Actions of Insulin in Fat Cells EFFECTS OF LOW TEMPERATURE, UNCOUPLERS OF OXIDATIVE PHOSPHORYLATION, AND RESPIRATORY INHIBITORS*

When isolated rat epididymal fat cells were incubated with [125I]iodoinsulin for 5 min at 37 degrees, radioactivity accumulated in the plasma membrane fraction (Peak 1) and an unidentified particulate fraction (Peak 2) as reported previously (Kono, T., Robinson, F.W., and Sarver, J.A. (1975) J. Biol. Chem. 250, 7826-7835). This accumulation of radioactivity in Peak 2 (but not that in Peak 1) was greatly impaired when cells were incubated with iodoinsulin in the presence of a variety of metabolic inhibitors that reduce the cellular content of ATP. The reduction in the ATP level coincided with a disappearance of the stimulatory effects of insulin on sugar transport and the hormone-sensitive phosphodiesterase. In contrast, ATP depletion had no significant effects, at least during a 5-to 15-min incubation, on the intracellular water space and on the basal sugar transport and phosphodiesterase activities. When cells once depleted on ATP by treatment with 2,4-dinitrophenol (1 mM; 10 min) were washed and suspended in fresh buffer, the ATP level was recovered almost fully in 10 min. This recovery coincided with the restoration of responsiveness to insulin. When cells were incubated with [125I]iodoinsulin or insulin for 5 min at 15 degrees instead of 37 degrees, a negligible quantity of radioactivity accumulated in Peak 2 and insulin failed to activate sugar transport. In contrast, under the same conditions, radioactivity accumulated in Peak 1 and insulin stimulated phosphodiesterase considerably. These results suggest that ATP, or some other compound metabolically related to ATP, may be necessary for the actions of insulin on sugar transport and phosphodiesterase. ATP, or some other related compound, may also be necessary in the formation of the radioactive Peak 2, although the physiological function and cellular location of this peak are yet to be ascertained.


7232
When isolated rat epididymal fat cells were incubated with [*2511iodoinsulin for 5 min at 37", radioactivity accumulated in the plasma membrane fraction (Peak 1) and an unidentified particulate fraction (Peak 2) as reported previously (Kono, T., Robinson, F. W., and Sarver, J. A. (1975)5. Biol. Chem. 250,[7826][7827][7828][7829][7830][7831][7832][7833][7834][7835]. This accumulation of radioactivity in Peak 2 (but not that in Peak 1) was greatly impaired when cells were incubated with iodoinsulin in the presence of a variety of metabolic inhibitors that reduce the cellular content of ATP. The reduction in the ATP level coincided with a disappearance of the stimulatory effects of insulin on sugar transport and the hormone-sensitive phosphodiesterase.
In contrast, ATP depletion had no significant effects, at least during a 5to 15-min incubation, on the intracellular water space and on the basal sugar transport and phosphodiesterase activities. When cells once depleted of ATP by treatment with 2,4dinitrophenol(1 mM; 10 min) were washed and suspended in fresh buffer, the ATP level was recovered almost fully in 10 min. This recovery coincided with the restoration of responsiveness to insulin.
When cells were incubated with [12511iodoinsulin or insulin for 5 min at 15" instead of 37", a negligible quantity of radioactivity accumulated in Peak 2 and insulin failed to activate sugar transport.
In contrast, under the same conditions, radioactivity accumulated in Peak 1 and insulin stimulated phosphodiesterase considerably. These results suggest that ATP, or some other compound metabolically related to ATP, may be necessary for the actions of insulin on sugar transport and phosphodiesterase. ATP, or some other related compound, may also be necessary in the formation of the radioactive Peak 2, although the physiological function and cellular location of this peak are yet to be ascertained.
Insulin stimulates sugar transport across the plasma membranes of certain cell types (1) and activates phosphodiesterase in fat cells (2, 3). The initial step in these and other physiological actions of insulin is probably the interaction of the hor-* This work was supported in part by United States Public Health Service Grants 5ROl AM 06725, 5POl AM 07462, and lP17 AM 17026.
$ Recipient of National Service Award AM 01033 from the United States Public Health Service. mone with its specific receptors on the plasma membrane (4,5). It has yet to be determined how this interaction modifies certain enzyme activities (6). Recently, we found that when fat cells were incubated with ['*"I]iodoinsulin, washed, and homogenized, radioactivity was recovered in an unidentified subcellular fraction as well as in the plasma membrane fraction (3). The work reported here was initiated to determine whether the accumulation of insulin (or its derivative) in this unidentified fraction is essential for activation of sugar transport and phosphodiesterase.
A preliminary account of this work has been published (7).

MATERIALS AND METHODS
Crude bovine serum albumin (Fraction V, Lot 53CO6703, crude bacterial collagenase (type II, Lot 15COO75), dried firefly lanterns, and carbonyl cyanide m-chlorophenylhydrazone were obtained from Sigma. Cyclic 8-[3HlAMP ' was purchased from SchwarzlMann and purified by chromatography on AG 5OW-X2 (3). Labeled ['251]iodoinsulin was obtained from Cambridge Nuclear and purified on a column (1.2 x 100 cm) of Sephadex G-50 using a solution containing 1 rnM HCl, 50 rnM NaCl, and 1 mg/ml of Fraction V albumin. Labeled [3H]inulin, ['*CJurea,mannitol (in 70% ethanol), and 3-0-methyl-n-[14C]glucose (in 90% ethanol) were purchased from New England Nuclear. The latter two preparations were dried in oac~o and redissolved in 0.145 M NaCl solution. All the labeled solutions were divided into small fractions and kept at -20". Insulin, 10 times recrystallized, was kindly supplied by Dr. J. Schlichtkrull of Novo Laboratories. Other reagents were either reagent grade or the best grade available.
Isolated fat cells were prepared by the collagenase method (8) from epididymal adipose tissue of Sprague-Dawley rats weighing 160 to 210 g. Unless stated otherwise, the binding of ['2511iodoinsulin, hormonal stimulation of sugar transport, and hormonal activation of phosphodiesterase were studied under the following conditions. Binding of ['25ZlZodoinsuZin -Approximately 400 mg of fat cells were incubated with 0.8 rnM ['251]iodoinsulin (approximately 0.5 pC!i/ ml) for 5 min at 37" in 4 ml of Krebs-Henseleit Hepes buffer, pH 7.4, containing 20 mg/ml of Fraction V albumin (4). The incubated cells were fractionated essentially as described previously (3). Briefly, cells were washed twice with 0.25 M sucrose, 10 rnM Tris/HCl (Buffer A) and homogenized in the same buffer with a Dounce tissue grinder (type B, 8 strokes). The homogenate was centrifuged for 2 min at 0" in a Beckman J-21B centrifuge (with a JA-20 rotor) set at 10,000 rpm. The supernatant, free of fat, was further centrifuged for 30 min at approximately 200,000 x g (50,000 rpm in a Beckman No. 65 rotor). The second supernatant was discarded, and the pellet was suspended in 2 ml of cold Buffer A containing 1 rnM EDTA. This suspension (of Actions of Insulin in Fat Cells 2227 crude microsomes) was well homogenized with a Swinny apparatus (9), placed on top of a linear sucrose density gradient (13 x 100 mm; see below), and centrifuged in a Beckman SW 41 rotor for 35 min at 40,000 rpm.2 The sucrose density gradient solution was 15 to 45%, w/ w, with respect to the sucrose concentration, contained 1 rnM EDTA(Na), and was buffered with 10 mM Tris/HCl, pH 7.4. After centrifugation, the sucrose solution was fractionated (0.7 ml/fraction) and analyzed for iE51 and 5'-AMPase (3) activities.
All the tests to be compared were performed with portions of a pooled cell preparation; the samples, including an appropriate control, were centrifuged at the same time. The overall recovery of the plasma membrane was assessed from the height of the peak of 5'-AMPase activity in the sucrose density gradient profile. The difference in recoveries observed among samples centrifuged at the same time was usually less than 20%. In some experiments, the recovery of NADH:ferricyanide oxidoreductase (a marker enzyme for endoplasmic reticulum (3)) was also monitored. The distribution of protein, a commonly used reference standard, was not determined since its concentration in the sucrose solution was usually too low to be measured accurately.
Hormonal Stimulation of Sugar Transport -Approximately 50 mg of cells were preincubated, or "conditioned," in 5 ml of Krebs-Henseleit Hepes buffer with albumin (see above) at 37"for 30 min (i.e. until the basal rate of transport was stabilized at a low level)." Subsequently, the cell suspension was supplemented with 1 nM insulin or indicated reagents, centrifuged 10 s, and concentrated to approximately 50 mg/ml by removing excess infranatant solution. Several samples of the concentrated cell suspension, 0.2 ml each, were transferred to small test tubes and kept at 37". The total time for insulin treatment, including the time for centrifugation, concentration, and dividing, was 5 to 10 min. In some experiments, insulin was added to the 0.2-ml cell suspension. The rate of sugar transport was determined by modification of the "oil flotation" method described by Gliemann et al. (10). Briefly, approximately 10 mg of cells in 0.2 ml of buffer were mixed at 0 s with 0.05 ml of a sugar solution consisting of 5 mM 3-@methyl-n-['4Clglucose (0.2 +Ci), 0.9 mg/ml of [3Hlinulin (0.2 &I), and 0.154 M NaCl. The mixture was shaken manually at 37". After 10 s, 0.2 ml of this incubation mixture was transferred to a small plastic centrifuge tube containing 0.1 ml of dinonylphthalate and placed in a Beckman microfuge (model 152). The centrifuge was turned on at 20 s and turned off at 50 s. Subsequently, the centrifuge tube was cut at the middle of the dinonylphthalate layer, and cells packed in the upper half were dispersed into a mixture of 1 ml of water and 10 ml of scintillation fluid containing Triton X-100 (11). In the evaluation of the transport data, it was assumed that 3-Omethyl-n-glucose in excess of inulin was intracellular. The so-called intracellular sugar space was calculated by dividing the intracellular quantity of 3-O-methyl-n-glucose by the extracellular concentration of the same sugar (12).
Hormonal Stimulation of Phosphodiesterase -Approximately 200 mg of cells were incubated with or without 2 nM insulin and the other indicated agents in 5 ml of Krebs-Henseleit Hepes buffer containing albumin (see above) for 5 or 15 min at 37". The cells were then washed with Buffer A (see above) and homogenized in a Dounce tissue grinder. Phosphodiesterase activity was determined in Fraction P-2 (crude microsomal fraction) as described previously (3).
Other Methods -The quantity of cells in a sample was estimated from the content of malate dehydrogenase as described previously (13). ATP was assayed by the luciferin-luciferase method (14). The activity of N-acetyl-/3-glucosaminidase was determined at pH 4.3 in the presence of 0.2% Triton X-100 (15, 16). "Insulinase" was assayed by measuring the formation of radioactive materials that were soluble in 5% trichloroacetic acid from ['2511iodoinsulin at pH 7.4 (17, 18). The immunoreactivity of '*SI-labeled derivatives of iodoinsulin was * Similar results were obtained when the gradient was centrifuged for 45 min at 30,000 rpm or for 35 min at 35,000 rpm. However, the separation of the plasma membrane  and endoplasmic reticulum (NADHferricyanide oxidoreductase) was not good when the centrifugation was carried out for 45 min or more at 40,000 rpm. 3 After the conclusion of the present work, it was found that the basal transport activity stabilized at a lower level (e.g. 10.8 ? 0.5 pl/ g) and a relatively larger plus insulin activity (e.g. 43.0 f 0.5 pi/g or 4.0-fold of the basal activity) was observed when cells were "aged" for 30 min at 37" in a small volume of buffer (10 to 20 mg of cells in 0.2 ml) rather than in a large volume (approximately 50 mg in 5 ml) as described here. The reason for this was unknown.
determined by a modification of the double antibody method (19). All the experiments with fat cells were repeated on at least two separate occasions (3).

On Binding of Zodoinsulin
-In confirmation of our previous results (31, when fat cells were incubated  with 1Yliodoinsulin  for several minutes, washed, homogenized, and fractionated on a sucrose density gradient, radioactivity was recovered in the plasma membrane fraction (Fractions 7 to 9 in Fig. IA, referred to as Peak 1) and in an unidentified particulate fraction (Fractions 13 and 14, referred to as Peak 2). When the incubation was terminated at 30 or 60 s, only Peak 1 was clearly distinguishable.
This suggested that there might be a short lag in the formation of Peak 2. Both peaks reached their maxima in 5 min (Peak 1 > Peak 2), which were maintained during the next 25 min (data not shown).
When 1 pM native insulin was added to the system in this steady state, the peaks declined almost in parallel during the subsequent 10 min (Fig. lB), presumably due to displacement of iodoinsulin by native hormone. The apparent competition between the two compounds was also seen when cells were incubated with 0.

Distribution of Y activity in a crude microsomal fraction of fat cells incubated with ['"SIliodoinsulin.
Fat cells were incubated with [12511iodoinsulin, and the cell homogenates fractionated through the sucrose density centrifugation step. The experiments were carried out as described under "Materials and Methods" with some modifications as specified below. A, cells were incubated with 0.8 nM [1251]iodoinsulin at 37" for 0.5 (01, 1 (xl, or 5 (0) min. B, after cells had been incubated with [12511iodoinsulin as above for 5 min, unlabeled insulin was added to a final concentration of 1 pM, and incubation continued at 37" for 0 CO), 5 (x), or 10 (0) min longer. C, cells were incubated for 5 min with 0. 2 were both suppressed slightly more than 50% by 10 nM native insulin and almost eliminated by 1 PM hormone. These data indicated that the apparent affinities of insulin for the hypothetical binding sites in Peaks 1 and 2 were similar and that the binding of radioactivity to both sites was almost fully displacable by the native hormone.
Peak 2, once formed, was partially lowered in 10 min when 1 mM 2,4-dinitrophenol was added to cells that had been treated with ['251]iodoinsulin ( Fig. 20 The rate of this lowering was apparently slower than that observed in the displacement experiment ( (1 mM, 10 min) had no detectable effect on the sedimentation characteristics of the subcellular particles in Peak 2 as well as those in Peak 1.
When incubated in a cell-free system, the subcellular particles corresponding to Peak 2, but prepared from cells that had not been exposed to ['2sI] (3). Peak 2 was distinct from the peaks of either N-acetyl-P-glucosaminidase or "insulinase" (Fig. 3). The former is generally considered to be a marker enzyme for lysosomes (15,16). Although a part of this enzyme was apparently fractionated into the soluble fraction (Fraction 15), most of its activity was found in Fractions 3 to 5. It should be noted that these fractions (Fractions 3 to 5) exhibited negligible lZ51 activity. Insulinase was thought to bind some iodoinsulin. However, it was not clear from this experiment ( Fig. 3) whether there was a special insulinase corresponding to Peak 2. The radioactive substances solubilized from Peaks 1 and 2 were still reactive with anti-insulin serum, but to lesser extents than the original iodoinsulin (Table I). No further information was obtained regarding the homogeneity and identity of these solubilized substances.
On Hormonal Stimulation of Sugar Transport-Similar to the finding of Livingston and Lockwood (231, our preliminary data (not shown) indicated that the transport of a nonmetabolizable sugar across the plasma membrane of fat cells could be measured with reasonable accuracy by the method used in the present study. Table II shows that the transport of 3-0methyl-n-glucose was stimulated approximately 2-fold by insulin treatment of fat cells and that this stimulation was reduced to a large extent when cells were incubated with insulin in the presence of 2,4-dinitrophenol, dicumarol, KCN, or NaN, at appropriate concentrations.
At the same concentrations, these inhibitors greatly depressed the ATP level. In 4 2,CDinitrophenol was used as Tris salt, which is highly watersoluble. When 2,bdinitrophenol in ethanol solution was added to a cell suspension (the final ethanol concentration, l%), recoveries of 5'-AMPase and lz51 activity in Peak 1 were considerably lower than the controls. were filtered with Millipore membrane (EGWP, 0.2 pm) and the 9 activity retained on the filter was determined (0). As a control, aliquots of cells were incubated with 0.8 nM [1Z511iodoinsulin for 5 min at 37" and fractionated as above. Their subcellular fractions were filtered (without any additional incubation with iodoinsulin), and the lz51 activity retained on the filter was determined (0). None of the inhibitors used in Experiments A and B had any significant effects on the recoveries of 5'-AMPase and NADH:ferricyanide oxidoreductase activities. See also the legend for Fig. 1. contrast, concentrations of inhibitors that did not significantly reduce the hormonal stimulation had negligible effect on the ATP level. None of the inhibitors had any significant effect on the basal transport activity. Neither 1 mM 2,4-dinitrophenol nor 10 mM NaN, had any detectable effect on the intracellular or extracellular water space (Table III). Alterations in these spaces would greatly change the apparent transport activity as determined by the present method. As it is well known, the transport in this system is by facilitated diffusion, regardless of the presence or absence of insulin (10). Fig. 4 shows that the ATP level was restored almost completely by 10 min in cells that had been treated with 2,4dinitrophenol, washed, and suspended in fresh buffer. Under the same conditions, the cellular responsiveness to insulin was also restored (Table IV, Experiment A). Experiment B shows that 2,4-dinitrophenol added to cells 5 min after insulin partially reduced the insulin-stimulated transport activity during the subsequent lo-min incubation.
This effect was quantitatively similar to that on Peak 2 observed under essentially the same conditions  Cells were incubated with 0.8 nM ['251Jiodoinsulin for 5 min at 37". Cell homogenates were fractionated under standard conditions as shown in Fig. L4. Samples of Peak 1, Peak 2, and ['~~Iliodoinsulin (each approximately 1000 cpm) were incubated for 1 h at 0" in 1% Triton X-100 at pH 1.2 (HCI) and subsequently neutralized with NaOH. The neutralized samples were incubated for 3 days at O-5" with guinea pig anti-insulin serum plus rabbit anti-guinea pig serum. The precipitate was separated by centrifugation and its lzSI activity counted. Results obtained in two separate experiments (A and B) are reported. The activity of lz51 precipitated in the absence of the two antisera was 3% or less of the total activity. insulin at 15" and their transport activity was determined at the same temperature, a negligible hormonal effect was recorded (Table V, Experiment B). Since the insulin-stimulated transport activity was clearly detectable at 15" (Experiment 0, these data imply that the stimulation process, rather than the transport itself, was particularly sensitive to the low temperature. As noted before, the formation of Peak 2 was also sensitive to low temperature (Fig. lD). On Hormonal Stimulation of Phosphodiesterase -When cells were incubated with graded concentrations of 2,4-dinitrophenol, the level of ATP and the hormonal stimulation of phosphodiesterase were depressed almost in parallel (Fig. 5). The stimulation was also depressed by dicumarol, carbonyl Cell suspensions, maintained at 37", were mixed with the indicated inhibitors at 0 min, with 1 nM insulin (where indicated) at 10 min, and with the sugar solution at 15 min. The uptake of 3-0methyl-n-glucose during a 20-s incubation was determined by the oil flotation method, and the results are shown in terms of the sugar distribution spaces as described under "Materials and Methods." Samples for ATP assay were taken at 10 min. The ATP assay by the luciferin-luciferase method was not affected by any of the inhibitors tested, except by 1 rnM dicumarol which suppressed the data by approximately 20%. (1 Mean value f S.E. (n = 3, except control, where n = 6 to 9). b DNP, 2,4-dinitrophenol. c Dinitrophenol, dicumarol, and KCN were neutralized with Tris, NaOH, and KH,PO,, respectively. of inhibitors were the same as those observed in the transport experiment (Table II). Arsenate, an uncoupler of certain nonoxidative phosphorylation (20), also reduced the hormonal effect. This agent, unlike the others, partially stimulated the basal enzyme activity.
The reason for this was unknown.
2,4-Dinitrophenol, in contrast to the SH compounds and metal-chelating agents tested in a previous study (31, had   The sugar transport activity was determined by standard methods after cells had been treated at 37" as specified below. In Experiment A, aliquots of cells were incubated with buffer alone for 10 min (untreated cells). Other portions of cells were incubated with 1 mM 2,4-dinitrophenol for 10 min (DNP-treated cells). A third set of cells was incubated with 1 mrvr dinitrophenol for 10 min as above; the cells were then washed and incubated in a fresh buffer for 10 min (recovered cells). The buffer used in this experiment was supplemented with 2 rnM glucose. All the cell preparations were incubated with or without 1 nM insulin for an additional 5 min before their transport activity was determined.
In Experiment B, samples of cells were incubated with or without 1 nM insulin for 5 min (control). A second aliquot of cells was incubated with 1 mM dinitrophenol and with or without 1 nM insulin for 10 min. Insulin and dinitrophenol were added at the same time (0 min). A third set of cells was first incubated with or without insulin and, 5 min later, mixed with 1 rnM dinitrophenol.
The mixture was further incubated for 10 min before its transport activity was determined.  Effects of reduced temperature on stimulation of sugar transport by insulin Sugar transport was determined as described under "Materials and Methods" at the indicated temperatures after cells had been incubated without or with insulin at the specified temperatures.
In Experiment A, cells were incubated with or without insulin for 5 min at 37", and the transport activity was determined at the same temperature. In Experiment B, cells were incubated with or without insulin for 5 min at 15", and the transport activity was determined at the same temperature.
In Experiment C, cells were incubated with or without insulin for 5 min at 37" and 5 more min in a 15" water bath. During this second incubation, the temperature of the cell suspension (0.2 ml in volume) was lowered to 15". The transport activity was determined at 15". Insulin and dinitrophenol were added at the same time. Standard procedures described under "Materials and Methods" were used for determination of ATP (A) and phosphodiesterase activities in the basal (0) and insulin-stimulated (0) states. The length of the small vertical bar indicates twice the standard error ( n values for ATP and phosphodiesterase determinations are 6 and 3, respectively). Cells were incubated with or without 2 no insulin in the presence or absence of a metabolic inhibitor as indicated. The treated cells were washed and homogenized. Phosphodiesterase was assayed in Fraction P-2 (crude microsomal fraction) as described under "Materials and Methods." In Experiment A, cells were incubated with or without the indicated metabolic inhibitors for 15 min at 37". The inhibitors and insulin were added at the same time. In Experiment B, samples of cells were incubated with or without insulin in the presence of 1 mM 2,4-dinitrophenol for 10 min at 37" and, subsequently, washed and homogenized. Other cell samples were incubated with or without insulin for 10 min at 37" and homogenized in the presence of 1 rnM dinitrophenol.
In Experiment C, aliquots of cells were incubated with buffer alone for 10 min at 37" (untreated cells). Other aliquots of cells were incubated with 1 rnM dinitrophe-no1 for 10 min at 37" (dinitrophenol-treated cells). A third set of aliquots were incubated with 1 rnM 2,4-dinitrophenol for 10 min as above; subsequently, the cells were washed and incubated with buffer alone for 10 min at 37" (recovered cells). All of the cell preparations were incubated with or without 2 nM insulin for an additional 5 min prior to washing followed by homogenization.
The buffer used in this experiment was supplemented with 2 rnM glucose. In Experiment D, cells were incubated with or without 2 nM insulin for 5 min at either 37" or 15" as indicated.

Experiments
Phosphodiesterase activity. Cells were incubated with or without 2 nM insulin for 15 min at 37". Dinitrophenol (Tris salt) was added to a final concentration of 1 mM at the indicated time intervals. Phosphodiesterase was assayed under standard conditions. The figure shows the basal phosphodiesterase activity (01, insulin-stimulated enzyme activity (O), basal activity after dinitrophenol treatment (A), and insulin-stimulated activity after dinitrophenol treatment (A). The length of the small vertical bar indicates twice the standard error (n = 3). V). Dinitrophenol added after insulin depressed the insulinstimulated phosphodiesterase activity almost to the basal level in 10 min (Fig. 6). During the same period of time, dinitrophenol depressed the insulin-stimulated sugar transport activity to a lesser extent (Table IV) and only partially lowered Peak 2 (Fig. 2C) in experiments that were carried out under identical conditions.

DISCUSSION
The present data show that the level of ATP and stimulation of sugar transport and phosphodiesterase by insulin are affected almost in parallel by graded concentrations of several metabolic inhibitors (Tables II and VI and Fig. 5). The data also indicate that the level of ATP and the cellular responsiveness to insulin are both restored rapidly and almost completely in cells that have been treated with dinitrophenol, washed, and resuspended in fresh buffer (Tables IV and VI and Fig. 4) (Tables IV and VI  and Fig. 5), (b) the formation of Peak 1, or the interaction of insulin with the plasma membrane (Fig. 2, A andB), and (c) the quantities of oil droplets (from broken cells) in the incubation mixtures (data not shown). The present results are apparently consistent with previous observations that (a) oligomycin (a respiratory inhibitor) blocked the insulin stimulation of phosphodiesterase in fat cells (24), (b) uncouplers of oxidative phosphorylation inhibited the insulin activation of sugar transport in muscle, although the same agents partially stimulated the basal transport activity (la), and (c) insulin stimulated the turnover of n2P (25) and facilitated the incorporation of '?'P into phospholipids (26) and proteins (27,28) in fat cells. On the other hand, it has been repeatedly suggested in the past that insulin might stimulate dephosphorylation. This line of reasoning is supported by the observations that sugar transport in muscle is stimulated by anoxia (29) or uncouplers of oxidative phosphorylation (12) and that the physiological actions of insulin are essentially antagonistic to those of catecholamines, which (as P-agonists) stimulate phosphorylation (30). However, sugar transport in fat cells is not stimulated by anoxia (data not shown) or uncouplers of oxidative phosphorylation (Table II). In addition, the earlier suggestion does not exclude a possibility that ATP, or some other compound metabolically related to ATP, might be necessary in some step(s) in the mechanism of insulin actions.
As noted earlier, the original aim of the present work was to examine whether the accumulation of insulin (or its derivative) in an unidentified subcellular fraction (Peak 2) is essential for the physiological actions of insulin. Our data show that, at 15", insulin apparently stimulates phosphodiesterase before Peak 2 is formed (Table VI and Fig. ID) and that, at 37", 2,4-dinitrophenol seems to deactivate the enzyme before Peak 2 has disappeared (Figs. 6 and 2C). These results imply that the accumulation of insulin in the unidentified fraction may not be involved in the stimulation of phosphodiesterase. On the other hand, both the formation of Peak 2 and the hormonal activation of sugar transport are (a) apparently dependent on ATP or some other related compound (Fig. 2 Table II), (6) highly sensitive to low temperature (Fig. lD and  Table V), and (cl slowly lowered by dinitrophenol added after insulin or iodoinsulin ( Fig. 2C and Table IV). Therefore, a correlation can still be postulated between these two processes.

, A and B and
As for the mechanism of formation of Peak 2, we think it highly unlikely that this peak is formed as an artifact of homogenization.
In fact, no Peak 2 is detectable in homogenates prepared by our standard method after cells have been incubated with iodoinsulin at a low temperature or in the presence of certain metabolic inhibitors, such as 2,4-dinitrophenol (Figs. v), 2A, and 2B). As noted earlier, dinitrophenol has no apparent effects on the formation of Peak 1 and on the sedimentation characteristics of Peaks 1 and 2 (Fig. 2, A and C). It is also doubtful that Peak 2 represents [12511iodoinsulin incorporated by fat cells. This view is supported by the observations that Peak 2 disappears as rapidly as Peak 1 with the addition of native insulin to fat cells (Fig. lB) and that very little lZ51 activity is found in the lysosomal fraction (Fig. 3). Therefore, it is tentatively considered that Peak 2 corresponds to [""Iliodoinsulin (or its derivative) localized in some specific regions on the cell surface. A mosaic nature of the plasma membrane has been described in the liver cell system (31)(32)(33). The effects of low temperature and metabolic inhibitors (Figs. lD, 2A, and 2B) as well as the apparent short delay in the formation of Peak 2 (Fig. lA) may be explained if this peak is formed as a result of a lateral migration of insulin or the insulin. receptor complex on the cell surface. It was previously reported that the lateral translocation of immunoglobulin observed on the lymphocyte surface was blocked by low temperature, 2,4-dinitrophenol, or NaN, (34,35).