Primary Culture of Isolated Adipocytes

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY by The American Society of BioIogicd Chemists, Inc. Primary Culture of Isolated Adipocytes A NEW MODEL TO STUDY INSULIN RECEPTOR REGULATION AND Vol. 259, No. 10, lame of May 25, pp Printed in U.S.A. INSULIN ACTION* (Received for publication, November 14,1983) Stephen Marshall$, W. Timothy Garvey, and Miriam Geller From the University of California at San Diego, Department of Medicine, M-OBE, Division of Endocrinology and Metabolism, La Jplla, California The mechanism by which insulin regulates cell surface insulin receptors was examined in primary cultured rat adipocytes. When cells were incubated in insulin-free medium, specific aai-insulin binding progressively increased over 3-4 days followed by a plateau of binding. Insulin prevented up-regulation at low doses (1 ng/ml), while higher doses (5-25 ng/ml) resulted in a net loss of surface receptors. A lag period of 4-6 h preceded insulin-induced changes in receptor number, and such a lag was seen prior to the inhibitory effect of insulin on the insertion of nascent receptors into the plasma membrane. Regulation of surface receptors continued after the removal of insulin, consistant with the idea that insulin generates a signal which can sustain receptor regulation in the absence of ligand. Thus, a 1-h insulin pulse (100 ng/ml) was sufficient to block up-regulation, whereas longer exposure times (4-12 h) produced a net loss of surface receptors. When cells were exposed to insulin for a fixed time (5 h), subsequent receptor loss was insulin dose-dependent. Thus, the net number of cell surface insulin receptors is determined by both insulin concentration and the duration of insulin exposure. Time course studies after a 12-h insulin pulse revealed a progressive loss of surface receptors for up to 36 h. At later times receptor number returned toward control values, thus demonstrating that triggering of receptor regulation is reversible. To determine whether insulin-induced down-regulation was mediated by receptor loss, or by receptor translocation to an intracellular site, we measured receptor distribution (cell surface and intracellular pools) in control and 72-h insulin-treated adipocytes. These data revealed that down-regulation was mediated by a net loss of receptors rather than by receptor redistribution. An early biological response of adipocytes to insulin is rapid endocytosis of insulin-receptor complexes, and we found that insulin plays an important role in regulating this endocytotic uptake rate. Thus, compared to freshly isolated cells, adipocytes cultured for 72 h in insulin-free medium had an enhanced ability to internalize both insulin and insulin receptors, whereas cells chronically treated with insulin endocytosed insulinreceptor complexes at a much slower rate. These data are interpreted as an insulin-mediated change in one * This work was supported by a New Investigator Research Award from the National Institutes of Health (AM-33647) and by the Gail Patrick Velde Feasibility Grant from the American Diabetes Association. 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. $ To whom correspondence should be addressed. of the cellular responses of adipocytes to the biological action of insulin. Over the past several years we have been actively investigating insulin-mediated regulation of cell surface insulin receptors in freshly isolated adipocytes (1-5). Such cells provided an excellent model since they represent a classical target tissue of insulin and express many of the biological actions of insulin. Results thus far have revealed that acute treatment of adipocytes with a high concentration of insulin leads to a rapid time-, temperature-, and energy-dependent loss of surface receptors with 50 and 75% decreases by 2 and 4 h, respectively (1, 2). This phenomenon, known as insulin-induced receptor down-regulation, is mediated by endocytotic uptake of insulin-receptor complexes, since conditions known to inhibit absorptive endocytosis, such as low temperature or depletion of metabolic energy, block both ligand uptake and down-regulation (2). Further support for this hypothesis is derived from the finding that in the presence of insulin, the extent of receptor loss and the amount of internalized insulin were highly correlated as functions of time and temperature (2), and that receptors lost from the cell surface could be recovered within the cell interior (6). Although these studies qualitatively agreed with the well documented human and animal studies (7-11) showing a negative correlation between hyperinsulinemia in vivo and the number of insulin receptors on several tissues, the actual time required for insulin to negatively modulate receptor number differed markedly under in vivo and in vitro conditions. For example, adipocytes incubated in vitro with high insulin concentrations had 50% fewer cell surface receptors after 2 h (1,2), whereas, in vivo, insulin-induced receptor loss occurred over a period of several days (12). This discrepancy was later resolved by our finding that tris(hydroxymethy1)- aminomethane, a constituent of the in vitro adipocyte incubation buffer, effectively prevented endocytosed receptors from recycling back to the plasma membrane (4). In adipocytes the number of surface receptors during short term insulin exposure represents an equilibrium between the rate of receptor uptake and the rate at which endocytosed receptors are recycled and reinserted back into plasma membrane. Disturbance of this equilibrium by blocking recycling allowed the former process to predominate and resulted in the rapid depletion of cell surface receptors. Despite the fact that our earlier adipocyte studies characterized and examined the kinetics of acute insulin-induced down-regulation (1, 2) and later led to the important discovery that insulin receptors are normally recycled after endocytotic uptake of insulin-receptor complexes (3), regulation of cell surface adipocyte receptors 6376

2 in vitro under physiological conditions remained to be demonstrated. With the realization that acute adipocyte studies were not suitable for investigating insulin-induced regulation of cell surface receptors under physiological conditions where receptor recycling was operative, we developed a method for maintaining isolated rat adipocytes in primary culture (13). Using this method, we now report that cell surface receptors on adipocytes can indeed be regulated by insulin, but, in contrast to our earlier experiments, this process occurs over several days rather than hours. Besides kinetic differences, the actual mechanism(s) of insulin-induced receptor regulation appears to differ markedly from that found in our acute adipocyte studies where down-regulation was studied in the absence of receptor recycling (1, 2). MATERIALS AND METHODS Porcine monocomponent insulin was generously supplied by Dr. Ronald Chance of the Lilly, Na'2'I was purchased from New England Nuclear, collagenase was from Worthington, Dulbecco's modified Eagle's medium was from GIBCO, and tris(hydroxymethy1)aminomethane, chloroquine, trypsin, and trypsin inhibitor were from Sigma. Iodination of Insulin-'261-Insulin was prepared to a specific activity of pci/mg as described by Freychet et al. (14). Preparation of Isolated Adipocytes-Male Sprague-Dawley rats weighing g were killed by cervical dislocation, and the epididymal fat pads removed under sterile conditions. Isolated adipocytes were obtained using a method modified from Rodbell (15) by shaking finely minced tissue (1-3 g) in 4-ounce sterile polypropylene containers at 37 "C for 1 h in Dulbecco's modified Eagle's medium (4 ml) containing 25 mm Hepes,' collagenase (2 mg/ml), and albumin (40 mg/ml). Cells were then filtered through nylon mesh (1000 pm), centrifuged at 100 rpm for 1 min, and washed twice in Dulbecco's MEM containing 10 mm Hepes, 2% fetal calf serum, 1% bovine serum albumin, penicillin (20 units/ml), and streptomycin (20 mg/ml). Adipocyte number was determined according to a modification of method 111 of Hirsch and Gallian (16), in which the cells are fixed in 2% osmium tetroxide in 0.05 M collidine buffer (made isotonic with saline) for 24 h at 37 "C and then taken up in a known volume of M NaCl for counting. Counting was performed with a model ZB Coulter counter with a 400 p~ aperture. Primary Culture-Isolated adipocytes (5 X lo' cells/ml) were sterilely incubated at 37 "C in Dulbecco's MEM, ph 7.4, in airtight 50- ml sterile polypropylene tubes (5-20 ml/tube) with cells floating on top of the medium in a thin cell layer. At the end of the incubation period adipocytes were transferred to polystyrene tubes (17 X 100 mm) and centrifuged, and the medium was removed and replaced with MEM buffer (ph 7.6), containing 10 mm Hepes and 1% bovine serum albumin. Adipocytes were again centrifuged, resuspended in medium (2-3 X 10' cells/ml), and then aliquoted (1 ml) into polystyrene tubes (17 X 100 mm) for measurement of specific '2'I-insulin binding. In experiments where adipocytes were exposed to insulin (down-regulation experiments) control and insulin-treated cells were washed three times with insulin-free medium at ph 7.0 to remove extracellular ligand, and then further incubated at this ph for an additional hour at 37 "C to remove any remaining receptor-bound insulin. Previous data (1, 2) have demonstrated that this method effectively removes all receptor-bound insulin and any internalized insulin (including subsequently generated degradation products). Cells were then resuspended in MEM buffer, ph 7.6, and specific '251-insulin binding was measured. Measurement of Cell Surface Receptors-Isolated adipocytes (2-3 X 10' cells) were incubated in a total volume of 1 ml of MEM buffer (ph 7.8) with 0.2 ng of '=I-insulin in the absence or presence of 50 pg of unlabeled insulin. Incubations were performed in polystyrene tubes (17 X 100 mm) in a shaking water bath at 16 'C for 2 h unless otherwise indicated. It should be noted that measurements of specific '=I-insulin binding to intact adipocytes were performed at 16 'C because insulin internalization is inhibited at this low temperature (17-20); thus, cell-associated '2SI-insulin reflects only binding to cell surface receptors. The binding reaction was terminated and free 12'1- ' The abbreviations used are: Hepes, 4-(2-hydroxyethyl)-I-piperazineethanesulfonic acid; MEM, modified Eagle's medium. Receptor Regulation in Cultured Adipocytes 6377 insulin was separated from cell-bound radioactivity by removing aliquots (300 pl) from the cell suspension and rapidly centrifuging the cells in plastic microtubes to which 100 pl of silicone oil had been added. Silicone oil has a specific gravity intermediate between buffer and cells; therefore, after centrifugation, three layers result: cells on top, oil in the middle, and medium on the bottom. The cells were then removed and the radioactivity was determined. Specific '%Iinsulin binding was determined in triplicate from each incubation tube, and each experiment was a representative example of at least three experiments unless otherwise indicated. Nonspecific Binding-Nonspecific binding is defined as the amount of '9-insulin remaining in the cell layer in the presence of a large excess (50 pg/ml) of unlabeled insulin. When cells equilibrate with a tracer concentration of '261-insulin( ng/ml), only 3-7% of the bound insulin represents nonspecific binding. For all samples, total and nonspecific binding were determined, and the total binding was corrected to reflect specific binding. Measurement of Insulin-mediated Receptor Internalization-During acute treatment of adipocytes with insulin (less than 4 h) a net loss of cell surface receptors is not seen because internalized receptors are rapidly recycled back to the plasma membrane (3). However, in the presence of a maximally effective concentration of tris(hydroxymethy1)aminomethane recycling is almost completely inhibited (4). Thus, a rapid and linear monoexponential loss of cell surface receptors is observed over 2 h (2). To estimate the rate at which insulin-receptor complexes are endocytosed after long term insulin treatment in primary culture, we measured the loss of cell surface insulin receptors after an acute 2-h treatment of adipocytes with 100 ng/ml of insulin and 35 mm Tris. Under conditions where adipocytes were chronically treated with insulin for 72 h, extracellular and receptor-bound ligand was first removed before the acute downregulatory phase of the experiment was performed. Trypsinization of Adipocytes-Adipocytes were exposed to 150 pg/ ml of trypsin for 15 min at 24 "C, before 300 pg/ml of soybean trypsin inhibitor was added to stop proteolysis. Cells were then washed twice in MEM buffer (ph 7.6) and incubated for 2 h at 16 "C in 1 ml of MEM buffer containing 0.2 ng of '=I-insulin in the presence or absence of 10 pg/ml of unlabeled insulin. Under these conditions trypsinization effectively inactivates greater than 95% of the surface receptors as measured by loss of specific '2'I-insulin binding (13). Cell Solubilization and Receptor Precipitation and Reconstitutwn- Adipocytes (2-80 X lo' cells) were solubilized (0.5-ml final volume) by the addition of 50 pl of solubilization fluid to achieve a final concentration of 0.4% Triton X-100 and 0.5 mg/ml of bacitracin. After vigorously vortexing cells for 1 min, solubilization of cells and receptors was allowed to continue for 30 min at 4 "C since these conditions were found to be optimal for extracting greater than 90% of the total and intracellular receptors with negligible receptor degradation. To precipitate receptors, we transferred 500 p1 of the solubilized cell mixture to 1.5-ml microfuge tubes and added 50 pl of IgG (2%) and 600 pl of polyethelene glycol (30%) as described by Cuatrecasas (21). The mixture was vortexed (1 rnin), and then centrifuged for 3 min at 11,OOO X g in a Beckman Microfuge 11 so that three layers formed cellular lipids on top; buffer in the middle containing detergent, proteases, and small M, proteins; and larger M, proteins (including insulin receptors) pelleted at the bottom. The supernatent containing the lipids and buffer was removed by aspiration, and the protein pellet was washed twice with phosphate-buffered saline and reconstituted in 400 pl of binding buffer (ph 7.8) containing Dulbecco's MEM, 35 mm Hepes, 2 mm EDTA, and 0.5% bovine serum albumin. Measurement of Total and Intracellular Receptors-The cellular distribution of adipocyte insulin receptors was assessed by measuring specific '%insulin binding to receptors solubilized from intact (total receptor pool) or trypsinized adipocytes (intracellular receptor pool). Total receptor content was measured in triplicate by adding 20 pl of '=I-insulin (0.2 ng/ml final concentration) with 80 pl of polyethylene glycol precipitated-reconstituted receptors and incubating in the absence or presence of 10 pg/ml of unlabeled insulin for 16 h at 4 "C. The binding reaction was terminated and free '"I-insulin was separated from receptor-bound ligand by the addition of 150 pl of polyethylene glycol (30%) and rapid microcentrifugation at 11,OOO X g for 2 min. The pellet was washed twice with phosphate-buffered saline (300 pl), and the tip of the microfuge tube was cut and counted. To measure intracellular receptors, we first trypsinized adipocytes to destroy almost all binding activity on the cell surface, and then measured '%I-insulin binding to solubilized reconstituted receptors

3 6378 Receptor Regulation in Cultured Adipocytes as described above. Nonspecific binding, defined as the amount of lz6i-insulin remaining in the pellet in the presence of excess unlabeled insulin (10 pg/mi), under all conditions was 5% of total radioactivity and probably represented simple trapping of radioactivity within the protein pellet. RESULTS As shown in Fig. L4, when freshly isolated rat adipocytes were maintained in primary culture for up to 2 weeks there was a progressive increase in the insulin binding capacity of these cells over the first 3-4 days, followed by a plateau of binding at times thereafter. Cell number remained constant throughout the experiment, and Scatchard analysis of insulin binding data at 3, 7, and 15 days revealed that the greater binding capacity was due to an increase in the number of receptors, with little or no change in receptor affinity (data not shown). It should also been noted from Fig. 1A that untreated adipocytes were maintained for up to 2 weeks in primary culture, which greatly exceeds our previous primary culture time of 2 days (13). Although unlikely, the possibility existed that a portion of cell surface receptors were proteolytically inactivated during the initial collagenase digestion so that subsequent recovery of binding activity was observed during the first few days in culture. Therefore, to test this idea we prepared isolated adipocytes as described under "Materials and Methods," and then further exposed cells to collagenase at 37 "C for various times up to 2 h. If receptor proteases were present in the collagenase preparation, then it would be expected that the insulin-binding capacity would progressively decrease over time during the second collagenase exposure. Since this was not observed (data not shown) we conclude that receptor proteolysis during cell preparation was negligible. To test the hypothesis that up-regulation of receptors in vitro is a consequence of culturing cells in insulin-free medium, we added various amounts of insulin to the medium (1-100 ng/ml) and cultured cells for up to 4 days. These data (Fig. 1B) revealed that as little as 1 ngjml of insulin (added every other day) markedly attenuated the progressive increase in receptor number, whereas treatment with 5, 25, and 100 ng/ml of insulin resulted in a dose-dependent loss of surface receptors. For the sake of clarity, it should be mentioned that binding to freshly isolated adipocytes (time zero) is used as a reference value, so that changes in binding above or below this value are considered receptor up-regulation or downregulation, respectively. It is also important to note that the decreased binding was not caused by carryover of free or receptor-bound ligand from the preincubation phase of the experiment, since acute insulin exposure for up to 4 h did not decrease binding compared to control cells similarly treated but not exposed to insulin (Fig. 4 and Ref. 1). Fig. 2A depicts insulin binding competition curves over a wide range of insulin concentrations using cells maintained for 48 h in either insulin-free medium (control) or medium containing 100 ng/ml of insulin and shows that the ability of insulin-treated cells to subsequently bind insulin is decreased $;:e Q cn C f e 5 - Insulin Concentration (na/rnl) Insulin Conc. (nslml) - 14 c : O' Time (days) FIG. 1. Effect of insulin preincubation on the ability of adipocytes to specifically bind 'P61-insulin. A, adipocytes (5 X lo' cells/ml) were incubated at 37 "C in insulin-free medium for up to 2 weeks. At the indicated times cells were washed, and specific '"Iinsulin binding was measured after a 2-h incubation at 16 "C. B, cells were preincubated for up to 4 days at 37 'C with 0, 1, 5, 25, or 100 ng/ml of insulin and then thoroughly washed to remove all extracellular and cell-hound insulin as described under "Materials and Methods." Binding to cell surface insulin receptors was measured after a 2-h incubation at 16 "C. The concentration of insulin during the preincubation phase of the experiment is noted to the right of each curve. 48 h Insulin-Treatment ' Insulin Bound (pg) FIG. 2. Ability of 48-h insulin-treated adipocytes to specifically bind '461-insulin at various insulin concentrations. A, adipocytes were preincubated either in the absence (control) or presence of 100 ng/ml of insulin for 48 h, after which time cells were washed to remove extracellular and receptor-bound ligand. Cells were then incubated for 2 h at 16 "C with 0.2 ng/ml of '=I-insulin plus unlabeled insulin at the indicated concentrations and specific binding was determined. B, Scatchard plots of these binding data are shown for control and 48-h insulin-treated cells with the ratio of hound to free insulin plotted on the ordinate, and bound insulin on the abscissa.

4 Receptor Regulation in Cultured Adipocytes 6379 at all points on the curve. Scatchard plots of these insulinbinding data are shown in Fig. 2B. Although precise interpretation of these curvilinear plots is difficult, the generally comparable shapes of the curves indicate that the major difference in binding capacity between up-regulated and down-regulated cells is due to differences in the number of receptors, rather than a change in receptor affinity. Shown in Fig. 3 are the rates at which primary cultured adipocytes degrade insulin at various ambient insulin concentrations. To monitor insulin catabolism, we incubated cells in medium containing 0.1 ng/ml of '251-insulin plus the indicated concentrations of unlabeled insulin (1-100 ng/ml). The per cent total extracellular radioactivity soluble in trichloroacetic acid was used as an index of degradation. As depicted, degradation of extracellular insulin was linear over time and related to the concentration of insulin in the medium. With greater insulin concentrations, a smaller proportion of the total insulin was degraded, suggesting that the degradative capacity of adipocytes is saturable. Since insulin is degraded relatively slowly under our culture conditions, the readdition of insulin every other day (Fig. 1B) probably resulted in a somewhat higher insulin concentration than at time zero. To explore the mechanisms by which insulin regulates cell surface receptors, we focused our attention on three basic questions. First, does insulin modulate insulin receptor number immediately, or is there a lag time before the onset of insulin-induced receptor regulation? Secondly, is the continued presence of insulin necessary to sustain changes in receptor concentration, or can insulin trigger biological events that can then lead to subsequent receptor regulation in the absence of ligand? And lastly, are insulin-induced changes in the concentration of plasma membrane receptors mediated by a redistribution of receptors between the cell surface and the cell interior? The question of the lag time preceding receptor regulation is addressed in Fig. 4. When '9-insulin binding to control and insulin-treated cells was measured at various times up to 24 h, no difference in binding was seen for the first 4 h, whereas at times thereafter differences in binding capacity between the two groups progressively increased. Thus, after the addition of insulin, there was a lag of about 4-6 h before the onset of receptor regulation. It should be noted that the net change in binding between the groups reflects a balance between two processes: up-regulation in the absence of ligand, and down-regulation in the presence of insulin. I ' Insulin Conc. Incubation Time (hrs) FIG. 3. Degradation of insulin by primary cultured adipocytes. Adipocytes (5 X lo' cells ml) were incubated with 0.1 ng/ml of IZ6I-insulin in the absence or presence of unlabeled insulin (0-100 ng/ml). After 16 and 24 b insulin degradation was determined by measuring trichloroacetic acid solubility of extracellular radioactivity. Each point represents the mean of three determinations. 25 nglml t Insulin 1 " Time (hrs) FIG. 4. Lag time preceding insulin-mediated changes in insulin binding. Cellswere preincubated at 37 "C in the absence (controls) or presence of 25 ng/ml of insulin. At the indicated time cells were washed to remove extracellular and receptor-bound ligand, and specific '=I-insulin binding was measured after a 2-h, 16 "C incubation period. B/F, ratio of bound to free insulin. 4.a x 3.0 s Y p 2.5 i3.e v) c T l , I Insulin Conc O'O 6 12 la 34 Recovery Time After Tryp.&izati&' (hrs) FIG. 5. Effect of insulin on the recovery of adipocyte '''Iinsulin binding following tryptic inactivation of cell surface receptors. Isolated adipocytes were trypsinized under sterile conditions, exposed to trypsin inhibitor, and after several washings incubated at 37 "C in the absence or presence of the indicated insulin concentrations. At the various times cells were washed and specific '=I-binding to cell surface receptors was measured after a 2-h, 16 "C incubation period. To confirm the existence of a lag preceding the effect of insulin on receptor metabolism, we used our method for following insulin receptor biosynthesis and membrane insertion (13) which entails inactivating cell surface insulin receptors on adipocytes with trypsin, and then maintaining these cells in primary culture at 37 "C for various times. When the insertion of nascent receptors onto the surface of adipocytes was monitored in insulin-free medium (by measuring the recovery of specific '261-insulin binding activity), incorporation proceeded at a linear rate for up to 24 h (Fig. 5 and Ref. 13). In the presence of 1 ng/ml of insulin insertion of receptors at later times was slowed, and at higher concentrations it appeared to actually stop (Fig. 5). However, during early times (up to 6 h) the rate of receptor appearance was identical for all groups of cells. In other words, there was again an initial lag period of about 6 h before insulin has a noticeable effect

5 6380 Receptor Regulation in Cultured Adipocytes on the net number of cell surface receptors. Since we have previously shown that recovery of binding activity is mediated by de mu0 receptor synthesis and insertion of receptors (13), these current findings could be interpreted to mean that insulin affects receptor biosynthesis. However, the alternative possibility that insulin influences receptor recovery by accelerating the turnover rate of nascent surface receptors cannot be excluded. Regardless of the actual mechanism it is apparent that a lag phase precedes insulin-induced changesincell surface receptor number. Presented in Fig. 6 are data supporting the hypothesis that insulin triggers down-regulation so that the continued presence of ligand is not required to sustain receptor loss. A depicts an experiment where adipocytes were incubated in insulin-free medium (control) or in the presence of 25 ngjml of insulin for either the first 12 h of the experiment or continuously. At the indicated times cells were thoroughly washed and insulin binding was determined after a 2-h, 16 "C incubation period. For both 12-h and continuously insulintreated cells the ability of adipocytes to subsequently bind '251-insulin progressively decreased in parallel during the first 36 h. At times thereafter, binding in the 12-h insulin-treated cells gradually returned toward control values, whereas the binding capacity of continuously treated cells remained low. Thus, it appears that insulin triggers down-regulation and that the down-regulatory effect of insulin persists for about 6' n 5 ' 0 z 4- X 3' 2' F P 1- g01.. a Control Insulin Treatment a. I 1.E cn f 12-8 I v) Insulin PulseTime - (h) -I 24 h following ligand removal. When cells were pulsed for various times with 25 ng/ml of insulin (Fig. 6B), and then incubated in insulin-free mediumfor the duration of the experiment, loss of binding capacity (measured at 24 h) was directly related to the insulin exposure time. It should be noted that compared to time zero, a net loss of receptors occurred in cells pulsed with insulin for 4-12 h, whereas a 1- h insulin pulse prevented subsequent up-regulation. Triggering of insulin-induced down-regulationis a very sensitive process as evidenced by the data presented in Fig. 7. Only a 5-h insulin treatment time with 1 ng/ml of insulin was required to completely prevent the receptor up-regulation normally seen at 27 h. At higher insulin concentration (5-100 ng/ml) a net loss in the number of cell surface receptors was observed. Thus, it is apparent that the length of insulin exposure as well as the concentration of insulin modulates the final concentration of receptors on the surface of adipocytes. To examine whether insulin-induced loss of cell surface receptors was mediated by a redistribution of receptors from the plasma membrane to the cell interior, or an actual loss in the total number of cellularreceptors, we performed the experiment shown in Fig. 8. First, control and insulin-treated adipocytes (100 ng/ml) were incubated at 37 "C for 48 h, and then extensively washed to remove all extracellular and cellassociated insulin. Binding to cell surface insulin receptors was then measured by incubating intact adipocytes for 2 h at 16 "C with '2sI-insulin. As seen in Fig. SA, insulin-treatedcells had 58% fewer cell surface receptors compared to up-regulated control cells. To measure total insulin receptors (cell surface + intracellular), adipocytes were solubilized and the insulin-binding capacity of soluble receptors was determined after a 16-h, 4 'C incubation period. When these results were plotted (Fig. 8B), a 49% decrease in total receptors was found. It should be noted that a direct comparisonbetween the number of cell surface receptors and total receptors cannot be made, since the assay conditions used to measure both pools of receptors were markedly different. However, when 10. A Time 0 B 21 h incubation v) t O Incubation Time (hrs) FIG. 6. Effect of pulsatile insulin treatment on subsequent '"I-insulin binding. A, adipocytes (5 X IO' cells/ml) were cultured in the presence of 25 ng/ml for either the first 12 h of the experiment, or continuously for the duration of the experiment. At the indicated times cells were thoroughly washed to remove extracellular and receptor-bound insulin, and specific '161-insulin binding was measured aftera 2-h, 16 "C incubationperiod.control cells weresimilarly treated but not exposed to insulin. B, cells were exposed to 25 ng/ml of insulin at 37 "C for the time indicated at the right of each curve, thoroughly washed to remove all ligand, and then incubated in insulin-free medium. At the indicated times cells were again washed and specific '=I-insulin bindingwasmeasured after a 2-h, 16 "C incubation period. 12 i I Treatment FIG. 7. Ability of various insulin concentrations to trigger subsequent receptor down-regulation. Cultured adipocytes were treated for 5 h with the indicated concentrations of insulin, washed to remove all extracellular and cell-bound ligand, and then incubated in insulin-free medium for an additional 22 h. After 27 h of incubation, all groups of cells were washed and the binding of '261-insulii to cell surface receptors was determined after 2-h, a 16 "C incubation period. Each bar represents the mean (+S.E.) of duplicate incubations sampled in triplicate.

6 Treated $ 4 12xTzszG- (Trypsinked) O Control Insulin- Control Insulin- Treated FIG. 8. Specific '"I-insulin binding to cell surface, total and intracellular receptors after 48 h of insulin treatment. Adipocytes (5 X 10' cells/ml) were incubated for 48 h at 37 "C in the absence (controls) or presence of 100 ng/ml of insulin, and then washed to remove extracellular and receptor-bound ligand. A, specific binding to cell surface insulin receptors was measured by incubating intact adipocytes (3.5 X lo6 cells/ml) for 2 h at 16 "C with 0.2 ng/ml of 'l-insulin in the absence (total binding) or presence of 10 pg/ml of unlabeled insulin (nonspecific binding). Data represent the mean (&.E.) of three incubations sampled in triplicate. B, the total pool of cellular insulin receptors (cell surface plus intracellular) was determined by measuring the binding of '%I-insulin to soluble receptors extracted and harvested from adipocytes. To measure the intracellular pool of insulin receptors, we treated intact adipocytes with 150 pg/ ml of trypsin to inactivate >95% of cell surface receptors (C), and then solubilized these cells with Triton X-100 and measured 'Tinsulin binding after harvesting receptors (Fig. 40). the extent of insulin-induced down-regulation of surface receptor (58%) and the loss of total receptors (49%) are expressed as a percentage of their respective controls, then the close agreement between these values, and the fact that the intracellular pool did not increase (Fig. 8D), indicates that down-regulation was mediated by an actual loss of cell surface receptors rather than by translocation of receptors from the cell surface to the cell interior. Since receptor distribution was determined at a single point in time, it is entirely possible that a complete time course would reveal that receptor translocation and intracellular accumulation precedes receptor loss at earlier times. The binding of insulin to receptors on the surface of adipocytes is rapidly followed by endocytotic uptake of insulinreceptor complexes, but a net loss of surface receptors is not observed because intracellular receptors are efficiently recycling back to the plasma membrane (3). However, this does not necessarily mean that internalization of receptor complexes is unrelated to receptor regulation. Insulin-induced uptake of insulin-receptor complexes may represent a necessary and early step in the sequence of events culminating in eventual receptor loss. If this were the case, then insulin could conceivably regulate receptor loss by altering the rate at which receptor complexes were internalized after the initial binding event. To test this idea, adipocytes were cultured for various times (0-3 days) in medium containing 0, 5, or 100 ng/ml of insulin. After these cells were washed to remove extracellular and cell-associated ligand, '251-insulin binding was measured after a 1-h, 37 "C incubation in either the absence (controls) or presence of 0.2 mm chloroquine. The rationale behind this protocol is that chloroquine has been shown to effectively Receptor Regulation in Cultured Adipocytes 6381 prevent the intracellular processing and degradation of insulin in adipocytes (17-20). This results in the trapping of intact '*'I-insulin within cells, and thus provides a good estimate of the amount of internalized insulin (2). Shown in the upper panel of Fig. 9 is the amount of cell-associated '=I-insulin after incubation of insulin-pretreated cells (0-3 days) for 1 h at 37 "C in the absence or presence of chloroquine. As shown, the chloroquine effect was greatest in cells preincubated for 3 days in insulin-free medium, and smallest in cells pretreated for 3 days with 100 ng/ml of insulin. To express changes in the amount of trapped intracellular *%I-ligand relative to the amount of insulin bound, the chloroquine data in Fig. 9 were replotted as a per cent increase in cell-associated '2SI-insulin above controls (Fig. 9B). It is clear when expressed in this manner that adipocytes cultured in insulin-free medium exhibited an enhanced ability to endocytose ligand, whereas after prolonged exposure to insulin there was a progressive slowing in the rate of insulin internalization. At 5 ng/ml of insulin, a concentration which maintains surface receptor number fairly constant (Fig. l), little change in the ligand uptake rate was observed. Since only endocytosis of the ligand was monitored in the above experiment, it remained to be demonstrated that there was a concomitant change in the rate at which the receptor itself was internalized. Consequently, after maintaining adipocytes for 3 days in insulin-free medium or medium containing 100ng/ml of insulin, we measured the rate at which occupied receptors were endocytosed by following the rapid loss of cell surface receptors over 2 h under conditions where receptor recycling was inoperative. Following Tris pretreatment to prevent recycling receptor, uptake was initiated by exposing cells to 100 ng/ml of insulin at 37 "C. At the times Incubation Time (Days) FIG. 9. Ability of adipocytes to internalize lssi-insulin after pretreatment with insulin. A, isolated adipocytes were pretreated with 0, 5, or 100 ng/ml of insulin for times up to 3 days. At the indicated times insulin was removed by washing, and each group of cells was incubated for 1 h at 37 "C with 0.2 ng/ml of '?-insuiin in the absence (control) or presence of 0.2 mm chloroquine. Specific cell-associated '%I-insulin is expressed as an absolute percentage of total added '%insulin. B, the relationship between insulin binding and the chloroquine-induced increase in cell-associated radioactivity is expressed by plotting the chloroquine values in A as per cent increase above their corresponding untreated control. 1 I

7 6382 Receptor Regulation in Adipocytes Cultured 80 Day 3 - Control Time (mid FIG. 10. Ability of adipocytes to internalize insulin-receptor complexes as a function of insulin pretreatment. Freshly isolated adipocytes (Day 0-Control) and cells cultured for 3 days in the absence (Day 3 ControZ) or presence of 25 ng/ml of insulin (Day 3-Zmulin-treated) were thoroughly washed, before the rate of acute insulin-induced down-regulation was measured (as described under Materials and Methods ). Briefly, each of the three groups of cells was treated with Tris (35 mm) to inhibit receptor recycling and then exposed to 100 ng/ml at 37 C so that almost all surface receptors were occupied by ligand. At the indicated times all extracellular and surface-bound insulin was removed, and the loss of receptors was followed by monitoring the loss of insulin-binding capacity (measured at 16 C) compared to insulin. cells similarly treated but not exposed to In 1982 (22) and 1983 (13) we described for the first time a method for maintaining isolated adipocytes in primary culture. Viability of such cultured cells was established by multiple criteria. For example, after 2 days in culture, adipocyte cell number and specific insulin-binding capacity remained unchanged, cellular metabolism was sufficient to supply the metabolic energy necessary to mediate absorptive endocytosis, cells remained impermeable to ~-[~H]glucose and trypan blue indicating integrity of the plasma membrane, and lastly, adipocytes were able to synthesize both protein and functional insulin receptors (13). With this culture technique we further demonstrated the utility of this method in examining the kinetics of insulin receptor biosynthesis and membrane inser- tion (13). Since these earlier reports Cushman and co-workers (23) have presented confirmatory data on adipocyte viability after 24 h in primary culture, thus supporting our original findings. In the present study, we now use primary cultured adipocytes to examine the other arm of receptor metabolism, namely, insulin-induced down-regulation of cell surface receptors. Although the basic phenomenon of insulin-induced down- regulation is now well documented and includes suqh cells as IM-9 lymphocytes (24), hepatocytes (25), fibroblasts (26), and 3T3 cells (27, 28), it is important to investigate receptor regulation for each cell type since the cellular mechanisms mediating this process appear to differ. For example, in studies by Krupp and Lane (29) insulin treatment of cultured chick liver cells for 18 h decreased cell surface receptors by 60%, without affecting the number of total cellular receptors. Moreover, neither the rates of receptor synthesis nor degradation were changed as measured by the heavy isotope density-shift technique. Thus, in these cells insulin treatment mediated down-regulation by causing a redistribution of receptors from the cell surface to the cell interior. In 3T3-Ll differentiated preadipocytes, however, Ronnett et al. (27) convincingly demonstrated that insulin-induced receptor loss resulted from an increase in the rate of receptor degradation, and this led to a loss of both cell surface and total insulin receptors. Still another mechanism by which cells appear to down-regulate their surface receptors was revealed by the studies of Kasuga and co-workers (30) and Berhanu and Olefsky (31). Using transformed human IM-9 lymphocytes the former group showed that insulin accelerated the turnover of biosynthetically and externally labeled insulin receptors by increasing the rate of receptor degradation, whereas the latter workers elaborated on this mechanism by demonstrating that photolabeled receptors on the cell-surface of IM-9 lymphocytes are shed into the medium where they were rapidly degraded. So it appears that the number of surface receptors on lymphocytes is decreased by insulin, not through intracellular sequestration or accelerated degradation, but rather by extrusion of receptors from the cell surface. Thus, insulin indicated in Fig. 10 the cells were thoroughly washed to remove all extracellular and cell-associated insulin, and specific 251-insulin binding to cell surface receptors was measured after a 2-h, 16 C incubation period. A semilog plot of these regulates receptor number by at least three mechanisms: data revealed that after 3 days in insulin-free medium (Day receptor degradation, receptor redistribution, and extracellu- 3 Control) adipocytes internalized plasma membrane insulin lar shedding of receptors. receptors (as measured by loss of surface receptors) at a faster In earlier adipocyte studies we reported that acute insulin rate than either freshly isolated cells (Day 0-Control) or cells treatment of adipocytes negatively modulated cell surface treated for 3 days with 25 ng/ml of insulin. It should be receptor number through rapid endocytotic uptake of insulinmentioned that when loss of surface receptors is plotted as a receptor complexes (1, 2); however, we later found that insupercentage of receptors initially on the surface of cells (At lin-induced down-regulation was caused by Tris inhibition of time zero -before insulin treatment) this corrects for dif- insulin receptor recycling (4). The current study was initiated ferences in the number of surface receptors so that endocytotic to assess whether insulin could regulate its own receptors in uptake rates for up-regulated and down-regulated cells can be adipocytes under more physiological conditions where recepdirectly compared. tor recycling was operative. Using our recently developed technique of maintaining isolated adipocytes in primary cul- DISCUSSION ture (13), we clearly demonstrated that insulin can indeed regulate the concentration of cell surface receptors, but more interestingly we found that the kinetics of receptor loss under physiological conditions was in marked contrast to our previous short term adipocyte studies (1-4). For example, rather than finding that rapid down-regulation begins almost immediately upon the addition of insulin and is essentially completed by about 4 h (75% receptor loss), we now find a lag period precedes down-regulation and that receptors are lost at a much slower rate requiring at least 2 days for completion. The most likely explanation for this kinetic difference is that two different mechanisms underlie the common observance of insulin-induced down-regulation. In the absence of recycling the rapidity of down-regulation appears to be a direct result of endocytotic uptake of insulin-receptor complexes (2, 6). However, this clearly is not the mechanism in our long term studies, since insulin was found to trigger down-regulation such that receptor loss continued in the absence of ligand. In other words, receptor loss still occurred despite the fact that insulin-receptor complexes were not formed. It should be mentioned that, although insulin does not appear to regulate cell surface receptors in cultured adipocytes directly through endocytotic uptake of insulin-receptor complexes, the possibility remains that uptake of these

8 Receptor Regulation i In Cultured Adipocytes 6383 complexes at early times before the observance of receptor mechanism that prevents complete receptor loss so that even loss may be involved in the actual triggering of this regulatory under extreme hyperinsulinemic conditions some receptors mechanism. remain to mediate the cellular actions of insulin. In earlier Insulin-induced translocation of receptors from the cell studies (1,2) we presented indirect evidence supporting such surface to the cell interior or loss of receptors via shedding a mechanism by showing that insulin-induced receptor downare two other potential mechanisms explaining down-regula- regulation plateaued after 75% of cell surface receptors were tion of adipocyte insulin receptors. However, the close agree- lost (at 4 h), and that the remaining receptors had a diminment between the loss of cell surface receptors and total ished ability to endocytose z51-insulin. These results sugreceptors (Fig. 8) argues against the translocation hypothesis, while the idea that insulin receptors are down-regulated via gested that insulin treatment caused a functional change in adipocytes, namely, a progressive impairment in ligand intershedding is remote based on earlier studies showing that nalization. However, the physiologicalsignificance of this receptors are internalized in adipocytes following ligand binding (6). Furthermore, it is unlikely that either mechanism is involved in receptor up-regulation, since cell surface receptors are increased not decreased (eliminating shedding), and since the number of intracellular receptors (10% of total) is insufficient to account for the extent of receptor up-regulation finding could be questioned because later studies showed that insulin-induced down-regulation was mediated by Tris inhibition of receptor recycling (3-5). In the current study insulininduced slowing of the ligand uptake rate was observed under more physiological conditions after 72 h of insulin treatment, and accelerated ligand uptake was seen after up-regulation in (eliminating receptor redistribution). insulin-free medium.moreover, concomitant changes were At this point we are left with two alternative, but not necessarily mutually exclusive, explanations. Insulin regulates cell surface receptors by either altering the rate of receptor found in the endocytotic uptake rate for the receptor itself. Considered together, these new findings add credence to our original idea that a regulatory mechanism exists to limit degradation/inactivation through effects on turnover of un- receptor loss and provide compelling evidence that prolonged occupied receptors, or insulin regulates the rate of de mu0 insulin treatment alters one of the insulin-sensitive biological receptor synthesis and/or maturation through effects on receptor transcription, translation, or post-translational receptor processing. Although insulin-induced down-regulation through effects on receptor biosynthesis has yet to be demactions of adipocytes, namely, endocytosis of insulin-receptor complexes. Overall, the present study using primary cultured rat adipocytes provides new information on the kinetics and mechonstrated, it is known that during differentiation of 3T3-C2 anisms by which insulin regulates the net number of cell preadipocytes to mature adipocytes (32) and after glucocorti- surface insulin receptors. In addition, the resulting data highcoid treatment of IM-9 lymphocytes (33) cell surface insulin light the importance of investigating the mechanism(s) of receptors are increased due to an accelerated rate of receptor insulin-mediated receptor regulation and receptor metabolism synthesis. under in vitro conditions which most closely approximate the Triggering of receptor regulation by insulin is a major actual processes occurring in uiuo. finding of the current study which provides new insights into the process of insulin-mediated down-regulation. Basically, REFERENCES we interpret both the triggering effect and the lag time preceding insulin-induced receptor loss as indicative that a 1. Marshall, S., and Olefsky, J. M. (1980) J. Clin. Znuest. 66, 763 signal is generated after the initial binding event which in 2. Marshall, S., and Olefsky, J. M. (1981) Diabetes 30, Marshall, S., Green, A., and Olefsky, J. M. (1981) J. Biol. Chem. turn mediates receptor regulation. Although the nature and 266, characteristics of this insulin-generated putative signal will 4. Marshall, S., and Olefsky, J. M. (1982) Biochem. Biophys. Res. require further investigation, such a postreceptor regulatory Commun. 102,646 mechanism is supported by both our adipocyte data and the 5. Marshall, S., and Olefsky, J. M. (1983) J. Cell. Physiol. 117, 195 earlier results of Car0 and Amatruda (34), who found that the 6. Green, A., and Olefsky, J. M. (1982) Proc. Natl. Acad. Sci. U. S. A. number of surface insulin receptors on primary cultured hep- 79, Soll, A. H., Kahn, C. R., and Neville, D. M., Jr. (1975) J. Biol. atocytes was reduced after treating cells with insulin mimick- Chem. 250, ers such as spermine or hydrogen peroxide. Since these agents 8. Freychet, P., Laudat, M. H., Laudat, P., Rosselin, G., Kahn, C. initiate the biological actions of insulin without interacting R., Gorden, P., and Roth, J. (1977) FEBS Lett. 25,339 with the receptor itself, these findings provided evidence that 9. Olefsky, J. M. (1976) J. Clin. Znuest. 57, 1165 down-regulation can be regulated at a postbinding or postre- 10. Bar, R. S., Gorden, P., Roth, J., Kahn, C. R., and DeMeyts, P. ceptor site, In a related study by Schwa11 and Erickson (35) (1976) J. Clin. Znuest. 68, 1123 on the mechanism(s) by which luteinizing hormone down- 11. Archer, J. A., Gorden, P., and Roth, J. (1975) J. Clin. Znuest. 66, 166 regulates receptors on the surface of cultured granulosa cells, 12. Kobayashi, M., and Olefsky, J. M. (1978) Am. J. Physiol. 236, they found that only a very brief exposure (2 min) to physio- E53 logical concentrations of luteinizing hormone was required to 13. Marshall, S. (1983) Diabetes 32, 319 produce a 50% loss of cell surface receptors 24 h later. It is 14. Freychet, P., Roth, J., and Neville, D. M., Jr. (1971) Biochem. interesting to note that both luteinizing hormone and insulin Bwphys. Res. Commun. 43,400 are hormones which are released episodically, the former 15. Rodbell, M. (1964) J. Biot. Chem. 239, Hirsch, J., and Gallian, E. (1968) J. Lipid Res. 9, 110 during the ovulatory cycle (36), and the later after feeding 17. Marshall, S., and Olefsky, J. M. (1979) J. Biol. Chem. 254, (37). Thus, among cell types and among the various hormones which are released as a pulse, there may be a common mech- 18. Marshall, S., and Olefsky, J. M. (1980) Endocrinology 107,1937- anism underlying regulation of surface receptors Insulin is known to regulate the sensitivity of adipocytes to 19. Hammons, G. T., and Jarett, L. (1980) Diabetes 29,475 the biological action of insulin by negatively modulating the 20. Suzuki, K., and Kono, T. (1979) J. Biol. Chem. 264, Cuatrecasas, P. (1972) Proc. Natl. Acad. Sci. U. S. A. 69, 318 number of cell surface insulin receptors (1). However, since 22. Marshall, S. (1982) Dinbetes 31, Suppl. 2,42A (abstr.) the role of insulin in carbohydrate, protein, and fat metabo- 23. Simpson, I. A., Hedo, J. A., and Cushman, S. W. (1984) Diabetes lism is SO integral to health, one can imagine a regulatory 33,13

9 6384 Receptor Regulation in Cultured Adipocytes 24. Gavin, J. R., 111, Roth, J., Neville, D. M., Jr., DeMeyts, P., and Yamada, K. (1981) Proc. Natl. Acad. Sci. U. S. A. 78,6917 Buell, D. N. (1974) Proc. Natl. Acad. Sci. U. S. A. 71, Berhanu, P., and Olefsky, J. M. (1982) Diabetes 31, Blackard, W. G., Guzelian, P. S., and Small, M. E. (1978) Endo- 32. Knutson, V. P., Ronnett, G.V., and Lane, M.D. (1982) Proc. crinology 103, Sci. Natl. Acad. U. S. A. 79, Matt, D. M., ~ ~ B. v., and ~ Bennett, ~ p. d H. (1979), J. ~ i ~ l 33-. Fantus, 1. G., Saviolakis, G. A.9 Hedo, J. A.9 and Gordon, p. (1982) Chem. J. Biol. Chem. 257, , Caro, J. F., and Amatruda, J. M. (1980) Science (Wash. D. C.) 27. Ronnett, G. V., Knutson, V. P., and Lane, M. D. (1982) J. Bwl. 210,1029 Chem. 257, Schwall, R. H., and Erickson, G. F. (1983) J. Biol. Chem. 258, 28. Chang, T-H., and Polakis, S. E. (1978) J. Bwl. Chem. 253, Hilary, H., and Elstein, M. (1969) J. Endocriml. 43, K~PP, M., and Lane, M. D. (1981) J. Bkl. hm 256, Greenfield, M., Kolterman, O., Olefsky, J. M., and Reaven, G. M. 30. Kasuga, M., Kahn, C. R., Hedo, J. A., Van Obberghen, E., and (1978) Metabolism 27,