Insulin controls the cyclic AMP-dependent phosphorylation of integral and peripheral proteins associated with the rat liver plasma membrane

Ablation of calcineurin Aβ reveals hyperlipidemia and signaling cross-talks with phosphodiesterases

Insulin resistance, hyperlipidemia, and cardiovascular complications are common dysregulations of metabolic syndrome. Transplant patients treated with immunosuppressant drugs such as cyclosporine A (CsA), an inhibitor of calcineurin phosphatase, frequently develop similar metabolic complications. Although calcineurin is known to mediate insulin sensitivity by regulating β-cell growth and adipokine gene transcription, its role in lipid homeostasis is poorly understood. Here, we examined lipid homeostasis in mice lacking calcineurin Aβ (CnAβ(-/-)). We show that mice lacking calcineurin Aβ are hyperlipidemic and develop age-dependent insulin resistance. Hyperlipidemia found in CnAβ(-/-) mice is, in part, due to increased lipolysis in adipose tissues, a process mediated by β-adrenergic G-protein-coupled receptor signaling pathways. CnAβ(-/-) mice also exhibit additional pathophysiological phenotypes caused by the potentiated GPCR signaling pathways. A cell autonomous mechanism with sustained cAMP/PKA activation is found in CnAβ(-/-) mice or upon CsA treatment to inhibit calcineurin. Increased PKA activation and cAMP accumulation in CnAβ(-/-) mice, however, are sensitive to phosphodiesterase inhibitor. Indeed, we show that calcineurin regulates degradation of phosphodiesterase 3B, in addition to phosphodiesterase 4D. These results establish a role for calcineurin in lipid homeostasis. These data also indicate that potentiated cAMP signaling pathway may provide an alternative molecular pathogenesis for the metabolic complications elicited by CsA in transplant patients.

Sub-family selective actions in the ability of Erk2 MAP kinase to phosphorylate and regulate the activity of PDE4 cyclic AMP-specific phosphodiesterases

Expressed in intact cells and in vitro, PDE4B and PDE4C isoenzymes of cyclic nucleotide phosphodiesterase (PDE), in common with PDE4D isoenzymes, are shown to provide substrates for C-terminal catalytic unit phosphorylation by the extracellular signal-regulated kinase Erk2 (p42(MAPK)). In contrast, PDE4A isoenzymes do not provide substrates for C-terminal catalytic unit phosphorylation by Erk2. Mutant PDE4 enzymes were generated to show that Erk2 phosphorylation occurs at a single, cognate serine residue located within the C-terminal portion of the PDE4 catalytic unit. PDE4 long-form isoenzymes were markedly inhibited by Erk2 phosphorylation. The short-form PDE4B2 isoenzyme was activated by Erk2 phosphorylation. These functional changes in PDE activity were mimicked by mutation of the target serine for Erk2 phosphorylation to the negatively charged amino acid, aspartic acid. Epidermal growth factor (EGF) challenge caused diametrically opposed changes in cyclic AMP levels in COS1 cells transfected to express the long PDE4B1 isoenzyme compared to cells expressing the short PDE4B2 isoenzyme. We suggest that PDE4 enzymes may provide a pivotal point for integrating cyclic AMP and Erk signal transduction in cells with 4 genes encoding enzymes that are either insensitive to Erk2 action or may either be activated or inhibited. This indicates that PDE4 isoenzymes have distinct functional roles, giving credence to the notion that distinct therapeutic benefits may accrue using either PDE4 subfamily or isoenzyme-selective inhibitors.

Co-transfection with protein kinase D confers phorbol-ester-mediated inhibition on glucagon-stimulated cAMP accumulation in COS cells transfected to overexpress glucagon receptors

Glucagon elicited a profound increase in the intracellular cAMP concentration of COS-7 cells which had been transiently transfected with a cDNA encoding the rat glucagon receptor and under conditions where cAMP phosphodiesterase activity was fully inhibited. This was achieved in a dose-dependent fashion with an EC50 of 1.8+/-0.4 nM glucagon. In contrast with previous observations made using hepatocytes [Heyworth, Whetton, Kinsella and Houslay (1984) FEBS Lett. 170, 38-42], treatment of transfected COS-7 cells with PMA did not inhibit the ability of glucagon to increase intracellular cAMP levels. PMA-mediated inhibition was not conferred by treatment with okadaic acid, nor by co-transfecting cells with cDNAs encoding various protein kinase C isoforms (PKC-alpha, PKC-betaII and PKC-epsilon) or with the PMA-activated G-protein-receptor kinases GRK2 and GRK3. In contrast, PMA induced the marked inhibition of glucagon-stimulated cAMP production in COS-7 cells that had been co-transfected with a cDNA encoding protein kinase D (PKD). Such inhibition was not due to an action on the catalytic unit of adenylate cyclase, as forskolin-stimulated cAMP production was unchanged by PMA treatment of COS cells that had been co-transfected with both the glucagon receptor and PKD. PKD transcripts were detected in RNA isolated from hepatocytes but not from COS-7 cells. Transcripts for GRK2 were present in hepatocytes but not in COS cells, whereas transcripts for GRK3 were not found in either cell type. It is suggested that PKD may play a role in the regulation of glucagon-stimulated adenylate cyclase.

Receptor-mediated stimulation of lipid signalling pathways in CHO cells elicits the rapid transient induction of the PDE1B isoform of Ca2+/calmodulin-stimulated cAMP phosphodiesterase

Chinese hamster ovary cells (CHO cells) do not exhibit any Ca2+/calmodulin-stimulated cAMP phosphodiesterase (PDE1) activity. Challenge of CHO cells with agonists for endogenous P2-purinoceptors, lysophosphatidic acid receptors and thrombin receptors caused a similar rapid transient induction of PDE1 activity in each instance. This was also evident on noradrenaline challenge of a cloned CHO cell line transfected so as to overexpress alpha 1B-adrenoceptors. This novel PDE1 activity appeared within about 15 min of exposure to ligands, rose to a maximum value within 30 min to 1 h and then rapidly decreased. In each case, the expression of novel PDE1 activity was blocked by the transcriptional inhibitor actinomycin D. Challenge with insulin of either native CHO cells or a CHO cell line transfected so as to overexpress the human insulin receptor failed to induce PDE1 activity. Reverse transcriptase-PCR analyses, using degenerate primers able to detect the PDE1C isoform, did not amplify any fragment from RNA preparations of CHO cells expressing PDE1 activity, although they did so from the human thyroid carcinoma FTC133 cell line. Reverse transcriptase-PCR analyses, using degenerate primers able to detect the PDE1A and PDE1B isoforms, successfully amplified a fragment of the predicted size from RNA preparations of both CHO cells expressing PDE1 activity and human Jurkat T-cells. Sequencing of the PCR products, generated using the PDE1A/B primers, yielded a novel sequence which, by analogy with sequences reported for bovine and murine PDE1B forms, suggests that the PDE1 species induced in CHO cells through protein kinase C activation and that expressed in Jurkat T-cells are PDE1B forms.

Activation of adenylate cyclase in human platelet membranes by guanosine 5'-[beta gamma-imido]triphosphate is inhibited by cyclic-AMP-dependent phosphorylation. Slow activation occurs in the absence of ATP

Incubation of platelet membranes with guanosine 5'-[beta gamma-imido]triphosphate causes a slow increase in GS (stimulatory GTP-binding protein) activation of adenylate cyclase. Mg2+ is necessary for this slow activation. This process is inhibited in the presence of ATP, and inhibition is greater if cyclic AMP is also included in the incubation. Adenosine 5'-[beta gamma-imido]triphosphate instead of ATP in the incubation facilitates the slow activation in the presence of cyclic AMP, and incubation of membranes with cyclic-AMP-dependent protein kinase inhibitor decreased inhibition of the slow activation of adenylate cyclase by ATP and cyclic AMP. A protein of 45 kDa in platelet membranes is phosphorylated in a cyclic-AMP-dependent manner. The transition from a reversibly activated form of GS to an irreversibly activated form is substantially slower in the presence of ATP and cyclic AMP. We propose that guanosine 5'-[beta gamma-imido]triphosphate-activated GS may exist in phosphorylated or non-phosphorylated forms, and that the non-phosphorylated form is the more active of the two species. The non-phosphorylated form of GS may correspond to the irreversibly activated state.

Evidence that insulin and guanosine triphosphate regulate dephosphorylation of the beta-subunit of the insulin receptor in sarcolemma membranes isolated from skeletal muscle

When sarcolemma membranes isolated from rat skeletal muscle were incubated with [gamma-32P]ATP, a membrane protein of apparent Mr 95,000 was rapidly phosphorylated, with the 32P content reaching a maximum within 2 s. On the basis of immunoprecipitation with anti-insulin-receptor antiserum, phosphoamino acid analysis and Mr, this protein probably represents the beta-subunit of the insulin receptor. Similarly, on incubation of the membrane with adenosine 5'-[gamma-[35S]thio] triphosphate the 95 kDa protein was thiophosphorylated, indicating thiophosphorylation of the beta-subunit of the insulin receptor on the basis of immunoprecipitation studies. The effect of insulin on the phosphorylation of this protein in the membrane was studied. Insulin induced a 20% decrease in the 32P labelling of the protein when the membranes were phosphorylated for 10 s. This insulin effect was dose-dependent, with half-maximal effect obtained at 2-3 nM-insulin. Addition of GTP, but not GDP or guanosine 5'-[beta, gamma-imido]triphosphate, enhanced the effect to 35% inhibition, with half-maximal effect of GTP obtained at 0.5 microM. GTP had no effect on the phosphorylation of the protein in the absence of insulin. Analysis of this insulin effect showed that insulin increased the rate of dephosphorylation of the 95 kDa protein in the membrane. In contrast, insulin had no effect on thiophosphorylation of the 95 kDa membrane protein after incubation with adenosine 5'-[gamma-[35S]thio]triphosphate. Since thiophosphorylated proteins are less sensitive to phosphatase action, these investigations suggest that insulin stimulated a protein phosphatase activity in a GTP-dependent manner. The possibility that GTP-regulatory proteins are involved in the action of insulin on the phosphorylation of the insulin receptor and other membrane proteins is discussed.

Insulin effect on translational diffusion of lipids and proteins in the plasma membrane of isolated rat hepatocytes

The effects of insulin (10(-10)-10(-8) mol/l) on lateral diffusion of three fluorescent lipid probes, 1-acyl-2-(N-4-nitrobenzo-2-oxa-1,3-diazole)aminocaproyl phosphatidylcholine (NBD-PC), 5-(N-hexadecanoyl)aminofluorescein (F-C16), 5-(N-dodecanoyl)aminofluorescein (F-C12), and of fluorescein isothiocyanate-labeled proteins in the plasma membrane of intact rat hepatocytes were studied by the technique of fluorescence recovery after photobleaching. The absolute lateral diffusion coefficients of the lipid analogues NBD-PC, F-C16 and F-C12 at 21 degrees C were 2.5 X 10(-9) cm2/s, 5.4 X 10(-9) cm2/s and 19 X 10(-9) cm2/s, respectively. The diffusion coefficient mean of proteins labeled with fluorescein isothiocyanate was 6.4 X 10(-10) cm2/s. Insulin at 10(-9) and 10(-8) mol/l reduced the lateral diffusion coefficient for F-C12- and F-C16-labeled cells by 20% and for NBD-PC-labeled cells by 30% (P less than 0.025). The insulin effect was specific as tested by cell incubation with proinsulin and desoctapeptide insulin (10(-8) mol/l) and was detectable after 7 min of insulin preincubation. In contrast to lateral diffusion of lipid probes, lateral mobility of unselected membrane proteins was not altered by insulin. The observed modulation of lipid dynamics in the plasma membrane of intact hepatocytes, by which a variety of membrane functions can be influenced, may be an important step in the mechanism of insulin action.

Insulin-stimulated high affinity cyclic AMP phosphodiesterase in rat mammary acini

High affinity cyclic AMP phosphodiesterase activity in preparations of acini isolated from mammary tissue of lactating rats is shown to be stimulated by the addition of physiological concentrations of insulin to incubations of acini in vitro. This effect is expressed specifically on membrane-associated phosphodiesterase and occurs in the absence of concurrent protein synthesis. The possible functional role of this aspect of insulin's action on mammary tissue is discussed and compared with the well-known reversal by this hormone of the effects of lipolytic agents in adipose tissue and liver.

Calcium-dependence of insulin receptor phosphorylation

Phosphorylation of the insulin receptor of isolated rat adipocytes in response to insulin has been studied. Immunoprecipitation of adipocyte membrane protein demonstrated increased incorporation of 32P after exposure to insulin for 15 min, but this was dependent on the presence of physiological concentrations of Ca2+ and Mg2+. Autoradiography of solubilized immunoprecipitated membrane protein after sodium dodecyl sulphate/polyacrylamide-gel electrophoresis revealed that most of the 32P incorporation occurred in a band corresponding to Mr 95 000, which has been identified previously as the beta-subunit of the insulin receptor. 32P incorporation was inhibited by 2,4-dinitrophenol and trifluoperazine. It is suggested that insulin-receptor phosphorylation is an energy-requiring process that is Ca2+-dependent and may be modulated by calmodulin. Phosphorylation may proceed independently of glucose transport.

Insulin and glucagon regulate the activation of two distinct membrane-bound cyclic AMP phosphodiesterases in hepatocytes

Glucagon (10 nM) caused a transient elevation of intracellular cyclic AMP concentrations, which reached a peak in around 5 min, and slowly returned to basal values in around 30 min. When 1 mM-3-isobutyl-1-methylxanthine (IBMX) was present, this process yielded a Ka of 1 nM for glucagon. The addition of insulin (10 nM) after 5 min exposure to glucagon (10 nM) caused intracellular cyclic AMP concentrations to fall dramatically, attaining basal values within 10 min. The regulation of this process was dose-dependent, exhibiting a Ka of 0.4 nM for insulin. If insulin and glucagon were added together to hepatocytes, then insulin decreased the magnitude of the cyclic AMP response to glucagon. IBMX (1 mM) prevented insulin antagonizing the action of glucagon in both of these instances. A gentle homogenization procedure followed by a rapid subcellular fractionation of hepatocytes on a Percoll gradient was developed. This was used to resolve subcellular membrane fractions and to identify cyclic AMP phosphodiesterase activity in both membrane and cytosol fractions. Glucagon and insulin only affected the activity of two distinct membrane-bound species, a plasma-membrane enzyme and a 'dense vesicle' enzyme. Glucagon (10 nM), insulin (10 nM), IBMX (1 mM), dibutyryl cyclic AMP (10 microM) and cholera toxin (1 microgram/ml) all elicited the activation of the 'dense vesicle' enzyme. The plasma-membrane enzyme was not activated by glucagon, IBMX or dibutyryl cyclic AMP, although insulin and cholera toxin both led to its activation. The degree of activation of the plasma-membrane enzyme produced by insulin was increased in the presence of IBMX or dibutyryl cyclic AMP. Glucagon pretreatment (5 min) of hepatocytes blocked the ability of insulin to activate the plasma-membrane enzyme. The activity state of these phosphodiesterases is discussed in relation to the observed changes in intracellular cyclic AMP concentrations. It is suggested that insulin exerts its action on the plasma-membrane phosphodiesterase through a mechanism involving a guanine nucleotide-regulatory protein.

Guanine nucleotides can activate the insulin-stimulated phosphodiesterase in liver plasma membranes

The insulin-stimulated cyclic AMP phosphodiesterase from liver plasma membranes is shown to be activated upon incubation with guanine nucleotides in the presence of ATP. The non-hydrolysable analogue of ATP, adenylyl imidodiphosphate failed to substitute for ATP in achieving activation. GTP, its non-hydrolysable analogues p[NH]ppG and GTP-gamma-S, as well as GDP, all elicited activation. It is suggested that guanine nucleotides, and probably insulin, exert their effect on this enzyme through a distinct species of guanine nucleotide regulatory protein.

The short term hormonal control of cytoplasmic protein phosphorylation in hepatocytes from fed rats

Exposure of 32P-prelabelled isolated hepatocytes to vasopressin affected the phosphorylation of nine of the 26 phosphoproteins resolved by sodium dodecyl sulphate gel electrophoresis. Glucagon (2 nM) or cyclic AMP elicited significant changes in the phosphorylation of only four phosphoproteins. A very high concentration of glucagon (1000 nM) affected additional phosphoproteins. Insulin alone significantly increased the phosphorylation of a single protein. Vasopressin, A23187, glucagon and cyclic AMP all induced the dephosphorylation of a single phosphoprotein of mol. wt 20,000. The significance of these results with respect to the short-term control of hepatic metabolism is discussed.

Characterization of the phosphorylated form of the insulin-stimulated cyclic AMP phosphodiesterase from rat liver plasma membranes

Incubation of intact purified rat liver plasma membranes with insulin, cyclic AMP and ATP led to the activation of the peripheral "low-Km" cyclic AMP phosphodiesterase. When (gamma-32P]ATP was included in the incubation mixture, after purification of this enzyme to homogeneity it was found to contain 1 mol of alkali-labile 32P/mol of enzyme. Treatment of the homogeneous phosphorylated enzyme with alkaline phosphatase released all of the 32P from the protein while restoring its activity to the native state. The reversibility of the activation that is achieved by the phosphorylation of this enzyme could also be demonstrated with a high-speed supernatant from rat liver. This restored the activity of the activated membrane-bound enzyme to its native state. The Ka for the cyclic AMP-dependence of this process (1.6 micrometer) was unaffected by a range of ATP concentrations (1-10 mM) and by a range of membrane protein concentrations (0.2-2 mg/ml). Adenylyl imidodiphosphate could not substitute for ATP, and concanavalin A could not substitute for insulin, as essential ligands in the activation process. The purified activated enzyme exhibited Km 0.6 microM, Vmax 10.9 units/mg of protein and Hill coefficient (h) 0.47. The Vmax. for this activated enzyme was much higher than that of the native enzyme, yet h was much lower.

Purification and properties of the insulin-stimulated cyclic AMP phosphodiesterase from rat liver plasma membranes

The peripheral high-affinity cyclic AMP phosphodiesterase from rat liver plasma membranes was purified to apparent homogeneity. The procedure used involved the initial purification of liver plasma membranes and the solubilization of the enzyme by using a high-ionic-strength medium. This was followed by chromatography of the enzyme on DEAE-cellulose, Affi-Gel Blue, a novel affinity column and Sephadex G-100. A 9500-fold purification of the enzyme with a 24% yield was achieved by this procedure. The purified enzyme was apparently monomeric (Mr 52000) as it exhibited identical molecular weights on analysis by gel filtration, sedimentation and sodium dodecyl sulphate/polyacrylamide-gel electrophoresis. It is suggested that the non-Michaelis kinetics exhibited by the enzyme are due to it obeying a mnemonical mechanism, where it displays Km 0.7 micrometer, Vmax. 9.1 units/mg of protein and Hill coefficient (h) 0.62. Cyclic GMP acts as a poor substrate for the enzyme, with Km 120 micrometer and Vmax. 0.4 unit/mg of protein, and also as an inhibitor of the enzyme, with I50 (concentration giving 50% inhibition) 150 micrometer when assayed at 0.4 micrometer-cyclic AMP. Inhibition by 5'-AMP is unlikely to be of physiological importance, as it is only a weak inhibitor of the enzyme (I50 47 mM assayed at 0.4 micrometer-cyclic AMP).