Direct comparison of inositol phosphoglycan with prostaglandylinositol cyclic phosphate, two potential mediators of insulin action.

Though insulin signalling is thought by many groups to function without second messenger action, others have provided evidence for the existence and action of such regulators. Chemically quite different compounds, however, have been proposed as mediators, such as various inositol phosphoglycans and prostaglandylinositol cyclic phosphate (cyclic PIP). In spite of marked structural differences, these compounds are reported to have the same regulatory properties, i.e. to activate protein ser/thr phosphatases and to inhibit protein kinase A. In order to clarify this discrepancy, the regulatory potency of these different compounds was assayed under identical conditions. It was found that in contrast to cyclic PIP, the synthetic inositol phosphoglycan PIG41 neither directly inhibited protein kinase A nor activated protein ser/thr phosphatases. However, when added to intact cells, such as primary adipocytes, PIG41 inhibited isoproterenol-stimulated lipolysis. This effect most likely results from tyrosine phosphorylation of insulin receptor substrates (IRSs) by PIG41. This tyrosine phosphorylation is not carried out by the insulin receptor tyrosine kinase but by cytosolic tyrosine kinases. This indicates that cyclic PIP, an intracellular regulator, which primarily acts on protein kinase A and on protein ser/thr phosphatases, operates more downstream in the signal transduction cascade as compared to the inositol phosphoglycan PIG41. Thus, cyclic PIP appears to be a suitable candidate to close the gap between IRSs and the protein kinases/phosphatases involved in the signal transduction of insulin.

Prostaglandylinositol cyclic phosphate (cPIP): a novel second messenger of insulin action. Comparative analysis of two kinds of "insulin mediators".

Insulin induces a broad spectrum of effects over a wide time interval. It also stimulates the phosphorylation of some cellular proteins, while decreasing the state of phosphorylation of others. These observations indicate the presence of different, but not necessarily mutually exclusive, pathways of insulin action. One well-known pathway represents a phosphorylation cascade initiated by the tyrosine kinase activity of the insulin receptor followed by involvement of different MAP-kinases. Another pathway suggests the existence of low molecular weight insulin mediators whose synthesis and/or release is initiated by insulin. Comparable analysis of two kinds of insulin mediators, namely inositolphosphoglycans and prostaglandylinositol cyclic phosphate (cPIP), has been carried out. It has been shown that the expression of a number of enzymes, such as phospholipase A(2), phospholipase C, cyclo-oxygenase and IRS-1-like enzyme, could regulate the biosynthesis of cPIP in both normal and diabetes-related conditions. Data on the activity of a key enzyme of cPIP biosynthesis termed cPIP synthase (IRS-1-like enzyme) in various monkey tissues before and twice during an euglycemic hyperinsulinemic clamp have been presented. It has been concluded that in vivo insulin increases cPIP synthase activity in both liver and subcutaneous adipose tissue of lean normal monkeys. It has been also suggested that abnormal production of cPIP could be related to several pathologies including glucocorticoid-induced insulin resistance and diabetic embryopathy. Further studies on cPIP and other types of insulin mediators are necessary to aid our understanding of insulin action.

Prostaglandylinositol cyclic phosphate synthase activity in the liver of insulin-resistant rhesus monkeys before and after a euglycemic hyperinsulinemic clamp.

Prostaglandylinositol cyclic phosphate (cPIP), functionally a cAMP antagonist, is a novel, low-molecular weight mediator of insulin action. Both essential hypertension and type 2 diabetes may be associated with a reduction of cPIP synthesis. In intact cells and in plasma membranes, cPIP synthesis is stimulated by insulin, which activates cPIP synthase by tyrosine phosphorylation. We measured the activities of cPIP synthase in the homogenates of freeze-clamped and then lyophilized liver samples from five insulin-resistant, adult rhesus monkeys, obtained under basal fasting conditions and again under maximal insulin stimulation during a euglycemic hyperinsulinemic clamp. The mean cPIP synthase activity in basal samples (0.33 +/- 0.09 pmol/min/mg protein) was not significantly different at the end of the clamp (0.24 +/- 0.11 pmol/min/mg protein). Basal cPIP synthase activityVoL 12, No. 1, 2001 was directly related to both basal cAMP content and basal fractional activity of cAMP-dependent protein kinase (PKA): r=0.85, p<0.05 and r=0.86, p<0.05, respectively. In turn, insulin-stimulated cPIP synthase activity was inversely related to both the insulin-stimulated fractional activity of PKA (r=0.89, p<0.02) and the insulin-stimulated total PKA activity: r=0.94, p<0.005. The findings suggest that in the liver of insulin-resistant rhesus monkeys, cPIP synthase activity, which leads to the synthesis of the low-molecular weight mediator cPIP, may oppose cAMP synthesis and PKA activity.

Prostaglandin deficiency promotes sensitization of adenylyl cyclase.

Inhibition of prostaglandin synthesis by the drug indomethacin suppresses the synthesis of the cyclic AMP antagonist, prostaglandylinositol cyclic phosphate (cyclic PIP), and leads to a metabolic state comparable to type II diabetes. It was of interest whether prostaglandin-deficiency likewise causes sensitization of adenylyl cyclase, as this has been reported for the diabetic state. In liver plasma membranes of indomethacin-treated male rats, basal and forskolin-stimulated cyclic AMP synthesis remained unchanged when compared to untreated control rats. In control rats, stimulation of cyclic AMP synthesis by fluoride (2.2-fold) or glucagon (3.5-fold) was much lower than stimulation by forskolin (6.6-fold). In contrast, in indomethacin-treated rats, stimulation of cAMP synthesis by fluoride (4.6-fold) or glucagon (5.2-fold) nearly matched the stimulation by forskolin (6.4-fold). The level of alpha1-adrenergic receptors was slightly reduced, from 450 to 320 fmol/mg protein, by the indomethacin treatment. Independent of the treatment by indomethacin, stimulation of cyclic AMP synthesis by adrenaline failed, in agreement with the low density of adrenergic beta-receptors. In conclusion, PGE deficiency sensitizes adenylyl cyclase in rat liver for G protein-coupled receptors (glucagon) and also for fluoride.

Two different mechanisms for activation of cyclic PIP synthase: by a G protein or by protein tyrosine phosphorylation.

The biosynthesis of the functional, endogenous cyclic AMP antagonist, prostaglandylinositol cyclic phosphate (cyclic PIP) is performed by the plasma membrane-bound enzyme cyclic PIP synthase, which combines prostaglandin E (PGE) and activated inositol phosphate (n-IP) to cyclic PIP. The Km values of the enzyme for the substrates PGE and n-IP are in the micromolar range. The plasma membrane-bound synthase is activated by fluoride, by the stable GTP analog GMP-PNP, by protamine or biguanide, by noradrenaline, and by insulin. The activation by protamine or biguanide and fluoride (10 mM) is additive, which may indicate the presence of two different types of enzyme, comparable to phospholipase Cbeta and phospholipase Cgamma. Plasma membrane-bound cyclic PIP synthase is inhibited by the protein tyrosine kinase inhibitor tyrphostin B46 with an IC50 of 1.7 microM. However, the solubilized and gel-filtrated enzyme is no longer inhibited by tyrphostin, indicating that the activity of cyclic PIP synthase is connected with the activity of a membrane-bound protein tyrosine kinase. Cyclic PIP synthase activity of freshly prepared plasma membranes is unstable. Upon freezing and rethawing of liver plasma membranes, this instability is increased about 2-fold. Protein tyrosine phosphatase inhibitors [vanadate, fluoride (50-100 mM)] stabilize the enzyme activity, but protease inhibitors do not, indicating that inactivation of the enzyme is connected with protein tyrosine dephosphorylation. Cyclic PIP synthase is present in all tissues tested, like brain, heart, intestine, kidney, liver, lung, skeletal muscle, spleen, and testis. Apart from liver, cyclic PIP synthase activity in most tissues is rather low, but it can be increased up to 5-fold when protein tyrosine phosphatase inhibitors like vanadate are present in the homogenization buffer. Preincubation of cyclic PIP synthase of liver plasma membranes with the tyrosine kinase src kinase causes a 2-fold increase of cyclic PIP synthase activity, though this is certainly not the physiological role played by src kinase in intact cells. The data indicate that cyclic PIP synthase can be activated by two separate mechanisms: by a G protein or by protein tyrosine phosphorylation.

Phosphatidylinositol 3-kinase and prostaglandylinositol cyclic phosphate (cyclic PIP), a mediator of insulin action, in the signal transduction of insulin.

It has been suggested that downstream signaling from the insulin receptor to the level of the protein kinases and protein phosphatases is accomplished by prosta-glandylinositol cyclic phosphate (cyclic PIP), a proposed second messenger of insulin. However, evidence points also to both phosphatidylinositol 3-kinase, which binds to the tyrosine phosphorylated insulin receptor substrate-1, and the Ras complex in insulin's downstream signaling. We have examined whether a correlation exists between these various observations. It was found that wortmannin, a specific inhibitor of phosphatidylinositol 3-kinase, prevented insulin-induced, as well as cyclic PIP-induced activation of glucose transport, indicating that PI 3-kinase action on glucose transport involves downstream signaling of both insulin and cyclic PIP. Wortmannin has no effect on cyclic PIP synthase activity nor on the substrate production for cyclic PIP synthesis either, indicating that the functional role of PI 3-kinase is exclusively downstream of cyclic PIP.

Degradation of the cyclic AMP antagonist prostaglandylinositol cyclic phosphate (cyclic PIP) by dephosphorylation.

The cAMP antagonist, prostaglandylinositol cyclic phosphate (cyclic PIP), is synthesized from prostaglandin E and activated inositol phosphate. From various tissues only that amount of cyclic PIP can be isolated that constitutes the difference between synthesis and degradation. In order to overcome this drawback, the cyclic PIP degrading enzyme or enzymes had to be characterized prior to searching for inhibitors. Cyclic PIP degrading activities have been found in all rat tissues tested, and are lowest in brain (380 pmol x min(-1) x g(-1) wet weight) and highest in liver (1460 pmol x min(-1) x g(-1) wet weight). They are associated primarily with particulate structures of the cells, but not with the plasma membrane. There appear to be at least two different enzymatic activities involved in the degradation of cyclic PIP, because there are two pH-optima, one between pH 7 and 8 and another between pH 4 and 5. It is assumed that these activities are located in microsomes and lysosomes. Because prostaglandylinositol is the final product obtained in the degradation of cyclic PIP, a phosphodiesterase and a phosphatase should be involved, which could not yet be identified individually. Like alkaline phosphatase, cyclic PIP-degrading enzymes require Mg2+ and they are inhibited by heavy metal ions such as mercuric and copper chloride, by sodium fluoride and interestingly, by prostaglandins.

Synthesis and action of the cyclic AMP antagonist, prostaglandylinositol cyclic phosphate (cyclic PIP), in Dictyostelium discoideum.

The cyclic AMP antagonist, prostaglandylinositol cyclic phosphate (cyclic PIP), is synthesized from prostaglandin E (PGE) and activated inositol phosphate (n-IP) in the presence of ATP by an enzyme of rat liver plasma membranes. Extracts of the slime mould Dictyostelium discoideum contain this activated inositol phosphate and D. discoideum cells convert [3H]PGE1 to [3H]cyclic PIP. This extracted polar [3H]product co-chromatographed with cyclic PIP from rat liver on gel filtration, anion exchange- and adsorption chromatography. Starving D. discoideum cells show cyclic AMP-induced oscillations, which can be inhibited by cyclic PIP (0.4 x 10(-7) M), but not by its phosphomonoester prostaglandylinositol phosphate (PIP) (1.4 x 10(-7) M). AMP and ADP at much higher concentrations (1 mM) antagonized these oscillations. The time needed for aggregation and fruiting body formation of starving D. discoideum cells is extended by cyclic PIP (1.4 x 10(-7) M) up to 3-fold, whereas its phosphomonoester (1.9 x 10(-7) M) showed a 9-fold weaker effect, and AMP and ADP even at 1 mM concentration showed no effect.

Insulin resistance, a result of reduced synthesis of prostaglandylinositol cyclic phosphate, a mediator of insulin action? Regulation of cyclic PIP synthetase activity by oral antidiabetic and antihypertensive drugs.

Reduced ability or failure to stimulate cyclic adenosinemonophosphate (AMP) synthesis on a second addition of hormone 30 min after a first stimulation was taken as an indirect indication of the synthesis of the cyclic AMP antagonist prostaglandylinositol cyclic phosphate (cyclic PIP). In diabetic rats, because of an increased possibility of restimulating cyclic AMP synthesis, the formation of cyclic PIP should be reduced. Additionally, severalfold increased basal cyclic AMP synthesis can be observed in diabetic hepatocytes in comparison with controls. Upon measuring cyclic PIP levels after hormonal stimulation in all organs of diabetic rats, it was found that stimulation of cyclic PIP synthesis by insulin decreased gradually in a time-dependent manner. Plasma membranes were prepared from diabetic Ksj db/db mice and from spontaneously hypertensive rats (SHR), and in a subsequent assay for cyclic PIP synthetase, an up to 60% decrease of enzyme activity was found. Cyclic PIP synthetase can be completely inhibited by preincubation with protein kinase A. It is most likely that this serine phosphorylation reaction by which the enzyme is inhibited also in vivo is a result of increased cyclic AMP levels. The addition of 10(-5)-10(-4) M sulfonylureas to the enzyme assay of liver plasma membrane causes full inhibition, and the addition of 10(-5)-10(-4) M biguanides, a two- to fourfold activation of the enzyme. Activation of cyclic PIP synthetase by biguanides can also be demonstrated in intact cells. It is a fast reaction and additive with respect to the activation by fluoride or guanylyl-imidodiphosphate (GMP-PNP), and it is most likely the effect with which the biguanides produce the correcting changes in metabolism. Furthermore, antihypertensive drugs like captopril, guanethidine, and dihydralazine also activate cyclic PIP synthetase. In contrast to the activation by the biguanides, this effect is not additive to the activation by fluoride. It appears that essential hypertension and type 2 diabetes are connected with or may be the result of a reduction in synthesis of the intracellular messenger cyclic PIP, whose synthesis is stimulated by hormones like insulin and noradrenaline (alpha-adrenergic action).

Biosynthesis of the endogenous cyclic adenosine monophosphate (AMP) antagonist, prostaglandylinositol cyclic phosphate (cyclic PIP), from prostaglandin E and activated inositol polyphosphate in rat liver plasma membranes.

The endogenous cyclic adenosine monophosphate (AMP) antagonist, cyclic PIP, has been identified as a prostaglandylinositol cyclic phosphate. It inhibits protein kinase A 100% and activates protein serine phosphatase about sevenfold. It is biosynthesized by an enzyme of the plasma membrane when the assay mixture contains adenosine triphosphate (ATP), Mg2+, prostaglandin E and a novel inositol polyphosphate, which cannot be substituted by commercially available inositol phosphates. This novel inositol polyphosphate is a very labile compound. On anion exchange chromatography it elutes in the range of ATP, which may indicate the presence of three phosphate groups. It adsorbs on charcoal, which suggests the presence of a hydrophobic component, possibly a guanosine. Pyrophosphates obtained from inositol 1,4- and inositol 2,4-bisphosphate are accepted by cyclic PIP synthetase for the synthesis of cyclic PIP. The biosynthesis is characterized by enzyme kinetic parameters like dependence on time, enzyme and substrate concentration. The pH optimum of the enzyme is in the range 7.5-8. The enzyme functions optimally with prostaglandin E and poorly with prostaglandin A as the substrate. The presence of fluoride in the assay causes a three- to fourfold increase in cyclic PIP synthesis, which may be correlated with activation via G proteins. These data support previous reports on the chemical structure and action of cyclic PIP. With respect to the possible isomers of cyclic PIP, these indicate that it is most likely the C4-hydroxyl group of the inositol which binds the C15-hydroxyl group of prostaglandin E. A model of hormone-stimulated synthesis of cyclic PIP is proposed: phospholipase A2 and phospholipase C, activated by G proteins upon alpha-adrenergic stimulation, liberate either unsaturated fatty acids or inositol phosphates, which are transformed to prostaglandins and to novel inositol polyphosphate with an energy-rich bond. The cyclic PIP synthetase combines these two substrates to cyclic PIP.

Indomethacin treatment causes loss of insulin action in rats: involvement of prostaglandins in the mechanism of insulin action.

Glucose tolerance tests in rats showed that after indomethacin treatment plasma insulin levels rose five-fold higher than in untreated controls. Accordingly, the pancreatic islets of indomethacin-treated rats secreted insulin at a threefold higher rate. Glucose tolerance tests additionally showed that indomethacin treatment led to a retarded disposal of the elevated blood glucose. Both effects appear to be caused by an attenuation of the hormone responsiveness for insulin and noradrenaline (alpha-adrenoceptor action) by indomethacin. The following observations support this view: insulin and adrenaline (alpha-adrenoceptor action) lost their ability to lower cyclic adenosine monophosphate (AMP) levels in hepatocytes; the glycogen content of liver and skeletal muscle was reduced by 95% and 65%, respectively; in adipocytes the stimulation of glucose transport by insulin was reduced by 60%. These effects of indomethacin can be reversed by the addition of exogenous prostaglandin E (PGE), as elevated cyclic AMP synthesis was again sensitive to alpha-adrenergic inhibition in the liver. These results indicate a relationship between prostaglandins and insulin action. These effects of indomethacin could result from reduced synthesis of cyclic PIP (prostaglandylinositol cyclic phosphate), a proposed second messenger for insulin and alpha-adrenoceptor action, whose synthesis was decreased by indomethacin treatment and increased by the addition of exogenous PGE. Stimulation of glucose transport by cyclic PIP was unaffected by indomethacin treatment, in contrast to the stimulation by insulin. Inhibition of PGE and cyclic PIP synthesis resulted in a metabolic state comparable to insulin resistance in non-insulin-dependent diabetes mellitus.

The endogenous cyclic AMP antagonist, cyclic PIP: its ubiquity, hormone-stimulated synthesis and identification as prostaglandylinositol cyclic phosphate.

This report shows that the cyclic AMP antagonist cyclic PIP is present in all organs and tissues of the rat so far examined: brain, heart, lung, intestine, kidney, liver, spleen, skeletal muscle and fat. The synthesis of cyclic PIP is stimulated by insulin or noradrenaline (alpha-adrenergic action) in a dose-dependent fashion. Increasing cyclic PIP synthesis with increasing insulin concentrations matches the insulin receptor binding curves. Cyclic PIP levels in blood serum remain low after hormonal stimulation and no cyclic PIP can be detected in urine. As an indication of its ubiquity, cyclic PIP was even detected in yeast. Prostaglandin E (as shown by incorporation of [3H]PGE into cyclic PIP and demonstration of a constant specific activity), myo-inositol (as shown by acid hydrolysis of the dephosphorylated cyclic PIP and mass spectrometric identification of the products) and one phosphate (as shown by the ionic nature of cyclic PIP and its inactivation by phosphodiesterase plus phosphatase) are components of cyclic PIP. Chemical derivatization experiments of cyclic PIP suggest the phosphate to be bound to myo-inositol and the myo-inositol phosphate to the prostaglandin E by its C15-hydroxyl group.

Prostaglandylinositol cyclic phosphate, an antagonist to cyclic AMP.

The existence of a low molecular weight intracellular regulator has been demonstrated. It is composed of PGE, myoinositol, and one phosphate and has been named cyclic PIP. Its synthesis is stimulated by hormones such as insulin in a dose-dependent manner, it is therefore suggested that cyclic PIP is a second messenger equivalent to cyclic AMP in its potency but antagonistic in its action.