Amino Acid Regulation of Insulin Action in Isolated Adipocytes

Transcription

1 THE JOURNAL OF BOLOCKXL CHEMSTRY Q 1989 hy The American Society for Biochemistry and Molecular Biology, nc. Vol. 264, No, 4, ssue of February 5, pp ,1989 Printed in U.S.A. Amino Acid Regulation of nsulin Action in solated Adipocytes SELECTVE ABLTY OF A~NO ACDS TO EN~AN~E BOTH NSULN NSULN RESPONSVENESS OF THE PROTEN SYNTHESS SYSTEM* Stephen Marshall$ and Ricardo Monzon From the Department of Biochemistry, University of Tennessee, Memphis, Tennessee S~N~TVTY AND MAXM^ (Received for publication, July 13, 1988) Using the number and concentration of amino acids in Dulbecco s modified Eagle s medium as reference (~MEM = loo%), we found that a maximally effective concentration of insulin (10 ng/ml) stimulated protein synthesis by 125% over basal rate in the presence of 50% amino acids (ECeo = 19%), but by only 48% in amino acid-free buffer. Moreover, time course experiments revealed that amino acid regulation of insulin action was very rapid (ty, of 9.5 min) and readily reversible (e30 min). This effect was specific in that basal rates of protein synthesis were unaltered by amino acids. A second effect of amino acids was to markedly enhance insulin sensitivity of the protein synthesis system in a dose-dependent manner. Thus, the halfmaximally effective concentrations of insulin required to stimulate protein synthesis fell from 0.43 to 0.25 to 0.15 ng/ml in the presence of 0, 50, and 150% amino acids. Neither insulin sensitivity nor maximal insulin responsiveness of the glucose transport system was altered by amino acids, nor did amino acids affect the insulin binding capacity of cells. When we divided the 14 amino acids found in DMEM into two groups, we found that one group of 7 amino acids had little or no effect on insulin sensitivity or responsiveness, whereas the other group was fully active (a 157% increase in insulin responsiveness, ED60of 0.21 ng/ml versus a 68% increase, ED50 of 0.51 ng/ml, with no amino acids). soleucine and serine together increased both insulin sensitivity and responsiveness to 60-70% of that seen with the full complement of amino acids. n conclusion: 1) amino acids modulate insulin action by enhancing maximal insulin responsiven~s and insulin sensitivity of the protein synthesis system, and the regulatory site of amino acid action appears to be distal to the common signal pathway, within the insulin action-protein synthesis cascade, and 2) the effects of amino acids are specific, in that basal rates of protein synthesis are unaffected, only certain amino acids influence insulin action, and amino acids fail to alter insulin binding or the insulin-responsive glucose transport system. These studies, together with those in the companion paper, demonstrate that the pleiotropic actions of insulin on enhancing glucose uptake and protein synthesis are mediated through divergent pathways that can be independently regulated. t is likely that such control mechanisms are involved in coordi- *This work was funded by Research Grant DK38754 from the National nstitutes 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 adu~rtisemen~ in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $Recipient of Research Career Deveopment Award DK from the National nstitutes of Health. To whom correspondence should be addressed nating the disposition of metabolic substrates during the daily fluctuations amino acids and glucose associated with dietary uptake and may be relevant to the etiology of certain insulin-resistant states such as Type 1 diabetes and obesity. n the companion paper (l), we demonstrated that glucose selectivity desensitizes the glucose transport system at a site downstream from the common signal pathway, within the specific insulin action-glucose transport cascade. Since glucose, an important metabolic substrate, can specifically regulate insulin sensitivity and responsiveness of the glucose transport system and since amino acids are known to directly influence protein turnover by suppressing protein degradation in liver, muscle, and heart (2-51, we examined whether amino acids could regulate the insulin-responsive protein synthesis system in isolated adipocytes. These results are presented in the current study. EXPERMENTAL PROCEDURES The procedures used are described in the accompanying paper (1). RESULTS Effects of Amino Acids on Maximal nsulin Responsiveness-To test the idea that amino acids regulate maximal insulin responsiveness and/or insulin sensitivity of the protein synthesis system, we preincubated adipocytes for 30 min in buffer containing various concentrations of amino acids (ranging from 5 to 150%) with 0 insulin (basal values), 0.3 ng/ml insulin, or 25 ngfml insulin (submaximal and maximally effective doses, respectively). nco~oration of [3Hf leucine into protein was then measured after 1 h. t is important to mention that the composition of amino acids included in our buffer is based on the number and concentration of amino acids found in Dulbecco s modified Eagle s medium (containing 25 mm Hepes and 1% bovine serum albumin; ~ ~ ~ M Thus, ). 100% amino acids is equivalent to the number and concentration of amino acids in DMEM. The concentration of amino acids in DMEM fmg/liter) is as follows: isoleucine, 105 mg; serine, 42 mg; histidine, 42 mg; arginine, 84 mg; t~~ophane, 16 mg; lysine, 146 mg; glycine, 30 mg; threonine, 95 mg; glutamine, 584 mg; valine, 94 mg; phenylalanine, 66 mg; methionine, 30 mg; tyrosine, 104 mg; leucine, 105 mg; cystine, 63 mg. As shown in Fig. 1, basal rates of protein synthesis were unaffected by the inclusion of amino acids, whereas maximal insulin responsiveness (25 ng/ml) was markedly enhanced in a dose-dependent manner: a 48% increase above basal was ~. The abbreviations used are: DMEM, Dulbecco s modified Eagle s medium; Hepes, 4-{2-hydroxyeth~l~-l-piperaz~neethanesulfonic acid.

2 Amino Acid Regulation of nsulin Action Amino Acid Concentration (%) FG. 1. Amino acid effects on maximal insulin responsiveness. Cells were preincubated for 30 min with the indicated concentrations of amino acids in the absence (basal) or presence of 0.3 or 25 ng/ml insulin. ncorporation of [3H]leucine into protein was then determined after 1 h. DMEM was used as reference so that 100% amino acids reflects the same total number and concentration of amino acids as found in DMEM (minus leucine) f0.25 ng/ml Amino Acid 2000 ' nsulin Concentration (ng ml) FG. 2. Amino acid effects on insulin sensitivity. Cells were preincubated for 30 min with the indicated concentrations of amino acids and various concentrations of insulin (1-2 ng/ml). nsulin doseresponse curves were constructed based on [3H]leucine incorporation into protein (1 h/37 "C), and estimates of insulin sensitivity were obtained by calculating the insulin ED,, values for each amino acid treatment group. seen at 0% amino acids, a 125% increase was seen at 50% amino acids, but little additional effect was observed between 50 and 150%. The half-maximally effective amino acid concentration was 19%, well within the physiological concentrations of amino acids found in rat plasma (6). When cells were exposed to a submaximally effective concentration of insulin (0.3 ng/ml), only a minimal stimulatory effect on protein synthesis was observed in the absence of amino acids, whereas a near maximal effect was seen at 150% amino acids = 61%). The finding that a suboptimal dose of insulin is more effective in enhancing protein synthesis in the presence of amino acids suggests that insulin sensitivity is concomitantly increased along with insulin responsiveness. Effects of Amino Acids on nsulin Sensitivity-To substantiate that amino acids can indeed enhance the sensitivity of adipocytes to insulin, we obtained complete insulin doseresponse curves using cells exposed to 0,50, and 150% amino acids (Fig. 2). Results from these experiments demonstrate that amino acids can markedly increase insulin sensitivity. Thus, insulin EDs0 values decreased from 0.43 ng/ml & 0.02 (n = 4) in the absence of amino acids, to 0.25 ng/ml (n = 10) at 50% amino acids. t is important to note that there was a further 40% increase in insulin sensitivity when the amino acid concentration was increased from 50 to 150% (from an insulin ED50 of ng/ml, respectively, (p < 0.01)). Hqwever, little or no difference was noted in maximal insulin responsiveness over this range of amino acids. Since a marked increase in sensitivity was observed without a corresponding change in maximal insulin responsiveness, this finding suggests that amino acid regulation of insulin sensitivity is not linked directly to changes in maximal responsiveness. Since we were using a new method to assess t3h]leucine incorporation into protein, it was important that we use the classical trichloroacetic acid method to confirm our basic observation that amino acids regulate insulin sensitivity and responsiveness. Fig. 3 depicts the results of such studies and demonstrates that nearly identical results are obtained using either method. nability of Amino Acids to Alter nsulin Binding-At least two mechanisms can account for enhancement of insulin sensitivity by amino acids, an increase in insulin binding or post-receptor regulation of insulin action. To discriminate between these possibilities, we assessed the effects of amino acids on insulin binding. As can be seen in Fig. 4, when adipocytes were suspended in 0 or 50% amino acids and then incubated at 16 "C with '251-insulin plus various concentrations of unlabeled insulin for 3 h, the resulting insulin binding competition curves were essentially superimposable (EDsO of 6 ng/ml). Thus, we conclude that amino acids do not influence the initial binding event. Similar results were obtained when cells were incubated with 150% amino acids (data not shown). nability of AminoAcidsto Alter nsulin Sensitivity or Responsiveness of the Glucose Transport System-To further localize the site of amino acid action, we assessed the specificity of amino acids by examining the insulin-responsive glucose transport system. On the leftpanel of Fig. 5, maximal insulin responsiveness of the glucose transport system is plotted as a function of amino acid concentration, whereas on the right panel, complete insulin dose-response curves are shown for cells treated with 0, 50, and 150% amino acids. t n Whole Cell Method TCA Method Yo 50 Yo 0 Yo 50 Yo Amino Acid Concentration (YO) FG. 3. Amino acid effects on insulin responsiveness and sensitivity: comparison of results assessing ['Hlleucine incorporation into protein using the whole cell method (Zeft pane0 or trichloroacetic acid (TCA) method (right panel). Cells were preincubated in the absence or presence of 50% amino acids for 30 min with various concentrations of insulin ( ng/ml). Open bars represent basal and solid bars depict maximally insulin-stimulated rates of [3H]leucine incorporation into protein. White values on black bars reflect the insulin ED5a values (calculated from complete insulin dose-response curves).

3 Amino Acid Regulation of nsulin Action K Amino Acids A50 X Amino Acids 8 J,.,.,.,.,., 1 a nsulin Concentration (ng / rnl) FG. 4. nability of amino acids to alter insulin binding. Cells were incubated for 30 min in the absence or presence of 50% amino acids and various concentrations of insulin. Specific '251-insulin binding was then assessed after a 3-h incubation at 16 "C. adipocytes with 25 ng/ml insulin for 30 min to maximally stimulate protein synthesis and then added amino acids (Fig. 6). When protein synthesis was measured, over a fixed 10- min period, throughout the 60-min time course, several interesting points emerged. First, it is apparent that amino acids are effective in increasing maximally insulin-stimulated rates of protein synthesis whether added before or after the addition of insulin. Second, as shown in Fig. 6, amino acid enhancement of the stimulatory effect of insulin on protein synthesis is relatively slow (tlh = 9.5 min) when compared to the rapidity of [3H]leucine uptake (<5 s, Ref. 6) and incorporation into protein (15-30 s, Ref. 6). To examine the question of reversibility, adipocytes were preincubated at 37 "C for 30 min in the absence or presence of 150% amino acids, washed twice, and then resuspended in the assay buffer as indicated in Fig. 7. After 30 min, complete insulin dose-response curves were obtained for each of the c nsulin Pretreated 1000 ~ %-Amino Acids 50 % '1 50 Yo Amino Acid Conc. (%) nsulin Conc. (ng ml) FG. 5. nability of amino acids to alter maximal insulin responsiveness or insulin sensitivity of the glucose transport system. A, cells were preincubated for 30 min with the indicated concentrations of amino acids in the absence (basal) or presence (maximal) of 25 ng/ml insulin. Glucose transport was then assessed during a 3-min assay. B, cells were preincubated for 30 min with 0, 50, or 150% amino acids plus various concentrations of insulin, after which time glucose uptake was measured. is clear from these results that amino acids do not influence the insulin-responsive glucose transport system. The inability of amino acids to alter the insulin-responsive glucose transport system is an important finding, since it indicates that the locus of amino acid action lies at a site downstream from the common signal pathway, within the insulin action-protein synthesis cascade. n other words, although insulin's pleiotropic actions are initiated by a common event (binding to cell surface receptors) adipocytes appear to possess the ability to specifically regulate the responsiveness and sensitivity of each effector system at individual postreceptor sites. t should also be noted that the inability of amino acids to alter the insulin sensitivity of glucose transport system is consistent with the finding that insulin binding capacity is unchanged by amino acids (Fig. 4). f amino acids had altered insulin binding capacity, then we would have seen changes in the insulin sensitivity of glucose transport. Kinetic Studies on the Onset and Reversibility of Amino Acid Action-Two important unanswered questions were: how quickly do amino acids enhance maximal insulin responsiveness and how quickly is this effect reversed upon removal of amino acids? To answer the first question, we pretreated 5 Control J m Amino Acid Treatment Time (Min) FG. 6. Time course of amino acid action on maximal insulin responsiveness. Adipocytes were incubated in the absence (basal) or presence of 25 ng/ml insulin for 30 min. After adding amino acids (100% final), [3H]leucine incorporation into protein was measured (10 min assay) at various times throughout the 60-min experiment. Open bars represent maximally insulin-stimulated (10 ng/ml) rates of protein synthesis. -.- C E Amino Acid Treatment 2 25od.. m nsulin Concentration (ng ml) FG. 7. Reversibility of amino acid action. Cells were preincubated at 37 "C for 30 min in the absence or presence of 150% amino acids, washed twice, and resuspended in buffer containing either 0 or 150% amino acids. Adipocytes were then incubated for 30 min with or without amino acids and the indicated concentrations of insulin. At the end of the 30-min incubation period, incorporation of [3H] leucine into protein was measured. From complete insulin doseresponse curves the insulin EDSO values for each group were calculated.

4 2040 Amino Acid Regulation of nsulin Action N~ Grouy 1 Brou~2!C Amino k Glu Ser Acids Ser G~Y lyr m: -~ m * Met Arg p&*' CY s* Thr His Val * FG. 8. Partial identification of regulatory amino acids. Cells were preincubated for 30 min at 37 "C in the absence (hatched bum) or presence of 25 ng/ml insulin (solid burs) plus the indicated number or amino acids (all at a concentration of 100%). The underlined amino acids are both glycogenic and ketogenic, whereas the remaining amino acids are exclusively glycogenic. Essential amino acids are indicated by an asterisk. Complete insulin dose curves were obtained for each of the four groups, and the insulin EDBo values were determined. These values are depicted above each bar and represent the mean k S.E. of several experiments (indicated in white at the bottom of each bar). Maximal rates of [3H]leucine incorporation are depicted by the total height of the bars (both solid and open). three groups. As can be seen, the stimulatory effect of amino acids on enhancing insulin responsiveness and sensitivity (upper curue) were almost completely reversed upon removal of amino acids (dashed curue). Partial dentification of Regulatory Amino Acids-n all of the preceding experiments, a buffer containing 14 individual amino acids was used. n the study shown Fig. in 8, we divided these amino acids into two groups of seven amino acids and arbitrarily named them groups 1 and 2. As can be seen, group 1 amino acids markedly increased maximal insulin responsiveness and insulin sensitivity (from an EDso of 0.54 ng/ml in amino acid-free buffer to 0.19 mg/ml), whereas the amino acids in group 2 had little or no effect on insulin action. Of the seven amino acids in group 1, we found that isoleucine and serine in combination were particularly effective in enhancing insulin sensitivity and responsiveness. Since the ability of group 1 and group 2 amino acids to alter insulin action differed so markedly, several conclusions can be drawn based on the characteristics of the amino acids in each group. First, based on the fact that the essential amino acids were almost equally divided between groups 1 and 2 (indicated by *), it appears unlikely that the ability of amino acids to enhance maximal insulin responsiveness is mediated by substrate availability. f this were the case, a maximal response would not be expected from only half of the essential amino acids. Second, it appears that amino acid action is not simply related to the glycogenic or ketogenic properties of the amino acids, since these amino acids are also equally divided between the two groups. Overall, we conclude that only selective amino acids exert a regulatory effect on insulin action, but we cannot at this time classify amino acids based on their metabolic characteristics. tors are tyrosine-specific protein kinases that can undergo ligand-induced autophosphorylation (7-12) and that a novel inositol-containing glycolipid may mediate many of insulin's biological actions (13), these are relatively early events in the insulin action cascade. Fewer studies have focused on the late events in the insulin action pathway, since investigating regulatory events downstream from insulin binding and release of insulin second messengers is inherently difficult. n this and the companion paper (), we have devised a strategy for assessing post-receptor regulation of insulin action at distal sites along the insulin action pathway. This was accomplished by comparing and contrasting two insulin effector systems in the same cell type under identical conditions. These studies have provided convincing evidence that additional control mechanisms are operative within the individual insulin-specific effector pathways. n the current study we have demonstrated that amino acids modulate insulin action by enhancing both maximal insulin responsiveness and insulin sensitivity of the protein synthesis system. When these findings are considered together with the results of the companion manuscript (1) on the selective ability of glucose to desensitize the insulinresponsive glucose transport system, a unique perspective relating to regulation of insulin action is obtained. This view is depicted schematically in Fig. 9. The most prominent feature of our proposed model is that the pleiotropic effects of insulin are mediated through divergent pathways that can be independently regulated and the metabolic substrates such as glucose and amino acids appear to play important roles in regulating insulin action. Moreover, the regulatory site of substrate action appears to be down-stream from the common signal pathway, within the specific effector cascade. Thus, glucose decreased insulin sensitivity and responsiveness of the glucose transport system without altering insulin's stimulatory effect on protein synthesis (l), whereas amino acids enhanced insulin sensitivity and responsiveness of the protein synthesis system without influencinglucose transport (Figs. 1,2, and 5). The identification of such regulatory mechanisms provides a reasonable physiological explanation to an important metabolic question. That is, how do cells regulate the ultimate responsiveness of each insulin effector system when all of insulin's actions are mediated by a common event, the binding of insulin to cell surface receptors? A likely answer is that metabolic substrates and cellular metabolites are intimately involved in the fine-tuning of the insu!in action cascade. Several important questions regarding substrate availability should be addressed in regard to basal and insulin-stimu- b+ij- Common.. & b &...-4 DSCUSSON An in-depth understanding of insulin action at the cellular and molecular level will require detailed knowledge of both early and late events in the insulin action pathway. Although recent and exciting studies have revealed that insulin recep- FG. 9. Proposed model for the regulation of insulin action by metabolic substrates.

5 Amino Regulation Acid of nsulin Action 2041 lated rates of protein synthesis. For example, how can adipocytes synthesize proteins when amino acids are omitted from the medium? The most likely answer to this question is that enough endogenous amino acids are generated during proteolysis to support protein synthesis during prolonged cells mediated by the rise in blood glucose levels following a meal. Less appreciated, however, has been the regulatory role of circulating amino acids. Among the studies that have shown a regulatory role for amino acids, two lines of investigation are particularly relevant to the current studies. First, tion times. Consequently, exogenous amino acids are not it is well established that certain amino acids such as leucine required to sustain optimal rates of protein synthesis. This view is supported by the finding that under basal conditions and arginine are potent secretagogues for the release of insulin from cells (18). Second, several studies have protein synthesis is linear from 0.5 to 6 h, both in the absence demonstrated that amino acids regulate the rate at which and presence of amino acids (data not shown) and that cellular protein is degraded in liver (4, 5, 19), muscle (2, 3, extracellular amino acids (from 0 to 150%) have little or no 20), and fibroblasts (21). Mechanistically, these events occur effect on the amount of [3H]leucine incorporated into protein primarily through inhibition of autophagy and reduced pro- (Fig. 1). teolysis. Moreover, these amino acid-regulated events precede Although amino acids appear not to be rate limiting under very rapidly, since the initial autophagic response rise occurs basal conditions, the intracellular concentration of amino acids could become rate limiting in insulin-treated cells such that uptake of amino acids from the medium is required to sustain maximal rates of protein synthesis. This potentially could explain the observed dose-dependent increase in maxi- mal insulin responsiveness when cells are incubated with amino acids. However, since we found that [3H]leucine incorporation into protein was linear in insulin-treated cells for up to 6 h in the absence of amino acids (l), this possibility seems unlikely. f amino acids were rate limiting, it is reasonable to assume that the full complement of amino acids would be required to sustain protein synthesis. This clearly was not observed, since we found that half of the full complement of amino acids were without effect on maximal insulin responsiveness, whereas the other half were fully active. Also, we observed that as few as 2 amino acids (isoleucine and serine) could produce 70-80% of the maximal effect, suggesting a ficity appeared to be more restrictive. Thus, in muscle, heart, and diaphragm, only leucine inhibited protein degradation (2, 3,20), whereas in fibroblasts phenylalanine and arginine were regulatory (21). selective regulatory effect of amino acids, rather than an effect Our current studies on isolated adipocytes have contributed on substrate availability. to this growing body of information by revealing a link be- A more likely site of amino acid action is within the insulin tween the biological actions of insulin on protein metabolism action cascade leading to enhancement of protein synthesis. and the ability of selective amino acids to modulate insulin This is supported by our finding that amino acids markedly responsiveness and sensitivity. Although the physiological increase insulin sensitivity. This effect on sensitivity is not significance of this interaction remains to be determined, our linked directly to changes in maximal insulin responsiveness, current hypothesis is that a regulatory triangle exists between since we found that amino acids significantly increased insulin dietary uptake of metabolic substrates, insulin release from sensitivity by 40% under conditions where insulin responsivethe endocrine pancreas, and regulation of insulin sensitivity ness was maximally elevated. Additionally, in preliminary and responsiveness at the level of peripheral target tissues, studies we found that amino acids profoundly shortened the such as adipose tissue. Thus, circulating levels of glucose and time course of insulin activation and that a subgroup of five amino acids would modulate cellular metabolism through two amino acids could selectively enhance insulin sensitivity by independent, but integrated, control systems. The first system 100% without affecting maximal insulin responsiveness. Tois the classical substrate-pancreatic axis, in which glucose and amino acids act as potent secretagogues for the release of gether, these data add credence to the idea that amino acids insulin. The second control system would act in tandem with modulate insulin action by influencing events within the the first, by enabling insulin target tissues to continuously insulin action cascade. Although no evidence is presented pertaining to the mechanism of amino acid action, it is known that certain amino acids can allosterically regulate enzymatic activity within cells. For example, the binding of isoleucine to threonine deaminase causes a confirmation change that inhibits enzymatic activity (14). Since it is widely held that insulin enhances protein synthesis through a cascade of phosphorylation events terminating with phosphorylation of ribosomal S6 (15-17), it is at least conceivable that isoleucine could modify one or more of the substrates, kinases, or phosphatases involved in the insulin action-phosphorylation cascade. The role of metabolic substrates such as glucose and amino acids in regulating insulin action at the level of the target cell is only now becoming evident; however, the concept that metabolic substrates can initiate a carefully orchestrated sequence of events that culminate in profound effects on cellular metabolism has been known for many years. Perhaps the best known example is the rapid release of insulin from pancreatic almost immediately after removal of amino acids and is followed shortly thereafter by a maximal increase in proteolysis (4). Particularly interesting and analogous to our own data is the finding that only certain amino acids regulate protein degradation rates. For example, Poso et al. (4) demonstrated that in perfused rat liver a combination of 12 amino acids exhibited no suppressive activity on protein degradation, whereas the other amino acids (Leu, Phe, Tyr, Gln, Pro, His, Trp, and Met) were inhibitory. Similar studies using primary cultures of rat hepatocytes revealed that a major portion of this inhibitory action was due to the presence of methionine, phenylalanine, and tryptophan (5). n other cell types, speci- monitor circulating levels of glucose and amino acids to modulate cellular responsiveness at distinct post-receptor sites within the insulin action pathways. Although we believe that this tandem control system coordinates the disposition of metabolic substrates during the daily fluctuations of substrates associated with dietary uptake, the regulatory mechanisms underlying the actions of glucose and amino on insulin action may also have important implications relating to the etiology and amelioration of certain insulin-resistant states such as Type 1 diabetes mellitus and obesity. Acknowledgment-We gratefully acknowledge the editorial assistance of Henri Kruse. REFERENCES 1. Marshall, S. (1989) J. Bid. Chem. 264, Buse, M. G., and Reid, S. S. (1975) J. Clin. fnuest. 56, Chua, B., Siehl, D. L., and Moran, H. E. (1979) J. Biol. Chem. 254,

6 2042 Amino Regulation Acid 4. Poso, A. R., Wert, J. J., Jr., and Mortimore, G. E. (1982) J. Biol. Chem. 257, Sommercorn, J. M., and Swick, R. W. (1981) J. Biol. Chem. 256, Minemura, T., Lacy, W. W., and Crofford, 0. B. (1970) J. Biol. Chem. 245, Kasuga, M., Karlsson, F. A. and Kahn, C. R. (1981) Science 215, Kasuga, M., Zick, Y., Blith, D. L., Karlsson, F. A., Haring, H. U., and Kahn, C. R. (1982) J. Biol. Chem. 257, Avruch, J., Nemenoff, R. A., Blackshear, P. J., Pierce, M. W., and Osathanondh, R. (1982) J. Biol. Chem. 257, Zick, Y., Kasuga, M., Kahn, C. R., and Roth, J. (1983) J. Biol. Chem. 258, Roth, R., and Cassell, D. J. (1983) Science 219, Yu, K-T., and Czech, M. P. (1984) J. Biol.Chem. 259, of nsulin Action 13. Low, M. G., and Saltiel, A. R. (1987) Science 239, Darnell, J., Lodish, H., and Baltimore, D. (1986) in Molecular Cell Biology, pp , Scientific American Books, New York 15. Smith, C. J., Rubin, C. S., and Rosen, 0. M. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, Hansson, A., and ngelman-sundberg, M. (1985) Eur. J. Biochern. 161, Trevillyan, J. M., Perisic, O., Traugh, J. A., and Byus, C. V. (1985) J. Biol. Chen. 260, Ganong, W. F. (1987) in Reuiew of Medical Physiology, pp , Appleton and Lange, Norwalk, CT 19. Hopgood, M. F., Clark, M. G., and Ballard, J. F. (1977) Bioehern. J. 164, Fulks, R. M., Li, J. B., and Goldberg, A. L. (1975) J. Bwl. Chern. 250, Bradley, M. 0. (1977) J. Biol. Chem. 252,