Effects of epinephrine and insulin on phosphopeptide metabolism in adipocytes.

Isolated adipocytes, incubated in the presence of extracellular 32Pi to steady state 32P incorporation into cellular phosphopeptides, were exposed to hormones for 5 min. Epinephrine (10(-6) M) stimulated 32P incorporation into at least 12 major phosphopeptides, distributed in the cytoplasm, endoplasmic reticulum, and plasma membrane. Quantitatively pre-eminent among these were peptides of molecular weight 123,000 and 69,000, each located both in the cytoplasm and endoplasmic reticulum. The effect of epinephrine (10(-7) M) on 32P incorporation into these two peptides was augmented by theophylline (10(-3) M) in a synergistic fashion. Norepinephrine, dibutyryl N6,O2'-dibutyryl adenosine 3':5'-monophosphate, adrenocorticotropic hormone (ACTH) (synthetic 1 to 24 fragment), and glucagon mimicked the effect of epinephrine. Insulin modified adipocyte peptide phosphorylation in two ways. When present as the sole hormone, insulin (100 microunits/ml) consistently and selectively stimulated the 32P incorporation into a peptide of molecular weight 123,000 (endoplasmic reticulum, cytoplasm) without significant alteration in the 32P content of any other major peptide. A second effect of insulin was evident when epinephrine (10(-6) M) was present simultaneously. Insulin significantly inhibited the epinephrine-stimulated phosphorylation of the molecular weight 69,000 (endoplasmic reticulum, cytoplasm) and 26,000 (plasma membrane) peptides. Nevertheless, persistence of insulin-stimulated phosphorylation of the 123,000 peptide in the presence of epinephrine was shown by a 32P content of this peptide that was greater in the presence of both hormones than with either individually. These findings indicate that in intact adipocytes: (a) epinephrine acutely alters the phosphorylation of a large number of adipocyte peptides, partly at least, via activation of adenosine 3':5'-monophosphate (cyclic AMP)-dependent protein kinase; (b) insulin opposes several epinephrine-stimulated phosphorylations in a manner consitent with its ability to lower epinephrine-stimulated intracellular cyclic AMP accumulation in adipocytes; and (c) insulin, in addition, exerts a unique stimulatory effect on adipocyte peptide phosphorylation that is independent of its effects on cyclic AMP metabolism and may be medicated by the generation of an as yet undefined intracellular "messenger" unique to insulin.

Isolated adipocytes, incubated in the presence of extracellular s*P, to steady state 3zP incorporation into cellular phosphopeptides, were exposed to hormones for 5 min. Epinephrine (10-O M) stimulated a*P incorporation into at least 12 major phosphopeptides, distributed in the cytoplasm, endoplasmic reticulum, and plasma membrane. Quantitatively pre-eminent among these were peptides of molecular weight 123,000 and 69,000, each located both in the cytoplasm and endoplasmic reticulum. The effect of epinephrine (lo-' M) on szP incorporation into these two peptides was augmented by theophylline ( 10ms M) in a synergistic fashion. Norepinephrine, dibutyryl W,O*'-dibutyryl adenosine 3':5'-monophosphate, adrenocorticotropic hormone (ACTH) (synthetic 1 to 24 fragment), and glucagon mimicked the effect of epinephrine.
Insulin modified adipocyte peptide phosphorylation in two ways. When present as the sole hormone, insulin (100 microunits/ml) consistently and selectively stimulated the szP incorporation into a peptide of molecular weight 123,000 (endoplasmic reticulum, cytoplasm) without significant alteration in the s*P content of any other major peptide. A second .effect of insulin was evident when epinephrine (10-O M) was present simultaneously.
Ins&n significantly inhibited the epinephrine-stimulated phosphorylation of the molecular weight 69,000 (endoplasmic reticulum, cytoplasm) and 26,000 (plasma membrane) peptides. Nevertheless, persistence of insulin-stimulated phosphorylation of the 123,000 peptide in the presence of epinephrine was shown by a *lP content of this peptide that was greater in the presence of both hormones than with either individually.
These findings indicate that in intact adipocytes: (a) epinephrine acutely alters the phosphorylation of a large number of adipocyte peptides, partly at least, via activation of adenosine 3':5'-monophosphate (cyclic AMP)-dependent protein kinase; (b) insulin opposes several epinephrine-stimulated phosphorylations in a manner consistent with its ability to lower epinephrine-stimulated intracellular cyclic AMP accumulation in adipocytes; and (c) insulin, in addition, exerts a unique stimulatory effect on adipocyte peptide phosphorylation that is independent of its effects on cyclic AMP metabolism and may be mediated by the generation of an as yet undefined intracellular "messenger" unique to insulin.
In an accompanying report (1) we described the major phosphopeptides generated by isolated, intact adipocytes incubated with szP, and catalogued them by molecular weight and subcellular distribution. Steady-state szP incorporation into adipocyte phosphopeptides was achieved by a P-hour incubation under the conditions described. This report will describe the perturbations of 3*P labeling of phosphopeptides from the steady state by brief exposure of *zP-labeled adipocytes to insulin, as well as to agents known to increase intracellular cyclic AMP.' These studies define the major substrates of cyclic AMP-mediated protein phosphorylation in the intact

RESULTS
When fat cells labeled in the presence of NaH,azPO, for 2 hours were exposed to epinephrine for an additional 5 min, a large number of phosphopeptides exhibited significantly enhanced incorporation of r*P ( Table  I). Phosphopeptides of molecular weight 123,000 and 69,000 exhibited the greatest increase in **P content. These were distributed in the cytosol and endoplasmic reticulum.
In addition, at least seven other phosphopeptides in the homogenate showed significantly enhanced phosphorylation; phosphopeptides with electrophoretic mobility essentially identical with the species observed in the homogenate were detected in one or more subcellular fractions, and these also, exhibited enhanced **P incorporation in response to epinephrine (Table I). These include over one-half of the major phosphopeptides in the endoplasmic reticulum and cytosol, as well as a major plasma membrane phosphopeptide of molecular weight 26,000. The s*P content of the major phosphopeptides of the mitochondria was not significantly altered. (Amounts of nuclei recovered in these experiments were inadequate to allow systematic analysis of this fraction.) The magnitude of the increase in phosphorylation was dependent on the dose of epinephrine and was clearly evident at lo-' M (Table II). In the presence of submaximal concentrations of both epinephrine (10-l M) and theophylline (lOeJ M), the increase in phosphorylation of 69,000-and 123,000-dalton peptides was greater than the sum of the increments due to either agent alone (Table II). Exposure of adipocytes to norepineph-  The effect of insulin on the phosphorylation of the 123,000 peptide appeared to be dose-dependent and was detectable with 5 microunits/ml of insulin, the lowest concentration tested (data not shown). The insulin-induced increase in phosphorylation apparent after 5 min persisted unaltered for 15 min (data not shown).
A second effect of insulin on the phosphorylation of adipocyte peptides was observed in the presence of epinephrine. When -'*P-labeled adipocytes were exposed to insulin (100 microunits/ml) and epinephrine (1 pM) simultaneously for 5 min, the epinephrine-stimulated phosphorylation of the 69,000-dalton peptide was partially inhibited by insulin, as was the epinephrine-stimulated phosphorylation of the plasma membrane phosphopeptide of molecular weight 26,000 (Table  IV). In marked contrast, the phosphorylation of the 123,000dalton peptide, which was augmented both by insulin and epinephrine individually, exhibited stimulation of r2P incorporation when both hormones were present simultaneously that significantly exceeded that induced by either hormone alone and was statistically indistinguishable from the sum of the increments induced by each agent alone (Table IV). DISCUSSION The results reported clearly demonstrate that it is possible to detect acute hormonally induced changes in protein phosphorylation in intact cells. The rapidity and selectivity of these changes strongly support the concept that they reflect alterations in the activity of protein kinases and protein phosphatases, or both, rather than alternative mechanisms, such as changes in '*P incorporation into mononucleotides, intracellular ATP levels, or protein synthesis and degradation (5,6). Three types of evidence indicate that the enhanced phosphorylation of at least the peptides of molecular weight 123,000 and 69,000 caused by epinephrine was mediated by increased intracellular cyclic AMP and stimulation of the cytoplasmic cyclic AMP-dependent protein kinase (7). First, qualitatively similar effects are reproduced by a variety of agents known to increase intracellular cyclic AMP (8,9); second, the stimulation of *?P incorporation obtained in the presence of submaximal concentrations of both epinephrine and theophylline is synergistic (IO); and finally, the observation that endogenous peptides of identical molecular weight are substrates in vitro for the endogenous cyclic AMP-dependent protein kinase. These arguments cannot at present be extended rigorously to the remainder of peptides showing epinephrine-stimulated phosphorylation in intact adipocytes. Nevertheless, the observation that epinephrine enhances the phosphorylation of a large number of adipocyte peptides is not due to analytic bias, as indicated by the absolute selectivity of insulin-stimulated phosphorylation in parallel experiments and is compatible with current concepts of the mode of action of epinephrine.
In fact, since our analysis is limited to the most prominent phosphopeptides, we may have overlooked entirely quantitatively minor phosphorylations of great functional significance. Furthermore, it is likely that our data underestimate the magnitude of the changes occurring in the intact cell. Soderling et al. (11) reported that the basal activity ratio of the cyclic AMP-dependent protein kinase in fat cells was 0.4, as opposed to 0.15 to 0.2 for fat pads. Thus, the preparation of isolated fat cells may itself engender a slight activation of this enzyme and diminish the apparent stimulation by hormone. In addition, the analytic techniques employed (i.e. polyacrylamide gel electrophoresis in sodium dodecyl sulfate) do not allow absolute resolution of epinephrine-responsive phosphopeptides from adjacent or superimposed unresponsive phosphopeptide moieties. Finally, and most important, the washing procedure employed in freeing cells from the medium after exposure to hormones requires 4 to 5 min, and this may be sufficient to allow the partial or complete decay of certain hormonally induced changes.
In summary, we have demonstrated that epinephrine stimulates the phosphorylation of a large number of peptides in the 123,000-dalton peptide is much less apparent. The ability of insulin alone to promote this phosphorylation suggests that cyclic AMP is not involved. Furthermore, the additive effects of epinephrine and insulin (present simultaneo&ly) on the phosphorylation of the 123,000-dalton peptide point to the continued expression of both the epinephrine-induced, cyclic AMP-stimulated phosphorylation and the insulin-stimulated phosphorylation, both occurring in the same subcellular fraction. This finding indicates strongly that the ability of insulin to stimula,te the phosphorylation of the 123,000-dalton peptide persist; in the presence of elevated intracellular cyclic AMP within the same intracellular compartment and is thus independent of insulin's effect on cyclic AMP metabolism. MdWd The intracellular localization of the 123,000-dalton peptide suggests by analogy that insulin may generate an as yet unidentified "second messenger" that modifies the activity of a protein kinase, protein phosphatase, or both. The list of potential candidates is of course quite large, and further speculation is not warranted. The existence of a "second messenger" unique to insulin was previously suggested from studies in liver (12) and muscle (13 In support of the present observations, Benjamin and Singer (14,15) recently reported that epinephrine stimulated the phosphorylation of an adipocyte peptide of molecular weight 60,000 to 65,000, an effect partly inhibited by insulin, whereas insulin alone stimulated the phosphorylation of a peptide of molecular weight 140,000. Taken together, these findings clearly document an action of insulin that is mediated independently of cyclic AMP (and thus analogous to certain other insulin effects, e.g. on glucose transport, pyruvate dehydrogenase activity, etc.) and that appears to proceed concurrently with actions mediated by a lowering of intracellular cyclic AMP. Certain effects of both classes are expressed at the level of protein phosphorylation. It appears that a fuller understanding of the mechanisms by which insulin regulates protein phosphorylation may yield insights of general' significance for the mechanism of this hormone's action.