Dexamethasone Causes Translocation of Glucose Transporters from the Plasma Membrane to an Intracellular Site in Human Fibroblasts*

To investigate the mechanism by which glucocorticoids inhibit glucose transport in peripheral tissues, we have used a monoclonal antibody directed against the human glucose transporter to measure the relative amounts of glucose transporter polypeptide in various cell fractions of human foreskin fibroblasts after treat- ment with and without dexamethasone. In cells treated for 4 h with 100 nM dexamethasone, a decrease of 48% in glucose transport was accompanied by a decrease of 40% in the amount of glucose transporter polypeptide in a plasma membrane fraction enriched 10-fold in 5’-nucleotidase activity and a 78% increase in the amount of transporter polypeptide in a fraction of putative intracellular membranes, designated P2. There was no significant change in the amount of transporter polypeptide in whole cell lysates. Insulin (200 nM) stimulated glucose transport in basal fibroblasts by only 9%. However, addition of insulin for 30 min to cells that had been treated for 4 h with dexamethasone completely reversed the dexa-methasone-induced decrease in glucose transport and also reversed the dexamethasone-induced changes in glucose transporter polypeptide content of the plasma membrane and P2 fractions. From these observations we conclude that dexamethasone decreases glucose transport by causing translocation of glucose transporters from the plasma membrane to an internal lo- cation and that insulin reverses the dexamethasone effect by reversing the translocation.

To investigate the mechanism by which glucocorticoids inhibit glucose transport in peripheral tissues, we have used a monoclonal antibody directed against the human glucose transporter to measure the relative amounts of glucose transporter polypeptide in various cell fractions of human foreskin fibroblasts after treatment with and without dexamethasone. In cells treated for 4 h with 100 nM dexamethasone, a decrease of 48% in glucose transport was accompanied by a decrease of 40% in the amount of glucose transporter polypeptide in a plasma membrane fraction enriched 10-fold in 5'nucleotidase activity and a 78% increase in the amount of transporter polypeptide in a fraction of putative intracellular membranes, designated P2. There was no significant change in the amount of transporter polypeptide in whole cell lysates.
Insulin (200 nM) stimulated glucose transport in basal fibroblasts by only 9%. However, addition of insulin for 30 min to cells that had been treated for 4 h with dexamethasone completely reversed the dexamethasone-induced decrease in glucose transport and also reversed the dexamethasone-induced changes in glucose transporter polypeptide content of the plasma membrane and P2 fractions. From these observations we conclude that dexamethasone decreases glucose transport by causing translocation of glucose transporters from the plasma membrane to an internal location and that insulin reverses the dexamethasone effect by reversing the translocation.
imide (11,14,17,22,23). Kinetic studies have established that the inhibition is due to a decrease in the maximal velocity of transport ( VmaJ with no change in the substrate concentration required for half-saturation of transport (6,11,18,19). Furthermore, it has been shown that the dexamethasone effect is retained in isolated plasma membranes from rat adipocytes, as indicated by both reduced D-glucose transport into vesicles and reduced photolabeling of the transporter with the high affinity ligand, cytochalasin B (24).
These findings suggest that the inhibitory action of glucocorticoids on glucose transport is mediated by a specific glucocorticoid-induced protein(s) whose effect is to either decrease the number of functional glucose transporters in the plasma membrane or to decrease the intrinsic activity (catalytic constant) of these transporters. The former situation could be the result of either inactivation of transporters that remain in the plasma membrane, the net translocation of transporters from the plasma membrane to intracellular membranes, or the net loss of transporters through protein degradation. The possibility of translocation is prompted by the studies showing that insulin increases glucose transport into rat adipocytes by causing a translocation of glucose transporters from an internal location to the plasma membrane (25, 26).
In this paper we offer evidence that glucocorticoids inhibit glucose transport by causing the translocation of transporters from the plasma membrane to an intracellular membrane fraction. With human fibroblasts we have used a monoclonal antibody specific for the human glucose transporter (27) to quantitate the amount of transporter polypeptide in various cell fractions prepared from dexamethasone-treated and basal cells. Although dexamethasone caused no change in the total cellular content of transporter, it reduced the amount of transporter in the plasma membranes and increased the amount in a putative intracellular membrane fraction. Furthermore, insulin reversed both the dexamethasone-induced decrease in glucose transport and the dexamethasone-induced increase in the internal glucose transporters.
This study thus for the first time provides direct evidence that glucocorticoids decrease glucose transport by causing a redistribution of the transporter from the plasma membrane to an intracellular location, and that the reversal by insulin of the physiological effect of glucocorticoids is accompanied by reversal of the glucocorticoid-induced translocation of transporters. Costar (3406). MEM' (320-1095), penicillin/streptomycin (600-5145), 0.05% trypsin, 0.02% EDTA phosphate-buffered saline (610-5300), L-glutamine (320-5030), BME vitamins (320-1040), and BME amino acids (320-1051) were from GIBCO. The fetal bovine serum was purchased from Hazeton, Lenexa, KS and ICN Biomedicals, Costa Mesa, CA. Dexamethasone was from Steraloids, Wilton, NH. Porcine insulin, bovine serum albumin, Triton X-100, dithiothreitol, 2',3'-AMP, trichloroacetic acid, phenylmethanesulfonyl fluoride, and pepstatin were from Sigma. Sodium lauryl sulfate was purchased from Pierce Chemical Co. The sucrose was ultrapure grade from Schwarz/ Mann and the nitrocellulose sheets (0.2 pm) were from Schleicher & Schuell. Cell Culture-Human diploid fibroblast cultures were established from foreskins using a described method (29). The cells were grown as monolayers in MEM supplemented with fetal bovine serum (10% v/v) and penicillin (50 units/ml) and streptomycin (50 pg/ml). Stock cultures were maintained in 150-cmZ flasks in an atmosphere of 5% COz and humidified air at 37 "C. The culture medium was changed once per week, or the cells were subcultured at a 1:2 dilution using 0.05% trypsin, 0.02% EDTA solution in phosphate-buffered saline to between the 8th and 20th passages. release the cells. Cultures were used for experimental procedures Measurement of Hexose Transport-Cells were subcultured in four wells of six-well Costar cluster dishes and grown to confluence, which was about 4 days. Each well contained about 3.5 x 10' cells and 300 pg of protein. Twelve to 18 h before the experiment the cells were washed two times with 2 ml of MEM, 10% fetal bovine serum. For the experimental procedures the cells were washed three times with 4 ml of Krebs-Ringer-phosphate buffer containing 136 mM NaCl, 4.7 mM KC1, 1.25 mM MgS04, 1.0 mM CaCIZ, 5.0 mM NaHzP04, pH 7.4 (KRP), at 37 "C. Two ml of KRP supplemented with 100-fold dilutions of the stock GIBCO L-glutamine (2 mM final), BME amino acids, and BME vitamins, pH 7.4 (KRPS) was added to each well. Dexamethasone was prepared by solution in 150 mM NaCI, 10 mM sodium phosphate, pH 7.4 (PBS), and its concentration determined from the absorbance at 242 nm using a value of 1.5 X lo' M" cm" for the molar absorbance coefficient. It was added in 20 pl to two of the four wells, whereas the other two wells received the vehicle. Incubation was in a shallow water bath at 37 'C, generally for 4 h. The pH of the medium did not change from 7.4 during this period. At the end of the incubation, the buffer was aspirated and replaced by 950 pl of KRPS containing either no addition, dexamethasone, insulin, or both at 37 "C. Hexose uptake at 37 "C was then initiated by adding 50 pl of PBS containing radiolabeled 2-deoxyglucose or 3-0-methylglucose to a final concentration of 50 p~ hexose and 1 pCi/ ml. Uptake was terminated by flicking the medium into a waste container and washing the cells by dipping each plate successively in a series of three baths of PBS at 0 'C. For 3-0-methylglucose transport, the transport inhibitor phloretin (100 p~) was included in the PBS. The cells were solubilized in 1.0 ml of 1.0% Triton X-100, and the radioactivity in a 6 0 0 4 aliquot was determined on a Packard Tri-Carb scintillation spectrometer.
determined. The uptake of 2-deoxyglucose in the presence of 20 pM cytochalasin B, a potent inhibitor of transport (30), was 5% of the value found in its absence. Because this percentage was so low, the values for the uptake of 2-deoxyglucose were not corrected for cytochalasin B-insensitive uptake. Values from wells on the same plate routinely agreed within +5% of the average. Typically, the value for 2-deoxyglucose uptake was about 100 pmol/min/35-mm dish of basal cells.
Relative Amounts of Total Glucose Transporter-Confluent monolayers were treated exactly as described for uptake. After washing with KRP the monolayers were resuspended in 2 ml of KRPS. PBS (20 pl) was added to two of the four wells and 20 pl of 10 p~ dexamethasone was added to the remaining two wells. The medium was swirled to ensure thorough mixing, and the cells were incubated at 37 "C. After 4 h the medium was aspirated and the cells were immediately chilled by immersion in PBS at 0 "C. The PBS was thoroughly aspirated from the wells and the cells were scraped off the dish with a rubber policeman in approximately 0.5 ml of PBS/well. Samples from replicate wells of ten dishes were pooled, centrifuged, and resuspended in 500 p1 of PBS, and the protein concentration was determined. Protease inhibitors pepstatin, EP-475, and diisopropyl fluorophosphate were added to the cell samples at final concentrations of 1 pg/ml, 10 p~, and 1 mM, respectively. These were stored at -80 ' C and subsequently thawed and diluted in SDS sample buffer to a final protein concentration of 2 mg/ml. Subcellular Fractionation-Human foreskin fibroblasts were fractionated by a slight modification of the method of Buchanan et al. (31). The fibroblasts were subcultured in 150-cm2 culture flasks and grown to confluence. The cells were given fresh medium on the evening before the experiment. The next day the cells were washed free of culture medium with KRP and subjected to hormone treatment at 37 "C in KRPS (see figure legends for details). For each experiment the subcellular fractionations of cells from six basal and six dexamethasone-treated 150-cm' flasks were carried out in parallel. All subsequent procedures were carried out in a 3 'C cold room. The cells were rapidly chilled with 150-200 ml of ice-cold PBS/flask. The fibroblasts were washed free of PBS with 150-200 ml of hypotonic lysis medium, 1 mM NaHC03, pH 8.25, and were bathed in 6 ml of lysis medium and rotated gently on an orbital shaker for 1 min. The lysis medium was decanted and the procedure repeated twice. The flasks were broken open with a wrench, and the cells were scraped off the flask bottom with a plastic cell scraper in 6 ml of lysis medium supplemented with the protease inhibitors, 100 p~ phenylmethanesulfonyl fluoride, 1 pg/ml pepstatin, and 10 p~ EP-475. Contents from six flasks were pooled and homogenized with five manual strokes in a glass homogenizer with a Teflon pestle (Thomas Scientific, model 3431 D70C). The pooled lysates (35-38 ml) were brought to a final concentration of 0.5 mM NazEDTA and centrifuged in two polycarbonate tubes at 28,000 X g. , for 20 min (20,000 rpm, Beckman Ti-60 rotor). The pellets (PI fraction) were resuspended in 3 ml of 10% (w/ v) sucrose by trituration with a pipette, layered on a discontinuous gradient at 27, 38, and 48% (w/v) sucrose and centrifuged at 76,000 X gave for 2 h (25,000 rpm, Beckman SW 41 rotor). The bands at the 10/27, 27/38, and 38/48% interfaces and the pellet were designated fractions F,-F,, respectively. These were removed in about 1 ml, diluted with 5 ml of PBS, and centrifuged at 70,000 X g. , for 25 min (32,000 rpm, Beckman Ti-50 rotor). The pelletsd (F1-F') from this centrifugation were resuspended in 150-200 pl of PBS. The supernatant from the first 28,000 x g centrifugation was centrifuged at 150,000 X gave for 90 min (45,000 rpm, Beckman Ti-60 rotor). The resultant pellet, designated Pa, was resuspended in 300 pl of PBS. The fractions were frozen in liquid nitrogen and stored at -80 "C.
Gel Electrophoresis and Zmmunoblotting-Final concentrations in sample buffer for SDS-polyacrylamide gel electrophoresis were 2% (w/v) SDS, 25 mM dithiothreitol, 10% (v/v) glycerol, 1 mM EDTA, 95 mM Tris-C1, pH 6.8, and 100 pg/ml bromphenol blue; the sample buffer also included pepstatin, EP-475, and diisopropyl fluorophosphate at final concentrations of 1 pg/ml, 10 p~, and 1 mM, respectively. The samples were not heated; those containing DNA were sheared by repeated passage through a 25-gauge needle. Polypeptides were separated on 10% polyacrylamide slab gels (9 x 6 x 0.05 cm) according to the method of Laemmli (32). For immunoblotting of the glucose transporter, the proteins were electrophoretically transferred from the gel onto nitrocellulose in 25 mM sodium phosphate buffer, pH 6.5. After transfer for 4 h at 225 mA, the nitrocellulose blot was blocked with 3% bovine serum albumin and then treated with the monoclonal antibody directed against the human erythrocyte glucose transporter at 2 pg/ml. The transporter-antibody complex was labeled with a second antibody, lZ5I-sheep anti-mouse F(ab'), at 1.2 X IO6 dpm/ml, and the complex was detected by autoradiography. The details of this immunoblotting procedure have been described (33). Labeled bands were excised from the nitrocellulose using the autoradiogram as a template. The bands were counted in a Packard gamma counter. Each value was corrected for background label by cutting and counting a region within the same lane that showed no labeled band. The amount of protein applied to each lane of the gel was always in a range such that the radioactivity in the transporter band on the blot was directly proportional to the protein applied. There was some variation in the amount of radioactivity found in the transporter band from blot to blot; separate blots of the same samples performed with the same reagents gave values that differed by as much as 25%. For this reason, unless noted otherwise, all comparisons of the relative amounts of transporter are based on values from lanes on the same blot.
Glycosidase Treatment of Intact Fibroblasts-Cells were grown as monolayers in 35-mm wells and treated as described above for the transport measurements. After the 4-h period in the absence or presence of 100 nM dexamethasone at 37 "C, the cells were rapidly cooled by dipping the dish in PBS at 0 "C. All subsequent procedures were done in a 3 'C cold room. The PBS was aspirated and the cells were washed in sodium acetate buffer (70 mM NaCl, 100 mM sodium acetate, pH 5.8). Individual wells of replicate control and dexamethasone-treated cells were covered with 1.0 ml of sodium acetate buffer with or without 100 milliunits of endo-&galactosidase. At various times the enzyme solution was removed and the digestion was stopped by dipping the dish in PBS containing 10 mM mercuric chloride (41). Excess buffer was aspirated, and the cells were solubilized in 150 pl of SDS sample buffer with the protease inhibitors. The mobility of the glucose transporter was examined by gel electrophoresis and immunoblotting. To determine whether the dexamethasone-induced decrease in glucose transport persisted under the above conditions, the cells were treated for 4 h at 37 "C in KRPS in the absence or presence of 100 nM dexamethasone and subsequently for 2 h at 3 "C in 1 ml of sodium acetate buffer, pH 5.8. 2-Deoxyglucose uptake was then measured at 3 "C in sodium acetate buffer, pH 5.8. The uptake of 2-deoxyglucose into cells treated in this way was linear for 15 min and prior dexamethasone treatment decreased the rate of uptake by 45%.
Other Procedures"5'-Nucleotidase activity was measured by the method of Avruch et al (34) in the presence of 0.1% Triton X-100 and 50 pg/ml bovine serum albumin; 5 mM 2',3'-AMP was included in the assay medium in order to inhibit nonspecific phosphatase activity. For determination of protein concentration, protein was precipitated with 10% (w/v) trichloroacetic acid and measured colorimetrically according to a modification of the Lowry method (351, with bovine serum albumin as a standard. All data are expressed as means f S.D.; their statistical significance was evaluated using Student's t test.

Time Course and Dose Response for the Dexamethasone Effect on Hexose Transport in Human
Fibroblasts-The monoclonal antibody that was available for quantitating the glucose transporter polypeptide by the immunoblot procedure does not cross-react with the glucose transporter of rodent cells, which have been used for most of the previous studies of glucocorticoid regulation of hexose transport (1-19). Consequently, a human cell line was chosen for investigation. It had been briefly reported that treatment of human fibroblasts with dexamethasone phosphate at a very high concentration (5 p~) for a prolonged period (48 h) reduced 2-deoxyglucose uptake by 34% (21). We have found that 100 nM dexamethasone reduced 2-deoxyglucose uptake by over 40% in 4 h, as shown in Fig. 1. A significant reduction in uptake occurred within 30 min. By 4 h a maximal inhibition of 45 & 8% is reached (range of inhibition between 32 and 59% in nine separate experiments). The half-time for development of the maximum effect is about 80 min. In a control experiment dexamethasone (1 p~) added to cells immediately prior to the uptake assay had no effect on 2-deoxyglucose uptake (not shown). The dose-response relationship for dexamethasone inhibition of hexose uptake is shown in Fig. 2. The cells were exposed for 4 h to various concentrations of dexamethasone. Some decrease in 2-deoxyglucose uptake was evident at concentrations as low as 0.1 nM, whereas the maximum suppressive effect was obtained at 100 nM. The half-maximum decrease occurred at about 5 nM, a value which is within the range of values reported for the dissociation constant for dexamethasone binding to the glucocorticoid receptor in human foreskin fibroblasts (3-25 nM) (36, 37) and is typical for physiological glucocorticoid effects.
The measurement of hexose transport by the uptake of 2deoxyglucose includes both the transport of this sugar and its phosphorylation by hexokinase (38). In this study the assumption has been that transport is the rate-limiting step. To test this assumption, the effect of dexamethasone (100 nM, 4 h, 37 "C) on the transport of 3-O-methylglucose, a glucose analogue that does not undergo any metabolic conversion (38), was compared to its effect on the uptake of 2-deoxyglucose. Dexamethasone increased the half-time for equilibration of cells with 3-0-methylglucose from 25 to 60 s; this effect thus corresponds to a 58% reduction in the rate of transport.
Total Amounts of the Glucose Transporter in Basal and Dexamethusone-treated Fibroblasts-Having established that dexamethasone caused a rapid decrease in hexose transport in these fibroblasts,/we examined the effect of the hormone on the amount of the glucose transporter protein. Cells were treated with 100 nM dexamethasone. A 39-44% decrease in transport was verified by measurement of 2-deoxyglucose uptake in cells carried in parallel incubations. Whole cells solubilized in SDS sample buffer were analyzed for glucose transporter polypeptide by SDS-PAGE and immunoblotting, as described under "Experimental Procedures." A single band with a mobility corresponding to 55 kDa was found (see Figs. 3 and 5); the transporter in human erythrocytes and Hep G2 cells has previously been shown to be this size (27,39). The radioactivity in the band was directly proportional to the amount of protein loaded on the gel over the range of 10-40 pg (data not shown), and thus this method enabled quantitation of the relative amount of transporter polypeptide in various samples. In five separate experiments, SDS samples from paired plates of basal and dexamethasone-treated cells were prepared and immunoblotted in triplicate on a single gel. The average value for the ratio of the amount of total cellular glucose transporter protein in dexamethasone-treated cells to that in basal cells was 0.95 & 0.10. This value is not significantly different from 1.0 at the level of p < 0.05. Thus, dexamethasone inhibition of hexose transport is unlikely to be due to net degradation of the transporter. Subcellular Fractionation of Fibroblasts-Cells were treated withd 100 nM dexamethasone for 4 h and were fractionated as described under "Experimental Procedures." In this procedure the method for fractionating human fibroblasts described by Buchanan et al. (31) was modified by changing the sucrose concentrations of the gradlent step to improve the yield of plasma membranes. The method involved an initial 28,000 X g. , centrifugation of the homogenate. The resulting pellet, PI, containing most of the membranous material, was fractionated on a discontinuous sucrose gradient. Cell organelles sedimenting at the 10/27, 27/38, and 38/48% sucrose interfaces and at the bottom of the gradient were designated F,-F,, respectively. F1 and F, were 8-10-fold enriched in 5'nucleotidase activity ( Table I). We conclude that these two fractions contain the plasma membranes. On the basis of the work of Buchanan et al. (31), fractions FS and F4 would be expected to contain primarily mitochondria and nuclei, respectively. In an extension of the procedure of Buchanan et al. (31), we drew upon a subcellular fractionation procedure for rat adipocytes described by Cushman and Wardzala (25). They isolated a microsomal membrane fraction which contains an internal pool of glucose transporters by high-speed centrifugation of the supernatant from an initial centrifugation of the homogenate at 12,000 X gave for 15 min. In this study, the initial supernatant was centrifuged at 150,000 X g. , for 90 min. This gave a pellet, designated P,, that was not enriched in 5'-nucleotidase and presumably comprises internal membranes. Table I shows the results of a typical subcellular fractionation. The data indicate that incubation of intact fibroblasts with a maximally suppressive concentration of dexamethasone had no significant effect of the distribution of protein or 5'-nucleotidase activity, with the possible exception of 5'nucleotidase activity in fraction F,. This entire fractionation experiment with basal and dexamethasone-treated cells was repeated five times. Ratios of protein and 5'-nucleotidase specific activity in each fraction from the dexamethasonetreated to those in the same fraction from basal cells were

Fig. 3). The corrected counts/min in the transporter band of the immunoblot was divided by the amount of protein loaded onto the lane of the SDS gel. Fractions F2 and
Pz were run on two blots, and the values from each are given.
In another experiment this fraction was assayed for 5"nucleotidase activity and found to contain enzyme at a specific activity of 0.9.
ND, not determined. e The yield of F, was low here due to loss. In four other experiments both basal and dexamethasone-treated cells yielded about 0.7 mg of F,.
calculated. The values of these ratios for each fraction were not significantly different from 1.0 (p < 0.05). This includes the ratio for 5'-nucleotidase specific activity in fraction F,, which averaged 1.10 k 0.16 (S.D.).
Glucose Transporter Content of Subcellular Fractions-The glucose transporter content of the subcellular fractions was determined by SDS-PAGE and immunoblotting. Fig. 3 shows a representative autoradiograph of the glucose transporter polypeptide in the subcellular fractions, with the samples from basal and dexamethasone-treated cells in paired lanes; the last two columns of Table I give the quantitative data from this experiment. As expected, the plasma membrane fractions F, and Fz were enriched in glucose transporter. Typically, the degree of enrichment was 4-5-fold. In addition, a considerable amount of transporter was located in the putative intracellular membrane fraction Pa.
The effect of dexamethasone on the glucose transporter content of the various fractions was to cause a decrease in the plasma membrane fraction Fz and an increase in the P, fraction ( Fig. 3 and Table I). Data from five separate experiments of this type are summarized in Fig. 4. Although in the single experiment shown in Fig. 3 and Table I, the ratio of transporter in the homogenate from dexamethasone-treated to basal cells was 1.24, the average value from the five experiments was not significantly different from 1.0 (Fig. 4), in agreement with the results for total transporter content presented above. The transporter content of the F, fraction was reduced to 60% of the basal value, whereas the content of the -TOP P2 fraction rose to 178% of the basal value. Dexamethasone treatment had no significant effect on the transporter content of the other plasma membrane fraction, FI. As would be expected, the hormone slightly reduced the transporter content of fraction PI, the pellet from which F, is derived. Also, a decrease in fraction Fs, which on the basis of 5"nucleotidase activity and transporter content has only a low content of plasma membrane (Table I), was almost certainly due to the contamination of this fraction with the F,-type plasma membrane. Deglycosylation of Glucose Transporters on the Surface of Intact Fibroblasts-The carbohydrate moiety of the glucose transporter from human erythrocytes is susceptible to partial digestion by the glycosidase endo-@-galactosidase, and the deglycosylated transporter migrates with a greater mobility on SDS-PAGE (40). These observations suggested that it might be possible to distinguish surface from intracellular transporters in human fibroblasts by selectively deglycosylating the surface transporters with endo-@-galactosidase and then examining the total cellular transporter pool by SDS-PAGE and immunoblotting. Basal and dexamethasonetreated cells were exposed to the glycosidase at 3 "C in order to prevent both endocytosis of the enzyme and recycling of the transporter, either of which might lead to deglycosylated intracellular transporters.
An autoradiograph from an experiment of this type is presented in Fig. 5. After 1 h of incubation with the glycosidase (lanes 4 and 5), most of the transporter in basal cells migrated as a sharp band of higher mobility as compared to the cells treated for only 1 min (lane I), whereas a substantial portion of the transporter in dexamethasone-tmated cells still migrated as the broad band of lower mobility (lane 5). Several experiments of this type consistently showed that a greater portion of the glucose transporters was protected from digestion in dexamethasone-treated cells as compared with basal cells. We presume that this protected population is the internal pool of transporters. Unfortunately, glycosidase treatment did not increase the mobility of the susceptible transporters sufficiently to allow clean separation from the unmodified transporters, and consequently this result, although consistent with the results from the fractionation experiments, is only qualitative.
Insulin Reversal of the Dexamethasone Effects-Germinario et a1 (21) found that, after reduction of hexose transport by 34% upon exposure to 5 PM dexamethasone phosphate for 48 h, treatment with 60 nM insulin for 2 h led to a %fold increase in the rate of transport, such that the rate was the same as that in basal cells treated with insulin. This observation and other reports demonstrating insulin reversal of the dexamethasone effect in rat adipocytes (42)(43)(44)    port and then exposed to 200 nM insulin for 15 and 30 min.
The results in Table I1 show that insulin treatment restored the rate of transport to that seen in basal fibroblasts. Fibroblasts that had not been exposed to dexamethasone exhibited a small, but statistically significant increase in uptake.
The effect of insulin on the subcellular fractionation of the glucose transporter was examined after treatment of cells under identical conditions to those used for the examination of transport. In one experiment complete subcellular fractionations of basal, dexamethasone-treated, insulin-treated, and dexamethasone plus insulin-treated fibroblasts were performed. Immunoblots of the plasma membrane fraction F, and the putative intracellular membrane fraction Pz from this experiment were obtained. In agreement with the effect on transport, insulin reversed the dexamethasone effect on the glucose transporter content of both fractions to the basal state. The values of the ratios of the amount of transporter polypeptide in the F, fraction from dexamethasone-treated, insulin-treated, and dexamethasone plus insulin-treated cells to that in this fraction from basal cells were 0.61, 0.97, and 0.94, respectively.
Because it was technically difficult to perform four complete subcellular fractionations at the same time, on repetition this experiment was simplified by isolating only the Pz fraction.
This simpler experiment was carried out twice. Mean values of the ratio of glucose transporter in the P, fraction from the treated to untreated fibroblasts for the three experiments were 1.44 f 0.10, 0.72 f 0.05, and 0.91 f 0.14 for dexamethasone alone, insulin alone, and dexamethasone plus insulin, respectively. The ratios for the dexamethasone and insulin treatments are significantly different from 1.0 at the level of p 0.005 and p < 0.05, respectively, whereas that for the combined treatment is not significantly different from 1.0 (p < 0.05). Thus, these results reinforce the finding from the single complete subcellular fractionation. The value of less than 1.0 for insulin alone is consistent with the view that the slight increase in transport in response to insulin alone is due to translocation of transporter from the internal pool to the plasma membrane.

DISCUSSION
From our finding that the dexamethasone-induced inhibition of hexose transport in human fibroblasts was not accompanied by a change in the total cellular content of the glucose transporter polypeptide, we conclude that the decrease in transport activity is not due to a change in the rate of biosynthesis or degradation of the transporter. Two alternative explanations are (a) a change in the activity of the transporter in the plasma membrane, and ( b ) translocation of transporters from the cell surface to an intracellular location. The latter explanation is supported by the following evidence: (i) Dexamethasone treatment decreased the transporter content of the plasma membrane fraction Fa and increased the content of the putative intracellular membrane fraction Pa.
(ii) Dexamethasone treatment increased the proportion of the glucose transporters that were resistant to exoplasmic digestion by endo-@-galactosidase. (iii) Insulin reversed the dexamethasone effect on both transport and subcellular localization of the transporter. In fat and muscle cells insulin is known to cause the translocation of glucose transporters from the interior to the plasma membrane (25, 26,45,46).
For a rigorous assessment of the stoichiometry of transporter translocation, both the plasma membrane and intracellular membranes should ideally be isolated as pure fractions in 100% yield. Then the total amount of glucose transporter lost from the plasma membrane should appear in intracellular membranes. Moreover, if translocation accounted for the entire effect on transport, the transporter content of the plasma membrane and transport activity should decrease by the same percentage. On the basis of 5"nucleotidase activity we recovered only 37% of the plasma membrane in fractions F1 and F,; the main losses probably occurred in the sucrose layers and during the sedimentation and resuspension of the fractions. Furthermore, the intracellular location of the glucose transporter is not known; hence, there is no known marker for assessing the fractionation characteristics and the recovery of the intracellular membranes containing the transporter. Thus, an analysis of the stoichiometry of the translocation process is not yet possible.
We propose the following semi-quantitative interpretation of our data. The plasma membrane-enriched fraction Fz showed a 40% reduction in glucose transporter in response to dexamethasone. This value is almost as large as the effect on transport itself, and therefore this finding suggests that translocation accounts entirely for the effect on transport. The other plasma membrane-enriched fraction, F1, showed no change in glucose transporter. An explanation for this apparent inconsistency is that F1 may also contain intracellular membranes, such that the decrease in the transporter content of the plasma membranes in this fraction is approximately offset by the increase in the transporter content of the intracellular membranes in this fraction. In addition, another portion of the intracellular membranes containing transporter appeared in the Pz fraction, which contained very little plasma membrane on the basis of its 5'-nucleotidase activity. This fraction may exhibit almost the full increase in glucose transporter content of the intracellular membranes, which would thus be about 80% greater than the basal content. For the five separate fractionation experiments, the actual increase in the transporter content of P,, expressed as counts/min from the immunoblot, averaged 95% of the decrease in transporter in the Fz fraction (Table I and data not shown).
Carter-Su and Okamoto (24) investigated the dexamethasone-induced decrease in glucose transport in isolated rat adipocytes. They showed that plasma membrane vesicles prepared from basal and dexamethasone-treated cells retained the decrease in transport activity observed in the intact adipocyte. These investigators identified the glucose transporter in the plasma membranes, intracellular membranes, and total cellular membranes from basal and dexamethasone-treated cells by photoaffinity labeling with the ligand, cytochalasin

Dexamethasone Causes Transk
B. The amount of photoaffinity-labeled glucose transporter found in the plasma membranes of dexamethasone-treated cells was less than that in basal cells, but there was no change in the amount found in the intracellular or total membranes. However, as the authors noted, in rat adipocytes the intracellular pool of transporters is so large that no increase would have been observed, even if the basis for the dexamethasone effect was translocation from the plasma membrane to the intracellular location. These results were thus consistent with translocation, but did not exclude the alternative interpretation that dexamethasone caused a change in the intrinsic activities of plasma membrane transporter such that each transporter exhibited a lower turnover number and a reduced susceptibility to photoaffinity labeling.
Our results indicate that a focus of further investigation should be the search for the dexamethasone-induced protein(s) that causes the redistribution of glucose transporters between the plasma membrane and intracellular membranes. Possibly this protein(s) functions by direct interaction with the glucose transporter.