Insulin Elicits a Redistribution of Transferrin Receptors in 3T3-Ll Adipocytes through an Increase in the Rate Constant for Receptor Externalization"

Incubation of 3T3-Ll adipocytes with insulin at 37 "C resulted in a 2-fold increase in specific binding of transferrin to cell-surface receptors, as measured by a subsequent incubation of cells at 4 "C with lZ5I-transferrin. The insulin concentration required for half-maximal effect was 10 nM, and the half-time for insulin action was 40 s. By comparison, insulin stimu- lated hexose transport in 3T3-Ll adipocytes with a half-maximal effect at 8 nM and a half-time of 105 s. Scatchard analysis of 1251-transferrin binding to cells at 4 "C showed that the insulin-induced increase in transferrin receptor binding was due to an increase in the number of surface transferrin receptors. When cells were incubated for 2 h at 37 "C with "'1-trans-ferrin to achieve steady-state binding and then exposed to insulin, there was a 1.7-fold increase in surface- bound transferrin (acid-sensitive) and a corresponding decrease in intracellularly bound transferrin (acid- insensitive). Thus, insulin elicits translocation of intracellular transferrin receptors to the plasma membrane. Concomitant with the 2-fold increase in surface receptors in response to insulin, there was a 2-fold increase in the rate of 58Fe3+ uptake from '@Fe3+-loaded trans- ferrin. The rate of externalization of the intracellular '"1-transferrin-receptor complex at 37 "C was deter- mined for basal and insulin-treated cells. Insulin in- creased the first-order rate constant for this


Insulin Elicits a Redistribution of Transferrin Receptors in 3T3-Ll
Adipocytes through an Increase in the Rate Constant for Receptor Externalization" ( Incubation of 3T3-Ll adipocytes with insulin at 3 7 "C resulted in a 2-fold increase in specific binding of transferrin to cell-surface receptors, as measured by a subsequent incubation of cells at 4 "C with lZ5Itransferrin. The insulin concentration required for half-maximal effect was 10 nM, and the half-time for insulin action was 40 s. By comparison, insulin stimulated hexose transport in 3T3-Ll adipocytes with a half-maximal effect at 8 nM and a half-time of 105 s. Scatchard analysis of 1251-transferrin binding to cells at 4 "C showed that the insulin-induced increase in transferrin receptor binding was due to an increase in the number of surface transferrin receptors. When cells were incubated for 2 h at 37 "C with "'1-transferrin to achieve steady-state binding and then exposed to insulin, there was a 1.7-fold increase in surfacebound transferrin (acid-sensitive) and a corresponding decrease in intracellularly bound transferrin (acidinsensitive). Thus, insulin elicits translocation of intracellular transferrin receptors to the plasma membrane. Concomitant with the 2-fold increase in surface receptors in response to insulin, there was a 2-fold increase in the rate of 58Fe3+ uptake from '@Fe3+-loaded transferrin. The rate of externalization of the intracellular '"1-transferrin-receptor complex at 37 "C was determined for basal and insulin-treated cells. Insulin increased the first-order rate constant for this process 1.7-fold. The effect of insulin on the rate of externalization is sufficient to account for the increase in surface transferrin receptors.
Transferrin is a serum glycoprotein involved in the transport of iron into cells (1). The cellular uptake of iron is initiated by the binding of diferric transferrin to a transferrin receptor on the plasma membrane. The ferrotransferrin-receptor complex is then internalized in a temperature-dependent process (2,3). Iron dissociates from transferrin in an acidic, nonlysosomal compartment and accumulates within the cell, whereas the resulting apotransferrin remains bound to its receptor (3)(4)(5)(6). The apotransferrin-receptor complex recycles back to the plasma membrane where, at pH 7.4, there is rapid release of intact apotransferrin into the media (2, 3). Unoccupied transferrin receptors also continuously recycle between the plasma membrane and intracellular membranes (7). *This work was supported by a fellowship from the Juvenile Diabetes Foundation (to L. I. T.) and Grant DK 25336 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The steady-state level of plasma membrane transferrin receptors is subject to hormonal regulation. Insulin elicits a %fold increase in the level of surface transferrin receptors in rat adipocytes due to the redistribution of intracellular receptors to the plasma membrane (8). The simplest kinetic cit scription of the steady-state condition for an unoccupied receptor that is recycling continuously between the plasma membrane and the intracellular membranes is (9): k,,[intracellular receptors] = %,[plasma membrane receptors], where ke, and ken are the first-order rate constants for externalization and endocytosis of the receptor, respectively. Thus, according to this analysis, the ratio of plasma membrane receptors to intracellular receptors is equal to kex/ken, and the insulin-induced increase in this ratio found for adipocytes could be due to either an increase in k,, or a decrease in ken.
In the present study, we show that insulin also causes net translocation of transferrin receptors to the plasma membrane in 3T3-Ll adipocytes, and we have examined the basis of this effect. By direct measurement of the rate of release of intracellular transferrin, we have established that the locus of insulin action is on the rate constant for externalization of the transferrin-receptor complex. In addition, the time course and concentration dependence of the stimulatory effect of insulin on surface transferrin receptors and hexose transport have been compared, since the latter process involves the translocation of glucose transporters from an intracellular site to the plasma membrane (10,11).
Mature 3T3-Ll adipocytes (2 X lo6 cells/well, 2 mg of protein/well) were used between 8 and 12 days after initiation of differentiation, at which time greater than 95% of the cells exhibited the adipocyte phenotype (12). Cells were incubated for 2 h in serum-free Dulbecco's modified Eagle's medium at the beginning of each experiment.
Hexose Transport-Hexose transport was assayed by the uptake of 2-deoxyglucose, a process for which transport has previously been ~ ~ The abbreviations used are: BSA, bovine serum albumin; IGF-11, insulin-like growth factor 11; KRP, Krebs-Ringer phosphate buffer. 8975 shown to be rate-limiting in 3T3-Ll adipocytes (12). Measurement of uptake at 4 "C after exposure of cells to vehicle or insulin at 37 "C washed at 37 "C with three 1-ml aliquots of KRP (128 mM NaCl, 4.7 (see Fig. 1) involved the following procedure. Cell monolayers were mM KCI, 1.25 mM MgSO,, 1.25 mM CaC12, 5 mM NaH,PO,, pH 7.4), and then 2.0 ml of KRP was added. Cells were then incubated at 37 "C with vehicle (50 p1 of 10 mM HCl) or insulin (final concentration, 100 nM). The pH of KRP after this addition was 7.28. At the desired time, a 6-well dish was transferred to ice, and buffer was replaced with a ml/well KRP at 4 "C. Each well was washed subsequently with two additional 1-ml aliquots of KRP at 4 "C and then incubated for 2 h at 4 "C with 0.95 ml of KRP containing 1 mg/ml BSA. Hexose uptake was initiated by addition of 50 pl of 2-de0xy[~Hj glucose (final concentration, 50 p~; 0.3 pCi/well) and stopped 20 min later by aspiration of the medium, followed by three rinses with KRP at 4 "C. The cells were solubilized with 1.0 ml of 1% Triton X-100, and radioactivity was measured by liquid scintillation spectrometry. The uptake of 2-deo~y[~H]glucose at 4 "C in basal and insulinstimulated cells is linear for at least 30 min. This protocol was adopted so that cells were treated in an identical manner in the comparison of the time courses of insulin action on surface transferrin receptors and hexose transport (see Fig. 1).
When hexose uptake at 37 "C was assessed (see Fig. 21, cell monolayers were first washed with three 1-ml aliquots of KRP and then incubated for 20 min at 37 "C with 0.95 ml of KRP and vehicle (0.01 ml of 10 mM HCI) or various concentrations of insulin. The pH of KRP after this addition was 7.35. Hexose uptake was initiated by addition of 50 p1 of 2-deo~y[~H]glucose (final concentration, 50 p~; 0.3 pCi/well) and stopped 5 min later by three rinses with phosphatebuffered saline at 4 "C. Cell-associated radioactivity was measured as described above. The uptake of 2-deoxyglucose at 37 "C in basal and insulin-treated cells is linear for 20 min.
1251-Trunsferrin Binding ut 4 "C-Cell monolayers were washed with three 1-ml aliquots of KRP a t 37 "C followed by addition of 2.0 ml of KRP and 50 pl of vehicle (10 mM HCl) or insulin (see the figure pH of KRP was 7.28. At the desired time, a 6-well dish was transferred legends for insulin concentration and length of exposure). The final to ice, and the buffer was replaced with 1 ml of KRP at 4 "C. Each well was subsequently washed with two additional 1-ml aliquots of KRP at 4 "C and incubated for 2 h with 1 ml of 3 nM '251-transferrin (unless stated otherwise) in KRP containing 1 mg/ml BSA at 4 "C (100,000-150,000 cpm/well). A t the end of the 2-h incubation, unbound ligand was aspirated, and nonspecific binding was reduced by three 1-ml washes (1 min each) with cold KRP. Cells were solubilized with 1 N NaOH, and the radioactivity was determined in a Beckman y-counter. A control experiment, in which the efficacy of three rapid washes was compared with three 1-min washes, indicated that none of the specifically bound '"I-transferrin was lost in the latter procedure, whereas the percent nonspecific binding was reduced. Nonspecific binding was taken as the '=I-transferrin bound in the presence of an excess (1 p~) of unlabeled diferric transferrin; it ranged from 20 to 50% of the total binding. The value for nonspecific binding probably consists of both unbound transferrin trapped with the cells and transferrin bound to unknown sites of low affinity present in relatively large amounts. All data have been corrected for nonspecific binding.
In control experiments, the time course of binding of 3 and 10 nM 'Z51-transferrin to cells at 4 "C was examined over a 5-h period. Specific binding of '251-transferrin exhibited two phases: a rapid increase and then a slow linear rise. The initial phase was complete within 30 min and 1 h for 10 and 3 nM transferrin, respectively, and constituted about 90% of the binding at 2 h. We have therefore defined the specific binding after 2 h at 4 "C as equilibrium binding to surface receptors. We have been unable to determine the cause of the slow second phase of specific binding. 1251-Trunsferrin Binding ut 37 "C-Cell monolayers were washed with three 1-ml aliquots of KRP at 37 "C and then incubated with 3 nM '251-transferrin in KRP containing 1 mg/ml BSA for 2 h at 37 "C. Vehicle or insulin was then added as described for Table I; and at the washes of 1-min duration with cold KRP. Surface-bound ligand was desired time, the cells were transferred to ice and given three 1-ml discriminated from intracellular ligand by an acid-stripping procedure (13). Cells were incubated for 8 min at 4 "C with 1 ml of 200 mM acetic acid, 500 mM NaCl, pH 2.5, and rinsed with another 1 ml of this acid wash; the radioactivity in the combined washes, containing the acid-releasable lZ5I-transferrin, was determined. The cell monolayers were then solubilized with 1 N NaOH, and the acid-insensitive '251-transferrin was measured. Nonspecific binding, '251-transferrin bound in the presence of 1 p~ unlabeled diferric transferrin, ranged from 15 to 50% of the total binding. All data have been corrected for nonspecific binding.
In a control experiment, the time course of acid-sensitive binding (to surface receptors) and acid-insensitive binding (to intracellular receptors) of 'Z51-transferrin (3 nM) at 37 "C was examined. The binding to the two components reached a plateau level within 15 min and 1.5 h, respectively, and remained constant over the next 2.5 h. Thus, the binding of 3 nM '"I-transferrin for 2 h at 37 "C reflects steady-state conditions.
In another control experiment, the acid-stripping procedure was applied to cells incubated at 4 "C with lZI-transferrin (3 nM) from 15 min to 5 h. Under these conditions, where only surface binding would be expected, 70% of the '261-transferrin specifically bound at each time point was released by the acid treatment. Apparently, the acid treatment does not release 30% of the surface-bound transferrin. The alternative explanation, that 30% of the transferrin is internalized at 4 "C, seems unlikely since this percentage would have been expected to increase with time. The data under "Results" have not been corrected for incomplete release and thus may underestimate the binding of transferrin to surface receptors at 37 "C by 30%. Analysis of the data with this correction would not significantly alter the conclusions.
Rate of Extermlizution of Intracellular '25Z-Trunsferrin-Ce11s were incubated with 3 nM '251-transferrin at 37 'C for 2 h, exposed to vehicle or insulin at 37 "C for 5 min, and washed on ice as described above. Measurement of the rate of externalization of '251-transferrin required the removal of surface-bound ligand. Control experiments revealed that exposure of cells to 200 m M acetic acid, 500 m~ NaC1, pH 2.5, abolished insulin responsiveness, as measured by stimulation of hexose transport. Therefore, an alternative method was developed based on the fact that apotransferrin binds with much lower affinity than diferric transferrin to transferrin receptor at pH 7.4 (4,5). Cells were incubated for 15 min at 4 "C in 25 mM acetic acid, 150 mM NaCl, 2 mM CaCI2, pH 5.0, with 50 p~ deferoxamine mesylate (Ciba Pharmaceuticals, Inc.) in order to release and chelate the iron from the surface-bound diferric lZ5I-transferrin (4). Cells were then washed with two 1-ml aliquots of KRP and incubated for 30 min at 4 "C in 1 ml of KRP with 50 p~ deferoxamine and 125 nM unlabeled transferrin. A control experiment showed that 30 min is sufficient time for complete dissociation of surface-bound Iz5I-apotransferrin at 4 "C. Moreover, basal cells that had been treated in this manner were fully as insulin-responsive as cells maintained in KRP at 37 'C, as measured by stimulation of hexose transport. Subsequently, cells that had been stripped of surface-bound transferrin were washed once with K R P and externalization of intracellular Iz5I-transferrin was initiated by the addition of KRP containing 1 mg/ml BSA, 1 p~ transferrin with or without 1 p~ insulin at 37 "C and immediate transfer of each 6-well dish to a 37 "C bath. Externalization was stopped at the desired time by transfer of the cells to ice, followed at once by one wash with KRP at 4 "C. Cells were solubilized with 1 N NaOH in order to measure the remaining lZ5I-transferrin. Parallel measurements were made with cells exposed to 1 p~ unlabeled transferrin during the initial 2-h incubation at 37 "C, and all values have been corrected for nonspecific binding.
Rate of Iron Uptake-Cell monolayers were washed with three 1ml aliquots of KRP at 37 "C, and then 0.94 ml of KRP containing 1 mg/ml BSA at 37 "C was added. Vehicle (0.01 ml of 10 mM HCl) or insulin (final concentration, 100 nM) was added for 5 min prior to addition of 50 pl of ["Feltransferrin (final concentration, 17 nM). The pH of KRP after addition of vehicle was 7.35. Iron uptake was stopped at the desired time by transfer of the 6-well dishes to an ice bath, followed a t once by three 1-ml washes with KRP at 4 "C. The cells were solubilized with 1% Triton X-100, and the radioactivity was determined by scintillation spectrometry. for the effect of insulin on hexose transport under the same conditions is also presented in Fig. 1. This process proceeded more slowly; the 7-fold increase in the uptake of 2-deoxyglucose in response to insulin occurred with a half-time of 105 s. The concentration dependence of the effect of insulin on surface transferrin binding was also examined (Fig. 2). The increase was half-maximal at 10 nM insulin (average of two experiments, with individual values of 9 and 11 nM). Similarly, hexose uptake was stimulated half-maximally by 8 nM insulin (identical values obtained in two experiments). Thus, insulin regulates surface transferrin binding and hexose uptake with the same concentration dependence. The concentration of insulin that elicited half-maximal stimulation of these two responses was very similar to that required to elicit stimulation of 2-deoxyglucose uptake in 3T3-Ll adipocytes in two previously published studies (7 nM (12) and 6 nM (16)).

Time Course and Concentration
Effect of Insulin on Surface and Intracellular Transferrin Receptors-The insulin-induced increase in binding of 3 nM 1251-tran~ferrin to surface receptors could be elicited by an increase in either the affinity or the number of surface transferrin receptors. Scatchard analysis of Iz5I-transferrin binding at 4 "C showed that insulin increased the number of surface transferrin receptors from 17 to 28 fmol/well (Fig. 3). The values for the dissociation constant in basal and insulintreated cells were 2.3 and 1.5 nM, respectively. In a second separate experiment, the number of surface receptors for basal and insulin-treated cells were 9 and 22 fmol/well, respectively, and the dissociation constants were 1.1 and 1.0 nM. Thus, insulin caused a %fold increase in the number of surface transferrin receptors without a significant effect on receptor affinity.
The rapidity of the effect of insulin on the number of surface transferrin receptors on 3T3-Ll adipocytes indicates that protein synthesis is not involved. An alternative mechanism, expected on the basis of results obtained in rat adipocytes (8), is an insulin-induced redistribution of transferrin receptors from an intracellular pool to the plasma membrane. Incubation of cells with 3 nM '251-transferrin for 2 h at 37 "C resulted in a steady-state level of 68 fmol bound per well, of which 11 fmol were acid-sensitive and thus represented binding to surface transferrin receptors (Table I). When cells were exposed to 100 nM insulin for 5 or 8 min, there was a 1.7-fold increase in acid-sensitive binding (19 fmol/well) and a corresponding decrease in acid-insensitive binding (representing intracellular receptors) from 56 to 49 fmol/well. There was no change in the total Iz5I-transferrin bound (acid-sensitive + acid-insensitive) in response to insulin. Therefore, insulin

Effect of insulin on transferrin binding to surface and intracellular receptors
Cells were incubated for 2 h at 37 "C with 3 nM '=I-transferrin and then exposed to vehicle or insulin (final concentration, 100 nM) for the indicated time period. Transferrin bound to surface and intracellular receptors was then determined by acid wash (200 mM acetic acid, 500 mM NaC1, pH 2.5) at 4 "C as described under "Experimental Procedures." Transferrin that is removed by acid wash is bound to surface receptors, whereas the remaining transferrin is bound to  Table I).
Effect of Insulin on Uptake of Iron-The functional consequence of an insulin-induced increase in the number of surface transferrin receptors was examined by comparison of the rate of 5sFe3+ accumulation in control and insulin-treated cells (Fig. 5). The rate of "Be3+ uptake was linear for at least 90 min and was increased from 12 to 21 fmol/min/well in the presence of 100 nM insulin. In a parallel experiment performed on the same day, insulin elevated surface transferrin binding from 7.5 to 14 fmol of '251-transferrin bound per well (protocol is in legend of Table I). Thus, a 1.8-fold increase in the rate of "Fe3+ accumulation was paralleled by a 1.9-fold increase in 1251-transferrin bound to surface receptors.

DISCUSSION
Our results on the effect of insulin on surface transferrin receptors of 3T3-Ll adipocytes are similar to results obtained in rat adipocytes (8). There are 8 and 30 fmol of surface transferrin receptors/2 X IO6 rat adipocyte cells in the basal and insulin-treated states, respectively, with a dissociation constant of 1-2 nM; similar values were obtained in the present study (Fig. 3). Subcellular fractionation and analysis of the receptor content in the plasma membrane and low density microsome fractions from basal and insulin-treated rat adipocytes showed that the increased number of surface transferrin receptors is due to translocation of receptors from an intracellular location to the plasma membrane (8). Insulin also elicits translocation of transferrin receptors in 3T3-Ll adipocytes since the insulin-induced increase in the surface transferrin-receptor complex was accompanied by a stoichio-FeTf + R. + FeTf. R. 9 Tf . Ri + s Fe Tf SCHEME 1. FeTf, ferrotransferrin; R., cell-surface receptor; FeTf. R, ferrotransferrin-surface receptor complex; Tf. R,, intracellular apotransferrin-receptor complex. metric decrease in the intracellular transferrin-receptor complex (Table I).
Determination of the site at which insulin acts to alter the steady-state distribution of receptors between the plasma membrane and intracellular membranes would contribute to an understanding of the mechanism of action of insulin. The following analysis of the results from Table I and Fig. 4 demonstrates that insulin increased the rate constant for externalization of transferrin receptors and had no significant effect upon the rate constant for receptor internalization. A kinetic model for the recycling of transferrin and its receptor in the human hepatoma cell line, HepG2, has been described (14) (Scheme 1). According to this scheme, ferrotransferrin binds to its receptor on the cell surface. The ferrotransferrinsurface receptor complex undergoes endocytosis by a firstorder process with the rate constant k'e,, and then releases its iron within an acidic compartment. The resulting apotransferrin-receptor complex returns to the surface (rate constant, k'ex), and apotransferrin dissociates into the extracellular medium at pH 7.4. It is assumed in Scheme 1 that the rate constants for release of iron from the internalized ferrotransferrin-receptor complex and for the release of apotransferrin from the surface receptor are relatively large so that these steps are not rate-limiting. The latter assumption has been validated for hepatoma cells (14) where dissociation of apotransferrin from the cell surface at pH 7.3 and at 37 "C occurs with a half-time of 16 s. In terms of Scheme 1, the observed first-order process for the release of intracellular transferrin (Fig. 4) corresponds to the externalization of the apotransferrin-receptor complex (k'ex). Thus, insulin increased the value of the rate constant for externalization by a factor of 1.7.
This kinetic model also allows analysis of the effect of insulin on the internalization rate of the ferrotransferrinreceptor complex. At steady state, the rate of internalization of the ferrotransferrin-receptor complex (FeTf. R,) equals the rate of externalization of the apotransferrin-receptor complex (Tf.Ri).  (Table I) and the values for k'ex ( Fig. 4) can he substituted into Equation 1 to solve for k'en. The values for k'en in basal and insulin-treated cells were 0.55 and 0.50 min", respectively. It is unclear whether this 10% decrease in k'en in response to insulin is significant; but, if so, it is a small effect compared to the 70% increase in k'ex. The 1.7-fold effect of insulin on k'ex accounts almost entirely for the observed 1.7-fold increase in the level of the ferrotransferrin-receptor complex at the surface (Table I).
Although the rate constants for internalization and externalization of the unoccupied transferrin receptor (ken and kex in the Introduction) have not been measured in this study, it seems likely that insulin regulates unoccupied transferrin receptors in the same manner. This conclusion is supported by the similar magnitudes of the effect of insulin on occupied (Table I) and unoccupied (Fig. 3) surface transferrin receptors. Insulin elicited a 2-fold increase in unoccupied surface transferrin receptors, the amounts of which were subsequently measured by Scatchard analysis of lZ5I-transferrin binding at 4 "C (Fig. 3).
The observation that the transferrin receptor-mediated uptake of 59Fe3' is stimulated l.&fold by insulin (Fig. 5) is consistent with the above analysis. At steady state, the rate of iron uptake is equal to k.,,[FeTf.R.]; as shown in Table I, insulin increased the concentration of [FeTf. RBI by a factor of 1.7.
Since insulin affects the rate of externalization of intracellular transferrin receptors, investigation of the biochemical basis for insulin regulation should be directed toward an understanding of the distribution and movement of the intracellular receptors. The actual process that determines the rate constant for externalization of the transferrin receptor and that is regulated by insulin is unknown. Internalized transferrin receptors are initially found in peripheral endosomes and then in juxtanuclear endosomes (15). The receptors probably return to the plasma membrane from both these sites by vesicular trafficking that has yet to be defined. Given the complexity of the process, it is remarkable that it follows first-order kinetics.
A previous study from this laboratory has shown that insulin elicits a 2-fold increase in the rates of both fluid-phase endocytosis and exocytosis in 3T3-Ll adipocytes (16); and thus, insulin stimulates overall membrane recycling. A detailed analysis of the kinetics of fluid-phase exocytosis indicates that insulin increases the size of the endosomal compartment by a factor of 1.6 and the rate constant for fluidphase exocytosis from this compartment by a factor of only 1.25. Thus, the effect of insulin on the rate constants for externalization of the transferrin receptor and of endosomal fluid are different, and the relationship between the two effects remains to be elucidated.
The effect of insulin on the transferrin receptors of 3T3-L1 adipocytes is similar to the effect of epidermal growth factor on the transferrin receptors of human fibroblasts (17). Epidermal growth factor elicits a 2-fold increase in the number of surface transferrin receptors on fibroblasts, which is associated with a 2-fold increase in the rate constant for externalization of the intracellular transferrin-receptor complex.
Two other adipocyte membrane proteins, the receptor for insulin-like growth factor I1 (IGF-11) (18,19) and the glucose transporter (10,11), undergo redistribution from intracellular sites to the plasma membrane in response to insulin. The IGF-I1 receptor, like the transferrin receptor, continuously recycles (20,21). Since insulin has no effect on the rate constant for internalization of photoaffinity-labeled IGF-I1 receptor (20), it is probable that the basis for the redistribution of this receptor is also an increase in kex. A proposed alternative explanation, that redistribution is due to an inhibition of the rate of internalization a.s the result of a change in the phosphorylation state of the plasma membrane IGF-I1 receptor (221, seems unlikely. It is not known whether the glucose transporter continuously recycles. The half-times for insulin activation of hexose transport at 37 "C in 3T3-Ll adipocytes (105 s, Fig. 1) and rat adipocytes (120 s, Ref. 23) are considerably longer than the half-times for redistribution of the transferrin receptor (40 s, Fig. 1) and the IGF-I1 receptor (45 s, Ref. 23). Thus, the translocation of the transporter may occur by a distinct pathway. Recent work from our laboratory has described the isolation of vesicles containing the insulin-responsive intracellular glucose transporters (24). It will be of interest to determine if the insulin-responsive intracellular transferrin receptors are also located in these vesicles.