Cell-free stimulation of the insulin-sensitive cAMP phosphodiesterase by the joint actions of ATP and the soluble fraction from insulin-treated rat liver

Expression and regulation of cyclic nucleotide phosphodiesterases in human and rat pancreatic islets

As shown by transgenic mouse models and by using phosphodiesterase 3 (PDE3) inhibitors, PDE3B has an important role in the regulation of insulin secretion in pancreatic β-cells. However, very little is known about the regulation of the enzyme. Here, we show that PDE3B is activated in response to high glucose, insulin and cAMP elevation in rat pancreatic islets and INS-1 (832/13) cells. Activation by glucose was not affected by the presence of diazoxide. PDE3B activation was coupled to an increase as well as a decrease in total phosphorylation of the enzyme. In addition to PDE3B, several other PDEs were detected in human pancreatic islets: PDE1, PDE3, PDE4C, PDE7A, PDE8A and PDE10A. We conclude that PDE3B is activated in response to agents relevant for β-cell function and that activation is linked to increased as well as decreased phosphorylation of the enzyme. Moreover, we conclude that several PDEs are present in human pancreatic islets.

A newly identified 50 kDa protein, which is associated with phosphodiesterase 3B, is phosphorylated by insulin in rat adipocytes

Phosphodiesterase 3B (PDE3B), a major PDE isoform in adipocytes, plays a pivotal role in the anti-lipolytic action of insulin. Insulin phosphorylates and activates PDE3B in a phosphatidylinositol 3-kinase-dependent manner. We identified a new 50 kDa protein that is phosphorylated by insulin and is co-immunoprecipitated with PDE3B by anti-PDE3B antibodies in rat adipocytes. The insulin-induced phosphorylation of the 50 kDa protein was also detected in a cell free system against the N-terminal and the catalytic regions, which are more than 700 amino acids apart recognize the 50 kDa protein, suggesting that it is not a proteolytic product, but an associated protein with PDE3B. Phosphoamino acid analysis indicated that both serine and threonine residues in the 50 kDa protein were phosphorylated, but only serine residues in PDE3B were phosphorylated. Therefore, it appears likely that this is a new protein which is associated with PDE3B.

Mitogen-activated protein kinase and p70 ribosomal protein S6 kinase are not involved in the insulin-dependent stimulation of cAMP phosphodiesterase kinase in rat adipocytes

To elucidate the mechanism of anti-lipolytic action of insulin in rat epididymal adipocytes, we explored the potential mechanism that might be involved in the hormone-dependent stimulation of cAMP phosphodiesterase (PDE) kinase. PDE kinase was assayed in a cell-free system. Both wortmannin and LY294002, highly specific inhibitors of phosphatidylinositol 3-kinase, almost completely blocked the hormonal effect not only on PDE kinase but also on mitogen-activated protein (MAP) kinase. Neither PD98059, a specific inhibitor of MAP kinase, nor rapamycin, a potent inhibitor of insulin-dependent stimulation of p70 ribosomal protein S6 kinase (p70S6K), had inhibitory effect on that of PDE kinase. These results are consistent with the view that (i) insulin-activated PDE kinase as well as MAP kinase and p70S6K are localized downstream of phosphatidylinositol 3-kinase, (ii) PDE kinase is distinct from either MAP kinase or p70S6K and (iii) PDE kinase does not exist downstream of either MAP kinase or p70S6K. It is suggested that PDE kinase and MAP kinase or p70S6K may be localized in separate branches of the cascade of insulin action. The branching point of the cascade could be either at or below the level of phosphatidylinositol 3-kinase.

Type III cGMP-inhibited cyclic nucleotide phosphodiesterases (PDE3 gene family)

Seven different but related cyclic nucleotide phosphodiesterase (PDE) gene families have been identified. Type III cGMP-inhibited (cGI) PDEs, the PDE3 gene family, are found in many tissues. cGI PDEs exhibit a high affinity for both cAMP and cGMP, and are selectively and relatively specifically inhibited by certain agents which augment myocardial contractility, promote smooth muscle relaxation and inhibit platelet aggregation. Adipocyte, platelet, and hepatocyte cGI PDE activities are regulated by cAMP-dependent phosphorylation. Insulin-induced phosphorylation/activation of adipocyte and hepatocyte cGI PDEs is thought to be important in acute regulation of triglyceride and glycogen metabolism by insulin. Two distinct cGI PDE subfamilies, products of distinct but related genes, have been identified. They exhibit the domain structure common to PDEs with a carboxyterminal region, conserved catalytic domain and divergent regulatory domain. In their catalytic domains cGI PDEs contain a 44 amino acid insertion not found in other PDE families. The expression of cGIP1 and cGIP2 mRNAs differs in different rat tissues, suggesting distinct functions for the two cGI PDE subfamilies, i.e., cGIP1 in adipose tissue, liver, testis and cGIP2 in myocardium, platelets and smooth muscle.

Multiple transcripts for the human cardiac form of the cGMP-inhibited cAMP phosphodiesterase

cDNAs for two distinct Type III cGMP-inhibited (cGI) cyclic nucleotide phosphodiesterases (PDE), designated cGIP1 and cGIP2, were previously cloned from rat adipose and human cardiac cDNA libraries, respectively. In this study, another cDNA (approximately 4.0 kilobase (kb)) encoding a cGI-PDE of 74 kDa (658 amino acids) was isolated from a human placental cDNA library. The nucleotide sequence of its open reading frame was virtually identical to a corresponding region in the 3' portion of the cardiac cGIP2 cDNA (approximately 7.6 kb) which encoded a approximately 125-kDa cGI-PDE (1141 amino acid). Northern blots and RNase protection assays revealed a prominent 4.4-kb transcript and a 7.6-kb transcript in human placenta. The transcription start site of the 4.4-kb transcript was assigned to cardiac cDNA nucleotide 1292, the putative beginning of exon 3 of the human cGIP2 gene, with a potential translation initiation site 183 bases downstream, as determined by RNase protection assay. The 5'-flanking region of the 4.4-kb transcript exhibited promoter activity in HeLa cells which expressed the 4.4-kb transcript, and contained a TATAA sequence 35 base pairs upstream from the tentative transcription start site. Recombinant cGI-PDEs, expressed in Sf9 cells from the 7.6- and 4.0-kb cDNA, exhibited differences in their subcellular localization and Km for cGMP. Thus, in human tissues, alternative transcription may contribute to generating at least two cGIP2 isoforms, cytosolic and membrane-associated cGI-PDEs with different Km values for cGMP.

Essential role of phosphatidylinositol 3-kinase in insulin-induced activation and phosphorylation of the cGMP-inhibited cAMP phosphodiesterase in rat adipocytes. Studies using the selective inhibitor wortmannin

Incubation of rat adipocytes with wortmannin, a potent and selective phosphatidylinositol 3-kinase (PI 3-kinase) inhibitor, completely blocked the antilipolytic action of insulin (IC50 = 100 nM), the insulin-induced activation and phosphorylation of cGMP-inhibited cAMP phosphodiesterase (cGI-PDE) as well as the activation of the insulin-stimulated cGI-PDE kinase (IC50 = 10-30 nM). No direct effects of the inhibitor on the insulin-stimulated cGI-PDE kinase, the cGI-PDE and the hormone-sensitive lipase were observed. These data suggest that activation of PI 3-kinase upstream of the insulin-stimulated cGI-PDE kinase in the antilipolytic insulin signalchain has an essential role for insulin-induced cGI-PDE activation/phosphorylation and anti-lipolysis.

Synergistic interactions between selective pharmacological inhibitors of phosphodiesterase isozyme families PDE III and PDE IV to attenuate proliferation of rat vascular smooth muscle cells

The interaction between selective inhibitors of 3',5'-cyclic-nucleotide phosphodiesterase (PDE) III (cyclic GMP inhibited phosphodiesterase) and selective inhibitors of PDE IV (Ro 20-1724 inhibited phosphodiesterase) to attenuate fetal bovine serum-stimulated incorporation of [3H]thymidine into DNA and cell proliferation was studied in a line (A10) of vascular smooth muscle cells (VSMC). The nonselective PDE inhibitors 3-isobutyl-1-methylxanthine (IBMX) and papaverine attenuated DNA synthesis with EC50 values (16 and 18 microM, respectively) in the same range as their published IC50 values (2-50 and 2-25 microM, respectively) as PDE inhibitors. The selective PDE III inhibitors CI-930 and cilostamide used alone attenuated DNA synthesis with EC50 values (> 300 and 5.3 microM, respectively) that were much higher than published IC50 values (0.15-0.46 and 0.005-0.064 microM, respectively) for inhibition of PDE III. In the presence of the PDE IV inhibitor rolipram (10 microM), their EC50 values were shifted (0.66 and 0.16 microM, respectively) much closer to their respective IC50 values. When the selective PDE IV inhibitors rolipram and Ro 20-1724 were used alone, they attenuated DNA synthesis with EC50 values (111 and > 100 microM, respectively) much higher than their IC50 values (0.6-2.6 and 2-13 microM, respectively) as inhibitors of PDE IV, but 10 microM CI-930 (PDE III inhibitor) shifted their EC50 values (0.56 and 1.5 microM, respectively) much closer to their IC50 values. In experiments that assessed VSMC proliferation using the MTT [3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide] method, IBMX and papaverine attenuated proliferation with EC50 values (27 and 58 microM, respectively) close to their IC50 values. CI-930 and cilostamide used alone did not cause 50% attenuation of proliferation at the highest concentrations tested (100 and 10 microM, respectively). In the presence of 5 microM rolipram, however, their effects were enhanced greatly with EC50 values (0.86 and 0.23 microM, respectively) that were close to their IC50 values as PDE III inhibitors. Similarly, rolipram and Ro 20-1724 attenuated VSMC proliferation with EC50 values close to their IC50 values in the presence (2.1 and 4.6 microM, respectively) but not in the absence (> 100 and > 10 microM, respectively) of 2 microM CI-930. The interactions between PDE III inhibitors and PDE IV inhibitors to attenuate DNA synthesis and VSMC proliferation were synergistic as determined by the combination index. The data demonstrate that the synergistic interactions that attenuate incorporation of [3H]thymidine into DNA are accompanied by synergistic attenuations of VSMC division.(ABSTRACT TRUNCATED AT 400 WORDS)

Evidence that protein kinase C may not be involved in the insulin action on cAMP phosphodiesterase: studies with electroporated rat adipocytes that were highly responsive to insulin

Partially permeabilized rat adipocytes with a high responsiveness to insulin were prepared by electroporation and used to study the effect of 1-(5-isoquinolinylsulfonyl)-2-methylpiperazine (H-7) on insulin actions in adipocytes. H-7 is a well-documented inhibitor of several protein kinases, including protein kinase C; however, it does not rapidly enter adipocytes protected with the intact plasma membrane. The cells were suspended in Buffer X [4.74 mM NaCl, 118.0 mM KCl, 0.38 mM CaCl2, 1.00 mM EGTA, 1.19 mM Mg2SO4, 1.19 mM KH2PO4, 25.0 mM Hepes/K, 20 mg/ml bovine serum albumin, and 3 mM pyruvate/Na, pH 7.4] and electroporated six times with a Gene-Pulser (from Bio-Rad) set at 25 microF and 2 kV/cm. In cells electroporated as above, insulin stimulated (a) membrane-bound, cAMP phosphodiesterase approximately 2.6-fold when the hormone concentration was 10 nM and (b) glucose transport activity approximately 4.5-fold when the hormone concentration was raised to 100 nM. H-7 strongly inhibited the actions of insulin on both glucose transport (apparent Ki = 0.3 mM) and cAMP phosphodiesterase (apparent Ki = 1.2 mM) in electroporated adipocytes. H-7 also inhibited lipolysis in adipocytes; the apparent Ki value for the reaction in intact cells was 0.45 mM, and that in electroporated cells was 0.075 mM. It is suggested that a certain protein kinase or kinases that are significantly sensitive to H-7 may be involved in the insulin-dependent stimulation of glucose transport and that of phosphodiesterase. However, protein kinase C (or Ca2+/phospholipid-dependent protein kinase) may not be involved, at least, in the hormonal action on phosphodiesterase since the apparent Ki value of H-7 for the reaction is too high.