Insulin stimulates actin comet tails on intracellular GLUT4-containing compartments in differentiated 3T3L1 adipocytes.

Incubation of isolated GLUT4-containing vesicles with Xenopus oocyte extracts resulted in a guanosine 5'-[gamma-thio]triphosphate (GTP gamma S) and sodium orthovanadate stimulation of actin comet tails. The in vitro actin-based GLUT4 vesicle motility was inhibited by both latrunculin B and a dominant-interfering N-WASP mutant, N-WASP/Delta VCA. Preparations of gently sheared (broken) 3T3L1 adipocytes also displayed GTP gamma S and sodium orthovanadate stimulation of actin comet tails on GLUT4 intracellular compartments. Furthermore, insulin pretreatment of intact adipocytes prior to gently shearing also resulted in a marked increase in actin polymerization and actin comet tailing on GLUT4 vesicles. In addition, the insulin stimulation of actin comet tails was completely inhibited by Clostridum difficile toxin B, demonstrating a specific role for a Rho family member small GTP-binding protein. Expression of N-WASP/Delta VCA in intact cells had little effect on adipocyte cortical actin but partially inhibited insulin-stimulated GLUT4 translocation. Taken together, these data demonstrate that insulin can induce GLUT4 vesicle actin comet tails that are necessary for the efficient translocation of GLUT4 from intracellular storage sites to the plasma membrane.

Incubation of isolated GLUT4-containing vesicles with Xenopus oocyte extracts resulted in a guanosine 5-[␥-thio]triphosphate (GTP␥S) and sodium orthovanadate stimulation of actin comet tails. The in vitro actinbased GLUT4 vesicle motility was inhibited by both latrunculin B and a dominant-interfering N-WASP mutant, N-WASP/⌬VCA. Preparations of gently sheared (broken) 3T3L1 adipocytes also displayed GTP␥S and sodium orthovanadate stimulation of actin comet tails on GLUT4 intracellular compartments. Furthermore, insulin pretreatment of intact adipocytes prior to gently shearing also resulted in a marked increase in actin polymerization and actin comet tailing on GLUT4 vesicles. In addition, the insulin stimulation of actin comet tails was completely inhibited by Clostridum difficile toxin B, demonstrating a specific role for a Rho family member small GTP-binding protein. Expression of N-WASP/⌬VCA in intact cells had little effect on adipocyte cortical actin but partially inhibited insulin-stimulated GLUT4 translocation. Taken together, these data demonstrate that insulin can induce GLUT4 vesicle actin comet tails that are necessary for the efficient translocation of GLUT4 from intracellular storage sites to the plasma membrane.
One of the major physiological responses of insulin is to increase glucose uptake in striated muscle and adipose tissue. This primarily results from the translocation of intracellular compartmentalized GLUT4 protein to the cell surface membrane (1)(2)(3)(4)(5). In unstimulated cells, GLUT4 continually cycles between the plasma membrane and various intracellular storage sites, whereas insulin stimulates a marked increase in the overall rate of GLUT4 exocytosis (6 -9). Multiple studies have begun to dissect the molecular machinery involved in this translocation process, and recent data have strongly implicated both the microtubule and actin cytoskeleton (10 -18). In particular, treatment of cells with the actin depolymerizing agent cytochalasin D or the actin monomer-binding toxins latrunculin A or B resulted in a marked inhibition of insulin-induced GLUT4 translocation (15)(16)(17)(18). Recently we have also observed that insulin induces dynamic actin remodeling at the inner surface of the plasma membrane (cortical actin) and in the perinuclear region in differentiated 3T3L1 adipocytes (19).
Actin has been proposed to play both a positive and negative role in the regulation of various vesicle trafficking events. In some systems, polymerized actin has been implicated as a barrier that undergoes depolymerization during vesicle trafficking (20 -24). On the other hand, actin is also thought to play a positive role by forming scaffolds for transport vesicles to move along during vesicle sorting decisions (21,25,26). Interestingly, rapid actin polymerization/depolymerization has been shown to function as a molecular motor in the intracellular trafficking and spreading of the infectious bacteria Listeria monocytogenes and Shigella flexnerii (27)(28)(29)(30)(31). In this regard, recent studies have reported that actin-based motility (comet tails) can be observed on small transport vesicles (26,32). In vitro studies using Xenopus egg extracts have observed GTP␥S 1 and sodium orthovanadate-stimulated actin comet tails on both endogenous vesicles or exogenous artificial vesicles containing phosphatidylinositol 4,5-bisphosphate or phosphatidylinositol 3,4,5-trisphosphate (26,33). Furthermore, the Rho family member of small GTP-binding proteins, Cdc42, can induce actin comet tails on the membrane vesicles in a N-WASP (neural Wiscott-Aldrich syndrome protein) and Arp2/3 (actin-related protein-2/3) complex-dependent manner (33)(34)(35)(36).
To explore the functional role of actin polymerization in GLUT4 translocation, we have reconstituted the formation of actin comet tails on GLUT4 vesicles using Xenopus egg extracts in vitro and in a broken adipocyte cell system. These data demonstrate that both GTP␥S plus sodium orthovanadate and insulin stimulation result in an N-WASP-dependent formation of comet tails on GLUT4-containing intracellular vesicles. Furthermore, expression of a dominant-interfering N-WASP mutant (N-WASP/⌬VCA) partially inhibited both insulin-and GTP␥S-induced GLUT4 translocation in 3T3L1 adipocytes.

EXPERIMENTAL PROCEDURES
Materials-Clostridum difficile toxin B was obtained from Techlab Inc. (Blacksburg, VA), and latrunculin B was purchased from Calbiochem (La Jolla, CA). The c-Myc epitope 9E10 monoclonal antibody was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Fluorescein isothiocyanate-conjugated donkey anti-mouse IgG and Texas Red-conjugated donkey anti-mouse IgG were purchased from Jackson Immunoresearch Laboratories (West Grove, PA). Rhodamine-phalloidin, streptolysin-O (SL-O), and rhodamine-actin were purchased from Sigma and Cytoskeleton (Denver, CO), respectively. The pGLUT4-enhanced green fluorescent protein (GLUT4-EGFP) was prepared as described previously (37). The Myc epitope-tagged GLUT4 was a generous gift from Dr. Amira Klip and was used to generate the double tagged GLUT4 fusion protein Myc-GLUT4-EGFP. A recombinant adenovirus encoding for GLUT4-EGFP (Ad-GLUT4-EGFP) was prepared by the University of Iowa Vector Core. N-WASP cDNA was prepared by PCR from Quick Clone rat cDNA (CLONTECH, Palo Alto, CA) and was ligated into pcDNA3 with a Myc-tagged sequence at the amino-terminal region. The N-WASP/⌬VCA cDNA was also generated by PCR and was cloned into pcDNA3. All other chemicals were reagent grade or the best quality commercially available.
Cell Culture and Transient Transfection-3T3L1 preadipocytes were cultured in Dulbecco's modified Eagle's medium containing 25 mM glucose, 10% calf serum at 37°C in a 8% CO 2 atmosphere, induced to differentiate into adipocytes, and transfected by electroporation as described previously (38). The cells were then allowed to adhere to tissue culture dishes for 24 h, and the adipocytes were then serum-starved for 2 h prior to experiments. In some experiments, the electroporated adipocytes were seeded on coverslips. Chinese hamster ovary cells were cultured in ␣-minimum essential medium containing 10% fetal bovine serum and were transfected by electroporation as described previously (39).
To quantitate the cell surface exposure of GLUT4, adipocytes were transfected with a Myc-GLUT4-EGFP cDNA, and nonpermeabilized cells were labeled with the Myc 9E10 antibody followed by Texas Red-labeled secondary antibody. The average red fluorescent and average green fluorescent pixel intensity of an entire image were calculated by the Zeiss LSM 510 software. The ratio for 10 cells was averaged as a measure for relative GLUT4 translocation. Nonspecific background fluorescence was determined by using cells transfected with GLUT4-EGFP and labeled in the same manner as cells transfected with Myc-GLUT4-EGFP.

FIG. 1. GTP␥S and sodium orthovanadate stimulate actin polymerization on GLUT4 vesicles in vitro.
A, 3T3L1 adipocytes were infected with adenovirus-GLUT4-EGFP, and 24 h later the cells were homogenized and centrifuged to obtain a high speed fraction enriched in GLUT4-EGFP-containing vesicles as described under "Experimental Procedures." The GLUT4-EGFP vesicle fraction was incubated with Xenopus oocyte extracts containing rhodamine-actin in the absence or in the presence of 100 M GTP␥S plus 1 mM sodium orthovanadate (panels a-e) and imaged at high magnification (ϫ100; zoom, ϫ7.5) every 22 s over a 2.2-min period (panels a-e). This is a representative experiment independently performed three to five times. The complete time lapse image is provided in the supplementary materials. B, the GLUT4-EGFP-expressing cells were gently sheared to obtain a broken adipocyte cell preparation as described under "Experimental Procedures." The broken cells were then mixed with Xenopus oocyte extracts containing rhodamine-actin in the absence (panel a) or in the presence of 100 M GTP␥S plus 1 mM sodium orthovanadate (panels b-d). Latruculin B (20 M) or GST-N-WASP/⌬VCA (25 g/ml) were added to the reaction mixtures in panels c and d, respectively. These are representative confocal fluorescent microscopy images taken at 10 -15 min following the initiation of the actin polymerization assay (magnification, ϫ100; zoom, ϫ3.0).

FIG. 2. Insulin stimulation induces actin polymerization on GLUT4 vesicles in a broken adipocyte preparation.
A, 3T3L1 adipocytes were infected with adenovirus-GLUT4-EGFP, and 24 h later the cells were either left untreated (panel a) or stimulated with 100 nM insulin for 5 min (panels b-e). Gently sheared broken cell preparations were isolated and then mixed with Xenopus egg extracts containing rhodamine-actin in the absence (panels a and b) or in the presence of 20 M latrunculin B (panel c), 25 g/ml GST-N-WASP/⌬VCA (panel d), or 0.5 g/ml C. difficile toxin B (panel e). These are representative confocal fluorescent microscopy images taken at 10 -15 min following the initiation of the actin polymerization assay (magnification, ϫ100; zoom, ϫ3.0). B, the broken adipocytes from insulin-stimulated cells were incubated with Xenopus egg extracts containing rhodamine-actin and imaged at high magnification (ϫ100; zoom, ϫ7.5) every 22 s over a 2.2-min period (panels a-f). This is a representative experiment independently performed three to five times. The complete time lapse image is provided in the supplementary materials.
Single Cell Microinjection-The microinjection and visualization of single 3T3L1 adipocytes was performed as described previously (40). Briefly, the cells were grown on coverslips, and prior to microinjection, the medium was changed to Lebovitz's L-15 medium containing 0.1% bovine serum albumin. Differentiated 3T3L1 adipocytes were impaled using the Eppendorf model 5171 micromanipulator, and nuclei were injected with 50 or 200 g/ml of cDNAs in 100 mM KCl, 5 mM Na 2 PO 4 , pH 7.2, with an Eppendorf model 5246 transjector. The cells were allowed to recover for 24 h and placed into a perfusion chamber maintained at 37°C and visualized by time lapse confocal fluorescence microscopy.
Preparation of Xenopus Egg Extract-Crude Xenopus egg extracts were prepared as described previously with slight modifications (33,35,41). Briefly, the Xenopus laevis eggs were dejellied in 2% cysteine, pH 7.8, washed with XB solution (100 mM KCl, 50 mM sucrose, 1 mM MgCl 2 , 0.1 mM CaCl 2 , 10 mM HEPES, pH 7.8, adjusted with KOH) and then washed twice with XB solution containing 5 mM EGTA and 10 g/ml of leupeptin, pepstatin A, and chymostatin. The eggs were homogenized at 4°C and centrifuged at 10,000 rpm in a SW41 rotor at 4°C for 15 min. The cytoplasmic layer was removed, and 1 ⁄20 volume of energy mix (150 mM creatine phosphate, 10 mM ATP, 1 mM EGTA, and 20 mM MgCl 2 ) and 1 ⁄1000 volume of leupeptin (10 mg/ml) were added. The crude egg extract was centrifuged at 55,000 rpm in a SW60 rotor at 4°C for 1 h. The supernatant was diluted 10 times with XB solution containing 5 mM EGTA and 10 g/ml of leupeptin, pepstatin A, and chymostatin and centrifuged at 60,000 rpm in a SW60 rotor at 4°C for 1 h. The clarified cytosol was concentrated to the original volume using a Centriprep 3 concentrator (Millipore) and snap frozen in separate aliquots for storage at Ϫ80°C.
Preparation of the Endosome and Homogenates from 3T3L1 Adipocytes-Differentiated 3T3L1 adipocytes were infected by Ad-GLUT4-EGFP adenovirus vector and allowed to express GLUT4-EGFP for 24 h. After 3 h of serum starvation, the adipocytes were treated without or with 100 nM insulin for 5 min and washed twice with ice-cold homogenized buffer (0.25 M sucrose, 5 mM MgCl 2 , 20 mM HEPES, pH 7.4, adjusted with NaOH) containing 10 g/ml of leupeptin, pepstatin A, and aprotinin. They were scraped and homogenized by passage through a 27 gauge needle (seven times for vesicle preparation and two times for broken cell preparation). The postnuclear supernatant, which contains GLUT4-EGFP vesicle, was obtained by a centrifugation at 4,000 rpm for 10 min, and the pellet was used for actin motility assay.
Actin-based Motility Assay-Actin-based motility was performed in 4 l of vesicle-free Xenopus egg extracts containing a final concentration of 3 M rhodamine labeled G-actin plus 1 l of postnuclear supernatant or 1 l of the cell pellet (resuspened in 10 volumes of homogenate buffer) of 3T3L1 homogenates. A 1-l aliquot of the cell-free motility reaction was placed on a glass slide under a coverslip, and rhodamine and EGFP images were acquired every 4 s using a LSM510 Zeiss confocal microscope. All digital images were subsequently cropped and annotated using Adobe Premiere 5.0 software on a Macintosh computer.
Streptolysin-O Permeabilization of 3T3L1 Adipocytes-Differentiated 3T3L1 adipocytes were permeabilized with SL-O as described previously (42) with minor modifications. Briefly, 3T3L1 adipocytes were washed three times with intracellular (ICR) buffer (140 mM potassium glutamate, 20 mM HEPES, pH 7.15, 7.5 mM MgCl 2 , 5 mM EGTA, 5 mM NaCl, 2 mM CaCl 2 ) and incubated in intracellular buffer containing 20 units/ml of activated SL-O for 5 min at 37°C. Under these conditions, more than 95% of the cells were permeabilized based upon incorporation of propidium iodide. Following SL-O permeabilization, the cells were washed twice with ICR buffer (intracellular buffer containing 1 mg/ml bovine serum albumin, 1 mM dithiothreitol and enriched with an ATP-regenerating system: 40 IU/ml creatine phosphokinase, 5 mM creatine phosphate, and 1 mM ATP). GLUT4-EGFP translocation assay was performed by incubating the cells in ICR buffer containing 100 M GTP␥S for 15 min at 37°C.

GTP␥S and Sodium Orthovanadate Induce Actin Polymerization and Vesicle Motility in Broken 3T3L1 Adipocytes-Pre-
vious studies have demonstrated that low speed X. laevis egg extracts that contain endogenous vesicles induce actin polymerization when incubated with GTP␥S and sodium orthovanadate (33,35). In contrast, high speed extracts that are depleted of endogenous vesicles are unable to induce actin polymerization unless exogenous membrane vesicles are added (33). To test whether GLUT4-containing vesicles obtained from 3T3L1 adipocytes have the ability to polymerize actin in the presence of GTP␥S and sodium orthovanadate, a GLUT4-EGFP-containing intracellular membrane fraction was obtained from serumstarved 3T3L1 adipocytes and was subjected to the in vitro actin polymerization assay system using the high speed vesiclefree egg extract supernatants. The addition of GTP␥S (100 M) and sodium orthovanadate (1 mM) to the reaction mixture resulted in an induction of actin comet tails on the GLUT4-EGFP vesicles concomitant with random GLUT4 motility by time lapse imaging (Fig. 1A and supplementary material), whereas the vesicle fraction itself did not induce actin polymerization (data not shown).
We have recently reported that although actin-yellow fluorescent protein expression can be used to observed gross changes in adipocyte actin structures, the resolution is insufficient to resolve small changes in actin polymerization such as comet tailing (19). Because living cells cannot be visualized using rhodamine-actin, we examined actin polymerization in gently sheared (broken) 3T3L1 adipocyte cell preparations incubated with the high speed vesicle-free egg supernatants (Fig.  1B). Gentle shearing of adipocytes expressing GLUT4-EGFP resulted in a random distribution of small vesicle compartments within the cell, and the addition of the vesicle-free egg extracts and rhodamine-actin only resulted in a small degree of discreet actin polymerization (Fig. 1B, panel a). In contrast, the addition of GTP␥S and sodium orthovanadate to the reaction mixture resulted in a large amount of actin polymerization on the GLUT4-EGFP-containing vesicles with enhanced motility (Fig. 1B, panel b). As a control, latrunculin B completely inhibited the formation of actin comet tails on GLUT4-EGFP-containing vesicles (Fig. 1B, panel c).
Recent studies have demonstrated that activated Cdc42, a structurally related Rho family homolog, can induce actin comet tails on membrane vesicles in an N-WASP-dependent manner (33)(34)(35)(36). N-WASP appears to function as a scaffolding protein that contains a central binding region for several Rho family small GTP-binding proteins and a carboxyl-terminal domain that interacts with both monomeric actin and the Arp2/3 complex (32,36,43). In this manner, N-WASP is thought to mediate upstream activation of actin nucleation and polymerization resulting in the formation of actin comet tails. To determine whether N-WASP plays a similar role in the formation of GLUT4 vesicle comet tails, we examined the effect of a dominant-interfering N-WASP GST fusion protein (N-WASP/⌬VCA) that lacks the monomeric actin-and Arp2/3binding sites. The addition of the GST N-WASP/⌬VCA fusion protein resulted in a marked inhibition of GTP␥S plus sodium orthovanadate-induced actin polymerization in the broken adipocyte preparation (Fig. 1B, panel d). In contrast, the addition of a wild type N-WASP GST fusion protein had no significant effect (data not shown). These data demonstrate that GLUT4 intracellular vesicles have the capability to nucleate actin and generate actin comet tails through an N-WASP-dependent mechanism.
Insulin Stimulates Actin Polymerization on Adipocyte Membrane Vesicles-Recently we and others have reported that insulin stimulation results in cortical actin remodeling followed by an increase in polymerized actin in the perinuclear region and that these actin dynamics are essential for insulin-induced GLUT4 translocation (19,44,45). To further investigate the functional role of actin remodeling in insulin-induced GLUT4 translocation, we next examined the ability of insulin to induce actin comet tails on GLUT4-containing vesicles (Fig.  2). As previously observed, broken adipocytes from control cells displayed the presence of GLUT4-EGFP with relatively low levels of polymerized actin ( Fig. 2A, panel a). In contrast, broken adipocytes isolated from insulin-stimulated cells resulted in a substantial amount of actin polymerization ( Fig. 2A,  panel b). In addition, many of the actin comet tail structures were found on GLUT4-EGFP-containing vesicles and could be observed to induce random GLUT4 motility by time lapse imaging ( Fig. 2B and supplementary materials). Latrunculin B completely inhibited the formation of actin comet tails on GLUT4-EGFP-containing vesicles ( Fig. 2A, panel c). Similarly, the addition of N-WASP/⌬VCA strongly inhibited insulin-induced actin polymerization ( Fig. 2A, panel d), whereas wild type N-WASP had no significant effect (data not shown).
Previously, we have reported that insulin does not induce Cdc42 activation in adipocytes but can activate a structurally related Rho family member, TC10 (37). Furthermore, TC10 has been reported to interact with N-WASP in a yeast two-hybrid system assay and to associate with N-WASP in vivo (46,47). Therefore, to assess a possible involvement of Rho family GTPbinding proteins on the formation of actin comet tails in the insulin-stimulated adipocytes, we utilized C. difficile toxin B, a specific inhibitor for Rho family small GTP-binding proteins, including TC10. Broken adipocytes from Toxin B-treated cells were devoid of any insulin-induced actin polymerization or GLUT4 vesicle comet tails ( Fig. 2A, panel e). These data demonstrate that insulin stimulation of intact cells activates the necessary cellular machinery to induced GLUT4 vesicle comet tails following in vitro reconstitution. In addition, insulin stimulation of actin polymerization is dependent upon both N-WASP and a Rho family member small GTP-binding protein, most likely TC10.
Dominant-interfering N-WASP/⌬VCA Inhibits GLUT4 Translocation-It is well established that both insulin and GTP␥S are potent stimulators of GLUT4 translocation in adipocytes (42, 48 -50). Because both of these agents also induce actin comet tails on GLUT4 vesicles in an N-WASP-dependent manner, we next examined the effect of the dominant-interfering N-WASP mutant on GLUT4 translocation (Fig. 3). Expression of GLUT4-EGFP in 3T3L1 adipocytes resulted in the typical GLUT4 distribution to the perinuclear region and in small vesicle compartments scattered throughout the cell. In contrast, insulin stimulation induced a robust redistribution of GLUT4-EGFP to the plasma membrane. As summarized in Fig. 3A, in the basal state 16.7 Ϯ 8.3% of the cells displayed a FIG. 4. Effect of N-WASP/⌬VCA on cortical actin structure in 3T3L1 adipocytes. 3T3L1 adipocyte nuclei were microinjected with cDNA encoding N-WASP/WT (panels a-c) or N-WASP/ ⌬VCA (panels d-f). The cells were allowed to recover for 24 h and double labeled with a monoclonal antibody directed against the Myc epitope tag (panels a and d) and rhodamine-labeled phalloidin (panels b and e) as described under "Experiment Procedures." The merged images are presented in panels c and f. These are a representative field obtained from three independent experiments. plasma membrane localization of GLUT4-EGFP. In contrast, following insulin stimulation plasma membrane-distributed GLUT4-EGFP was observed in 78.7 Ϯ 4.3% of the cells. Expression of N-WASP/WT did not alter the insulin stimulation of GLUT4-EGFP translocation, whereas expression of N-WASP/ ⌬VCA resulted in a partial but significant inhibition of insulininduced GLUT4-EGFP translocation (56.6 Ϯ 5.1%). Similarly, GTP␥S treatment of SL-O permeabilized adipocytes resulted in 64.1 Ϯ 8.9% of the cells displaying a plasma membrane GLUT4-EGFP distribution compared with 14.7 Ϯ 4.6% in the basal state. As observed with insulin stimulation, expression of N-WASP/WT did not have any significant effect of GTP␥Sstimulated GLUT4-EGFP translocation, whereas N-WASP/ ⌬VCA reduced translocation to 38.7 Ϯ 4.9% of the cells. To confirm the inhibitory effect of N-WASP/⌬VCA, we utilized a quantitative assay of translocation in which the relative surface exposure of a Myc-GLUT4-EGFP fusion protein is compared with the total amount of GLUT4 expression by comparing the ratio Myc label intensity to that of EGFP intensity (51). Consistent with the counting of plasma membrane rim fluorescence, expression of N-WASP/⌬VCA inhibited the insulin-stimulated cell surface exposure of the Myc epitope to 72% Ϯ 7.0% compared with the control cells. Together, these data demonstrate that disruption of actin polymerization and GLUT4 vesicle comet tails by expression of a dominant-interfering N-WASP mutant partially inhibits both insulin-and GTP␥Sstimulated GLUT4 translocation.
Dominant-interfering N-WASP/⌬VCA Does Not Disrupt Cortical Actin Structures-In contrast to fibroblasts, adipocytes do not have significant amounts of stress fiber actin but instead display polymerized actin that is restricted to the perinuclear region and cortical actin that is juxtaposed to the plasma membrane (19). To assess the effect of N-WASP on cortical actin, 3T3L1 adipocyte nuclei were microinjected with a cDNA encoding a Myc epitope-tagged N-WASP/WT or N-WASP/⌬VCA followed by rhodamine-phalloidin staining (Fig. 4). As previously observed, rhodamine-phalloidin labeling primarily detected a strong ring of cortical actin along the inner surface of the plasma membrane (Fig. 4, panels b and e). The expressed N-WASP/WT and N-WASP/⌬VCA proteins were localized in a small punctate distribution dispersed throughout the cell cytoplasm (Fig. 4, panels a and d). In addition, N-WASP/⌬VCA was also found to localize juxtaposed to the plasma membrane (Fig.  4, panel d). In any case, cortical actin labeling was not affected by expression of either N-WASP/WT or N-WASP/⌬VCA. These data demonstrate that the maintenance of cortical actin structure in 3T3L1 adipocytes is not N-WASP-dependent.
In conclusion, several studies have reported that insulinstimulated actin remodeling occurs both beneath the plasma membrane (cortical actin) and in the perinuclear regions that co-localizes with the major intracellular GLUT4 storage sites (19). In addition, alterations in actin polymerization affect multiple membrane transport processes and, in particular, insulinstimulated GLUT4 translocation (15)(16)(17)(18)(19)45). Our data demonstrate that insulin can induce actin comet tails on GLUT4containing intracellular storage vesicles. This process is dependent upon both a Rho family member GTP-binding protein and the neural Wisckott-Aldrich syndrome protein, N-WASP. Because insulin does not function to activate Cdc42 in adipocytes (37), these data further support a role for another member of this family such as TC10. This is consistent with the insulin activation of TC10, the interaction of active TC10 with N-WASP, and the fact that expression of TC10 has marked effects on GLUT4 translocation in adipocytes. The observation that dominant-interfering N-WASP only partially inhibited insulin-stimulated GLUT4 translocation is consistent with a study examining general membrane trafficking. In this case, actin polymerization was not essential but increased the efficiency of membrane transport (52). Thus, although the precise functional role of GLUT4 vesicle actin comet tails on the translocation process remains to be established, these data are most consistent with a role in GLUT4 trafficking out of the perinuclear storage sites.