Nocodazole inhibits insulin-stimulated glucose transport in 3T3-L1 adipocytes via a microtubule-independent mechanism.

Insulin stimulates glucose transport in adipocytes and muscle cells by triggering redistribution of the GLUT4 glucose transporter from an intracellular perinuclear location to the cell surface. Recent reports have shown that the microtubule-depolymerizing agent nocodazole inhibits insulin-stimulated glucose transport, implicating an important role for microtubules in this process. In the present study we show that 2 microm nocodazole completely depolymerized microtubules in 3T3-L1 adipocytes, as determined morphologically and biochemically, resulting in dispersal of the perinuclear GLUT4 compartment and the Golgi apparatus. However, 2 microm nocodazole did not significantly effect either the kinetics or magnitude of insulin-stimulated glucose transport. Consistent with previous studies, higher concentrations of nocodazole (10-33 microm) significantly inhibited basal and insulin-stimulated glucose uptake in adipocytes. This effect was not likely the result of microtubule depolymerization because in the presence of taxol, which blocked nocodazole-induced depolymerization of microtubules as well as the dispersal of the perinuclear GLUT4 compartment, the inhibitory effect of 10-33 microm nocodazole on insulin-stimulated glucose uptake prevailed. Despite the decrease in insulin-stimulated glucose transport with 33 microm nocodazole we did not observe inhibition of insulin-stimulated GLUT4 translocation to the cell surface under these conditions. Consistent with a direct effect of nocodazole on glucose transporter function we observed a rapid inhibitory effect of nocodazole on glucose transport activity when added to either 3T3-L1 adipocytes or to Chinese hamster ovary cells at 4 degrees C. These studies reveal a new and unexpected effect of nocodazole in mammalian cells which appears to occur independently of its microtubule-depolymerizing effects.

Insulin stimulates glucose transport in muscle and fat cells by regulated vesicular transport (1)(2)(3)(4). In the absence of insulin GLUT4, the major glucose transporter isoform expressed in these cells, is stored in an intracellular tubulovesicular com-partment. Insulin stimulates the exocytosis of GLUT4 from this compartment, leading to increased GLUT4 levels at the cell surface. This process occurs with a t 1/2 of 2-5 min.
Several recent studies have suggested an important role for microtubules in insulin-regulated GLUT4 trafficking. First, tubulin-␣ and vimentin have been identified as major components of intracellular GLUT4 vesicles in adipocytes (5). Second, it has been shown that insulin stimulates the long range movement of GLUT4 vesicles along linear tracks in adipocytes (6). Third, the microtubule-depolymerizing drug nocodazole inhibited insulin-stimulated glucose transport in adipocytes (7,8). Finally, nocodazole caused a dispersal of the perinuclear GLUT4 compartment in adipocytes (7,9). These studies have led to the notion that microtubules may play a fundamental role in GLUT4 trafficking and insulin action. It has been proposed that insulin may stimulate GLUT4 exocytosis, at least in part, by increasing the rate of association of GLUT4-containing membranes with microtubules. This step likely occurs in the perinuclear region of the cell, implicating an important role for the perinuclear localization of GLUT4 in insulin action (5,7).
In the present studies we have examined the dose response effects of nocodazole on microtubule integrity and insulin action in 3T3-L1 adipocytes. Using morphological and biochemical methods we observed maximal effects of nocodazole on microtubule depolymerization at a concentration of 2 M. Functional consequences of microtubule depolymerization were observed at 2 M nocodazole in that the Golgi apparatus and the perinuclear GLUT4 compartment were dispersed throughout the cytoplasm of the cell. Despite these effects there was no significant inhibition of insulin-stimulated glucose transport or GLUT4 translocation at 2 M nocodazole. At higher nocodazole concentrations (10 -33 M) we observed inhibition of glucose transport consistent with previous studies (7,8). We present data to show that the inhibitory effects of high concentrations of nocodazole are the result of an inhibitory effect of the drug on glucose transport activity rather than on the insulin-dependent recruitment of GLUT4 to the cell surface. We conclude that nocodazole does not inhibit insulin-stimulated GLUT4 translocation in 3T3-L1 adipocytes.

EXPERIMENTAL PROCEDURES
Reagents and Antibodies-All tissue culture media were purchased from Life Technologies, Inc., except fetal calf serum, which was obtained from Trace Biosciences (Clayton, Australia). Insulin, nocodazole, and taxol were obtained from Calbiochem. Bicinchoninic acid reagent, used in protein assays, and ECL Supersignal reagent were obtained from Pierce (Rockford, IL Cell Culture and Transfection-3T3-L1 fibroblasts obtained from the American Type Culture Collection (Rockville, MD) were cultured and differentiated into adipocytes as described previously (10). Chinese hamster ovary (CHO) cells were cultured as described previously (11). 116 (HA-GLUT4) 3T3-L1 adipocytes were obtained by transfection of 3T3-L1 fibroblasts with the retroviral expression vector pBabepuro containing GLUT4 HA-tagged at the extracellular loop, as described by Shewan et al. (12). Briefly, cells were infected with the relevant virus for 3-5 h in the presence of 4 g/ml Polybrene (Sigma). After a 48-h recovery period, infected cells were selected in Dulbecco's modified Eagle's medium containing 5% fetal calf serum supplemented with 2 g/ml puromycin (Sigma) and then differentiated as described above.
Glucose Uptake Assays-Cells cultured in 6-or 12-well plates were serum starved in KRP buffer (25 mM HEPES, pH 7.4, 120 mM NaCl, 6 mM KCl, 1.2 mM Mg SO 4 , 1 mM CaCl 2 , 0.4 mM NaH 2 PO 4 , 0.6 mM Na 2 HPO 4 ) containing 0.2% bovine serum albumin for 2 h at 37°C. Cells were incubated in 950 l of KRP buffer with the corresponding agents at the times and doses indicated in the figure legends. 2-Deoxy-[ 3 H]glucose (2-DOG) uptake was measured as described previously (13). Briefly, the assay was initiated by the addition of 50 l of 1 mM 2-DOG (20 Ci/mmol). After 4 min, the assay was terminated by washing the cells rapidly three times with ice-cold PBS. Cells were subsequently solubilized in 1% Triton X-100, and 3 H was quantified by scintillation counting (Packard 1900CA liquid scintillation analyzer, Packard Instrument Co.). Nonspecific 2-DOG uptake was determined in the presence of 50 M cytochalasin B.
Indirect Immunofluorescence Microscopy-Cells cultured on coverslips in 6-or 12-well dishes were starved for 2 h in Dulbecco's modified Eagle's medium containing 0.2% BSA or in KRP buffer containing 0.2% BSA buffer and incubated with the appropriate treatments. Cells were fixed with acetone for 5 min, washed with PBS, washed again with PBS containing 0.15 M in glycine, and incubated with 1% BSA containing PBS for 30 min. Cells were incubated with primary antibodies and diluted in PBS containing 1% BSA for 1 h. Cells were then washed with PBS containing 0.1% BSA and incubated for 30 min with phalloidinfluorescein isothiocyanate or with the corresponding Alexa 488-or Alexa 594-conjugated secondary antibody diluted in PBS containing 1% BSA. Normal rabbit serum was used as a negative control. Coverslips were washed with PBS, mounted onto glass microscope slides, and viewed using an X63/1.4 Zeiss oil immersion objective on a Zeiss Axiovert fluorescence microscope, equipped with a Bio-Rad MRC-600 laser confocal imaging system. In experiments designed for staining of cytoskeletal structures, cells were fixed and permeabilized simultaneously in cytoskeleton-stabilizing buffer (10 mM PIPES, pH 6.9, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl 2 , 1 mM EGTA) as described by Arcangeletti et al. (14).
The preparation of plasma membrane (PM) lawns was performed as described by Robinson et al. (15). Briefly, after incubating cells on coverslips with the appropriate treatment, adipocytes were sonicated yielding a lawn of PM fragments attached to the coverslip. Coverslips were then incubated with the relevant antibodies followed by Alexa 594-conjugated secondary antibody. The fluorescence intensity of individual PM fragments (100 -200) from five random fields for each experimental condition was quantified using NIH 1.62 software. The assay and quantitation were performed by separate investigators in a double blind fashion. In addition, parallel experiments measuring tubulin depolymerization and glucose transport were performed.
In experiments designed for cell surface staining, 3T3-L1 adipocytes expressing HA-tagged GLUT4 (12) were fixed in 2% paraformaldehyde in PBS for 20 min but not permeabilized. Cells were incubated with the anti-HA antibody for 1 h, washed, and incubated with the corresponding Alexa 488-conjugated secondary antibody, as described above, and visualized using an X63/1.4 Zeiss oil immersion objective on a Zeiss Axiovert fluorescence microscope, equipped with a Bio-Rad MRC-600 laser confocal imaging system. In some cases we also quantified surface labeling using a colorimetric detection assay as previously described (16). Briefly, adipocytes expressing HA-GLUT4 were grown in 24-well plates. Following the appropriate treatment cells were washed twice in PBS and fixed in 2% paraformaldehyde in PBS for 5 min. Cells were incubated with 2.5% normal swine serum and PBS for 30 min then incubated with anti-HA antibody in 1% normal swine serum in PBS or in 1% normal swine serum in PBS alone as a control for 60 min. After washing in 0.1% BSA in PBS (3 ϫ 10 min) cells were incubated with anti-mouse:horseradish peroxidase conjugate (1:5,000) in 2.5% normal rat serum in PBS for 30 min. Cells were then washed three times for 10 min each in PBS followed by incubation in o-phenylenediamine dihydrochloride reagent, made up according to the manufacturer's instructions (Sigma) for 30 min in the dark. The absorbance of the supernatant was measured at 450 nM.
Extraction of Polymerized Tubulin and Western Blotting-Polymerized tubulin was extracted from 3T3-L1 adipocytes cultured on 10-cm dishes as described previously (17). Briefly, cells were serum starved for 2 h by incubation in the presence of KRP buffer containing 0.2% BSA prior to different treatments. At the conclusion of the treatment period, cells were washed with PBS and incubated with 1 ml of Triton X-100 extraction buffer (2 M glycerol, 0.1 M PIPES, pH 7.1, 1 mM MgSO 4 , 1 mM EGTA, 0.1% Triton X-100, 10 g/ml aprotinin, 10 g/ml leupeptin, 0.25 mM phenylmethylsulfonyl fluoride) for 30 min at 37°C. The supernatant, containing the nonpolymerized tubulin, was discarded. One ml of SDS buffer (25 mM Tris-HCl, pH 7.4, 0.4 M NaCl, 0.5% SDS) was added to the dish for 5 min at 37°C to solubilize the remaining Triton X-100insoluble protein containing polymerized tubulin. 20-g aliquots of the Triton X-100-insoluble protein was resolved by SDS-polyacrylamide gel electrophoresis (18) and transferred overnight to polyvinylidene difluoride membranes. Membranes were immunoblotted using a specific anti-tubulin monoclonal antibody (DM1A) followed by incubation with horseradish peroxidase-conjugated goat anti-mouse secondary antibody. Immunoreactive bands were visualized by enhanced chemiluminescence using Supersignal reagent. Autoradiograms were quantified using a model GS-670 imaging densitometer (Bio-Rad).
Statistical Analysis-Statistical analyses were performed using Sigma Plot software. Statistical significance was established by Student's t test (p Ͻ 0.05).

Microtubule Integrity Is Not Required for Insulin-stimulated
Glucose Transport-Treatment of adipocytes with 0.2 M nocodazole for 1 h decreased the amount of polymerized tubulin by 65%, and a maximal effect of nocodazole (Ͼ95% depolymerization) was observed at a nocodazole concentration of 2 M (Fig. 1). Time course experiments revealed that these effects were rapid, occurring within 15 min of treatment (data not shown). Even though we observed a maximal effect of nocodazole on microtubule depolymerization at a concentration of 2 M we observed no significant inhibition of insulin-stimulated glucose transport or GLUT4 translocation under the same experimental conditions. Studies examining the effects of nocodazole on microtubules frequently use this drug at concentrations Ͼ10 M. At a concentration of 33 M, nocodazole caused an 80% decrease in insulin-stimulated 2-DOG uptake (Fig. 1), consistent with previous reports (7,8). Basal 2-DOG uptake was also inhibited significantly in the presence of 33 M nocodazole (ϳ75%). Basal and insulin-stimulated glucose transport were completely restored after removal of nocodazole from the culture medium (data not shown), ruling out possible cytotoxic effects of the drug.
To confirm the dose response effects of nocodazole on the integrity of microtubules in 3T3-L1 adipocytes we performed immunofluorescence microscopy on 3T3-L1 adipocytes using an anti-tubulin antibody. In the absence of nocodazole we observed an elaborate microtubule network in adipocytes. Microtubules radiated from the microtubule-organizing center to the cell periphery where they formed a tight cortical network (Fig.  2a). After incubation of adipocytes with 2 M nocodazole microtubules were completely absent (Fig. 2b), corroborating the biochemical data (Fig. 1). Treatment with 2 M nocodazole also promoted reorganization of vimentin-containing intermediate filaments into thick bundles localized beneath the PM and encapsulating the lipid droplets (Fig. 2, c versus d). The actin-based microfilaments, which presented as a cortical rim underlying the PM, were unaffected by nocodazole treatment (Fig. 2, e and f). Comparable results were achieved when cells were incubated with higher concentrations (10 -33 M) of nocodazole (data not shown).
Dispersal of the Perinuclear GLUT4 Compartment Does Not Inhibit Insulin-stimulated Translocation-We next examined the effects of 2 M nocodazole on the localization of GLUT4 in basal and insulin-stimulated adipocytes using indirect immunofluorescence microscopy. Insulin induced the translocation of GLUT4 from a perinuclear location to the cell surface (Fig. 3, basal versus ins). Treatment with 2 M nocodazole resulted in complete dispersal of the GLUT4 perinuclear compartment throughout the cytoplasm (Fig. 3, basal versus noc). Nocodazole (2 M) had no apparent effect on insulin-stimulated translocation of GLUT4 to the PM (Fig. 3, ins versus nocϩins), consistent with the glucose transport data presented above (Fig. 1). These data suggest that an intact GLUT4 perinuclear compartment is not required for efficient insulin-stimulated GLUT4 translocation. One possibility that has been proposed (7) is that microtubules regulate the movement of GLUT4 from its perinuclear compartment to the cell cortex whereupon they may be escorted to the surface membrane by actin filaments. Microtubule depolymerization may therefore affect the kinetics but not the magnitude of GLUT4 delivery to the cell surface. Extensive time course experiments were performed to look at both GLUT4 translocation and glucose transport in the presence of 2 M nocodazole. As shown in Fig. 4, we were unable to observe any significant change in the kinetics of either insulin-stimulated glucose transport or GLUT4 translocation to the PM in response to nocodazole.
Taxol Does Not Rescue the Inhibition of Insulin-stimulated Glucose Transport Induced by High Doses of Nocodazole-One possibility to explain the dose response effects of nocodazole on glucose transport (Fig. 1) is that at high doses nocodazole may inhibit insulin-stimulated glucose transport via a microtubuleindependent effect. To test this hypothesis we employed the anti-tumor agent taxol, which stabilizes microtubules (19). We reasoned that taxol may prevent any effects of nocodazole arising from microtubule depolymerization but not indirect effects of the drug. Treatment of adipocytes with 40 M taxol for 1 h promoted an increase in the level of polymerized tubulin (Fig.  5A). Moreover, taxol completely prevented the depolymerization of microtubules observed following incubation of cells with 33 M nocodazole. The presence of an intact microtubule network in cells treated simultaneously with 33 M nocodazole and 40 M taxol was confirmed by indirect immunofluorescence microscopy (data not shown). To confirm that taxol restored the function of the microtubule network we examined the localization of GLUT4 as well as two Golgi markers, syntaxin 6 and GS-15. As indicated in Fig. 5B, taxol completely prevented the nocodazole-induced redistribution of GLUT4 from its perinuclear location. Similar results were obtained for both syntaxin 6, a trans-Golgi network marker (20), and GS-15, a protein present in the Golgi apparatus (21) (data not shown).
We next examined the effects of taxol on 2-DOG uptake in adipocytes. Taxol alone had no significant effect on either basal or insulin-stimulated 2-DOG uptake (Fig. 5C) or GLUT4 translocation (data not shown). However, taxol was unable to overcome the inhibitory effect of 33 M nocodazole on basal or insulin-stimulated 2-DOG uptake (Fig. 5C).
Nocodazole Inhibits Glucose Transport Activity Independently of Effects on Signaling or Trafficking-Even though 33 M nocodazole inhibited insulin-stimulated glucose transport by 80%, we were unable to detect any significant decrease in insulin-stimulated GLUT4 translocation in the presence of high doses of nocodazole, as measured by the PM lawn assay (Fig. 6, A and B). Moreover, the movement of GLUT4 to the cell surface in the presence of 33 M nocodazole resulted in its productive fusion with the PM (Fig. 6, C and D), as determined by surface labeling of a GLUT4 construct bearing an HA epitope in one of its exofacial domains (22).
The most likely explanation for these data is that nocodazole somehow impairs the transport properties of GLUT4 after it has fused with the PM. To test this hypothesis we performed two separate experiments. First, we examined the reversal of nocodazole treatment on glucose transport under conditions in which membrane trafficking was inhibited (i.e. at 4°C). Cells were pretreated with 33 M nocodazole for 1 h, and insulin was added during the last 15 min. The cells were cooled rapidly to 4°C to immobilize GLUT4 at the cell surface, by washing with ice-cold KRP/BSA buffer in the presence or absence of 33 M nocodazole. After a 5-min incubation at 4°C, 2-DOG uptake was measured. As shown in Fig. 7A, when 2-DOG uptake was measured at 4°C the stimulatory effect of insulin was maintained. Similarly, the inhibitory effects of 33 M nocodazole on glucose transport were also maintained when 2-DOG assays were performed at 4°C. In contrast, when nocodazole was withdrawn during the 5-min incubation at 4°C there was a dramatic reversal in insulin-dependent glucose transport approaching values observed in control cells. These data provide further evidence that nocodazole does not interfere with trafficking of GLUT4 but somehow impairs the activity properties of the transporter itself. The second experiment that we performed was to determine whether nocodazole could inhibit glucose transport after the translocation process was complete. 3T3-L1 adipocytes were stimulated with insulin for 15 min at 37°C to promote translocation of GLUT4 to the cell surface. Cells were then cooled rapidly to 4°C, to prevent further trafficking, and then incubated with nocodazole on ice for 30 min prior to measurement of 2-DOG uptake, which was also performed at 4°C. The low dose of nocodazole (2 M) had no significant effect on glucose transport when added at 4°C, whereas the high dose of nocodazole (33 M) resulted in Ͼ80% inhibition of insulin-stimulated 2-DOG uptake (Fig. 7B). Similar effects of 33 M nocodazole were also observed after a 2-min incubation with the drug at 4°C, suggesting that this inhibitory effect was rapid (data not shown). As a control for these experiments we also examined the effects of the phosphatidylinositol 3-kinase inhibitor, wortmannin, at 4°C. Although this drug inhibits insulin-stimulated glucose transport when added prior to insulin treatment at 37°C we observed no inhibitory effect on insulin action when it was added after insulin at 4°C (Fig. 7B).
To determine whether the inhibition of glucose transport by nocodazole was specific to GLUT4 or common to other facilitative glucose transporters we examined the effects of nocodazole on glucose uptake in CHO cells, which only express the GLUT1 isoform (23). Treatment of CHO cells with 33 M nocodazole at Յ 4°C inhibited 2-DOG uptake by 60% (Fig. 7C). These data indicate that nocodazole inhibits glucose uptake by both GLUT4 and GLUT1 transporters in a rapid, dose-dependent manner, which is independent of effects on signaling or trafficking.

DISCUSSION
Nocodazole is frequently used by cell biologists to examine the role of microtubules in vesicle transport. Although causing complete fragmentation of the Golgi apparatus, nocodazole has limited effects on vesicle transport between the Golgi and the FIG. 5. Taxol does not rescue the inhibition of glucose transport by nocodazole. 3T3-L1 adipocytes were brought to basal conditions in KRP buffer containing 0.2% BSA buffer and then incubated in the absence (B, basal) or presence of 33 M nocodazole (N, noc) and/or 40 M taxol (T, tax) for 1 h. A, polymerized tubulin was extracted as described under "Experimental Procedures" and visualized by Western blot using anti-tubulin-␣ antibodies. A representative blot of three separate experiments is shown. B, cells cultured on coverslips were treated with nocodazole and/or taxol as described above. Cells were fixed in acetone, and GLUT4 labeling was visualized by indirect immunofluorescence as described under "Experimental Procedures." C, cells were treated with nocodazole and/or taxol as described above, and where indicated, 100 nM insulin was added for the last 15 min. 2-DOG uptake assays were performed as described under "Experimental Procedures." Error bars correspond to the S.E. from three independent experiments. * indicates the absence versus presence of nocodazole in insulin-stimulated cells, p Ͻ 0.00002.
FIG. 6. Nocodazole does not inhibit GLUT4 translocation to the cell surface. A and B, 3T3-L1 adipocytes were grown on glass coverslips and brought to basal conditions in KRP buffer containing 0.2% BSA buffer and then incubated in the absence (con) or presence of either 2 M or 33 M nocodazole (noc) for 1 h at 37°C. 100 nM insulin (ins) was added during the last 15 min, and the PM lawn assay was performed as described. Representative fields of PM lawns from different treatment groups (A) and quantitation of five different experiments (B) are shown. C and D, 3T3-L1 adipocytes expressing an HA-tagged GLUT4 molecule were incubated as described above. At the conclusion of the experiment cells were fixed, and cell surface-expressed GLUT4 was determined by labeling with an HA monoclonal antibody (panel C). Representative fields of surface labeled cells from two separate experiments are shown. D, quantitation of surface HA-GLUT4 labeling was performed by a colorimetric assay as described under "Experimental Procedures." Error bars correspond to the S.E. from three independent experiments. cell surface and on endosomal recycling. Nocodazole has little effect on recycling of the transferrin receptor (24), membrane protein processing in the Golgi (25), or retrieval of membrane proteins from the early endosome back to the trans-Golgi network (26). On the other hand, nocodazole perturbs membrane traffic between early endosomes and lysosomes (26,27). Similar results have been observed using alternate strategies to disrupt microtubule function. For example, overexpression of dynamitin mutants, which disrupt the function of microtubule motors, also perturbs trafficking to the lysosome, whereas endosomal recycling is unaffected (27). Several recent studies have reported that nocodazole inhibits insulin-stimulated glucose transport in adipocytes (7,8). This is of interest because it has been suggested that intracellular GLUT4-containing vesicles may represent specialized recycling vesicles (1). Hence, these effects may be consistent with a unique mode of transport for GLUT4 to the cell surface compared with other recycling proteins such as the transferrin receptor.
In the present study we have examined the effects of nocodazole on GLUT4 trafficking and glucose transport in detail. Initially, we studied the dose response effects of nocodazole on microtubule depolymerization and observed maximal effects of the drug at a concentration of 2 M. At this concentration of nocodazole we observed marked fragmentation of the Golgi apparatus and the perinuclear GLUT4 compartment. Despite these effects, the ability of insulin to stimulate GLUT4 translocation and glucose transport was unaffected by nocodazole. Hence, these data suggest that maintenance of the perinuclear GLUT4 compartment is not essential for insulin-mediated traffic to the cell surface (5)(6)(7)(8)(9). These observations are perhaps not surprising because this perinuclear compartment of GLUT4 may not represent the major insulin responsive pool (28). Previous immunoelectron micrographic studies in 3T3-L1 adipocytes indicate that GLUT4 is distributed between the trans-Golgi network and small tubulovesicular elements scattered throughout the cytoplasm (28). The trans-Golgi network pool, which likely constitutes the perinuclear labeling typically observed in 3T3-L1 adipocytes by immunofluorescence microscopy, does not represent the major intracellular insulin-responsive compartment. Although depolymerization of microtubules may not effect the magnitude of the GLUT4 translocation, it is possible that nocodazole may slow the rate of movement of these vesicles to the cell surface, as proposed previously (9). To investigate this possibility we studied the kinetics of GLUT4 translocation and glucose transport in the presence of nocodazole. However, we could find no evidence to indicate that the kinetics of GLUT4 translocation or glucose transport activation by insulin was impeded in the presence of nocodazole.
The observation that nocodazole had no effect on insulinstimulated glucose transport was surprising in light of previous studies (7,8). However, in each of these previous studies concentrations of nocodazole in excess of 10 M were used to study effects on glucose transport. Consistent with these previous studies, we observed a dose-dependent inhibition of nocodazole on insulin-stimulated glucose transport (Fig. 1). At 33 M nocodazole we observed Ͼ80% inhibition of insulin-stimulated glucose transport. This effect was not likely to be the result of depolymerization of microtubules because at high doses of nocodazole we were able to reverse the microtubule-depolymerizing effects with taxol but not the inhibition of glucose transport (Fig. 5). Strikingly, although nocodazole had a pronounced inhibitory effect on insulin-stimulated glucose transport we were unable to detect any inhibitory effect on GLUT4 translocation to the cell surface (Fig. 6). This was not due to an effect of nocodazole on fusion of the GLUT4 vesicles with the PM because we also observed normal GLUT4 translocation using a surface binding assay that only measures transporters that have been incorporated into the PM (Fig. 6, C and D). These results contrast with two other reports in which nocodazole was shown to inhibit GLUT4 translocation, as determined by the PM lawn assay (8,9). The basis for this discrepancy is not clear. Cells were cooled rapidly by washing with KRP buffer (Ϫ/ϩnocodazole) preequilibrated to 4°C and incubated in the same buffer (Ϫ/ϩnocodazole) for 5 min at 4°C. 2-DOG uptake measurements were performed at 4°C. Data shown represent the mean Ϯ S.D. of two experiments. B, 3T3-L1 adipocytes were brought to basal conditions in KRP buffer containing 0.2% BSA buffer and then incubated in the absence (basal) or presence of 100 nM insulin (ins) for 15 min. Where indicated, cells were incubated with 100 nM wortmannin 10 min prior to insulin treatment (wϩins). Cells were then rapidly cooled to Յ 4°C by washing in ice-cold buffer and incubated at Յ 4°C in buffer with 100 nM wortmannin (insϩwort), 2 M nocodazole (ins ϩ 2 M noc), or 33 M nocodazole (insϩ 33 M noc) for 30 min. 2-DOG uptake was measured for 90 s as described under "Experimental Procedures." Error bars correspond to the S.D. from two separate experiments. C, CHO cells, grown in 12-well plates, were cooled rapidly to Յ 4°C and incubated with 33 M nocodazole for the indicated times. 2-DOG uptake was measured for 90 s as described under "Experimental Procedures." Error bars correspond to the S.D. from two separate experiments.
In the present study we have used three different approaches to determine the extent of GLUT4 translocation to the cell surface, and all three methods indicate that nocodazole at either low or high doses does not inhibit this process significantly. In parallel, we observed a dramatic effect of nocodazole on the integrity of microtubules and the perinuclear GLUT4 compartment as well as on cellular glucose transport. Hence, it is unlikely that nocodazole was inactive in our assay systems.
The observation that nocodazole inhibited glucose transport independently of translocation of GLUT4 to the cell surface argues strongly in favor of a direct effect of this drug on the transport activity of the protein. In support of this hypothesis we observed that nocodazole inhibited glucose transport in adipocytes even when it was added to cells at 4°C (Fig. 7B). Under these conditions activation of glucose transport does not rely upon vesicle transport or signal transduction. Furthermore, similar effects were observed in cells that do not express the GLUT4 transporter (Fig. 7C). Moreover, these inhibitory effects were rapid (Fig. 7C). The precise nature of this inhibitory effect of nocodazole remains to be defined. However, nocodazole may either bind directly to the transporter, or it may modify the characteristics of the PM thus impairing the normal function of the transporter.
The present studies do not exclude the possibility that GLUT4 vesicles are transported along microtubules en route to the PM. However, these studies clearly show that the insulindependent recruitment of GLUT4 vesicles to the cell surface can occur relatively unimpaired after depolymerization of the microtubule network. Hence, our data suggest that it is unlikely that microtubule-dependent transport is a major ratelimiting step for the insulin-dependent delivery of GLUT4 to the cell surface. Most importantly, these studies reveal an effect of nocodazole, apparently unrelated to its microtubuledepolymerizing effects, to interfere with the transport properties of facilitative glucose transporters. This effect may involve many members of the glucose transport family, and so the use of this drug to study the role of microtubules is probably best confined to low doses where this nonspecific effect does not occur.