help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Knight, J. B.
Right arrow Articles by Olson, A. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Knight, J. B.
Right arrow Articles by Olson, A. L.
Right arrowPubmed/NCBI databases
*Compound via MeSH
*Substance via MeSH
Endocrinology Vol. 141, No. 1 208-218
Copyright © 2000 by The Endocrine Society

ARTICLES

Expression of a Prenylation-Deficient Rab4 Interferes with Propagation of Insulin Signaling through Insulin Receptor Substrate-11

John B. Knight, Kim T. Cao, G. Victor Gibson and Ann Louise Olson

Department of Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma 73190

Address all correspondence and requests for reprints to: Dr. Ann Louise Olson, Department of Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center, P.O. Box 26901, Room 853-BMSB, Oklahoma City, Oklahoma 73190. E-mail: ann-olson{at}ouhsc.edu


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Rab proteins are small GTP-binding proteins of the Ras superfamily that function in the regulation of vesicle transport processes. The Rab4 isoform has been implicated in insulin action. For instance, overexpression of a prenylation-deficient form of Rab4 has been shown to inhibit insulin-dependent GLUT4 translocation. Other steps affected by Rab4 in the cascade of events resulting from insulin receptor activation have not been elucidated. In the present studies, we measured effects on insulin-signaling proteins in 3T3-L1 adipocytes transiently expressing cytoplasmic forms of Rab4 and Rab5. Expression of a mutant Rab4 lacking a prenylation site resulted in reduced insulin-dependent phosphorylation of cytoplasmic and internal membrane-associated insulin receptor substrate-1, leading to decreased insulin receptor substrate-1-associated phosphatidylinositol 3'-OH kinase activation and decreased Akt activation. These effects were not observed upon introduction of a similar mutant form of Rab5. These data indicate that Rab4 or a Rab4-associated protein is involved at one or more steps in propagating the insulin signal, in addition to any role it may play in the regulation of GLUT4 vesicle translocation. Our results support models of insulin signaling in which regulation of internal membrane trafficking plays a role in transduction of the insulin signal.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
MEMBERS OF THE Rab family of small guanosine triphosphatases play a central role in regulation of membrane-sorting events. To date, over 40 mammalian Rab isoforms have been identified, with specific isoforms localized to discrete membrane compartments (for recent reviews, see Refs. 1, 2). The majority of Rabs possess a geranyl-geranylation moiety at their carboxyl-terminus that presumably functions in membrane anchoring (3, 4, 5). The specificity of membrane binding is directed by elements in the hypervariable regions of the various Rab isoforms (6, 7).

Rab proteins cycle between an active GTP-bound form and an inactive GDP-bound form, with the GTP-bound form found associated with membranes. After GTP hydrolysis, the GDP-bound Rabs are removed from the membrane by cytosolic GDP disassociation inhibitor proteins and return to their membrane of origin, where a new cycle of GTP binding is initiated.

A wide range of intracellular membrane functions have been shown to be regulated by specific Rab proteins, including SNARE complex formation in yeast (8), endosome fusion (9), endoplasmic reticulum to Golgi export in yeast (10), and vesicle docking to the actin cytoskeleton (11). Rab proteins appear to function in these processes in the regulated assembly of multiprotein complexes required for specific intracellular trafficking events.

Insulin-responsive tissues express several Rab isoforms, including Rab3b, Rab4, Rab5, and Rab8 (12, 13). Of these isoforms, only Rab4 has been shown to play a role in mediating insulin actions within the cell (14, 15, 16). Expression of wild-type Rab4 at high levels or expression of a prenylation-deficient Rab4 possessing a deletion of the geranyl-geranylation site have each been shown to decrease GLUT4 translocation in response to insulin (15). In those studies, overexpression of wild-type Rab4 caused a marked increase in the cytosolic pool of Rab4, which under normal conditions is predominantly membrane associated (12, 15). It was postulated that the increased pool of cytoplasmic Rab4 may interfere with GLUT4 translocation by sequestering proteins necessary for this process. Microinjection of an antibody to the hypervariable C-terminal region of Rab4 into 3T3-L1 adipocytes inhibited not only insulin-mediated GLUT4 translocation, but also actin filament rearrangement, an insulin-mediated process that is not linked to GLUT4 translocation (16, 17). This indicates that Rab4 plays multiple roles in insulin-mediated cellular processes, including its role in GLUT4 trafficking in adipocytes.

Insulin signaling begins with activation of the intrinsic tyrosine kinase activity of the insulin receptor upon ligand binding, leading to both mitogenic and metabolic effects of the cascade. A proximal event in insulin signaling is tyrosine phosphorylation by the insulin receptor tyrosine kinase of intracellular substrates in the insulin receptor substrate (IRS) family (18). The phosphotyrosine residues on the IRS isoforms act as binding sites for specific SH2 domain molecules, including molecules such as Grb-2, Nck, SHPTP-2, and phosphatidylinositol 3'-OH kinase (PI 3-kinase) (18, 19). Activation of PI 3-kinase is required for most if not all intracellular actions of insulin, including GLUT4 translocation (20). The interaction of PI 3-kinase with IRS-1 performs two functions. First, binding of the regulatory subunits of PI 3-kinase to phosphotyrosine residues in IRS-1 increases the specific activity of the enzyme (21, 22, 23). Second, IRS-1 localizes activated PI 3-kinase to physiologically relevant compartments to form a signaling complex (24, 25).

In both rat adipocytes and 3T3-L1 adipocytes, PI 3-kinase binds predominantly to IRS-1 (26, 27). The PI 3-kinase activity immunoprecipitated with either anti-IRS-1 or anti-phosphotyrosine antibodies is increased in both intracellular membranes and cytosol after insulin stimulation (24, 26, 27, 28). Because the lipid substrates of PI 3-kinase activity are membrane phosphoinositides, it is thought that the intracellular membrane-associated PI 3-kinase is the physiologically relevant pool. Taking these observations together, the phosphorylation of membrane-associated IRS-1 and its subsequent association with PI 3-kinase are probably important for propagation of the insulin signal.

The mechanism by which IRS-1 associates with the activated insulin receptor is currently speculative. At issue is the fact that the insulin receptor, which is activated at the plasma membrane surface and internalized to recycling endosomes, is found in a separate membrane-associated pool from IRS-1 based on flotation on sucrose density gradients (24, 29). Previous studies have shown that it is the membrane-associated pool of IRS-1 that serves as the initial site of IRS-1 phosphorylation (27, 30). In this report we confirm and extend previous findings that IRS-1 and activated insulin receptor reside in distinct subcellular membrane fractions. We also show that overexpression of a mutant form of Rab4 protein lacking the prenylation site prevents optimal insulin-mediated IRS-1 phosphorylation. Our data are consistent with a model in which insulin-dependent phosphorylation of membrane-associated IRS-1 occurs by a mechanism requiring regulated vesicular trafficking.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cell culture
3T3-L1 fibroblasts were obtained form American Type Culture Collection repository (Manassas, VA) and were cultured at 37 C in 8% CO2 and maintained in DMEM containing 25 mM glucose and 10% calf serum. Confluent cultures were induced to differentiate by incubation of the cells with DMEM plus 25 mM glucose, 10% FBS, 175 nM insulin, 1 µM dexamethasone, and 0.5 mM isobutyl-1-methylxanthine. After 4 days, the medium was changed to DMEM containing 25 mM glucose, 10% FBS, and 175 nM insulin, with the incubation period continuing for an additional 3 days. Under these conditions, greater than 95% of the cell population morphologically differentiated into adipocytes.

Virus production and infections
Recombinant vaccinia viruses were constructed to express mutant human Rab4 and human Rab5 proteins using previously described methods (31). The Rab4 (Rab4{Delta}CT) and Rab5 (Rab5{Delta}CT) complementary DNAs (cDNAs) were inserted into the SalI and KpnI sites of the recombination vector after PCR amplification of the cDNA with primers containing the appropriate restriction endonuclease recognition sequence. Rab4{Delta}CT carried a C-terminal truncation of the last three amino acids (upstream primer, ccgtcgacATGTCCGAAACCTACG; downstream primer, ccggtaccCTACTCCTGAGCGTTCGGGGGC). Rab5{Delta}CT carried a C-terminal truncation of the last four amino acids (upstream primer, ccgtcgacATGGCTAGTCGAGGC; downstream primer, ggggtaccCTGATTCCTGGTTGG). The fidelity of the PCR amplification was verified by sequencing the insert in the recombination vector. Underlined sequences show restriction endonuclease sites on the oligonucleotide primers.

Quantitative infection of the differentiated 3T3-L1 adipocytes was performed on day 7 postdifferentiation by infecting cells at a multiplicity of infection of 10 for 4 h in Ham’s F-12 medium without additions. Cells were treated with human insulin (Novalin) as described in the figure legends. Infection of 100% of the cell population was confirmed by an in situ assay for ß-galactosidase activity.

Plasma membrane sheet assay
Preparation of plasma membrane sheets was carried out as previously described (31). Briefly, cultured adipocytes grown in a six-well cluster dish were treated as described in the figure legends. After experimental treatment, cells are washed with ice-cold PBS and attached to the plate with 0.5% poly-D-lysine. The cells are swollen in hypotonic buffer and then sonicated to open cells and release intracellular contents. Pure plasma membrane fragments remaining attached to the plastic dish were scraped into 120 µl solubilization buffer [1% SDS, 20 mM HEPES (pH 7.5), 150 mM NaCl, and 1 mM EDTA]. The protein content of the plasma membrane sheets was measured in a 10-µl aliquot using a sensitive spectrofluorometric assay (32). For this assay, the solubilized protein sample was diluted to 60 µl in solubilization buffer. The diluted sample was mixed with 0.25 ml 0.2 M sodium borate buffer, pH 9.0, at room temperature for 5 min. A 20-µl aliquot of fluorescamine solution (0.2 mg/ml in acetonitrile) was added with vigorous vortexing. After a 20-min incubation at room temperature, fluorescence was measured using 395 nm excitation and 460 nm emission at high sensitivity using a Shimadzu RF5000U spectrofluorophotometer. A standard curve was generated using BSA ranging from 0.1–15 µg/ml.

Whole cell detergent lysates and immunoprecipitation
One-hundred-millimeter plates of treated 3T3-L1 adipocytes were washed twice with ice-cold Tris-buffered saline (TBS) followed by freezing in liquid nitrogen. The plates were thawed on ice and scraped into 1 ml 1% Nonidet P-40 lysis buffer [1% Nonidet P-40, 20 mM HEPES (pH 7.4), and 2 mM EDTA] containing phosphatase inhibitors (100 mM sodium fluoride, 10 mM sodium pyrophosphate, 1 mM sodium vanadate, and 1 mM molybdate) and protease inhibitors (10 µM leupeptin, 10 µg/ml aprotinin, 1.5 mM pepstatin A, and 1 mM phenylmethylsulfonylfluoride). The cells were lysed on ice for 20 min, and insoluble material was removed by microcentrifugation for 10 min at 4 C. The protein concentrations of the detergent lysates were determined by a Bradford protein assay (Pierce Chemical Co., Rockford, IL) using the manufacturer’s specifications. The detergent-soluble material was transferred to a fresh tube and incubated for 3 h with primary antibody preloaded on protein A-conjugated Sepharose beads (rabbit antibodies) or protein G-conjugated Sepharose beads (mouse or goat antibodies). Immune complexes were washed five times with lysis buffer before immunoblotting. For enzymatic assays, washes were carried out as described below.

Electrophoresis and immunoblotting
Samples were fractionated using SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore Corp.) in transfer buffer (25 mM Tris and 193 mM glycine, pH 8.5) for 3–4 amp hours at 4 C. Membranes blotted for insulin receptor ß-subunit, IRS-1, GLUT4, Rab4, or Rab5 were blocked with 7–10% dried milk and 0.3% Tween-20 in TBS. Antiphosphotyrosine blots were blocked with 2% BSA and 0.1% Tween-20 in TBS. Antiphosphotyrosine blots were carried out using either PY20 or PY99 monoclonal antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) as indicated in the figure legends. GLUT4 antiserum was a gift from Dr. Jeffrey Pessin, and Rab5 antiserum was a gift from Dr. Guangpu Li. EEA1 antiserum was provided by Dr. Andreas Brech. All other antibodies were commercially obtained (Santa Cruz Biotechnology, Inc.). Immunoblots were detected using an enhanced chemiluminescence system (Pierce Chemical Co.) and quantified using scanning laser densitometry.

PI 3-kinase assay
PI 3-kinase activity was measured in IRS-1 immune complexes obtained as described above. Immune complexes were washed and subjected to the in vitro PI 3-kinase assay as previously described (33), using bovine phosphotidylinositol (Sigma, St. Louis, MO) and [{gamma}-32P]ATP as substrates. Phosphorylated lipids were separated by TLC, visualized by autoradiography, and quantified by collecting and counting radiolabeled phosphotidylinositol 3-phosphate.

Akt protein kinase assay
Akt protein kinase activity was determined using total Akt immune complexes obtained by immunoprecipitation with anti-Akt 1 antibody conjugated to agarose beads (Santa Cruz Biotechnology, Inc.). Immune complexes were washed times times in Nonidet P-40 lysis buffer [25 mM HEPES (pH 7.4), 150 mM NaCl, and 0.1% Nonidet P-40], followed by two washes in basal kinase buffer [20 mM Tris (pH 7.5) and 10 mM MgCl2]. The assay was carried out in 30 µl basal kinase buffer supplemented with 30 µM cold ATP, 30 µM Crosstide peptide (Transduction Laboratories, Lexington, KY), 1 mM dithiothreitol, and 5 µCi [{gamma}-32P]ATP, which was incubated at room temperature for 30 min with constant agitation. The phosphorylated peptide was measured by spotting on a 2.5-cm square of phosphocellulose paper (cellulose grade P81 cation exchanger) and washing four times in 75 mM phosphoric acid. The washed paper was rinsed with acetone, and radioactivity was counted. Background measurements were determined by spotting the supplemented kinase buffer that was not incubated with immune complexes.

Subcellular fractionation of 3T3-L1 adipocytes
3T3-L1 adipocytes were fractionated by ultracentrifugation to obtain discrete internal membrane fractions, plasma membranes, and a cytosolic fraction, using a modification of previous methods (30, 34). Briefly, 100-mm dishes of treated cells were washed twice with cold PBS and scraped into 2.25 ml TES [20 mM Tris (pH 7.5), 255 mM sucrose, 2 mM EDTA, 100 mM sodium fluoride, 10 mM sodium pyrophosphate, 1 mM sodium vanadate, 1 mM molybdate, 10 µM leupeptin, 10 µg/ml aprotinin, 1.5 mM pepstatin A, and 1 mM phenylmethylsulfonylfluoride]. Samples were homogenized (20 strokes in a Dounce homogenizer, pestle A), and centrifuged at 16,000 x gmax (Beckman Coulter, Inc., Palo Alto, CA; JA-17 rotor at 10,000 rpm) for 15 min. The pellet was resuspended in TES, and centrifugation was repeated as described above. The 16,000 x gmax pellet consisted of nuclei, mitochondria, and plasma membrane. Plasma membrane fragments were isolated by resuspending the 16,000 x gmax pellet in TES buffer, homogenized (20 strokes in a Dounce homogenizer, pestle B), and layered over a 1.15 M sucrose cushion (TES buffer containing 1.15 M sucrose). The plasma membrane fragments were isolated from the interface between the sucrose cushion and homogenate. The plasma membrane fractions were diluted in TES buffer and pelleted by centrifugation at 48,000 x gmax for 45 min. The 16,000 x gmax supernatants were combined and centrifuged at 50,000 x gmax (Beckman Coulter, Inc.; SW60 rotor at 19,000 rpm) for 20 min to produce a low speed pellet (LSP). A high speed pellet was sedimented from the second supernatant by centrifugation at 180,000 x gmax (SW60 rotor at 37,000 rpm) for 60 min. The resulting supernatant was centrifuged for 90 min at 360,000 x gmax (SW60 rotor at 52,000 rpm) to produce a very high speed pellet (VHSP). The pellets were resuspended in homogenization buffer and solubilized in Laemmli buffer. For some experiments, a combined internal membrane pool was obtained by centrifugation of the original postnuclear supernatant for 90 min at 360,000 x gmax (SW60 rotor at 52,000 rpm). The total protein concentration of the fractions was determined by Bradford assay. The cytosolic fraction (the supernatant obtained after sedimentation of the VHSP) was concentrated by centrifuging through 30-kDa cut-off microconcentrators (Amicon, Danvers, MA). The protein concentration was determined by Bradford assay and adjusted to 2 mg/ml.

Insulin receptor internalization
Internalization of insulin receptors was assayed by measuring acid-dissociable prebound [125I]insulin (NEX 196, NEN Life Science Products, Boston, MA). Thirty-five-millimeter dishes of infected or mock-infected cells were prelabeled with [125I]insulin (0.02 nM, 375 µCi/µg) in ice-cold Krebs-Ringers-HEPES binding buffer (130 mM NaCl, 5 mM KCl, 1.3 mM MgSO4, and 50 mM HEPES, pH 7.4) containing 0.1% BSA for 3 h at 4 C. Unbound ligand was removed by washing twice with ice-cold binding buffer. Internalization was initiated by replacing cold buffer with binding buffer prewarmed to 37 C. Endocytosis was terminated by replacing prewarmed buffer with ice-cold binding buffer. To confirm that radiolabeled insulin was not dissociating from the cell surface during incubation at 37 C, 125I was counted in the medium after its removal (data not shown). Remaining surface-associated ligand was stripped by incubating in 1 ml acidified PBS (pH 5.2) for 5 min on ice and quantified by counting the 125I in the acid wash. Ligand internalization was determined by measuring disappearance of radioligand from the cell surface relative to that in prelabeled cells not undergoing endocytosis. Nonspecific binding was determined by carrying out identical experiments in duplicate plates containing 1 µM cold insulin and was subtracted for each time point.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Localization of mutant Rab proteins
A role for Rab4 in insulin-induced GLUT4 translocation has been demonstrated by several groups, using approaches based on interference with normal Rab function (14, 15, 16). The methods for disrupting Rab4 function in each of these reports rely on manipulation of small numbers of cells, limiting the biochemical analyses that could be performed. Because of this limitation, effects of Rab4 on insulin action other than GLUT4 translocation were not examined. To overcome this limitation, we used a recombinant vaccinia virus vector that has previously been shown to infect 3T3-L1 adipocytes with greater than 95% efficiency without interfering with insulin-dependent GLUT4 translocation (31). Recombinant viruses were engineered to express either a mutant Rab4 (Rab4{Delta}CT) or a mutant Rab5 (Rab5{Delta}CT). Each mutant protein carries a deletion of the geranyl-geranylation site, presumably removing the membrane-anchoring domain. As a control for viral infection, we used a recombinant virus expressing the lacZ gene of Escherichia coli, which codes for ß-galactosidase activity.

Rab4 and Rab5 protein distribution was measured in combined internal membranes (M) and in a cytosolic fraction (C) in cells treated or not treated with 100 nM insulin after infecting the cells with control virus, Rab4 {Delta}CT virus, or Rab5 {Delta}CT virus (Fig. 1Go). In control virus-infected cells, we observed redistribution of endogenous Rab4 from internal membranes to the cytosol upon insulin stimulation (Fig. 1AGo, lanes 1–4), whereas no change in distribution of endogenous Rab5 was detected under these conditions (Fig. 1BGo, lanes 1–4). In a previous study, Rab5 was shown to be reduced in high density membranes after insulin treatment (13). We did not expect to see this redistribution in the current experiment, because the total internal membrane fractions were pooled, and the small redistribution from high density microsomes would not be detectable. Cells infected with Rab4 {Delta}CT virus showed a marked increase in the mutant Rab4 protein in both cytosolic and internal membrane fractions (Fig. 1AGo, lanes 5–8) even though the membrane-anchoring domain of Rab4 was removed. This membrane association may reflect a secondary interaction of the Rab4 {Delta}CT protein with another membrane-bound protein, which reduces or eliminates the requirement for the C-terminal membrane anchor. Cells infected with Rab5 {Delta}CT showed marked overexpression of mutant protein in the cytosolic fraction only (Fig. 1BGo, lanes 5–8), indicating that C-terminal amino acids (presumably the prenylation signal) are indispensable for Rab5 association with internal membranes.



View larger version (40K):
[in this window]
[in a new window]
  		Figure 1. Expression of Rab4{Delta}CT and Rab5{Delta}CT in cytoplasmic and membrane fractions of 3T3-L1 adipocytes. 3T3-L1 adipocytes were infected with control virus (Control) expressing only ß-galactosidase or viruses encoding for a mutant Rab4 (Rab4{Delta}CT; A) or a mutant Rab5 (Rab5{Delta}CT; B) for 4 h in starvation medium. Cells were stimulated with or without 100 nM insulin for 20 min at 37 C and fractionated into cytosol (C) or pooled internal membranes (M) as described in Materials and Methods. Fifty micrograms of each fraction in basal and insulin-stimulated cells were fractionated using 12% SDS-PAGE and transferred to PVDF membranes. The blots were labeled with antibodies directed against Rab4 (A) or Rab5 (B).

 
Effects of mutant Rabs on insulin-induced GLUT4 translocation
We verified our experimental system by confirming that Rab4 {Delta}CT protein introduced by vaccinia viruses interfered with GLUT4 translocation, as has been previously reported (14, 15, 16). Plasma membrane sheets were prepared from mock-infected cells or cells infected with control, Rab5{Delta}CT, or Rab4{Delta}CT viruses treated without or with 100 nM insulin for 20 min at 37 C. The plasma membrane sheets were assayed for the accumulation of GLUT4 by immunoblotting (Fig. 2Go). The abundance of GLUT4 in the plasma membrane fraction was the same in mock and control infected cells (Fig. 2AGo, compare lanes 2 and 4), confirming that vaccinia virus infection does not interfere with insulin-dependent GLUT4 translocation. In agreement with previous reports (14, 15, 16), we observed that Rab4{Delta}CT reduced GLUT4 translocation by approximately 40% (Fig. 2AGo, compare lanes 2, 4, and 8). In contrast, Rab5{Delta}CT did not affect GLUT4 translocation. The results of multiple experiments are quantified in Fig. 2BGo. These data indicate that the expression of mutant Rabs on GLUT4 translocation is specific for the Rab4 isoform.



View larger version (45K):
[in this window]
[in a new window]
  		Figure 2. Vaccinia-virus expression of the Rab4{Delta}CT, but not Rab5{Delta}CT, mutant protein inhibits GLUT4 translocation. 3T3-L1 adipocytes were either uninfected (Mock, lanes 1 and 2) or infected with control virus (Con, lanes 3 and 4) or viruses expressing a mutant rab5 (Rab5{Delta}CT, lanes 5 and 6) or a mutant Rab4 (Rab4{Delta}CT, lanes 7 and 8) protein for 4 h in starvation medium. Cells were treated with (lanes 2, 4, 6, and 8) or without (lanes 1, 3, 5, and 7) 100 nM insulin for 20 min at 37 C. The cells were rapidly cooled, and plasma membrane sheets were prepared as described in Materials and Methods. Equivalent amounts (200 ng) of plasma membrane sheets were fractionated using 10% SDS-PAGE, transferred to PVDF membrane, and immunoblotted using anti-GLUT4 antibodies. The experiment was performed five times, and Western blots were quantified using scanning laser densitometry. The quantified experiments are summarized in the histogram. Values for insulin-mediated GLUT4 translocation were analyzed using two-tailed Student’s t test. The asterisk indicates a significant decrease compared with all mock and control infected values (P < 0.05).

 
Effects of mutant Rabs on insulin-induced IRS-1 phosphorylation
Rab4 has been shown to associate with GLUT4 vesicles immunoprecipitated from adipocyte lysates; however, this pool accounts for less than one fourth of the total complement of endogenous Rab4 (12, 13, 15). Immunohistochemical staining of adipocytes has shown that GLUT4 and Rab4 in part colocalize; however, there are large pools of GLUT4 and Rab4 that do not overlap (35). This suggests the possibility that Rab4 performs functions in addition to any it may play in GLUT4 translocation. In support of this, it has been demonstrated that interfering with Rab4 function impairs insulin actions such as actin rearrangement that can occur independently of GLUT4 translocation (16, 17). Actin rearrangement has been shown to be dependent on IRS-1 phosphorylation (17, 36). Taken together, these observations suggest the possibility that overexpression of Rab4{Delta}CT may affect insulin signaling at points very proximal to receptor activation.

We examined the effect of the mutant proteins on IRS-1 phosphorylation, as this is an early event in insulin signaling. Other IRS isoforms were not studied, because neither IRS-2 nor IRS-3 could be detected in our 3T3-L1 adipocytes (data not shown), and IRS-1 is the predominant IRS isoform in adipocytes (26, 27). In the experiments shown in Fig. 3Go, cells infected with control, Rab4{Delta}CT, or Rab5{Delta}CT viruses were treated with or without 100 nM insulin for 15 min at 37 C. Subcellular fractionation was performed by differential centrifugation to obtain combined internal membranes and cytosol. Fifteen micrograms of internal membranes and 50 µg of cytosol were immunoblotted using anti-IRS-1 or anti-phosphotyrosine (PY99). Although the total amount of IRS-1 in internal membranes and cytosolic fractions was not affected by either Rab4{Delta}CT or Rab5 {Delta}CT (Fig. 3Go), the level of insulin-dependent phosphorylation of IRS-1 in both fractions was reduced in cells expressing Rab4{Delta}CT. As Rab4{Delta}CT expression inhibited glucose uptake, it was possible that cellular ATP levels were lower in these cells, which, in turn, decreased tyrosine phosphorylation of proteins. To test this, we measured cellular ATP levels in mock, control, and Rab4{Delta}CT-infected cells under basal and insulin stimulation. No difference in total cellular ATP was found under any of these conditions (data not shown). Rab5{Delta}CT had no effect on tyrosine phosphorylation of IRS-1. Decreased mobility of IRS-1 in both internal membranes and cytosol was observed in all insulin-treated cells. Decreased mobility, presumably resulting from serine-threonine phosphorylation of IRS-l, was not altered by expression of Rab4 {Delta}CT.



View larger version (38K):
[in this window]
[in a new window]
  		Figure 3. Tyrosine phosphorylation of IRS-1 is reduced by Rab4{Delta}CT, but not Rab5{Delta}CT. 3T3-L1 adipocytes were infected with either control virus (Control) or virus expressing a mutant Rab4 (Rab4{Delta}CT) or a mutant Rab5 (Rab5{Delta}CT) protein for 4 h in starvation medium. Cells were treated with or without 100 nM insulin for 20 min at 37 C and then immediately scraped into 255 mM sucrose homogenization buffer as described in Materials and Methods. The combined internal membrane fractions were resuspended in sucrose homogenization buffer and solubilized in Laemmli sample buffer. The cytosolic fraction was concentrated using Centricon 30 filters (Amicon). Fifteen micrograms of solubilized internal membranes (lanes 1–6) or 50 µg cytosolic fraction (lanes 7–12) were fractionated using 7.5% SDS-PAGE, transferred to PVDF membranes and immunoblotted using antiphosphotyrosine antibodies or anti-IRS-1 antibodies (A). Three independent experiments were quantified using scanning laser densitometry. The ratio of tyrosine phosphate/IRS-1 was plotted in the histogram (B). The results of three independent experiments were analyzed using two-tailed Student’s t test. The asterisk indicates a significant decrease compared with all mock and control infected values (P < 0.05).

 
We did not observe changes in the total IRS-1 content of the internal membrane fraction or in the cytosolic fraction after insulin treatment, in contrast with previously reported findings (27, 30). This discrepancy may be due to the fact that we used a much higher speed centrifugation (360,000 x gmax vs. 220,000 x gmax) to isolate the membrane-associated IRS-1.

We next determined whether there was a physiological consequence of decreased IRS-1 phosphorylation in the Rab4 {Delta}CT-infected cells. It is well established that a proximal event in insulin signal transduction is the activation of phosphatidylinositol 3-kinase. Activation occurs by recruitment of the p85 regulatory unit of PI 3-kinase to phosphorylated IRS isoforms (18, 20). In 3T3-L1 cells, tyrosine-phosphorylated IRS-1 is the major docking protein for PI 3-kinase (26, 27), and this prompted us to determine the PI 3-kinase activity in IRS-1-immunoprecipitated complexes from control, Rab4{Delta}CT-infected, and Rab5{Delta}CT-infected cells. Insulin-dependent PI 3-kinase activity was increased to similar levels (7.5-fold) in both mock-infected and control-virus infected cells (Fig. 4AGo). In contrast, Rab4{Delta}CT significantly reduced insulin-induced PI 3-kinase activation to 53% of control values (P < 0.02). Rab4{Delta}CT also reduced basal PI 3-kinase to 60% of control values; however, this reduction was not statistically significant. Thus, the fold activation of PI 3-kinase activity by insulin was similar in controls and Rab4{Delta}CT-infected cells, but the absolute magnitude of activation was decreased due to reduced phospho-IRS-1 able to bind PI 3-kinase after insulin stimulation. Rab5{Delta}CT had no statistically significant effect on the activity of PI 3-kinase (Fig. 4AGo).



View larger version (36K):
[in this window]
[in a new window]
  		Figure 4. IRS-1-associated PI 3-kinase activity and total cellular Akt activity are inhibited by Rab4{Delta}CT, but not Rab5{Delta}CT. A, 3T3-L1 adipocytes were either uninfected (Mock) or infected with control virus (Control) or viruses expressing a mutant rab4 (Rab4{Delta}CT) or a mutant Rab5 (Rab5{Delta}CT) protein for 4 h in starvation medium. Cells were treated with or without 100 nM insulin for 20 min at 37 C and snap-frozen in liquid nitrogen. IRS-1 was immunoprecipitated from a 1% Nonidet P-40 whole cell detergent lysate. The immune complex was assayed for PI3-kinase activity as described in Materials and Methods, using bovine phosphotidylinositol and [{gamma}-32P]ATP as substrates. The phosphorylated products were separated by TLC, and the band corresponding to phosphoinositol 3-phosphate was scraped and counted. The results of three independent experiments were analyzed using two-tailed Student’s t test. The asterisk indicates a significant decrease compared with all mock and control infected values (P < 0.02). B, 3T3-L1 adipocytes were either uninfected (Mock) or infected with control virus (Control) or viruses expressing a mutant Rab5 (Rab5{Delta}CT) or a mutant Rab4 (Rab4{Delta}CT) protein for 4 h in starvation medium. Cells were treated with or without 100 nM insulin for 20 min at 37 C and snap-frozen in liquid nitrogen. Akt-1 was immunoprecipitated from a 1% Nonidet P-40 whole cell detergent lysate. The immune complex was assayed for kinase activity as described in Materials and Methods using the Crosstide peptide and [{gamma}-32P]ATP as substrates. The phosphorylated peptide was bound to phosphocellulose paper, washed extensively, and counted. The results of five independent experiments were analyzed using two-tailed Student’s t test. The asterisk indicates a significant decrease compared with all mock and control infected values (P < 0.05).

 
To further characterize the physiological significance of decreased PI 3-kinase activity, we measured activation of Akt-1 (Fig. 4BGo). Akt-1 has been shown to be wortmannin sensitive and activated by PI 3-kinase (36, 37, 38, 39, 40). We immunoprecipitated Akt-1 from whole cell detergent lysates, and enzymatic activity was measured. Akt-1 activity was significantly reduced by 40% (P < 0.05) in cells overexpressing Rab4{Delta}CT relative to mock-infected or control virus-infected cells (Fig. 4BGo). As observed for PI 3-kinase activity, basal Akt activity was also reduced in Rab4{Delta}CT-infected cells, but this was not statistically significant. The absolute magnitude of Akt activation in Rab4{Delta}CT-infected cells, basally and under insulin stimulation, reflects the decreased level of PI 3-kinase activity in those cells. Rab5{Delta}CT expression caused an overall small decrease in Akt activity that was not statistically significant. It should be noted that in individual experiments, Akt activity in whole cell lysates from Rab5{Delta}CT cells was often higher than that in controls, contributing to a high variability in this measurement. This contrasted with Akt activity in Rab4{Delta}CT-infected cells, which was consistently reduced with respect to control values.

Subcellular fractionation of 3T3-L1 adipocytes by differential centrifugation
IRS-1 is a hydrophilic protein that is thought to interact with membranes as a peripheral membrane protein. The precise localization of membrane-associated IRS-1 has not been defined, although studies suggest that membrane association might be mediated through cytoskeletal elements or through the AP3 adaptin complex (28, 41). To more precisely define the IRS-1 compartments, especially their relationship to those fractions containing internalized insulin receptor, we performed subcellular fractionation of 3T3-L1 adipocytes using a modification of standard procedures (30, 34). Our procedure differs from the standard adipocyte fractionation methods in that it incorporates an additional very high speed centrifugation step (360,000 x gmax), based on the fractionation protocol for rat adipocytes (24). In many procedures, the fraction referred to as high density microsomes was a 48,000 x gmax pellet, and low density microsomes were either a 180,000 or 220,000 x gmax pellet. We collected the 48,000 x gmax pellet( referred to here as the LSP) and a 180,000 x gmax pellet (referred to as the HSP). We then subjected the 180,000 x gmax supernatant to additional centrifugation at 360,000 x gmax to obtain a third pellet, referred to as the VHSP.

We performed this procedure on cells that were treated or not treated with 100 nM insulin for 2 or 20 min at 37 C (Fig. 5Go). Aliquots equivalent to 15% of each pelleted fraction and 5% of the total cytosolic fraction were analyzed by Western blot analysis. Figure 5Go depicts fractions probed by immunoblotting for tyrosine-phosphorylated proteins (PY99), insulin receptor ß-subunit, IRS-1, GLUT4, and EEA1 (a known marker for the endosomal compartment). Under our conditions, the insulin receptor was identified in the plasma membrane, LSP, and HSP (Fig. 5Go, lanes 1–9), whereas IRS-1 was excluded from these fractions and was found in the cytosol and VHSP (Fig. 5Go, lanes 10–13).



View larger version (26K):
[in this window]
[in a new window]
  		Figure 5. Subcellular fractionation of 3T3-L1 adipocytes by differential centrifugation. 3T3-L1 adipocytes were treated with or without 100 nM insulin for 2 or 20 min at 37 C and then immediately scraped into 255 mM sucrose homogenization buffer as described in Materials and Methods. Membrane fractions were resuspended in sucrose homogenization buffer and solubilized in Laemmli sample buffer. The cytosolic fraction was concentrated using Centricon 30 filters. Aliquots consisting of 15% of solubilized pellets corresponding to the plasma membrane (PM), 48,000 x gmax pellet (LSP), 180,000 x gmax pellet (HSP), 360,000 x gmax pellet (VHSP), and 5% of the cytosolic fraction (CYT) were fractionated using 7.5% or 10% SDS-PAGE, transferred to PVDF membranes and immunoblotted using antibodies directed against phosphotyrosine (PY99), insulin receptor ß-subunit (IR), IRS-1, GLUT4, and EEA1. The experiment was performed three times with identical results.

 
Analysis of Western blots indicated that 30% of the total IRS-1 is associated with the VHSP, and 60% is cytosolic, in agreement with a previous estimate (29). We observed that although significantly less IRS-1 was detected in the VHSP relative to cytosol, the level of phosphorylated IRS-1 in the VHSP was much higher than that in cytosol. This is expected based on previous stoichiometric analysis of insulin-induced IRS-1 phosphorylation in 3T3-L1 adipocytes (26, 27).

The use of {alpha}-EEA1 antibody to identify the endosomal vesicle compartment (42) allowed the observation that the membrane-associated IRS-1 pool is distinct from endosomes (Fig. 5Go, lanes 7–12). The endosomal compartment does, however, overlap with the major insulin-responsive pool of GLUT4 (Fig. 5Go, lanes 7–9). The relatively small amount of GLUT4 present in the VHSP is not insulin responsive under our conditions (Fig. 5Go, lanes 10–12).

Insulin receptor internalization in Rab4{Delta}CT-infected cells
Based on studies of transferrin receptor trafficking, it was possible that decreased insulin signaling in Rab4{Delta}CT-infected cells could result from changes in insulin receptor trafficking. To address this, we performed insulin receptor internalization studies using [125I]insulin. Internalization of a cohort of [125I]insulin-bound insulin receptors was measured over several time points up to 30 min in mock, control, and Rab4{Delta}CT-infected cells (Fig. 6AGo). Internalization proceeded rapidly over the first 5 min and reached steady state by 10 min. Neither virus infection nor expression of Rab4{Delta}CT had any effect on internalization of ligand-bound receptors compared with that in mock-infected cells. Although the internalization of ligand-bound insulin receptors was unaffected, it was possible that activated insulin receptors within an internal membrane pool were mistargeted to a degradation pool in cells overexpressing Rab4{Delta}CT, effectively reducing the amount of time that activated insulin receptor is able to phosphorylate intracellular substrates. It has been shown previously that internalized insulin receptors are highly kinase active (29). To determine whether Rab4{Delta}CT affected the abundance of activated (phosphorylated) internalized insulin receptors, the internal membranes from control and Rab4{Delta}CT-infected cells treated with or without 100 nM insulin for 20 min at 37 C were fractionated by differential centrifugation. The plasma membranes, LSP, and HSP were immunoblotted for the ß-subunit of the insulin receptor and for antiphosphotyrosine (Fig. 6BGo). Immunoblot analysis indicated that there was no change in the abundance of either insulin receptor or phosphorylated insulin receptor in plasma membranes or internal membranes isolated from Rab4{Delta}CT-infected cells after insulin treatment. The VHSP and cytosolic fractions were immunoblotted for IRS-1 and antiphosphotyrosine, confirming that the Rab4{Delta}CT decreased IRS-1 phosphorylation (Fig. 6CGo) even though IR phosphorylation was unaffected (Fig. 6BGo). Based on these data, we conclude that neither receptor internalization nor trafficking of activated internalized insulin receptors is affected by the mutant Rab4. In addition, the rate of dephosphorylation of activated insulin receptors after removal of extracellular insulin by low pH wash was not affected by Rab4{Delta}CT expression, suggesting that insulin receptor dephosphorylation was not significantly modified by the experimental treatment (data not shown).



View larger version (29K):
[in this window]
[in a new window]
  		Figure 6. Insulin receptor trafficking is not affected by Rab4{Delta}CT. A, 3T3-L1 adipocytes were either uninfected (Mock) or infected with control virus or mutant Rab4 (Rab4{Delta}CT) virus for 4 h in starvation medium. The cells were prelabeled with [125I]insulin in binding buffer at 4 C for 4 h. Cells were washed with ice-cold TBS, and prewarmed medium was added for various times ranging from 0–30 min. Acid-dissociable [125I]insulin was quantified. The percentage of receptor internalization was calculated from the ratio of cell surface-associated radioactivity remaining after incubation in 37 C medium. The internalization curve was generated from two independent experiments. B, 3T3-L1 adipocytes were either uninfected (Mock) or infected with control virus or a mutant Rab4 (Rab4{Delta}CT) virus for 4 h in starvation medium. Cells were treated with or without 100 nM insulin for 20 min at 37 C and then immediately scraped into 255 mM sucrose homogenization buffer as described in Materials and Methods. Membrane fractions were resuspended in sucrose homogenization buffer and solubilized in Laemmli sample buffer. The cytosolic fraction was concentrated using Centricon 30 filters. Aliquots consisting of 15% solubilized pellets corresponding to the plasma membrane (PM), 48,000 x gmax pellet (LSP), 180,000 x gmax pellet (HSP), 360,000 x gmax pellet (VHSP), and cytosolic fraction (CYT) were fractionated using 7.5% SDS-PAGE, transferred to PVDF membranes, and immunoblotted using antiphosphotyrosine antibodies (PY99) or antiinsulin receptor ß-subunit antibodies (B) and PY-99 or anti-IRS-1 antibodies (C). The experiment was performed three times with identical results.

 
Subcellular localization of membrane-bound rab4{Delta}CT
Removal of the C-terminal geranyl-geranylation site of Rab4 incompletely inhibited membrane association, suggesting that Rab4 {Delta}CT may be targeted to a specific membrane fraction. A more detailed analysis of the membrane distribution of Rab4 protein in control and Rab4 {Delta}CT-infected cells was performed in individual fractions derived from infected cells with or without insulin stimulation (Fig. 7Go). Equivalent portions (15%) of each fraction were analyzed for Rab4 by immunoblotting. Endogenous Rab4 was found most abundantly in the 180,000 x gmax pellet (HSP; Fig. 7Go, lane 7). A ratio of mutant Rab4{Delta}CT to endogenous rab4 was calculated to determine whether the mutant colocalized to subcellular fractions containing endogenous Rab4. By this analysis, cytosol and the 360,000 x gmax pellet (VHSP) were specifically enriched for Rab4{Delta}CT. Thus, expression of Rab4{Delta}CT results in unusually high accumulation of the mutant protein in the VHSP and the cytosol, the two fractions that contain IRS-1 (Fig. 5Go).



View larger version (19K):
[in this window]
[in a new window]
  		Figure 7. Expression of Rab4{Delta}CT-specific membrane fractions of 3T3-L1 adipocytes. 3T3-L1 adipocytes were infected with either control virus (C) or Rab4{Delta}CT virus (R) protein for 4 h. Cells were scraped into 255 mM sucrose homogenization buffer as described in Materials and Methods. Membrane fractions were resuspended in sucrose homogenization buffer and solubilized in Laemmli sample buffer. The cytosolic fraction was concentrated using Centricon 30 filters. Aliquots consisting of 30% for control and 5% for Rab4{Delta}CT-infected cells of fractions corresponding to the plasma membrane (PM), 48,000 x gmax pellet (LSP), 180,000 x gmax pellet (HSP), 360,000 x gmax pellet (VHSP), and cytosolic fraction (CYT) were fractionated using 12% SDS-PAGE, transferred to PVDF membranes, and immunoblotted using anti-Rab4 antibodies. The results of three experiments were analyzed using scanning laser densitometry. The ratio of mutant Rab4 to endogenous Rab4 (from control infected cells) was calculated from these scans.

 
Effect of reduced temperature on IRS-1 phosphorylation in Rab4{Delta}CT-infected cells
Based on the currently known functions of Rab4, it was likely that overexpression of Rab4{Delta}CT perturbed a membrane-sorting event, which could, in turn, lead to decreased IRS-1 phosphorylation. We reasoned that if Rab4{Delta}CT functioned in our assay by disrupting membrane trafficking, then disruption of this process by another method should not further decrease the effect of Rab4{Delta}CT on the insulin-dependent signaling processes we measured. Most pharmacological methods for inhibiting trafficking also reduce cellular ATP concentrations. We therefore reduced trafficking by stimulating cells at reduced temperature (15 C), a condition in which many membrane trafficking processes have been shown to be partially inhibited (43) and one in which insulin receptor kinase activity remains viable (30, 44, 45). IRS-1 was immunoprecipitated from whole cell detergent lysates isolated from control and Rab4{Delta}CT-infected cells that were stimulated with 100 nM insulin at either 37 or 15 C. Aliquots containing equivalent protein concentrations were immunoblotted with antiphosphotyrosine or anti-IRS-1 antibodies. At 37 C, tyrosine phosphorylation of IRS-1 in control cells was maximal at 2 min and decreased over time (Fig. 8AGo, lanes 1–4, and Fig. 8CGo), in agreement with previous studies (27). Tyrosine phosphorylation of IRS-1 in Rab4{Delta}CT-infected cells was reduced at all time points relative to that in control cells (Fig. 8AGo, lanes 5–8, and Fig. 8CGo) and did not show the time-dependent decrease in IRS-1 phosphorylation evident in control infected cells. When insulin stimulation was performed at 15 C, no difference in insulin-dependent IRS-1 tyrosine phosphorylation was observed in control vs. Rab4{Delta}CT-infected cells (Fig. 8Go, B and C). The level of tyrosine phosphorylation of IRS-1 in Rab4{Delta}CT-infected cells was the same whether cells were stimulated at 15 or 37 C (Fig. 8CGo). In contrast, control-infected cells stimulated at 15 C exhibited IRS-l phosphorylation levels that were reduced to those observed in Rab4{Delta}CT-infected cells. Thus, reduced temperature did not further inhibit the ability of the activated insulin receptor kinase to phosphorylate IRS-1 in Rab4{Delta}CT-infected cells, but did inhibit this process in control infected cells. These data are consistent with the idea that optimal IRS-1 phosphorylation requires membrane trafficking, and that this function is inhibited by both overexpression of Rab4{Delta}CT and reduced temperature.



View larger version (25K):
[in this window]
[in a new window]
  		Figure 8. Rab4{Delta}CT interferes with IRS-1 phosphorylation at 37 C, but not at 15 C. 3T3-L1 adipocytes were infected with control virus (Control) or mutant Rab4 (Rab4{Delta}CT) virus for 4 h in starvation medium. The cells were incubated with 100 nM insulin for 0, 2, 5, or 20 min at 37 C (A) or 0, 5, or 20 min at 15 C (B). IRS-1 was immunoprecipitated from equivalent amounts of a whole cell 1% Nonidet P-40 detergent lysate. The immunoprecipitated pellets were divided into two portions, fractionated by 7.5% SDS-PAGE, and transferred to PVDF membranes. The membranes were immunoblotted with antiphosphotyrosine antibodies or anti-IRS-1 antibodies. The ratio of phosphotyrosine to IRS-1 was calculated for three independent experiments. The quantification is shown in the histogram in C. The results were analyzed using two-tailed Student’s t test. The asterisk indicates a significant decrease at 15 C compared with 37 C (P < 0.05).

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In this study we have presented evidence that is consistent with a model for insulin-dependent phosphorylation of IRS-1, in which this process is carried out through regulated vesicle trafficking. We believe that vesicular trafficking is required for optimal insulin-mediated phosphorylation for several reasons. First, IRS-1 and the insulin receptor localize to separate subcellular fractions, suggesting that these two molecules reside in distinct cellular compartments. Communication between separate membrane-associated receptor and substrate compartments would require a mechanism to promote a close association of IRS-1 with activated insulin receptors, the simplest explanation invoking vesicle trafficking. Second, we interfered with insulin-mediated IRS-1 phosphorylation by introducing a mutant version of a Rab4 protein, a protein intimately involved with regulation of membrane trafficking. Additional support for the proposal that the mutant Rab4{Delta}CT interferes with a membrane trafficking event is provided by our results showing that insulin-dependent IRS-1 phosphorylation decreased when cells were stimulated at 15 C, a condition known to inhibit many vesicle trafficking processes (43), but no further reduction was observed when Rab4{Delta}CT-infected cells were stimulated with insulin at 15 C. This suggests that the process causing a decrease in insulin-dependent IRS-1 phosphorylation by each of these experimental interventions may occur by the same or a similar mechanism.

There are several steps in which control of membrane trafficking might regulate insulin signal transduction. It is well established that ligand-bound insulin receptors are rapidly internalized, potentially providing a means by which the activated insulin receptors access downstream effectors (29, 45, 46). However, it has recently been shown that introduction of a dominant negative dynamin protein that interferes with internalization of the insulin receptor does not inhibit IRS-1 phosphorylation, activation of Akt, or insulin-dependent GLUT4 translocation (47, 48, 49). These studies suggest that receptor endocytosis is not the primary membrane trafficking event involved in regulating insulin signaling through IRS-1. Our experimental observations that insulin receptor internalization is not affected by expression of the mutant Rab4 indicate that receptor endocytosis is not the event in which Rab4 plays a role. As Rab proteins generally perform functions in intracellular membrane trafficking, we propose that we are interfering with the interaction between the membrane-associated IRS-1 compartment and the activated insulin receptor. It is possible that a rapid and transient assembly of a specific protein complex may be required to bring the tethered IRS-1 molecules into close proximity with the membrane compartments containing activated insulin receptors, whether the activated receptor is at the plasma membrane or is trafficking through early endosomes. Such a step would provide an additional level of specificity to insulin signal transduction by providing a means for regulated assembly of specialized signaling compartment.

The trafficking between IRS-1 and insulin receptor may be a constitutive process, in which case IRS-1 is constantly presented to the insulin receptor, but only when the insulin receptor is activated is there significant phosphorylation of IRS-1. Consistent with this possibility is our observation that IRS-1-associated PI 3-kinase activity was reduced in the Rab4{Delta}CT basal cells. If we assume that basal association of PI 3-kinase with IRS-1 occurs through the low basal level of tyrosine-phosphorylated insulin receptor, then it is possible that Rab4{Delta}CT interferes with IRS-1 trafficking in the basal state as well as under insulin stimulation.

Explanations not invoking trafficking could also explain membrane-associated IRS-1 phosphorylation in response to insulin. For instance, phosphorylated, membrane-associated IRS-1 may be derived from cytosolic IRS-1 that is phosphorylated by insulin receptor as the free pool of IRS-1 randomly diffuses through the cell. Cytoplasmic phosphorylated IRS-1 could then associate with the intracellular membrane. Several lines of evidence argue against this possibility. First, in unstimulated cells, IRS-1 is associated with intracellular membranes, and the abundance of IRS-1 in those membranes does not increase after insulin stimulation (27, 29, 30). Second, intracellular membrane-associated IRS-1 has been shown to be phosphorylated before phosphorylation of the cytosolic IRS-1 pool in 3T3-L1 adipocytes treated with insulin at 4 C (30). Finally, Inoue et al. (27) have provided evidence that phosphorylated IRS-1 and IRS-2 translocate from intracellular membranes to the cytosol in response to insulin stimulation and that the intracellular membrane pool of IRS-1 is the initial site of IRS-1 phosphorylation by the activated insulin receptor. Our proposed model for IRS-1 phosphorylation is consistent with these observations. We also observed that Rab4{Delta}CT-infected cells had reduced IRS-1 phosphorylation in both the membrane-associated pool and the cytosolic pool. This would be expected if the cytosolic pool of phosphorylated IRS-1 were derived from the membrane-associated pool as suggested by other studies (27, 30).

Overexpression of a Rab4 mutant lacking the geranyl-geranlyation site resulted in an increase in this protein in the cytosolic fraction and all particulate fractions that were analyzed. We cannot distinguish whether the mutant protein is disrupting normal Rab4 function or interfering with other cellular functions that do not normally involve endogenous Rab4. This lack of specificity is a limitation in all studies that use overexpressed mutant proteins.

Rab4 has previously been tied to insulin action. Disruption of Rab4 function has been shown to interfere with GLUT4 trafficking, and this effect has been confirmed in the present study (14, 15, 16). In addition, insulin-induced rearrangement of the actin cytoskeleton has been shown to be disrupted by microinjection of an antibody directed against the C-terminal hypervariable region of Rab4 (16). These results indicate that Rab4 has pleiotropic effects on insulin action.

Rab4 itself has been shown to be a target of insulin action at several steps in the Rab4 cycle. Insulin stimulates guanine nucleotide exchange on Rab4 via a wortmannin-sensitive pathway (50). The geranyl-geranlyltransferase II enzyme that modifies Rab4 by addition of a C-terminal lipid moiety is activated by phosphorylation in response to insulin stimulation (51). Insulin also stimulates the translocation of Rab4 from internal membranes to cytosol in adipocytes (12, 15, 52) and skeletal muscle (53). Finally, Rab4 is phosphorylated by insulin-activated ERKI kinase (54). From these studies, it is clear that insulin stimulation results in several modifications of the Rab4 cycle, any or all of which may be important in a role for Rab4 in insulin action.

The importance of IRS-1 in insulin-mediated GLUT4 translocation is under debate. Complete inhibition of insulin-mediated IRS-1 phosphorylation by overexpression of PTB domains or SAIN domains of IRS-1 blocks mitogenic effects of insulin, but does not effect GLUT4 translocation or Akt activity (17, 36). On the other hand, ablation of IRS-1 in transgenic mice and primary adipocytes inhibits glucose uptake and GLUT4 translocation (55, 56, 57). In view of these discrepant data regarding the role for IRS-1 in GLUT4 translocation, we cannot conclude that the effect of interference of Rab4 function on IRS-1 phosphorylation is the mechanism by which Rab4 mutants in this and other studies affect GLUT4 translocation. We can, however, conclude that Rab4 may be involved in multiple insulin-mediated signaling pathways.

In the present study, we have expressed a mutant form of Rab4 in which the geranyl-geranlyation site has been removed. This mutation concentrates Rab4 protein in the cytoplasm of the cell, as previously reported (15). Surprisingly, this form of Rab4, unlike the form of Rab5 possessing a similar modification, was also associated with membrane fractions. Thus, although the geranyl-geranylation site is absolutely required for Rab5 to associate with membranes, it is not required for Rab4 to association with membranes in 3T3-L1 adipocytes. It is possible that the prenylation-deficient mutant Rab4 associates with membranes through one or more interactions with other membrane-bound proteins. It is not clear whether these putative receptor proteins are physiological receptors for endogenous Rab4. The fact that the mutant Rab4 was somewhat enriched in the VHSP along with IRS-1 suggested the possibility that the mutant Rab4 may bind IRS-1, thus inhibiting phosphorylation. This was discounted because mutant Rab4 protein did not coimmunoprecipitate with IRS-1 in Rab4{Delta}CT-infected cells (data not shown).

Interestingly, overexpression of Rab5 was without significant effect in our experiments, suggesting that the putative factors interact specifically with cytosolic Rab4. It will be necessary to identify proteins that bind cytosolic Rab4 to determine the specific role that Rab4 function plays in insulin signaling, and we are currently pursuing identification of these proteins.


   Acknowledgments
 
We thank Dr. Guangpu Li for providing Rab4 and Rab5 cDNAs and helpful discussions throughout the course of these studies, and Alan R. Trumbly and Katherine Oshel for excellent technical assistance.


   Footnotes
 
1 This work was supported by Grant 196085 from the Juvenile Diabetes Foundation, NIH Grant DK-47894, and Grant RA0064 from the American Diabetes Association. Back

Received August 2, 1999.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Novick P Zerial M 1997 The diversity of rab proteins in vesicle transport. Curr Opin Cell Biol 9:496–504[CrossRef][Medline]
  2. Schimmoller F, Simon I, Pfeffer SR 1998 Rab GTPases, directors of vesicle docking. J Biol Chem 273:22161–22164[Free Full Text]
  3. Rossi G, Yu JA, Newman AP, Ferro-Novick S 1991 Dependence of Ypt1 and Sec4 membrane attachment on Bet2. Nature 351:158–161[CrossRef][Medline]
  4. Khosravi-Far R, Lutz RJ, Cox AD, Conroy L, Bourne JR, Sinensky M, Balch WE, Buss JE, Der CJ 1991 Isoprenoid modification of rab proteins terminating in CC or CXC motifs. Proc Natl Acad Sci USA 88:6264–6268[Abstract/Free Full Text]
  5. Kinsella BT, Maltese WA 1992 Rab GTP-binding proteins with three different carboxyl-terminal cysteine motifs are modified in vivo by 20-carbon isoprenoids. J Biol Chem 267:3940–3945[Abstract/Free Full Text]
  6. Chavrier P, Gorvel J-P, Stelzer E, Simons K, Gruenberg J, Zerial M 1991 Hypervariable C-terminal domain of Rab proteins acts as a targeting signal. Nature 353:769–772[CrossRef][Medline]
  7. Stenmark H, Valencia A, Martinez O, Ullrich O, Gould B, Zerial M 1994 Distinct structural elements of Rab5 define its functional specificity. EMBO J 13:575–583[Medline]
  8. Sogaard M, Tani K, Ye R, Geromanos S, Tempst PKT, Rothman JE, Sollner T 1994 A Rab protein is required for the assembly of SNARE complexes in the docking of transport vesicles. Cell 78:937–948[CrossRef][Medline]
  9. Rybin V, Ullrich O, Rubino M, Alexandrov K, Simon I, Seabra MG, Goody R, Zerial M 1997 GTPase activity of Rab5 acts as a timer for endocytic membrane fusion. Nature 383:266–269
  10. Nuoffer C, Davidson HW, Matteson J, Meinkoth J, Balch WE 1994 A GDP-bound form of Rab1 inhibits protein export from the endoplasmic reticulum and transport between Golgi compartments. J Cell Biol 125:225–237[Abstract/Free Full Text]
  11. Peranen J, Auvinen P, Virta H, Wepf R, Simons K 1996 Rab8 promotes polarized membrane transport through reorganization of actin and microtubules in fibroblasts. J Cell Biol 135:153–167[Abstract/Free Full Text]
  12. Cormont M, Tanti JF, Zahraoui A, Van Obberghen E, Tavitian A, Le Marchand-Brustel Y 1993 Insulin and okadaic acid induce rab4 redistribution in adipocytes. J Biol Chem 268:19491–19497[Abstract/Free Full Text]
  13. Cormont M, Van Obberghen E, Zerial M, Le Marchand-Brustel Y 1996 Insulin induces a change in rab5 subcellular localization in adipocytes independently of phosphatidyl 3-kinase activation. Endocrinology 137:3408–3415[Abstract]
  14. Shibata H, Omata W, Suzuki Y, Tanaka S, Kojima I 1996 A synthetic peptide corresponding to the rab4 hypervariable carboxyl-terminal domain inhibits insulin action on glucose transport in rat adipocytes. J Biol Chem 271:9704–9709[Abstract/Free Full Text]
  15. Cormont M, Bortoluzzi MN, Gautier NM, Van Obberghen E, Le Marchand-Brustel Y 1996 Potential role of Rab4 in the regulation of subcellular localization of GLUT4 in adipocytes. Mol Cell Biol 16:6879–6886[Abstract]
  16. Vollenweider P, Martin SS, Haruta T, Morris AJ, Nelson JG, Cormont M, Le Marchand-Brustel Y, Rose DW, Olefsky JM 1997 The small guanosine triphosphate-binding protein rab4 is involved in insulin-induced GLUT4 translocation and actin filament rearrangement in 3T3–L1 cells. Endocrinology 138:4941–4949[Abstract/Free Full Text]
  17. Morris AJ, Martin SS, Haruta T, Nelson JG, Vollenweider P, Gustafson TA, Mueckler M, Rose DW, Olefsky JM 1996 Evidence for an insulin receptor substrate 1 independent insulin signaling pathway that mediates insulin-responsive glucose transporter (GLUT4) translocation. Proc Natl Acad Sci USA 93:8401–8406[Abstract/Free Full Text]
  18. White MF, Kahn CR 1994 The insulin signaling system. J Biol Chem 269:1–4[Free Full Text]
  19. Sun XJ, Rothenberg P, Kahn CR, Backer JM, Araki E, Wilden PA, Cahill DA, Goldstein BJ, White MF 1991 Structure of the insulin receptor substrate IRS-1 defines a unique signal transduction protein. Nature 352:73–77[CrossRef][Medline]
  20. Czech MP Corvera S 1999 Signaling mechanisms that regulate glucose transport. J Biol Chem 274:1865–1868[Free Full Text]
  21. Giorgetti S, Balloti R, Kowalski-Chauvel A, Tartare S, Van Obberghen E 1993 The insulin and insulin-like growth factor-I receptor substrate IRS-1 associates with and activates phosphatidylinositol 3-kinase in vitro. J Biol Chem 268:7358–7364[Abstract/Free Full Text]
  22. Myers Jr MG, Backer JM, Sun X-J, Shoelson S, Hu P, Schlessinger J, Yoakim M, Schaffhausen B, White MF 1992 IRS-1 activates phosphatidylinositol 3'-kinase by associating with src homology 2 domains of p85. Proc Natl Acad Sci USA 89:10350–10354[Abstract/Free Full Text]
  23. Backer JM, Myers Jr MG, Shoelson SE, Chin DJ, Sun X-J, Miralpeix M, Hu P, Margolis B, Skolnik EY, Schlessinger J, White MF 1992 Phosphatidylinositol 3'-kinase is activated by association with IRS-1 during insulin stimulation. EMBO J 11:3469–3479[Medline]
  24. Kelly KL, Ruderman NB 1993 Insulin-stimulated phosphatidylinositol 3-kinase. Association with a 185-kDa tyrosine phosphorylated protein (IRS-1) and localization in a low density membrane vesicle. J Biol Chem 268:4391–4398[Abstract/Free Full Text]
  25. Shepherd PR, Withers DJ, Siddle K 1998 Phosphoinositide 3-kinase: the key switch mechanism in insulin signalling. Biochem J 333:471–490
  26. Anai M, Ono H, Funaki M, Fukushima Y, Inukai K, Ogihara T, Hideyuki S, Onishi Y, Yazaki Y, Kikuchi M, Oka Y, Asano T 1998 Different subcellular distribution and regulation of expression of insulin receptor substrate (IRS)-3 from those of IRS-1 and IRS-2. J Biol Chem 273:29686–29692[Abstract/Free Full Text]
  27. Inoue G, Cheatham B, Emkey R, Kahn CR 1998 Dynamics of insulin signaling in 3T3–L1 adipocytes. J Biol Chem 273:11548–11555[Abstract/Free Full Text]
  28. Clark SF, Martin S, Carozzi AJ, Hill MM, James DE 1998 Intracellular localization of phosphatidylinositide 3-kinase and insulin receptor substrate-1 in adipocytes: potential involvement of a membrane skeleton. J Cell Biol 140:1211–1225[Abstract/Free Full Text]
  29. Kublaoui B, Lee J, Pilch PF 1995 Dynamics of signaling during insulin-stimulated endocytosis of its receptor in adipocytes. J Biol Chem 270:59–65[Abstract/Free Full Text]
  30. Heller-Harrison RA, Morin M, Czech MP 1995 Insulin regulation of membrane-association insulin receptor substrate 1. J Biol Chem 270:24442–24450[Abstract/Free Full Text]
  31. Olson AL, Knight JB, Pessin JE 1997 Syntaxin 4, VAMP2, and/or VAMP3/cellubrevin are functional target membrane and vesicle SNAP receptors for insulin-stimulated GLUT4 translocation in adipocytes. Mol Cell Biol 17:2425–2435[Abstract]
  32. Sogawa K, Takahashi K 1978 Use of fluorescamine-labeled casein as a substrate for assay of proteinases. J Biochem 83:1783–1787[Abstract/Free Full Text]
  33. Holt KH, Olson AL, Moye-Rowley WS, Pessin JE 1994 Phosphatidylinositol 3-kinase activation is mediated by high-affinity interactions between distinct domains within the p110 and p85 subunits. Mol Cell Biol 14:42–49[Abstract/Free Full Text]
  34. Simpson IA, Yver DR, Hissin PJ, Wardzala LJ, Karnieli E, Salans LB, Cushman SW 1983 Insulin-stimulated translocation of glucose transporters in the isolated rat adipose cells: characterization of subcellular fractions. Biochim Biophys Acta 763:393–407[Medline]
  35. Malide D, Cushman SW 1997 Morphological effects of wortmannin on the endosomal system and GLUT4-containing compartments in rat adipose cells. J Cell Sci 110:2795–2806[Abstract]
  36. Sharma PM, Egawa K, Gustafson TA, Martin JL, Olefsky JM 1997 Adenovirus-mediated overexpression of IRS-1 interacting domains abolishes insulin-stimulated mitogenesis without affecting glucose transport in 3T3–L1 adipocytes. Mol Cell Biol 17:7386–7397[Abstract]
  37. Franke TF, Kap