter IGF-I-induced 3T3L1 or NWTb3 proliferation but significantly impaired the ability of IGF-I to protect these cells from apoptosis. Immunoblotting revealed that methyl-cyclodextrin had no effect on IGF-I-induced activation of the IGF-IR or insulin receptor substrate 1 but increased and decreased the phosphorylation of MAPK and protein kinase B, respectively. In caveolae-negative HepG2 cells, the effect of methyl-cyclodextrin on IGF signaling and cellular function was similar to that observed in caveolae-positive 3T3L1 and NWTb3 cells. Furthermore, transfecting caveolin into HepG2 cells to give morphologically identifiable caveolae made no difference to IGF action, despite a demonstrable interaction between caveolin and the IGF-IR. This suggests that although IGF-IR localizes to caveolin-rich subcellular fractions and coimmunoprecipitates with caveolin, caveolae may not be obligatory for IGF signaling.
Introduction
CAVEOLAE ARE INVAGINATIONS of the plasma membrane suggested to be important for many aspects of cell function including cellular development, growth, and survival. Caveolae-negative mice are lean (1) and insulin resistant (2, 3) and have reduced lifespan (4) in addition to hyperproliferative (5) and cardiomyopathic (6) phenotypes. Caveolin, the primary protein marker for caveolae, binds to and modulates the functions of a broad range of molecular species including the tyrosine kinase receptors for EGF (7) and platelet-derived growth factor (8), nonreceptor tyrosine kinase c-src (9), the G protein-coupled receptor for endothelin (10), and the membrane estrogen receptor (11).
Evidence suggests that the insulin receptor is also localized to caveolae through an association with caveolin (12). Residence within this specialized subset of lipid rafts is reported to have a complex regulatory action on downstream insulin signaling, where caveolae are proposed to sort between mitogenic and metabolic signaling pathways (13, 14).
Components of the closely related IGF system are important mediators of various aspects of cellular function. With actions mediated primarily through the type 1 IGF receptor (IGF-IR), the IGFs exert effects on the control of cellular growth (15), differentiation (16), migration (17), metabolism (18), and survival (19).
Studies have proposed that the IGF-IR is also localized to caveolin-containing subcellular fractions, through direct binding to caveolin (20). However, the functional relevance of this finding has been questioned because a recent report suggested that caveolae were not required for IGF-I-stimulated differentiation and clonal expansion of 3T3L1 adipocytes (21).
Cholesterol modification, using the lipid-binding drug methyl-cyclodextrin (22), disrupts lipid raft domains, including caveolae (23), and thus alters the interaction between lipid raft-associated proteins. If the IGF-IR is a resident of lipid rafts, cholesterol depletion may have ramifications for IGF-mediated signaling and cellular function. Consequently, we have examined the effect of cholesterol depletion on IGF-induced phosphorylation of the putative signaling molecules and determined how this impacts on two other functions of IGFs, cellular mitogenesis and survival, in three cell systems: one that expresses endogenous caveolin and IGF-IR (3T3L1), one that expresses endogenous caveolin but stably overexpresses the IGF-IR (NWTb3), and one that expresses the IGF-IR with (HepG2cav) and without (HepG2) caveolin/caveolae. Using cells that all possess lipid rafts but differing expression of caveolin in relation to the IGF-IR allowed us to examine the role of caveolae/lipid rafts in the regulation of the IGF system.
Materials and Methods
Unless otherwise stated, all chemical reagents were purchased from Sigma (Dorset, UK).
Primary antibodies
Anti--actin (1:1000) and antivimentin (1:1000) were from Sigma; anticaveolin (1:2000) was from BD Transduction Laboratories, (Oxfordshire, UK); anti-IGF-IR (1:2000), anti-caveolin-1 (1:2000), and anti-caveolin-2 (1:1000) were from Santa Cruz Biotechnology (Santa Cruz, CA); anticlathrin (1:2000) was a kind gift of Dr. Martin Lowe, University of Manchester, Manchester, UK; anti-PY162IRS-1 (no. 44-816; 1:1000) and anti-PY1162/1163-IGF-IR (no. 44-804; 1:1000) were from Biosource International (Camarillo, CA); anti-IRS-1 (1:1000) was from Upstate (Milton Keynes, UK); and anti-PKB (1:1000), anti-PS473-PKB (no. 9217; 1:1000), anti-MAPK (1:1000), and anti-PT202/PY204-MAPK (no. 9101; 1:1000) were from Cell Signaling Technology (Beverly, MA).
Secondary antibodies
Horseradish peroxidase (HRP)-conjugated antimouse (1:5000) and HRP-conjugated antirabbit (1:2000 to 1:5000) were from Amersham (Little Chalfont, UK), and HRP-conjugated antigoat (1:10,000) was from Santa Cruz Biotechnology.
Cell culture
3T3L1 mouse fibroblasts (European Collection of Cell Cultures, Wiltshire, UK) were cultured in DMEM supplemented with 10% fetal calf serum (Pierce, Rockford, IL), 1 mM pyruvate, 4 mM glutamine, 0.01% streptomycin, and 0.0005% gentamicin.
NIH-3T3 mouse fibroblasts that stably overexpress the human IGF-IR [NWTb3 cells; a kind gift from Professor D. LeRoith, National Institutes of Health, Bethesda, MD (24)] were cultured in DMEM supplemented with 10% fetal calf serum, 1 mM pyruvate, 4 mM glutamine, 0.01% streptomycin, 0.0005% gentamicin, and 0.05% geneticin.
The human hepatocyte carcinoma line, HepG2 (European Collection of Cell Cultures) was maintained in MEM containing Earls salts supplemented with 10% fetal calf serum, 1 mM pyruvate, 2 mM glutamine, 0.01% streptomycin, 0.0005% gentamicin, and 1% nonessential amino acids.
The HepG2cav cell line was maintained in the same medium as wild-type cells (above) with the addition of the selective antibiotic neomycin (0.1%).
All cells were cultured in a humidified atmosphere of 5% carbon dioxide at 37 C.
Generation of a stable HepG2cav cell line
HepG2 cells were cotransfected (fugene 6 reagent; Roche Diagnostics, Basel, Switzerland) with pcDNA3.1-human caveolin-1 and pcDNA3.1-human caveolin-2 (a kind gift of Dr. Toyoshi Fujimoto, Nagoya University, Japan). Twenty-four hours after transfection, cells were transferred to antibiotic selective media and clones selected, expanded, and subjected to immunoblot analysis for caveolin-1/2 expression. A single clone cell line was selected and used for subsequent experiments.
Electron microscopy
Cells were centrifuged at 1000 x g for 2 min, resuspended in 1 ml human serum, centrifuged at 1500 x g for 3 min, and then fixed (2.5% glutaraldehyde/0.1 M sodium cacodylate buffer, pH 7.3) for 3 h. The resulting solid cell pellet was cut into small cubes, postfixed (1% osmium tetroxide, 0.05 M sodium cacodylate buffer, pH 7.3) for 1 h at room temperature, dehydrated in a graded alcohol series and then propylene oxide, and embedded in Taab epoxy resin (Taab Laboratories Equipment Ltd., Aldermaston, UK). The 0.5-μm sections were stained (1% toluidine blue, 1% borax) and representative areas selected for ultrathin sectioning. Pale gold sections were mounted on copper grids, contrasted with uranyl acetate and Reynold’s lead citrate, and examined under a Philips 301 electron microscope at an accelerating voltage of 60 kV.
Sample preparation
Lysates.
Cells were serum starved for 24 h, treated with 5 mM methyl-cyclodextrin for 1 h, and/or incubated with 5 nM human recombinant IGF-I for 15 min (determined through a series of preliminary time-course experiments) at 37 C, 5% CO2. Lysates were prepared by scraping cells washed with PBS (Oxoid, Hampshire, UK) into RIPA lysis buffer [50 mM Tris-Cl (pH 7.4), 1% Nonidet P40, 0.25% Na-deoxycholate, 150 mM NaCl, 1 mM EDTA] containing protease (Calbiochem, La Jolla, CA) and phosphatase inhibitors before centrifugation for 30 min at 10,000 x g and 4 C. The supernatants were harvested, assayed for protein content (Bio-Rad Laboratories, Hercules, CA), and diluted in reducing loading buffer [0.125 M Tris-Cl (pH 6.8), 0.1% SDS, 20% glycerol, 0.2% -mercaptoethanol, 0.001% bromphenol blue] before boiling for 5 min.
Immunoprecipitates.
Whole-cell lysates (500 μg protein) were precleared with 50 μl protein A-coated Sepharose beads (Zymed Laboratories, South San Francisco, CA) for 30 min at room temperature. These beads were pelleted by centrifugation at 5000 x g for 20 sec and discarded. In the test samples, supernatant was incubated with 2 μg primary antibody and 50 μl protein A-coated Sepharose beads overnight at 4 C. The control sample supernatants were incubated with 50 μl protein A-coated Sepharose beads alone. After this incubation, test and control samples were centrifuged (5000 x g for 20 sec) to pellet the beads, and the supernatants were collected, precipitated with 10% trichloracetic acid (TCA), and boiled in reducing loading buffer for 5 min. The pellets were washed three times in ice-cold PBS and boiled for 5 min in reducing loading buffer, and the beads were removed before electrophoresis by centrifugation at 5000 x g for 20 sec. The precipitated (pellet) and nonprecipitated (supernatant) fractions harvested from test and control samples were analyzed by immunoblot to demonstrate that in lysates known to contain the protein of interest, precipitation was dependent on the presence of the antibody.
Immunoblotting
Whole-cell lysates (50 μg protein) or immunoprecipitates were electrophoresed (Protean II Xi; Bio-Rad) on SDS-acrylamide (10%) gels and transferred to 0.2 μM nitrocellulose membranes (Bio-Rad). Membranes were blocked for 6 h (0.15 M NaCl, 1% dried milk, 0.1% Tween 20) and incubated with primary antibodies (diluted in blocking buffer) overnight at 4 C. After three 10-min washes [88 mM Tris (pH 7.8), 0.25% dried milk, 0.1% Tween 20], membranes were incubated with a species-specific HRP-conjugated secondary antibody (diluted in wash buffer) for 1 h at room temperature and then washed another three times for 10 min each. Immunoreactive proteins were visualized using enhanced chemiluminescence (Supersignal West Pico, West Femto Maximum Sensitivity Substrate; Pierce), and the protein molecular masses were established using calibrated full-range (7–201 kDa) kaleidoscope standards (Bio-Rad).
Subcellular fractionation
Cells were fractionated using a protocol adapted from the previously reported method of Smart et al. (25). This method takes advantage of the relatively light buoyant density of lipid rafts and caveolae and through a series of density gradient centrifugation steps purifies lipid rafts including caveolar membranes. Briefly, cells were washed with PBS, scraped, and pelleted by centrifugation for 5 min at 1500 x g. The pellet was resuspended (0.25 M sucrose, 1 mM EDTA, 20 mM Tricine, pH 7.8), and after 12 strokes with a loose-fitting dounce homogenizer, the homogenate was centrifuged at 1000 x g for 10 min. This step was repeated, and the postnuclear supernatants were combined and assayed for protein content. Two milligrams of protein were then overlaid onto 25 ml 30% Percoll (diluted in homogenization buffer) and centrifuged at 84,000 x g for 30 min in a 70 Ti fixed angle rotor (Beckman Coulter, Fullerton, CA).
After centrifugation, the plasma membrane fraction, which was visible as an opaque band, was extracted (2 ml) through the sidewall of the tube. An aliquot was saved on ice (positive control) and the rest combined with 0.16 ml homogenization buffer and 1.84 ml 50% OptiPrep (0.25 M sucrose, 6 mM EDTA, 120 mM Tricine, pH 7.8). A linear 20–10% density gradient was achieved by the addition of 4 ml 20% OptiPrep, followed by 4 ml 10% OptiPrep (50% OptiPrep diluted accordingly in homogenization buffer) and then centrifuged (Beckman SW40Ti rotor) at 52,000 x g for 90 min at 4 C.
The light fractions (top 6 ml) were removed, mixed with 4 ml 50% OptiPrep, and overlaid with 2 ml 5% OptiPrep (50% OptiPrep diluted in homogenization buffer). After a final centrifugation (Beckman SW40Ti rotor) for 90 min at 52,000 x g and 4 C, 12 1-ml fractions were collected, and the protein was precipitated with 10% TCA and boiled for 5 min in reducing loading buffer for immunoblotting.
[3H]Thymidine incorporation
Cells were serum starved for 24 h, pretreated for either 1) 1 h with 5 mM methyl-cyclodextrin or 2) 30 min with 100 nM LY294002 or 500 nM PD98059, and then incubated with 5 nM IGF-I. Twenty hours later, methyl-[3H]thymidine was added to a final concentration of 0.25 μCi/ml. After an additional 4 h at 37 C and 5% CO2, cells were washed with PBS and incubated with 10% TCA for 2 h at 4 C. After solubilization with 0.1 M NaOH, thymidine incorporation was assessed using a -counter (Packard Tri-Carb LS counter; Global Medical Instrumentation Inc., Ramsey, MN) and OptiPhase HiSafe liquid scintillant. Each condition was examined in triplicate; means of triplicate counts were calculated and values expressed as fold induction compared with serum-free control (mean ± SEM). The concentrations for methyl-cyclodextrin, LY294002, and PD98059 do not modify basal cell activity as determined by preliminary dose-response experiments.
Apoptosis assays
Cells were serum starved for 24 h and either treated for 1) 1 h with 5 mM methyl-cyclodextrin or 2) 30 min with 100 nM LY294002 or 500 nM PD98059 before incubation with 5 nM IGF-I. Twenty hours later, the medium (containing dead cells) was aspirated and combined with trypsinized (live) cells, which were pelleted at 1400 x g and fixed in 3% paraformaldehyde for 2 h at 4 C. Cells were cytospun onto poly-L-lysine-coated slides (5 min at 1400 x g), nuclear DNA stained with 4',6-diamidino-2-phenylindole (DAPI) (Vectashield; Vector Laboratories, Burlingame, CA) and examined under a fluorescence microscope (Leica Microsystems, Wetzlar, Germany). Chromatin fragmentation was used as an indicator of apoptosis and more than 1000 single cells per replicate were scored for apoptosis. Each condition was assessed in triplicate; means of the triplicate counts were calculated and values expressed as the number of apoptotic cells compared with serum-free control (mean ± SEM).
Statistical analysis
For [3H]thymidine uptake and apoptosis assays, triplicate counts of three or four independent experiments were analyzed by nonparametric Student’s t test (Mann-Whitney, SPSS) where P < 0.05 was considered statistically significant.
Results
IGF-IR and insulin receptor substrate 1 (IRS-1) are associated with caveolin in 3T3L1 and NWTb3 cells
Coimmunoprecipitation studies were used to examine whether caveolin interacted with the IGF-IR and IRS-1. Figure 1 demonstrates that incubation of 3T3L1 (Fig. 1A) and NWTb3 (Fig. 1C) whole-cell lysates with an antibody against either the IGF-IR or caveolin resulted in the precipitation of both molecules, although neither the receptor nor caveolin were precipitated by incubation with protein A-Sepharose alone. Duplicate experiments using an antibody against IRS-1, the immediate downstream substrate of the IGF-IR, suggested that IRS-1 was also directly associated with caveolin (Fig. 1B, 3T3L1 cells; Fig. 1D, NWTb3 cells).
Subcellular fractionation was then used to examine whether this relationship is also evident within the plasma membrane of 3T3 L1 and NWTb3 cells. This method purifies lipid rafts and caveolae through three centrifugation steps such that proteins associated with lipid rafts are found in the 12 fractions obtained from the final round of density gradient centrifugation. Immunoblotting of these 12 fractions demonstrated that the IGF-IR was present in the same fractions as caveolin in NWTb3 cells (Fig. 2A, fraction 3) and 3T3L1 cells (data not shown). Proteins such as clathrin and vimentin, which are thought not to reside within caveolae in situ, localized to fractions (6–12, bulk plasma membrane) distinct from those containing caveolin (Fig. 2A).
Although together these data suggest that the IGF-IR is localized to caveolar domains through an association with caveolin, such findings are not necessarily indicative of a modulatory action of caveolae on IGF signaling. We next investigated whether IGF function was affected after treatment with methyl-cyclodextrin, a cholesterol-binding drug that has been proven to disrupt lipid rafts, including caveolae (23).
Methyl-cyclodextrin attenuates IGF-I mediated survival of 3T3L1 and NWTb3 cells but has no affect on IGF-I stimulated proliferation
Initial studies confirmed that treatment with methyl-cyclodextrin (5 mM, 1 h) is sufficient to disrupt the interaction of IGF-IR with lipid rafts in NWTb3 cells because, when lipid rafts/caveolae were purified, IGF-IR was no longer detectable in the fractions obtained from the final round of density centrifugation (Fig. 2B). Similar results were obtained after methyl-cyclodextrin treatment of 3T3L1 cells (data not shown).
In subsequent studies, serum-starved 3T3L1 and NWTb3 cells were treated with methyl-cyclodextrin and/or IGF-I before assessment of cellular proliferation and survival. Figure 3A demonstrates that treatment with IGF-I caused a 2.6-fold increase (P < 0.01, n = 4) in [3H]thymidine uptake of 3T3L1 cells and that neither IGF-I-stimulated nor basal proliferation was significantly affected after pretreatment with methyl-cyclodextrin (IGF, 2.6 ± 0.1-fold; IGF plus methyl-cyclodextrin, 2.3 ± 0.3-fold; P < 0.01, n = 4).
Assessment of nuclear morphology demonstrated that IGF-I was also able to rescue 3T3L1 cells from apoptosis induced by serum withdrawal (Fig. 3B) (serum starved, 33.2 ± 0.2%; IGF, 4.7 ± 0.1% apoptosis; P < 0.01; n = 4). Although there was no independent effect of methyl-cyclodextrin on cell survival, pretreatment with this cholesterol-binding compound significantly attenuated IGF-I mediated survival (Fig. 3B) (IGF, 4.7 ± 0.1%; IGF plus methyl-cyclodextrin, 12.7 ± 0.2% apoptosis; P < 0.01; n = 4). Comparable effects on proliferation and survival were also obtained after methyl-cyclodextrin treatment of NWTb3 cells (Fig. 3C shows proliferation: IGF-I, 2.1 ± 0.1-fold; IGF plus methyl-cyclodextrin, 2.1 ± 0.2-fold; Fig. 3D shows apoptosis: IGF-I 6.1 ± 0.4%; IGF plus methyl-cyclodextrin, 27.1 ± 1.0% apoptosis; P < 0.01; n = 4).
These data suggest that lipid rafts/caveolae are not essential for IGF-I to have an effect on cellular mitogenesis, yet they are important in the transmission of IGF-I signals mediating cell survival.
Methyl-cyclodextrin alters the activation of IGF signaling molecules
It has been hypothesized that lipid rafts/caveolae sort between signaling pathways by differentially modulating the activity of intracellular signaling molecules (13). We have demonstrated that cholesterol removal has no effect on IGF-mediated cell proliferation but markedly disrupts cell survival. To examine this more closely, we next investigated the effect of methyl-cyclodextrin on the activation status of signaling proteins thought to be important for promoting proliferation and survival in response to IGF.
Serum-starved 3T3L1 and NWTb3 cells were treated with 5 mM methyl-cyclodextrin (1 h) and/or 5 nM IGF-I (15 min) before assessing the activation of IGF-IR, IRS-1, protein kinase B (PKB) and MAPK by immunoblotting with antibodies specific for the phosphorylated isoform of each molecule.
As illustrated in Fig. 4, IGF-I stimulated the phosphorylation of IGF-IR, IRS-1, PKB, and MAPK in 3T3L1 (Fig. 4A) and NWTb3 (Fig. 4B) cells. Although IGF-induced activation of both IGF-IR and IRS-1 remained unaltered, pretreatment with methyl-cyclodextrin decreased the phosphorylation of PKB and increased the phosphorylation of MAPK in response to IGF in both 3T3L1 and NWTb3 cell lines.
Having demonstrated that methyl-cyclodextrin treatment induced changes in IGF-stimulated protein phosphorylation and cell survival but not proliferation, we next used chemical inhibitors acting upstream of MAPK (PD98059) and PKB (LY294002) to assess the contribution of each pathway to cellular survival and mitogenesis.
In 3T3L1 and NWTb3 cells, pretreatment with either LY294002 or PD98059 attenuated IGF-I-stimulated proliferation (P < 0.01; n = 4), which suggests that both phosphatidylinositol-3-kinase (PI-3K)/PKB and MAPK pathways contribute to IGF-I-induced proliferation in these cells (data not shown). However, only PI-3K/PKB is necessary for IGF-I-mediated survival of 3T3L1 and NWTb3 cells because the actions of IGF-I were reduced after pretreatment with LY294002 (P < 0.01; n = 4) but not PD98059 (data not shown). This finding suggests that reduced levels of survival in response to methyl-cyclodextrin treatment may be therefore attributable to the reduced activity of PKB.
These data suggest a potential role for caveolae in mediating IGF-induced protein phosphorylation and cell survival, but a fact often overlooked when using methyl-cyclodextrin is that modulating cellular cholesterol levels disrupts all lipid rafts and not specifically the caveolar subset. Consequently, we sought to validate our results by investigating the effect of cholesterol depletion on IGF signaling and function in a caveolae/caveolin-negative cell line.
Methyl-cyclodextrin impacts on IGF signaling and function in a caveolin/caveolae-deficient HepG2 cell line
Cells commonly used to study IGF signaling were screened for the presence of caveolin. Although characterization by immunoblot and electron microscopy demonstrated that caveolin was expressed by 3T3L1 and NWTb3 fibroblasts (Fig. 5A) and that flask-like structures resembling caveolae were evident at the plasma membrane (Fig. 5, B and C), caveolin could not be detected by immunoblot analysis of the human hepatic carcinoma cell line HepG2 (Fig. 5A). These cells also lacked morphologically identifiable caveolae at the plasma membrane (Fig. 5D), suggesting that they were both caveolin and caveolae deficient. However, HepG2 cells are not reported to be devoid of cholesterol or sphingolipid, the main components of lipid rafts. We were therefore able to examine the impact of caveolae on IGF signaling because HepG2 cells also express the IGF-IR and associated intracellular signaling molecules (Fig. 5E).
After serum starvation and thymidine uptake analysis, data showed that 5 nM IGF-I induced a significant increase in proliferation of HepG2 cells (1.4 ± 0.0-fold; P < 0.01; n = 4) and that this increased rate of proliferation was unaltered after pretreatment with 5 mM methyl-cyclodextrin (Fig. 6A) (IGF-I, 1.4 ± 0.0-fold; IGF-I plus methyl-cyclodextrin, 1.2 ± 0.1-fold; P < 0.01; n = 4). IGF-I also significantly reduced the number of HepG2 cells undergoing apoptosis in response to serum withdrawal (Fig. 6B) (serum starved, 4.8 ± 0.0%; IGF, 1.3 ± 0.2% apoptosis; P < 0.01; n = 4). This effect was partially reversed after treatment with methyl-cyclodextrin (Fig. 6B) (IGF-I, 1.3 ± 0.2%; IGF-I plus methyl-cyclodextrin, 2.6 ± 0.2% apoptosis; P < 0.01; n = 4), even though methyl-cyclodextrin alone had no significant effect on HepG2 cell survival.
Phosphoimmunoblot analysis demonstrated that whereas methyl-cyclodextrin did not alter the IGF-induced phosphorylation of either the IGF-IR or its immediate substrate IRS-1, phosphorylation of PKB was markedly reduced and MAPK phosphorylation enhanced in response to IGF (Fig. 6C).
The fact that methyl-cyclodextrin had a similar affect on IGF function in cells with and without caveolae suggests that caveolae do not impact on IGF-I-mediated proliferation or survival; we then sought to confirm this hypothesis by assessing IGF signaling and function in HepG2 cells engineered to overexpress human caveolin-1 and human caveolin-2.
Caveolin associates with the IGF-IR and IRS-1 in HepG2cav cells
HepG2 cells were transfected with caveolin-1 and caveolin-2 and five stable clones selected for screening. Immunoblot analysis (Fig. 7A) demonstrated that the transfected HepG2 cells express both caveolin-1 and caveolin-2 to a similar degree as the NWTb3 cells (clone 5 was selected for study and clone 4 used as a transfected control), and by electron microscopy, we were able to show that the transfected proteins form morphologically identifiable caveolae at the plasma membrane (Fig. 7C).
Moreover, coimmunoprecipitation studies demonstrated that caveolin associates with IGF-IR (Fig. 7D) and IRS-1 (Fig. 7E) in HepG2cav cells. Taken together, these data suggest that in HepG2cav cells, the exogenous caveolin-1 and -2 can mimic the action of endogenous caveolin in the two fibroblast cell lines.
We next examined whether caveolin expression could impact IGF function by comparing IGF-mediated proliferation and survival in caveolin-negative and -positive HepG2 cells; Fig. 8 shows that functionally, the two cell lines respond in a similar manner and that the presence of caveolin has no significant effect on the phosphorylation of IGF-IR, IRS-1, PKB, or MAPK (Fig. 8C) after treatment with 5 nM IGF-I for 15 min. This reinforces the data presented after methyl-cyclodextrin treatment and suggests that caveolae are not absolute modulators of IGF signaling. Our data therefore indicate that the broader class of lipid rafts and not the caveolar subset are required for proper transmission of IGF signals in these cell types.
Discussion
The IGF-IR is proposed to localize to caveolin-rich domains of the plasma membrane, and as a result, it has also been suggested that the caveolins, a family of small cholesterol-binding proteins found in caveolae (26), are necessary for IGF function (20, 27). In this study, we too have shown that through a direct association with caveolin, the IGF-IR is present within caveolar membrane fractions. However, our results suggest that these membrane domains are not an essential component in the mediation of IGF actions.
The suggestion that caveolae were required for IGF signaling originated from a study in which IGF-I-induced differentiation of preadipocytes was impaired by treating cells with methyl-cyclodextrin to deplete cellular cholesterol (20). The formation of caveolae is critically dependent on surrounding cholesterol, and it has been shown that changes in the level of cellular cholesterol results in disrupted function of caveolin (23). However, the use of cholesterol-binding drugs such as methyl-cyclodextrin also leads to the disruption of other liquid-ordered domains within the plasma membrane, making it difficult to draw conclusions about the relative importance of caveolar vs. noncaveolar lipid rafts. Consequently, in this study, we determined the effect of cholesterol depletion on IGF signaling and function in three cell systems that all have the cholesterol and sphingolipid component of lipid rafts but different caveolin and IGF-IR expression; one expresses endogenous caveolin and IGF-IR (3T3L1), one expresses endogenous caveolin but stably overexpresses the IGF-IR (NWTb3), and one expresses the IGF-IR with (HepG2cav) and without (HepG2) caveolin/caveolae.
Treatment with methyl-cyclodextrin had no affect on the ability of IGF-I to activate the IGF-IR or the immediate downstream signaling molecule IRS-1 in 3T3L1, NWTb3, or HepG2 cells. Yet the phosphorylation of more distal components of IGF-I signaling pathways (which are also used by a variety of other hormones and growth factors) were differentially altered after cholesterol depletion. This finding may suggest that the altered activation of MAPK and PKB may be mediated, at least in part, through input from additional pathways such as those involving shc or src. Indeed, there is evidence to suggest that IGF stimulates protein kinase C translocation to lipid rafts, which may provide an alternative mechanism for activation of PKB (28).
The opposing effect on MAPK and PKB activity might have been anticipated because a differential activation of proteins after methyl-cyclodextrin treatment has been reported for the insulin receptor (29). It is thought that the integration of changes in the activity of signaling molecules may be the means by which lipid rafts/caveolae can sort downstream cell function.
In all three cell types, IGF-I-stimulated phosphorylation of PKB was markedly reduced after exposure to methyl-cyclodextrin; this supports data from other studies that have shown PI-3K is lost from lipid rafts after methyl-cyclodextrin treatment (30). Our inhibitor data indicate that this molecule is critical to the prosurvival role of IGF-I, and therefore it is not surprising that the capacity for IGF-I to rescue cells from apoptosis was also significantly impaired after methyl-cyclodextrin treatment (summarized in Fig. 9). However, cholesterol depletion had no effect on IGF-mediated proliferation even though we have shown through the use of specific inhibitors that the PI-3K/PKB pathway is required for IGF-I to affect this cellular response. This may reflect enhanced IGF-I activation of MAPK after cyclodextrin treatment, because an increase in the activity of one pathway (Figs. 4, A and B, and 6C; increased MAPK phosphorylation) coupled to a decrease in the activity of the second pathway (Figs. 4, A and B, and 6C; reduced PKB phosphorylation) could dampen any change in proliferation that may be otherwise observed (Fig. 9). This is consistent with existing data to suggest that the contribution of PI-3K and MAPK to IGF-mediated cell proliferation depends on the exact physiological setting; for example, although PI-3K is implicated in the survival but not the proliferative actions of IGF in myogenic cells (31), both PI-3K and MAPK have been reported to mediate IGF-I-induced proliferation in osteoblasts (32) and mammary cancer cells (33).
Most importantly, because methyl-cyclodextrin had a similar effect in caveolae-positive and caveolae-negative cells, our results indicate that cholesterol, but not caveolin, is important for transmission of IGF-I signals through the IGF-IR/ PI-3K/PKB pathway. A recent study of EGF-induced survival of prostate cancer cells (34) suggested that this might also be the case for activation of PKB by other tyrosine kinase receptors.
Similarly, the enhanced MAPK phosphorylation observed in 3T3L1, NWTb3, and HepG2 cells after methyl-cyclodextrin treatment indicates that caveolin is not necessary for IGF-I activation of MAPK but instead may suggest a negative regulatory role for lipid rafts. Indeed, other groups have observed enhanced basal and stimulated MAPK phosphorylation in cells treated with methyl-cyclodextrin, leading to the suggestion that after cholesterol depletion, MAPK becomes hypersensitive to activation (35). In contrast, some studies show reduced growth factor activation of MAPK after lipid raft disruption (36, 37, 38). These inconsistencies may be a product of the interaction between PKB and MAPK signaling pathways (39, 40) in some cell types. There is evidence to suggest that PKB can negatively regulate MAPK, and so the hyperactivation observed in response to methyl-cyclodextrin treatment could reflect attenuated inhibition of MAPK as a result of reduced PKB activation.
Initially we were concerned that the similarities between IGF function in caveolae-positive and caveolae-negative cell lines may reflect fundamental differences in the mouse and human cell signaling pathways. However, after stable transfection of caveolin-1 and caveolin-2 to give morphologically identifiable caveolae at the plasma membrane of HepG2 cells, we were able to show that caveolin has no significant effect on the ability of IGF-I to act as a growth or survival factor.
Our data therefore demonstrate that caveolae, and their resident marker protein caveolin, may not be obligatory for facilitating IGF effects on cellular proliferation and survival. Together with a recent study indicating that caveolae are not involved in IGF-induced differentiation of adipocytes (21), this suggests that IGF signaling and function may be dependent on noncaveolar lipid microdomains. The generality of these findings to additional IGF-responsive cell lines and, indeed, other membrane-associated receptor signaling pathways, requires investigation.
The role of the broader class of lipid rafts is often ignored, because experimentally they are difficult to define with no specific marker to distinguish them from caveolae. This study highlights the necessity for caution when interpreting data from experiments using methyl-cyclodextrin as a model of caveolar dysfunction and the potential importance for the broader class of lipid rafts in signal modulation.
Nonetheless, the role of caveolae must not be completely dismissed. We and others (20, 41) have shown a direct association between caveolin and the IGF-IR and also between caveolin and IRS-1; thus it is possible that although caveolae and caveolin are not essential for mediating IGF effects through the IGF-IR, they do have a role in modulating the cellular response to IGF. A study of caveolae function in relation to the structurally similar insulin receptor suggested that the presence of caveolae may facilitate the activation of signaling molecules that favor a metabolic response to insulin (29). The receptors for insulin and IGF have common downstream effectors, and thus caveolae may be involved in directing which signaling molecules are stimulated after binding of IGF to the IGF-IR.
It is speculated that caveolae may function as platforms for cross-talk between different receptor signaling pathways; Ushio-Fukai et al. (42) demonstrated that although transactivation of EGF receptor (EGFR) by angiotensin II was cholesterol dependent, cholesterol was not required for activation of the EGFR by its cognate ligand. IGF-I-mediated activation of IGF-IR is known to cause ligand-independent activation of the EGFR (43, 44), and phosphorylation of unbound IGF-IR has been observed after activation of the angiotensin (45) and EGF (46) receptors. It is possible that caveolin may play a key role in mediating this transactivation.
The structure of caveolae may underlie their ability to organize receptor cross-talk because they could represent an internalization pathway for lipid raft domains. The function of caveolae may therefore be analogous to that proposed for clathrin-coated pits, where intracellular trafficking of endocytic vesicles to endosomes serves to amplify and propagate signals (47). Indeed, it has been shown that stimulation with EGF recruits caveolin to early endosomal compartments (48); the coalescence of multiple internalized caveolae to form so-called caveosomes may thus provide an additional compartment for intracrine signaling.
The function of caveolae may therefore be 2-fold, first to direct cell function by sorting the downstream signals of a particular receptor and second to orchestrate a much larger-scale integration of signals from different receptor species; methyl-cyclodextrin may therefore mediate its effect by uncoupling cross-talk between different signaling molecules.
Our data suggest that caveolae are not essential for cellular mitogenesis and survival in response to IGF-I, but these microdomains may still be important for fine-tuning and integrating IGF signals with those received from other growth factors and cytokines within the cellular environment.
Acknowledgments
We gratefully acknowledge Derek Le Roith, Toyoshi Fujimoto, and Martin Lowe for kind donation of cells, plasmids, and antibodies, respectively, and the Biotechnology and Biological Sciences Research Council for financial support, and we thank Dr. Carolyn J. P. Jones and Michael Mackness for technical assistance.
Footnotes
First Published Online September 15, 2005
Abbreviations: DAPI, 4',6-Diamidino-2-phenylindole; HRP, horseradish peroxidase; IGF-IR, type 1 IGF receptor; IRS-1, insulin receptor substrate 1; PI-3K, phosphatidylinositol-3-kinase; PKB, protein kinase B; TCA, trichloracetic acid.
Accepted for publication September 10, 2005.
References
Razani B, Combs TP, Wang XB, Frank PG, Park DS, Russell RG, Li M, Tang B, Jelicks LA, Scherer PE, Lisanti MP 2002 Caveolin-1-deficient mice are lean, resistant to diet-induced obesity, and show hypertriglyceridemia with adipocyte abnormalities. J Biol Chem 277:8635–8647
Cohen AW, Razani B, Wang XB, Combs TP, Williams TM, Scherer PE, Lisanti MP 2003 Caveolin-1-deficient mice show insulin resistance and defective insulin receptor protein in adipose tissue. Am J Physiol 285:C222–C235
Oshikawa J, Otsu K, Toya Y, Tsunematsu T, Hankins R, Kawabe JI, Minamisawa S, Umemura S, Hagiwara Y, Ishikawa Y 2004 Insulin resistance in skeletal muscles of caveolin-3-null mice. Proc Natl Acad Sci USA 101:12670–12675
Park D, Cohen AW, Frank PG, Razani B, Lee H, Williams TM, Chandra M, Shirani J, De Souza AP, Tang B, Jelicks LA, Factor SM, Weiss LM, Tanowitz HB, Lisanti MP 2003 Caveolin-1 null mice show dramatic reductions in life span. Biochemistry 42:15124–15131
Razani B, Engelman JA, Wang XB, Schubert W, Zhang XL, Marks CB, Macaluso F, Russell RG, Li M, Pestell RG, Di Vizio D, Hou H, Kneitz B, Lagaud G, Christ GJ, Edelmann W, Lisanti MP 2001 Caveolin-1 null mice are viable but show evidence of hyperproliferative and vascular abnormalities. J Biol Chem 276:38121–38138
Park DS, Woodman SE, Schubert AL, Cohen AW, Frank PG, Chandra M, Shirani J, Razani B, Tang B, Jelicks LA, Factor SM, Weiss LM, Tanowitz HB, Lisanti MP 2002 Caveolin-1/3 double-knockout mice are viable, but lack both muscle and non-muscle caveolae, and develop a severe cardiomyopathic phenotype. Am J Pathol 160:2207–2217
Mineo C, James GL, Smart EJ, Anderson RGW 1996 Localization of epidermal growth factor-stimulated Ras/Raf-1 interaction to caveolae membrane. J Biol Chem 271:11930–11935
Yamamoto M, Toya Y, Jensen RA, Ishikawa Y 1999 Caveolin is an inhibitor of platelet-derived growth factor receptor signaling. Exp Cell Res 247:380–388
Li S, Couet J, Lisanti MP 1996 Src tyrosine kinases, G subunits, and H-Ras share a common membrane-anchored scaffolding protein, caveolin. J Biol Chem 271:29182–29190
Chun M, Liyanage U, Lisanti MP, Lodish HF 1994 Signal transduction of a G protein-coupled receptor in caveolae: colocalization of endothelin and its receptor with caveolin. Proc Natl Acad Sci USA 91:11728–11732
Schlegel A, Wang C, Pestell RG, Lisanti MP 2001 Ligand-independent activation of oestrogen receptor by caveolin-1. Biochem J 359:203–210
Gustavsson J, Parpal S, Karlsson M, Ramsing C, Thorn H, Borg M, Lindroth M, Peterson KH, Magnusson KE, Stralfors P 1999 Localization of the insulin receptor in caveolae of adipocyte plasma membrane. FASEB J 13:1961–1971
Parpal S, Karlsson M, Thorn H, Stralfors P 2001 Cholesterol depletion disrupts caveolae and insulin receptor signalling for metabolic control via insulin receptor substrate-1, but not for mitogen-activated protein kinase control. J Biol Chem 276:9670–9678
Yamamoto M, Toya Y, Schwenke C, Lisanti MP, Myers MG, Ishikawa Y 1998 Caveolin is an activator of insulin receptor signaling. J Biol Chem 273:26962–26968
Coppola D, Ferber A, Miura M, Sell C, D’Ambrosio C, Rubin R, Baserga R 1994 A functional insulin-like growth factor I receptor is required for the mitogenic and transforming activities of the epidermal growth factor receptor. Mol Cell Biol 14:4588–4595
Benito M, Valverde AM, Lorenzo M 1996 IGF-I: a mitogen also involved in differentiation processes in mammalian cells. Int J Cell Biol 28:499–510
Leventhal PS, Feldman EL 1997 Insulin-like growth factors as regulators of cell motility. Trends Endocrinol Metab 8:1–6
Baudry A, Lamothe B, Bucchini D, Jami J, Montarras D, Pinset C, Joshi RL 2001 IGF-I receptor as an alternative receptor for metabolic signaling in insulin receptor-deficient muscle cells. FEBS Lett 488:174–178
O’Connor R, Kauffmann-Zeh A, Liu Y, Lehar S, Evan GI, Baserga R, Blattler WA 1997 Identification of domains of the insulin-like growth factor I receptor that are required for protection from apoptosis. Mol Cell Biol 17:427–436
Huo H, Guo X, Hong S, Jiang M, Liu X, Liao K 2003 Lipid rafts/caveolae are essential for insulin-like growth factor-1 receptor signaling during 3T3–L1 preadipocyte differentiation induction. J Biol Chem 278:11561–11569
Hong S, Huo H, Xu J, Liao K 2004 Insulin-like growth factor-1 receptor signalling in 3T3–l1 adipocyte differentiation requires lipid rafts but not caveolae. Cell Death Differ 11:714–723
Christian AE, Haynes MP, Phillips MC, Rothblat GH 1997 Use of cyclodextrins for manipulating cellular cholesterol content. J Lipid Res 38:2264–2272
Hailstones D, Sleer LS, Parton RG, Stanley KK 1998 Regulation of caveolin and caveolae by cholesterol in MDCK cells. J Lipid Res 39:369–379
Blakesely VA, Kato H, Roberts CT, LeRoith D 1995 Mutation of a conserved amino acid residue (tryptophan 1171) in the tyrosine kinase domain of the IGF-I receptor abolishes autophosphorylation but does not eliminate biologic function. J Biol Chem 270:2764–2769
Smart EJ, Ying Y, Mineo C, Anderson RGW 1995 A detergent-free method for purifying caveoolae membrane from tissue culture cells. Proc Natl Acad Sci USA 92:10104–10108
Murata M, Peranen J, Schreiner R, Wieland F, Kurzchalia TV, Simons K 1995 VIP21/caveolin is a cholesterol-binding protein. Proc Natl Acad Sci USA 92:10339–10343
Podar K, Tai YT, Cole CE, Hideshima T, Sattler M, Hamblin A, Mitsiades N, Schlossman RL, Davies FE, Morgan GJ, Munshi NC, Chauhan D, Anderson KC 2003 Essential role of caveolae in IL-6- and IGF-I-triggered Akt-1-mediated survival of multiple myeloma cells. J Biol Chem 278:5794–5801
Partovian C, Simons M 2004 Regulation of protein kinase B/Akt activity and Ser473 phosphorylation by protein kinase C in endothelial cells. Cell Signal 16:951–957
Cohen AW, Combs TP, Scherer PE, Lisanti MP 2003 Role of caveolin and caveolin in insulin signaling and diabetes. Am J Physiol Endocrinol Metab 285:E1151–E1160
Peres C, Yart A, Perret B, Salles JP, Raynal P 2003 Modulation of phosphoinositide 3-kinase activation by cholesterol level suggests a novel positive role for lipid rafts in lysophosphatidic acid signalling. FEBS Lett 534:164–168
Halevy O, Cantley LC 2004 Differential regulation of the phosphoinositide 3-kinase and MAP kinase pathways by hepatocyte growth factor vs. insulin-like growth factor-I in myogenic cells. Exp Cell Res 297:224–234
Radcliff K, Tang TB, Lim J, Zhang Z, Abedin M, Demer, LL, Tintut Y 2005 Insulin-like growth factor-I regulates proliferation and osteoblastic differentiation of calcifying vascular cells via extracellular signal-regulated protein kinase and phosphatidylinositol 3-kinase pathways. Circ Res 96:398–400
Moelling K, Schad K, Bosse M, Zimmermann S, Schweneker M 2002 Regulation of Raf-Akt crosstalk. J Biol Chem 277:31099–31106
Zhuang L, Lin J, Lu ML, Solomon KR, Freeman MR 2002 Cholesterol-rich lipid rafts mediate akt-regulated survival in prostate cancer cells. Cancer Res 62:2227–2231
Furuchi T, Anderson RGW 1998 Cholesterol depletion of caveolae causes hyperactivation of extracellular signal-regulated kinase (ERK). J Biol Chem 273:21099–21104
Hua H, Munk S, Whiteside CI 2003 Endothelin-1 activates mesangial cell ERK1/2 via EGF-receptor transactivation and caveolin-1 interaction. Am J Physiol Renal Physiol 284:F303–F312
Karlsson M, Thorn H, Danielsson A, Stenkula KG, Ost A, Gustavsson J, Nystrom FH, Stralfors P 2004 Colocalisation of insulin receptor and insulin receptor substrate-1 to caveolae in primary human adipocytes. Cholesterol depletion blocks insulin signalling for metabolic and mitogenic control. Eur J Biochem 271:2471–2479
Nystrom F, Chen H, Cong L, Li Y, Quon MJ 1999 Caveolin-1 interacts with the insulin receptor and can differentially modulate insulin signaling in transfected Cos-7 cells and rat adipose cells. Mol Endocrinol 13:2013–2024
Zimmermann S, Moelling K 1999 Phosphorylation and regulation of Raf by Akt (protein kinase B). Science 286:1741–1744
Rommel C, Clarke BA, Zimmermann S, Nunez L, Rossman R, Reid K, Moelling K, Yancopoulos GD, Glass DJ 1999 Differentiation stage-specific inhibition of the Raf-MEK-ERK pathway by Akt. Science 286:1738–1741
Panetta D, Biedi C, Repetto S, Cordera R, Maggi D 2004 IGF-I regulates caveolin 1 and IRS-1 interaction in caveolae. Biochem Biophys Res Commun 316:240–243
Ushio-Fukai M, Hilenski L, Santanam N, Becker PL, Ma X, Griendling KK, Alexander RW 2001 Cholesterol depletion inhibits epidermal growth factor receptor transactivation by angiotensin II in vascular smooth muscle cells. J Biol Chem 276:48269–48275
Pietrzkowski Z, Sell C, Lammers R, Ullrich A, Baserga R 1992 Roles of insulinlike growth factor 1 (IGF-1) and the IGF-1 receptor in epidermal growth factor-stimulated growth of 3T3 cells. Mol Cell Biol 12:3883–3889
Gilmore AP, Valentijn AJ, Wang P, Ranger AM, Bundred N, O’Hare MJ, Wakeling A, Korsmeyer SJ, Streuli CH 2002 Activation of BAD by therapeutic inhibition of epidermal growth factor receptor and transactivation by insulin-like growth factor receptor. J Biol Chem 277:27643–27650
Zahradka P, Litchie B, Storie B, Helwer G 2004 Transactivation of the insulin-like growth factor-I receptor by angiotensin II mediates downstream signaling from the angiotensin II type 1 receptor to phosphatidylinositol 3-kinase. Endocrinology 145:2978–2987
Roudabush FL, Pierce KL, Maudsley S, Khan KD, Luttrell LM 2000 Transactivation of the EGF receptor mediates IGF-I-stimulated Shc phosphorylation and ERK1/2 activation in COS-7 cells. J Biol Chem 275:22583–22589
Di Fiore P, Gill GN 1999 Endocytosis and mitogenic signaling. Curr Opin Cell Biol 11:483–488
Pol A, Lu A, Pons M, Peiro S, Enrich K 2000 Epidermal growth factor-mediated caveolin recruitment to early endosomes and MAPK activation. J Biol Chem 275:30566–30572(Laura C. Matthews, Michael J. Taggart an)