| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Howard Hughes Medical Institute, Cox Institute, Cell and Molecular Biology Graduate Group (E.L.W., M.J.B.), and the Department of Medicine (J.J.C., M.J.B.), University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104
Address all correspondence and requests for reprints to: Morris J. Birnbaum, Howard Hughes Medical Institute, University of Pennsylvania School of Medicine, 415 Curie Boulevard, 322 Clinical Research Building, Philadelphia, Pennsylvania 19104. E-mail: birnbaum{at}mail.med.upenn.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
An experimental system often used to address this problem has been the cultured murine adipocyte 3T3-L1 model, in which both insulin and platelet-derived growth factor (PDGF) activate phosphatidylinositol 3-kinase (PI3K) comparably as measured by in vitro assay, yet only insulin stimulates glucose uptake and GLUT4 translocation (1, 2). In 3T3-L1 adipocytes as well as in vivo target tissues, the activated insulin receptor tyrosine kinase phosphorylates a soluble docking molecule, insulin receptor substrate (IRS), on tyrosine residues (reviewed in Ref.3). Phosphorylated IRS associates with a number of Src homology 2 (SH2) domain-containing proteins, of which the most important in regard to metabolic actions appears to be PI3K (4). Activation of this lipid kinase, which generates 3'-phosphoinositide species, is an absolute requirement for the stimulation of glucose uptake (5, 6). The PI3K-dependent enzymes that direct translocation of the specialized GLUT4 glucose transporter to the cell surface are still not clearly defined (7, 8), but a substantial amount of research suggests that activation of the serine/threonine protein kinase Akt (PKB) plays a significant role in this process (Refs.9, 10 , and reviewed in Ref.11). In addition, an alternative, parallel pathway has been implicated recently in insulin action (12). The ability of PDGF to activate many of these same signaling pathways, and in particular PI3K, in many undifferentiated cells stands in marked contrast to the failure of this growth factor to elicit GLUT4 translocation in 3T3-L1 adipocytes.
Recently we reported that the inability of PDGF to stimulate glucose uptake into 3T3-L1 adipocytes correlated with a lack of phosphorylation of Akt and, moreover, the loss of PDGF-dependent Akt activation coincided with differentiation of preadipocytes into mature fat cells (2). We interpreted this result as indicative of a differentiation-dependent suppression of signaling to Akt because PDGF was still able to promote phosphorylation of p70S6K and phosphorylated heat and acid-stable protein regulated by insulin (PHAS)-I/4E-binding protein 1 in adipocytes. The latter data are compatible with prior reports that PDGF stimulates serine/threonine phosphorylation of IRS-1 in 3T3-L1 adipocytes, thus inhibiting the actions of subsequently added insulin (13, 14). In contrast, a recent study suggested that effective PDGF receptors are completely lacking in adipocytes and that growth factor-dependent signaling events reported in the past derive from contaminating, undifferentiated fibroblasts (15). Resolving this issue is of some importance because the inability of PDGF to activate GLUT4 translocation has been cited as an argument in support of the idea that a unique early component of insulin signaling, possibly the translocation of IRS to a critical intracellular locale, is required for the metabolic actions of the hormone (Refs.16, 17 , and reviewed in Ref.18). One compromise might be that PDGF receptors are expressed at low levels in 3T3-L1 adipocytes, possibly below the limits of detection by immunofluorescence but sufficiently high to couple to some but not all signaling events. Here we address several questions related to this problem. First, we determined the capacity of PDGF receptors to signal in parental 3T3-L1 adipocytes. Next, by overexpressing wild-type and mutant human PGDF receptor (hPDGFR) in 3T3-L1 adipocytes, we determined the minimal receptor elements required to promote glucose transport and GLUT4 translocation. In addition, we assayed the effects of PDGF on key signaling intermediates such as IRS, phosphatidylinositol-3,4,5-trisphosphate (PIP3), Akt, and c-cbl and correlate them to both the insulin effect and the ability of the hPDGFR to elicit glucose uptake. In this way, we can gain insight into those signaling components that are required for receptor tyrosine kinase-mediated glucose transport and test existing ideas regarding the signaling modules that confer specificity to the insulin signaling cascade.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
/ß antibodies, polyclonal rabbit anti-IRS-1 antibodies, polyclonal rabbit anti-IRS-2 antibodies, and polyclonal anti-Grb2-associated binding protein (Gab)-1 antibodies, Upstate Biotechnology (Lake Placid, NY); Cy2-donkey-antisheep and Cy3-donkey-antirabbit antibodies, Jackson Immunoresearch (West Grove, PA); polyclonal anti-c-cbl antibodies, Transduction Laboratories (Lexington, KY); and all horseradish peroxidase (HRP)-conjugated secondary antibodies, antiphosphotyrosine monoclonal HRP-PY99, and antiphosphotyrosine monoclonal PY20 antibodies, Santa Cruz Biotechnology (Santa Cruz, CA). The sheep anti-GLUT4 antibodies were raised to the carboxy terminus of GLUT4 (2). The green fluorescent protein (GFP)-Grp1 pleckstrin homology (PH) fusion cDNA (provided by Michael P. Czech, University of Massachusetts, and Mark A. Lemmon, University of Pennsylvania) was inserted into replication-incompetent adenovirus and amplified by the Adenovirus Core Facility (University of Pennsylvania).
Wortmannin was obtained from Sigma (St. Louis, MO), porcine insulin was a gift from Eli Lilly (Indianapolis, IN), [1,2-3H]-2-deoxy-D-glucose was bought from NEN Life Science Products (Boston, MA), and PDGF-BB was purchased from Life Technologies, Inc. (Rockville, MD).
Cell culture, PDGFR constructs, retroviral infection, and adenoviral infection
3T3-L1 fibroblasts were maintained and differentiated into adipocytes as described previously (19). Before analysis, 3T3-L1 adipocytes were serum starved as described in the figure legends and subsequently stimulated with 100 nM insulin or 50 ng/ml PDGF.
The constructs encoding human PDGFR-ß and its variants in the retroviral vector pLXSN (CLONTECH, Palo Alto, CA) were generously provided by Andrius Kazlauskas (Schepens Eye Research Institute, Harvard Medical School, Boston, MA) and have been characterized (20). Briefly, we created 3T3-L1 cell lines stably expressing wild-type human PDGFR-ß and mutant variants F5, F740/751, Y740/751, which have key tyrosine residues changed to phenylalanine. The F5 receptor is mutated at tyrosines 740/751 (PI3K-binding site), 771 (Ras GTPase-activating protein site), 1009 (src homology-containing phosphotyrosine phosphatase-2 site), and 1021 (phospholipase C
site). The Y740/751 version is mutated similarly at each of these sites but retains a functional PI3K-binding site, and the F740/751 receptor contains mutations only at the PI3K site.
Stable cell lines were created by retrovirus-mediated gene transfer. Two pantropic retroviral packaging constructs, pVSV G and pCgp (provided by Michael H. Malim, Guy’s Hospital, London, UK), were transiently cotransfected into 293T cells with the PDGFR construct of interest (21). After 24 h, the cell-free supernatants containing the retrovirus encoding PDGFR were used to infect 3T3-L1 fibroblasts (22). Stably transduced cells were selected 48 h later in media containing 600 µg/ml G418, and colonies of neomycin-resistant cells were pooled. The cells were maintained in G418-containing media until initiation of the differentiation protocol.
The fusion protein GFP-Grp1 PH was expressed in 3T3-L1 adipocytes by adenoviral infection (multiplicity of infection, 1600) according to the optimized method of Orlicky and Schaack (23). Following infection, the cells were allowed to recover in complete media (DMEM with 10% fetal bovine serum). Cells were starved overnight in DMEM supplemented with 0.5% BSA and then transferred just before experimentation to Leibovitz’s L-15 media containing 0.2% BSA. Live cell imaging experiments (see below) were performed 36–48 h after infection.
Single-cell immunofluorescence assay for MAPK activation
3T3-L1 adipocytes were grown and differentiated on coverslips and stimulated with growth factor as described in the figure legends. Immunocytochemistry was performed as described (24). Samples were incubated with primary antibodies (rabbit antiphospho-MAPK diluted 1:50 and sheep anti-GLUT4 diluted 1:100) overnight at 4 C. Coverslips were incubated in secondary antibodies (Cy3-donkey-antirabbit diluted 1:2000 and Cy2-donkey-antisheep diluted 1:200) for 20 min to 1 h at room temperature. A solution of bis-benzimide (0.002%) was used to stain the nuclei. The coverslips were mounted on slides using a media containing 0.2% n-propylgallate and 80% glycerol in PBS and affixed with nail polish. Slides were imaged using a x40 objective. Digital image acquisition, processing, and automated quantitation of fluorescence intensity were performed using the MetaMorph imaging system (Universal Imaging Corp., West Chester, PA). The automated quantitation was performed by using the signal obtained by staining the nuclei of bona fide adipocytes (i.e. those displaying distinct, perinuclear GLUT4 staining) to create a mask. The mask was then transferred to the corresponding image stained with phospho-MAPK antibody and the average fluorescence intensity measured. Three randomly selected fields of view were imaged and processed in this manner for each slide, and experiments were typically done using duplicate samples. Approximately 15 verified adipocytes were measured in each field of view.
Glucose transport and GLUT4 translocation assays
Methods for measuring glucose uptake rates and GLUT4 translocation by plasma membrane sheets assay have been described (9, 25). Wortmannin pretreatment was carried out as done previously (2).
Lysate preparation, immunoprecipitation, and protein immunoblotting
Cells were lysed under nondenaturing conditions in high-salt lysis buffer (1% Triton X-100, 10% glycerol, 20 mM Tris, 200 mM NaF, 150 mM NaCl, 5 mM EDTA, 1 mM EGTA, with protease and phosphatase inhibitors, pH 7.5).
IRS-1 and IRS-2 were immunoprecipitated from adipocyte lysates in the high-salt lysis buffer described above (26). Approximately 500 µg protein were incubated overnight with 4 µg antibody at 4 C. Protein A-agarose was added to collect the immune complexes, and the antibody conjugates were washed three times with lysis buffer. All immune complexes were solubilized in Laemmli buffer and resolved by SDS-PAGE. Using the same immunoprecipitation methods, 300 µg protein was incubated with 1 µg Gab-1 antibody and 1.5 mg protein was incubated with 3 µg c-cbl antibody.
Western blots of total cell lysates and immunoprecipitated proteins were analyzed as described previously (27). In brief, nitrocellulose membranes were blocked for at least 1 h in 5% milk-TBST (Tris-buffered saline with 0.2% Tween 20), incubated in primary antibody at least 1 h in 3% milk-TBST, washed three times with TBST, incubated in secondary HRP-conjugated antibody 45 min to 2 h in 3% milk-TBST, washed three times in TBST, and the protein bands detected by enhanced chemiluminescence (Amersham Pharmacia, Uppsala, Sweden). The IRS-1, IRS-2, PDGFR, Gab-1, c-cbl, phospho-tyrosine, and phospho-MAPK antibodies were used at concentrations recommended by the manufacturer.
The standard Western blotting procedure was adjusted for detection of phospho-Akt. First, PBS was substituted for TBST in each wash and incubation. In addition, incubations with primary antibody were performed overnight at 4 C in 3% milk-PBS. Phospho-Akt (T308) was diluted 1:500, and phospho-Akt (S473) 1:1000.
Densitometric analysis of the immunoblots was performed by scanning each film and then quantitating the intensity of each band using NIH Image software.
PI3K assays
Assay of PI3K activity associated with IRS-1 or IRS-2 immunoprecipitates was performed essentially as described previously (28). The only modification was the use of high-salt lysis buffer for cell lysis and in the first set of three washes of the immune complex.
Real-time confocal microscopy
The 3T3-L1 adipocytes infected with GFP-Grp1 PH adenovirus were maintained in the 
TC3 open-dish system, which consists of individual 0.15-mm-thick tissue culture dishes optimized for use with oil immersion microscope objectives, an objective heater, and a stage warmer to maintain the living cells at 37 C throughout the experiment (Bioptechs, Inc., Pittsburgh, PA). Images were acquired using an Ultraview LCI Nipkow disc confocal microscope (Perkin-Elmer, Inc., Norwalk, CT) attached to a Nikon TE300 inverted microscope (Nikon, Melville, NY) fitted with a x60 oil immersion objective. The GFP-Grp1 PH was visualized using the 488-nm line of an argon laser, and the combination set of dichroic mirror (488 nm) and emission barrier filter (cutoff at 510 nm) was optimized to collect the GFP signal. Images were collected at 7.5-sec intervals before and after the addition of either 50 ng/ml PDGF-BB or 100 nM insulin.
Recruitment of GFP-Grp1 PH to the plasma membrane was quantitated using line intensity profiles drawn across each cell. These profiles were used to measure average pixel intensities at the plasma membrane (in two places), in the cytosol, and in the background (in two places). The amplitude of the plasma membrane fluorescence intensity is defined as (plasma membrane - cytosol), and the total fluorescence intensity is given by (plasma membrane + cytosol).
We determined the value of (plasma membrane - cytosol)/(plasma membrane + cytosol) at 0 and 2 min following stimulation with growth factor. Then, we calculated (plasma membrane - cytosol)/(plasma membrane + cytosol) to measure the effects of PDGF and insulin on GFP-Grp1 PH translocation and in vivo PIP3 production.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). In unstimulated cells, approximately 6% of the adipocytes displayed MAPK activation. Adipocytes treated with PDGF or insulin demonstrated 19 or 50% cells staining positive for MAPK activation, respectively, revealing that endogenous PDGFRs were expressed and were functional in parental 3T3-L1 adipocytes.
|
, Ref.2 , and reviewed in Ref.4). We considered two possibilities to explain these observations: First, endogenous PDGF receptors are expressed below a threshold necessary to activate glucose uptake; or second, the stimulation of the PDGF signaling pathway is incapable of activating a requisite intermediate, such as IRS, that is critical to the transmission of the signals that promote glucose transport. To test these alternative explanations, we overexpressed the panel of wild-type human PDGFR-ß and mutant receptors depicted in Fig. 2A in 3T3-L1 cells by retroviral infection. Extracts prepared from adipocytes were probed using a polyclonal antibody to hPDGFR, and each cell line showed roughly equivalent hPDGFR expression (Fig. 2B). In addition, stimulation with PDGF-BB revealed an electrophoretic mobility shift, confirming that the receptors were expressed on the cell surface and were capable of appropriate autophosphorylation.
|
). We measured immunoreactivity with phosphotyrosine antisera of the PDGFR in control (pLXSN) and wild-type hPDGFR adipocytes treated with PDGF as in Fig. 2C. We calculated a 5.6-fold increase in phosphotyrosine (PDGFR) signal in the wild-type hPDGFR-expressing adipocytes, compared with the control cells. Based on an earlier study that reports about 27,000 PDGF-binding sites per 3T3-L1 adipocyte, we estimated that our wild-type hPDGFR adipocytes contain approximately 151,000 PDGF-binding sites per cell (29). Interestingly, 3T3-L1 preadipocytes (fibroblasts) contain roughly 158,000 PDGF-binding sites per cell (29). Thus, hPDGFR adipocytes express approximately equivalent levels of PDGFR as observed in parental 3T3-L1 cells before initiation of the adipogenic program and concomitant down-regulation of PDGFR expression.
The activation of PI3K and subsequent generation of PIP3 are required for the insulin-dependent stimulation of GLUT4 translocation and glucose transport (reviewed in Ref.30). Accordingly, we chose to measure PIP3 generation using a single-cell assay in intact adipocytes overexpressing different PDGFRs. We expressed GFP-Grp1 PH, a high-affinity binding protein selective for PIP3, in adipocytes by adenoviral infection, and visualized the translocation of the fluorescent fusion protein to the plasma membrane in response to growth factor treatment by real-time, confocal microscopy (Fig. 3A). In all cell lines, insulin augmented PIP3 accumulation as indicated by GFP-Grp1 PH translocation to the plasma membrane, forming rims at the cell periphery. In parental adipocytes (pLXSN) and those overexpressing hPDGFR lacking a functional PI3K activation site (F740/751), PDGF weakly stimulated the redistribution of GFP-Grp1 PH. However, adipocytes expressing hPDGFR containing an intact PI3K-binding site (wild-type PDGFR and Y740/751) showed significant GFP-Grp1 PH translocation on addition of PDGF. Quantitation of the GFP-Grp1 PH plasma membrane rims performed by fluorescence intensity measurements across each cell demonstrated that PDGF-stimulated PIP3 accumulation in wild-type or Y740/751 hPDGFR adipocytes was quite similar to that produced in response to insulin in these cells (Fig. 3B).
|
). In each cell line, insulin treatment resulted in phosphorylation at both sites, whereas PDGF stimulated substantial Akt phosphorylation in the wild-type and Y740/751 hPDGFR expressing cell lines but not in the control cells (pLXSN) or those cells overexpressing hPDGFR without a functional PI3K binding site (F5 and F740/751).
|
, A and B). Moreover, PDGF-stimulated hexose uptake occurred to an equivalent extent as the insulin effect. Similarly, we assayed GLUT4 translocation by the plasma membrane sheets assay. The results correlated perfectly with the glucose transport data: PDGF and insulin stimulated to an equivalent degree GLUT4 recruitment to the plasma membrane in the wild-type and Y740/751 hPDGFR expressing cells, but only insulin stimulated GLUT4 translocation in the pLXSN, F5, and F740/751 adipocytes (Fig. 5, C and D). Simultaneous addition of PDGF and insulin to wild-type or Y740/751 hPDGFR adipocytes did not further increase GLUT4 translocation above the levels measured for each growth factor individually. In addition, both PDGF and insulin-dependent GLUT4 translocation were blocked by pretreatment with the PI3K inhibitor wortmannin (Fig. 6).
|
|
). As a more sensitive assay to determine whether PDGF is altering downstream signaling via intermediary IRS proteins, we measured PI3K activity associated with IRS-1 and IRS-2 using an in vitro enzymatic assay. Insulin activated PI3K associated with IRS-1 and IRS-2 in both control adipocytes and hPDGFR adipocytes (Fig. 7, B and C). There is no detectable increase in PI3K activity in the IRS-1 or IRS-2 immune complexes in response to PDGF in either cell type. This is consistent with the idea that PDGF can effectively mimic insulin action in 3T3-L1 adipocytes by a pathway that does not require insulin receptor substrates.
|
Finally, we examined the effects of PDGF on the tyrosine phosphorylation of c-cbl, the multivalent adapter protein shown recently to be an important component of an alternative, PI3K-independent pathway required for glucose transport in adipocytes (12). Hence, we immunoprecipitated c-cbl from control (pLXSN) and hPDGFR adipocytes treated with PDGF or insulin and probed the immunoblots with antiphosphotyrosine antibodies (Fig. 8A). We noted that although insulin specifically activates c-cbl tyrosine phosphorylation in control adipocytes as reported previously (32), PDGF strongly evoked the tyrosine phosphorylation of this protein in hPDGFR adipocytes. These results were quantitated by densitometry (Fig. 8B). PDGF strongly stimulated the phosphorylation of cbl in hPDGFR adipocytes capable of initiating glucose uptake (wild type, Y740/751) as well as in those cells which do not demonstrate robust glucose transport in response to PDGF (F740/751).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In 3T3-L1 adipocytes, insulin, but not PDGF, activates glucose transport (1, 2). Although recently it has been suggested that PDGFRs are lacking in cultured murine fat cells (15), other data would suggest that even though the expression of PDGFR is dramatically reduced during differentiation (29), the mitogen is still capable of signaling (Fig. 1 and Refs.2, 13, 14). Thus, two alternative models might explain the specificity of insulin in stimulating glucose transport in fat cells, i.e. that the pathway from PDGFR is selectively suppressed or that the strength of the PDGF-dependent signal is inadequate to produce a metabolic response. To test the latter hypothesis, we overexpressed PDGFR in 3T3-L1 adipocytes, reasoning that if the specificity were conferred by dosage, then increasing the density of PDGFR would enable the mitogen to elicit GLUT4 translocation. Although previous studies investigating glucose uptake used overexpressed, constitutively active PI3K to circumvent the requirement for insulin addition (33, 34, 35, 36, 37, 38), our system relies on the activation of endogenous pools of PI3K. Evidence that the degree of PDGFR overexpression was not excessive is that the reactivity of PDGFR with antiphosphotyrosine antisera was comparable with that of endogenous IRS-1/2 (Fig. 2C) and that PDGFR levels in hPDGFR adipocytes, compared closely with 3T3-L1 preadipocytes, suggesting physiological activation of downstream signaling molecules like PI3K. Nonetheless, moderate overexpression of PDGFR in 3T3-L1 adipocytes confers upon PDGF the capacity to elicit insulin-like effects such as robust generation of PIP3 in vivo, Akt activation, glucose uptake, and GLUT4 translocation. Moreover, we show that these signaling events depend on the presence of a functional PI3K activation site on the PDGFR. Because simultaneous addition of insulin and PDGF does not increase GLUT4 translocation above levels observed with either hormone alone, the data suggest that insulin and PDGF each maximally stimulate transport by the receptor tyrosine kinase-activated, PI3K-dependent signaling pathways that exist in fat cells. The simplest interpretation of these data is that the specificity achieved on differentiation into fat cells is accomplished by a reduction in the number of PDGFR, such that coupling to some downstream events such as MAPK activation is preserved, but the signal is insufficient to stimulate glucose transport.
The present study complements and extends previous reports investigating the effects of receptor tyrosine kinase overexpression on glucose transport in several ways. Overexpression of epithelial growth factor receptors in 3T3-L1 adipocytes results in the enhanced stimulation of glucose uptake and GLUT4 translocation to the plasma membrane without the involvement of insulin receptor or IRS-1 (39). The overexpression of PDGFR in CHO-GLUT4myc cells demonstrates that PDGF may activate PI3K, glucose uptake, and GLUT4myc translocation and that these effects are wortmannin sensitive and require the presence of a functional PI3K docking site on the overexpressed receptor (40). Finally, a recently published study (15) demonstrated that overexpression of PDGFR in 3T3-L1 adipocytes conferred the capacity for significant PDGF-stimulated Akt phosphorylation and GLUT4-EGFP translocation.
In contrast to our present study, however, Shigematsu et al. (15) find that mature adipocytes expressing well-characterized markers of the differentiated state do not express detectable PDGFR as assayed by single-cell immunofluorescence assay. However, we find that MAPK is activated by PDGF in bona fide adipocytes expressing GLUT4 and showing appropriate adipocyte morphology. Moreover, PDGF pretreatment of 3T3-L1 adipocytes results in retardation in the electrophoretic mobility of IRS-1 proteins because of serine phosphorylation as well as the loss of insulin-stimulated IRS-1 tyrosine phosphorylation and IRS-1-associated PI3K activity (13, 14). Because IRS-1 protein levels are increased 10- to 20-fold during adipogenesis, this protein is expressed preferentially in adipocytes (41). Therefore, several lines of evidence suggest that the PDGF signaling pathway can be activated to some extent in mature, 3T3-L1 adipocytes.
In addition to the demonstration that the inability of PDGF to induce GLUT4 translocation is due to a relative lack of receptors, the second important finding presented above is that in cells overexpressing sufficient PDGFR, PDGF-stimulated glucose uptake is driven independently of IRS-1 and IRS-2, the two IRS isoforms found in 3T3-L1 adipocytes (42, 43). The IRS proteins are obligate intermediates in the maintenance of glucose homeostasis in vivo (44, 45, 46). In recent years, the notion that IRS escorts PI3K to the appropriate subcellular compartments and thus imparts upon the insulin signaling cascade the unique ability to drive GLUT4 translocation gained wide acceptance (reviewed in Ref.18). In 3T3-L1 adipocytes, insulin-activated PI3K is localized to the low-density microsome fraction, but PDGF-stimulated PI3K is found primarily in the plasma membrane and endosomal fractions (16, 17). Yet our results demonstrate that both PDGF and insulin can stimulate PIP3 accumulation at or near the plasma membrane (Fig. 3). In agreement with this observation, a previous study revealed that insulin can stimulate the tyrosine phosphorylation of IRS proteins closely apposed to the plasma membrane, and some IRS-associated PI3K activity is also present in that locale (47). Most importantly, we show conclusively using a novel strategy that the involvement of IRS-1/2 is absolutely not required for hormone-stimulated GLUT4 translocation and glucose uptake. In the present report, we not only show that IRS-associated PI3K cannot be detected in response to PDGF but also provide an alternative scaffold for PI3K: the PDGF receptor itself. It is the combination of both results that makes the current report a compelling demonstration that IRS is not required for PDGF-stimulated GLUT4 translocation.
We also addressed the possibility that our overexpressed hPDGFR are interacting with another soluble, scaffolding molecule related to the IRS proteins, such as Gab-1. We found that PDGF stimulates the tyrosine phosphorylation of Gab-1 in each hPDGFR adipocyte line tested, and the level of Gab-1 phosphorylation achieved is substantially greater than that observed in the control adipocytes (data not shown). Because Gab-1 is comparably phosphorylated in hPDGFR adipocytes that experience PDGF-stimulated PIP3 accumulation and glucose uptake (WT hPDGFR, Y740/751 hPDGFR) and in hPDGFR adipocytes that do not (F740/751 hPDGFR), we believe that it is highly unlikely that Gab-1 acts as the critical intermediary docking protein associated with robust PI3K activity leading to glucose uptake.
Finally, we investigated the tyrosine phosphorylation of c-cbl in adipocytes overexpressing hPDGFR. We found that expression of hPDGFR significantly increased the tyrosine phosphorylation of this adapter protein in response to PDGF above levels observed following stimulation with insulin (Fig. 8, A and B). Like activation of Akt and glucose transport, we now show that this specificity is only related to the strength of the signal and not to any fundamental differences in the signaling capabilities of the two receptors.
Thus, we can make three broad conclusions from our observations: 1) activation of PDGFR in parental 3T3-L1 adipocytes leads to stimulation of various pathways in a dose-dependent manner; 2) PDGF-stimulated Akt phosphorylation and GLUT4 translocation depend absolutely on the presence of an adequate number of PDGFR containing an intact PI3K-binding site, although carboxy-terminal sites are expendable; and 3) PDGF-stimulated Akt phosphorylation and glucose transport do not require coupling of PI3K to IRS. By exploiting this novel system in which glucose transport may be driven independently of insulin, insulin receptor, and IRS proteins, we may be able to distinguish the most fundamental, basic requirements for successful GLUT4 translocation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: Gab-1, Grb2-associated binding protein; GFP, green fluorescent protein; GLUT4, insulin-responsive glucose transporter; Grp, general receptor for 3'-phosphoinositides; hPDGFR, human PDGFR; HRP, horseradish peroxidase; IRS, insulin receptor substrate; PDGF, platelet-derived growth factor; PDGFR, PGDF receptor; PH, pleckstrin homology; PHAS-I, phosphorylated heat and acid-stable protein regulated by insulin; PI3K, phosphatidylinositol 3 kinase; PIP3, phosphatidylinositol-3,4,5-trisphosphate; TBST, Tris-buffered saline with Tween 20.
Received April 17, 2003.
Accepted for publication May 20, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1 and phosphatidylinositol 3 kinase are the downstream mediators of the PDGF receptor’s mitogenic signal. Cell 73:321–334[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  M. Umahara, S. Okada, E. Yamada, T. Saito, K. Ohshima, K. Hashimoto, M. Yamada, H. Shimizu, J. E. Pessin, and M. Mori Tyrosine Phosphorylation of Munc18c Regulates Platelet-Derived Growth Factor-Stimulated Glucose Transporter 4 Translocation in 3T3L1 Adipocytes Endocrinology, January 1, 2008; 149(1): 40 - 49. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. S.L. Thong, P. J. Bilan, and A. Klip The Rab GTPase-Activating Protein AS160 Integrates Akt, Protein Kinase C, and AMP-Activated Protein Kinase Signals Regulating GLUT4 Traffic Diabetes, February 1, 2007; 56(2): 414 - 423. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Huang, T. Imamura, J. L. Babendure, J.-C. Lu, and J. M. Olefsky Disruption of Microtubules Ablates the Specificity of Insulin Signaling to GLUT4 Translocation in 3T3-L1 Adipocytes J. Biol. Chem., December 23, 2005; 280(51): 42300 - 42306. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. S. L. Thong, C. B. Dugani, and A. Klip Turning Signals On and Off: GLUT4 Traffic in the Insulin-Signaling Highway Physiology, August 1, 2005; 20(4): 271 - 284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Yoshino, S. R. G. Setty, C. Poynton, E. L. Whiteman, A. Saint-Pol, C. G. Burd, L. Johannes, E. L. Holzbaur, M. Koval, J. M. McCaffery, et al. tGolgin-1 (p230, golgin-245) modulates Shiga-toxin transport to the Golgi and Golgi motility towards the microtubule-organizing centre J. Cell Sci., May 15, 2005; 118(10): 2279 - 2293. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Steppan, J. Wang, E. L. Whiteman, M. J. Birnbaum, and M. A. Lazar Activation of SOCS-3 by Resistin Mol. Cell. Biol., February 15, 2005; 25(4): 1569 - 1575. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. D. Andl, T. Mizushima, K. Oyama, M. Bowser, H. Nakagawa, and A. K. Rustgi EGFR-induced cell migration is mediated predominantly by the JAK-STAT pathway in primary esophageal keratinocytes Am J Physiol Gastrointest Liver Physiol, December 1, 2004; 287(6): G1227 - G1237. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Yuasa, R. Kakuhata, K. Kishi, T. Obata, Y. Shinohara, Y. Bando, K. Izumi, F. Kajiura, M. Matsumoto, and Y. Ebina Platelet-Derived Growth Factor Stimulates Glucose Transport in Skeletal Muscles of Transgenic Mice Specifically Expressing Platelet-Derived Growth Factor Receptor in the Muscle, but It Does Not Affect Blood Glucose Levels Diabetes, November 1, 2004; 53(11): 2776 - 2786. [Abstract] [Full Text] [PDF]
|
|
0.05). Magnification, x400.
X signify a Y
F mutation and the loss of the activation site for the associated molecules. At right, the name of each hPDGFR variant is listed because it will be referred to throughout the paper. B, Adipocytes were serum starved overnight in DMEM with 0.5% BSA. Cells were starved an additional 2 h in Leibovitz’s L-15 media with 0.2% BSA and then either unstimulated (-) or stimulated with 50 ng/ml PDGF-BB for 2 min (P). Adipocyte lysates were resolved by SDS-PAGE, transferred to nitrocellulose, and immunoblotted with a polyclonal antibody raised to human PDGFR-ß. The approximate position of the PDGFR is indicated at the right. C, Adipocytes were serum starved as above and then either left unstimulated (-) or treated with 50 ng/ml PDGF-BB (P) or 100 nM insulin (I) for 2 min. Lysates were resolved by SDS-PAGE and transferred to nitrocellulose. Immunoblots were probed with monoclonal antiphosphotyrosine antibodies. The approximate positions of PDGFR and IRS are indicated at the right.
