PP2A contributes to endothelial death in high glucose: inhibition by benfotiamine

Endothelial death is critical in diabetic vascular diseases, but regulating factors have been only partially elucidated. Phosphatases play important regulatory roles in cell metabolism, but have not previously been implicated in hyperglycemia-induced cell death. We investigated the role of the phosphatase, type 2A protein phosphatase (PP2A), in hyperglycemia-induced changes in signaling and death in bovine aortic endothelial cells (BAEC). We explored also the influence of benfotiamine on this phosphatase. Activation of PP2A was assessed in BAEC by the extent of methylation and measurement of activity, and the enzyme was inhibited using selective pharmacological (okadaic acid, sodium fostriecin) and molecular (small interfering RNA) approaches. BAECs cultured in 30 mM glucose significantly increased PP2A methylation and activity, and PP2A inhibitors blocked these abnormalities. PP2A activity was increased also in aorta and retina from diabetic rats. NF-κB activity and cell death in BAEC were significantly increased in 30 mM glucose and inhibited by PP2A inhibition. NF-κB played a role in the hyperglycemia-induced death of BAEC, since blocking its translocation with SN50 also inhibited cell death. Inhibition of PP2A blocked the hyperglycemia-induced dephosphorylation of NF-κB and Bad, thus favoring cell survival. Incubation of benfotiamine with BAEC inhibited the high glucose-induced activation of PP2A and NF-κB and cell death, as well as several other metabolic defects, which likewise were inhibited by inhibitors of PP2A. Activation of PP2A contributes to endothelial cell death in high glucose, and beneficial actions of benfotiamine are due, at least in part, to inhibition of PP2A activation.

Ceramide induces mitochondrial abnormalities in insulin-secreting INS-1 cells: potential mechanisms underlying ceramide-mediated metabolic dysfunction of the beta cell

C2-ceramide, a cell permeable analogue of ceramide [CER] markedly reduced mitochondrial membrane potential [MMP] in insulin-secreting INS cells, which was followed by a significant accumulation of cytochrome c [Cyt c] into the cytosolic compartment. In a manner akin to CER, exposure of these cells to interleukin-1beta [IL-1beta] also resulted in reduction in MMP and cytosolic accumulation of Cyt c. Further, long-term exposure of these cells to either CER [but not its inactive analogue] or IL-1beta caused a marked reduction in their metabolic viability. However, unlike IL-1beta, which increased nitric oxide [NO] release, CER-treatment of INS cells had no effects of CER on NO release were demonstrable. Together, these findings suggest that CER-induced mitochondrial effects may not be mediated via iNOS gene expression and NO production. CER also activated an okadaic acid -sensitive protein phosphatase [CAPP] in the purified mitochondrial fraction, suggesting that CAPP might represent one of the target proteins for CER in the beta cell mitochondria. Together, our findings suggest direct detrimental effects of CER on mitochondrial function in beta cells leading to their dysfunction and demise via apoptosis. Moreover, our findings provide evidence for a potential difference in the mechanisms underlying CER- and IL-1beta-induced mitochondrial defects and apoptotic demise of the effete beta cell.

Regulatory roles for histone deacetylation in IL-1beta-induced nitric oxide release in pancreatic beta-cells

Histone (de)acetylases control gene transcription via modification of the chromatin structure. Herein, we investigated potential roles for histone deacetylation (or hypoacetylation) in interleukin-1β (IL-1β)-mediated inducible nitric oxide synthase (iNOS) and nitric oxide (NO) release in insulin-secreting INS 832/13 (INS) cells. Western blot analysis suggested localization of members of Class 1 and Class 2 families of histone deacetylases (HDACs) in these cells. Trichostatin A (TSA), a known inhibitor of HDACs, markedly reduced IL-1β-mediated iNOS expression and NO release from these cells in a concentration-dependent manner. TSA also promoted hyperacetylation of histone H4 under conditions in which it inhibited IL-1β-mediated effects on isolated β cells. Rottlerin, a known inhibitor of protein kinase Cδ, also increased histone H4 acetylation, and inhibited IL-1β-induced iNOS expression and NO release in these cells. It appears that the putative mechanism underlying the stimulatory effects of rottlerin on acetylation status of histone H4 are distinct from the HDAC inhibitory property of TSA, since rottlerin failed to inhibit HDAC activity in nuclear extracts isolated from INS cells. These data are suggestive of potential regulatory effects of rottlerin at the level of increasing the histone acetyltransferase activity in these cells. Together our studies present the first evidence to suggest a PKCδ-mediated signalling step, which promotes hypoacetylation of candidate histones culminating in IL-1β-induced metabolic dysfunction of the isolated β cell.

GTP-binding protein-independent potentiation by mastoparan of IL-1 -induced nitric oxide release from insulin-secreting HIT-T15 cells

Our recent data implicated small molecular weight G-proteins (e.g., H-Ras) in interleukin 1beta (IL 1beta)-induced metabolic dysfunction and apoptotic demise of the islet beta cell (Tannous et al., Biochem Pharmacol 2001; 62:1459-1468, Kowluru and Morgan, Biochem Pharmacol, 2002; 63:1027-1035, Chen et al. Biochem Pharmacol, 2003; 66:1681-1694). Recently, we have shown that mastoparan, a tetradecapeptide from wasp venom, has been shown to directly activate islet endogenous G-proteins and regulate islet function (Amin et al., Endocrinology 2003; 144: 4508-4518). Herein, we investigated potential contributory roles, if any, of mastoparan (Mas)-sensitive G-proteins in IL-induced nitric oxide (NO) release from insulin-secreting HIT-T15 cells. While, ineffective by itself, Mas significantly potentiated IL-induced NO release from HIT-T15 cells. Interestingly, Mas-17, an inactive analog of Mas, also potentiated IL-induced NO release, suggesting that the potentiating effect of Mas may not involve activation of specific G-proteins. Such potentiating effects on IL-induced NO release were also demonstrable in the presence of another polycationic compound, melittin. Together, these findings suggest that Mas-induced potentiation of IL-induced NO release may in part be due to its amphiphilic and polycationic nature. These data also warrant caution in the use of Mas to study its regulation of cellular function without the use of an appropriate negative control, such as Mas-17.

IL-1beta-induced nitric oxide release from insulin-secreting beta-cells: further evidence for the involvement of GTP-binding proteins

Recently, we have demonstrated regulatory roles for G-proteins (e.g., H-Ras) in IL-1beta induced NO release from HIT-T15 cells. Herein, we report a similar regulatory mechanism for IL-1beta induced NO release from RIN5F and INS-1 cells. Our data indicate that functional inactivation of Ras, either by Clostridial toxins or by specific inhibitors of Ras function, results in a significant inhibition in IL-1beta induced NO release, suggesting that activation of specific G-proteins is essential for IL-1beta induced NO release. In the present study, we report possible loci where IL-1beta treatment might result in functional activation of these G-proteins. For example, IL-1beta treatment resulted in significant reduction in (high-and low-affinity) GTPase activities in lysates derived from normal rat islets; such a scenario might lead to retention of candidate G-proteins in GTP-bound, active conformation. Further, IL-1beta treatment increased the G-protein carboxyl methyl transferase activity as well as carboxyl methylation of endogenous beta-cell proteins; such a modification has been shown to increase the membrane association and interaction of these G-proteins with their respective effector proteins. Also, we report immunologic localization of H-Ras regulatory proteins including its nucleotide exchange factor (GRF-1) and its effector protein (eg., Raf-1) in isolated beta-cells. Together, our data indicate localization, and regulation by IL-1beta, of specific enzymes that are critical to activation of G-proteins. Based on these preliminary findings, we propose a model for the involvement of G-proteins in IL-1beta induced NO release and subsequent demise of the pancreatic beta-cell.

Inhibitors of post-translational modifications of G-proteins as probes to study the pancreatic beta cell function: potential therapeutic implications

It is well established that glucose-induced insulin secretion involves generation of intracellular second messengers. Using specific inhibitors of guanosine triphosphate [GTP] biosynthesis [e.g., mycophenolic acid; MPA], we have identified a permissive role for GTP in glucose-stimulated insulin secretion. While the exact site of action for GTP within the islet beta cell remains to be identified and defined, recent evidence from several laboratories, including our own, indicate that it could involve activation of GTP-binding proteins [G-proteins]. These studies have identified both trimeric and monomeric forms of G-proteins within the pancreatic beta cell. Recent data also indicate that these G-proteins, specifically the monomeric G-proteins and the gamma subunits of trimeric G-proteins undergo a series of posttranslational modifications at their C-terminal cysteine. Such modifications include, isoprenylation, carboxyl methylation and palmitoylation. These modification steps appear to be essential for translocation of these proteins to the membrane sites for interaction with their respective effector proteins. This review primarily focuses on recent findings that clearly support the viewpoint that these posttranslational modification steps not only play obligatory roles in fuel-induced insulin secretion, but also in cytokine-mediated apoptotic demise of the beta cell. In this review, we also attempted to describe those findings involving the use of specific inhibitors for each of these pathways, and it is our hope that these aspects of beta cell metabolism and function generate interest in development of therapeutic intervention modalities to states of perturbed insulin release.

Functional inactivation by interleukin-1beta of glyceraldehyde-3-phosphate dehydrogenase in insulin-secreting cells

AIMS/HYPOTHESIS

It is well established that long-term exposure of isolated beta cells to cytokines [e.g., IL-1beta] results in increased expression of inducible nitric oxide synthase and subsequent release of nitric oxide, which in turn, has been shown to mediate a wide array of effects, including alterations in cellular high-energy metabolism. In this context, several extant studies have demonstrated significant reduction in adenine and guanine nucleotide triphosphate levels in beta cells exposed to IL-1beta. Herein, we examined the functional status of glyceraldehyde-3-phosphate dehydrogenase [GAPDH] in insulin-secreting cells exposed to IL-1beta, since it represents the first enzyme in the glycolytic pathway that is involved in the generation of ATP.

METHODS

GAPDH was assayed spectrophotometrically in the cytosolic fraction derived from control and IL-1beta -treated [300 pM for 24 hrs] insulin-secreting cell lines [HIT-T15 and RINm5F].

RESULTS

IL-treatment resulted in marked attenuation of GAPDH activity in HIT and RIN cells; such a reduction in this activity was not due to inhibition of its expression by IL-1. Instead, we observed that incubation of HIT and RIN lysates with peroxynitrite, a reactive intermediate of nitric oxide with superoxide anion, resulted in significant reduction in the GAPDH activity.

CONCLUSION/INTERPRETATION

These results identify a GAPDH as one of the biochemical loci for the effects of IL-derived peroxynitrite in the islet beta cell. The previously reported reduction in high-energy phosphate levels in an IL-treated beta cell may, in part, be due to inhibition of GAPDH activity, and subsequent reduction in the glycolytic efficiency of the beta cell.

Positive modulation by Ras of interleukin-1beta-mediated nitric oxide generation in insulin-secreting clonal beta (HIT-T15) cells

In the present study, we have shown that exposure of insulin-secreting clonal beta (HIT-T15) cells to interleukin-1beta (IL-1beta) results in a time- and concentration-dependent increase in nitric oxide (NO) release. These effects by IL-1beta on NO release were mediated by induction of inducible nitric oxide synthase (iNOS) from the cells. Preincubation of HIT cells with Clostridium sordellii lethal toxin-82, which irreversibly glucosylates and inactivates small G-proteins, such as Ras, Rap, Ral, and Rac, but not Cdc42, completely abolished IL-1beta-induced NO release. Pre-exposure of HIT cells to C. sordellii lethal toxin-9048, which monoglucosylates and inhibits Ras, Cdc42, Rac, and Rap, but not Ral, also attenuated IL-1beta-mediated NO release. These data indicate that activation of Ras and/or Rac may be necessary for IL-1beta-mediated NO release. Preincubation of HIT cells with C. difficile toxin-B, which monoglucosylates Rac, Cdc42, and Rho, had no demonstrable effects on IL-mediated NO release, ruling out the possibility that Rac may be involved in this signaling step. Further, two structurally dissimilar inhibitors of Ras function, namely manumycin A and damnacanthal, inhibited, in a concentration-dependent manner, the IL-1beta-mediated NO release from these cells. Together, our data provide evidence, for the first time, that Ras activation is an obligatory step in IL-1beta-mediated NO release and, presumably, the subsequent dysfunction of the pancreatic beta cell. Our data also provide a basis for future investigations to understand the mechanism of cytokine-induced beta cell death leading to the onset of insulin-dependent diabetes mellitus.

Activation of acetyl-CoA carboxylase by a glutamate- and magnesium-sensitive protein phosphatase in the islet beta-cell

Acetyl-CoA carboxylase (ACC) catalyzes the formation of malonyl-CoA, a precursor in the biosynthesis of long-chain fatty acids, which have been implicated in physiological insulin secretion. The catalytic function of ACC is regulated by phosphorylation (inactive)-dephosphorylation (active). In this study we investigated whether similar regulatory mechanisms exist for ACC in the pancreatic islet beta-cell. ACC was quantitated in normal rat islets, human islets, and clonal beta-cells (HIT-15 or INS-1) using a [(14)C]bicarbonate fixation assay. In the beta-cell lysates, ACC was stimulated by magnesium in a concentration-dependent manner. Of all the dicarboxylic acids tested, only glutamate, albeit ineffective by itself, significantly potentiated magnesium-activated ACC in a concentration-dependent manner. ACC stimulation by glutamate and magnesium was maximally demonstrable in the cytosolic fraction; it was markedly reduced by okadaic acid (OKA) in concentrations (<50 nmol/l) that inhibited protein phosphatase 2A (PP2A). Furthermore, pretreatment of the cytosolic fraction with anti-PP2A serum attenuated the glutamate- and magnesium-mediated activation of ACC, thereby suggesting that ACC may be regulated by an OKA-sensitive PP2A-like enzyme. Streptavidin-agarose chromatography studies have indicated that glutamate- and magnesium-mediated effects on ACC are attributable to activation of ACC's dephosphorylation; this suggests that the stimulatory effects of glutamate and magnesium on ACC might involve activation of an OKA-sensitive PP2A-like enzyme that dephosphorylates and activates ACC. In our study, 5-amino-imidazolecarboxamide (AICA) riboside, a stimulator of AMP kinase, significantly inhibited glucose-mediated activation of ACC and insulin secretion from isolated beta-cells. Together, our data provide evidence for a unique regulatory mechanism for the activation of ACC in the pancreatic beta-cell, leading to the generation of physiological signals that may be relevant for physiological insulin secretion.

Adenine and guanine nucleotide-specific succinyl-CoA synthetases in the clonal beta-cell mitochondria: implications in the beta-cell high-energy phosphate metabolism in relation to physiological insulin secretion

AIMS/HYPOTHESIS

Succinyl-CoA synthetase catalyses the substrate level phosphorylation of ADP or GDP. It also supplies succinyl-CoA for heme synthesis. Recently, two distinct mitochondrial succinyl-CoA synthetase activities, one specific for ATP and the other for GTP, have been characterized in various tissues of pigeon. Because of the relative importance of mitochondrial high-energy phosphate metabolism in physiological insulin secretion and the few data available on mitochondrial succinyl-CoA synthetase in the beta cell, this study examined whether ATP-specific and GTP-specific succinyl-CoA synthetase activities are localized in the clonal beta-cell mitochondria.

METHODS

Using the mitochondrial extracts from clonal beta [INS-1 and HIT-T15] cells, we measured the formation of succinyl-CoA from succinate, CoA and ATP or GTP. To confirm the identity of these two enzymes, individual subunits of ATP-specific and GTP-specific to succinyl-CoA synthetase were identified by Western blot analysis.

RESULTS

Both ATP-and GTP activities of succinylCoA synthetase were observed in the mitochondrial fractions from these cells. The ratios of GTP to ATP activities of succinyl-CoA synthetase were near unity in both of the cell types studied. Using affinity-purified antisera directed specifically against individual (alpha and beta) subunits of succinyl-CoA synthetase, we also identified both ATP-specific and GTP-specific forms of succinyl-CoA synthetase in HIT and INS cell mitochondria. Furthermore, using [gamma-32P]ATP as a phosphoryl donor, we observed that the alpha subunit of succinyl-CoA synthetase undergoes autophosphorylation at a histidine residue; co-provision of exogenous succinate and CoA resulted in pronounced dephosphorylation of the phosphorylated alpha subunit of succinyl-CoA synthetase.

CONCLUSION/INTERPRETATION

We provide evidence for the localization of two distinct activities of succinyl-CoA synthetase in the beta cell mitochondria. Whereas it is well established that ATP is critical for the beta cell mitochondrial metabolism, we propose that GTP generated by the activation of succinylCoA synthetase could promote key functional roles in the mitochondrial metabolism leading to insulin secretion.

Evidence for the carboxyl methylation of nuclear lamin-B in the pancreatic beta cell

Lamins are intermediate filament proteins that constitute the main components of the lamina underlying the inner-nuclear membrane and serve to organize chromatin. Lamins (e.g., lamin-B) undergo posttranslational modifications (e.g., isoprenylation and methylation) at their C-terminal cysteine. Such modifications are thought to render optimal association of lamins with the nuclear envelop. Herein, we examined whether nuclear lamin-B undergoes carboxyl methylation in islet beta cells. A 65- to 70-kDa protein was carboxyl methylated in intact rat islets and clonal beta (HIT or INS) cells or in homogenates which could be immunoprecipitated using lamin-B antiserum. Incubation of purified HIT cell-nuclear fraction with [(3)H]S-adenosyl methionine yielded a single carboxyl methylated protein peak (ca. 65-70 kDa); this protein was immunologically identified as lamin-B. Several methylation inhibitors, including acetyl farnesyl cysteine, a competitive inhibitor of protein prenyl cysteine methylation, inhibited the carboxyl methylation of lamin-B, indicating that the carboxyl-methylated amino acid is cysteine. These findings, together with our recent observations demonstrating that inhibition of protein isoprenylation causes apoptotic death of the pancreatic beta cell, raise an interesting possibility that inhibition of C-terminal cysteine modifications of lamin-B might result in disruption of nuclear assembly, leading to further propagation of apoptotic signals, including DNA fragmentation and chromatin condensation.

A defect late in stimulus-secretion coupling impairs insulin secretion in Goto-Kakizaki diabetic rats

A widely accepted genetically determined rodent model for human type 2 diabetes is the Goto-Kakizaki (GK) rat; however, the lesion(s) in the pancreatic islets of these rats has not been identified. Herein, intact islets from GK rats (aged 8-14 weeks) were studied, both immediately after isolation and after 18 h in tissue culture. Despite intact contents of insulin and protein, GK islets had markedly deficient insulin release in response to glucose, as well as to pure mitochondrial fuels or a non-nutrient membrane-depolarizing stimulus (40 mmol/l K+). In contrast, mastoparan (which activates GTP-binding proteins [GBPs]) completely circumvented any secretory defect. Basal and stimulated levels of adenine and guanine nucleotides, the activation of phospholipase C by Ca2+ or glucose, the secretory response to pertussis toxin, and the activation of selected low-molecular weight GBPs were not impaired. Defects were found, however, in the autophosphorylation and catalytic activity of cytosolic nucleoside diphosphokinase (NDPK), which may provide compartmentalized GTP pools to activate G-proteins; a deficient content of phosphoinositides was also detected. These studies identify novel, heretofore unappreciated, defects late in signal transduction in the islets of our colony of GK rats, possibly occurring at the site of activation by NDPK of a mastoparan-sensitive G-protein-dependent step in exocytosis.

Inosine monophosphate dehydrogenase: A molecular switch integrating pleiotropic GTP-dependent beta-cell functions

Studies of pancreatic islet function in the pathogenesis of type 2 diabetes mellitus have tended to focus on the short-term control of insulin secretion. However, the long-term control of beta-cell mass is also relevant to diabetes, since this parameter is reduced substantially even in non-insulin-dependent diabetes in humans. In animal models of type 2 diabetes, the normal balance between beta-cell proliferation and programmed cell death is perturbed. We take the perspective in this overview that inosine monophosphate dehydrogenase (IMPDH; EC 1.1.1. 205) may represent a previously neglected molecular integrator or sensor that exerts both functional (secretory) and anatomical (proliferative) effects within beta-cells. These properties reflect the fact that IMPDH is a rate-limiting enzyme in the new synthesis of the purine guanosine triphosphate (GTP), which modulates both exocytotic insulin secretion and DNA synthesis, as well as a number of other critical cellular functions within the beta-cell. Alterations in the expression or activity of IMPDH may be central to beta-cell replication, cell cycle progression, differentiation, and maintenance of adequate islet mass, effects that are probably mediated both by GTP directly, and indirectly via low molecular mass GTPases. If GTP becomes depleted, a hierarchy of beta-cell functions becomes progressively paralyzed, until eventually the effete cell is removed via apoptosis.

Prolonged depletion of guanosine triphosphate induces death of insulin-secreting cells by apoptosis

Inhibitors of IMP dehydrogenase, such as mycophenolic acid (MPA) and mizoribine, which deplete cellular GTP, are used clinically as immunosuppressive drugs. The prolonged effect of such agents on insulin-secreting beta-cells (HIT-T15 and INS-1) was investigated. Both MPA and mizoribine inhibited mitogenesis, as reflected by [3H]thymidine incorporation. Cell number, DNA and protein contents, and cell (metabolic) viability were decreased by about 30%, 60%, and 80% after treatment of HIT cells with clinically relevant concentrations (e.g. 1 microg/ml) of MPA for 1, 2, and 4 days, respectively. Mizoribine (48 h) similarly induced the death of HIT cells. INS-1 cells also were damaged by prolonged MPA treatment. MPA-treated HIT cells displayed a strong and localized staining with a DNA-binding dye (propidium iodide), suggesting condensation and fragmentation of DNA, which were confirmed by detection of DNA laddering in multiples of about 180 bp. DNA fragmentation was observed after 24-h MPA treatment and was dose dependent (29%, 49%, and 70% of cells were affected after 48-h exposure to 1, 3, and 10 microg/ml MPA, respectively). Examination of MPA-treated cells by electron microscopy revealed typical signs of apoptosis: condensed and marginated chromatin, apoptotic bodies, cytosolic vacuolization, and loss of microvilli. MPA-induced cell death was almost totally prevented by supplementation with guanosine, but not with adenosine or deoxyguanosine, indicating a specific effect of GTP depletion. An inhibitor of protein isoprenylation (lovastatin, 10-100 microM for 2-3 days) induced cell death and DNA degradation similar to those induced by sustained GTP depletion, suggesting a mediatory role of posttranslationally modified GTP-binding proteins. Indeed, impeding the function of G proteins of the Rho family (via glucosylation using Clostridium difficile toxin B), although not itself inducing apoptosis, potentiated cell death induced by MPA or lovastatin. These findings indicate that prolonged depletion of GTP induces beta-cell death compatible with apoptosis; this probably involves a direct impairment of GTP-dependent RNA-primed DNA synthesis, but also appears to be modulated by small GTP-binding proteins. Treatment of intact adult rat islets (the beta-cells of which replicate slowly) induced a modest, but definite, death by apoptosis over 1- to 3-day periods. Thus, more prolonged use of the new generation of immunosuppressive agents exemplified by MPA might have deleterious effects on the survival of islet or pancreas grafts.

Subcellular localization and characterization of nucleoside diphosphate kinase in rat retina: effect of diabetes

Nucleoside diphosphate kinase (NDP kinase) catalyzes the transfer of terminal phosphate from nucleotide triphosphates (e.g. ATP) to nucleotide diphosphates (e.g. GDP) to yield nucleotide triphosphates (e.g. GTP). Since guanine nucleotides play critical role(s) in GTP-binding protein (G-protein)-mediated signal transduction mechanisms in retina, we quantitated NDP kinase activity in subcellular fraction-derived from normal rat retina. A greater than 85% of the total specific activity was present in the soluble fraction, which was stimulated (up to 7 fold) by 2 mM magnesium. NDP kinase exhibited saturation kinetics towards di- and tri-phosphate substrates, and was inhibited by known inhibitors of NDP kinase, uridine diphosphate (UDP) or cromoglycate (CRG). We have previously reported significant abnormalities in the activation of G-proteins in streptozotocin (STZ)-diabetic rat retina (Kowluru et al. Diabetologia 35:624 631, 1992). Since NDP kinase has been implicated in direct interaction with and/or activation of various G-proteins, we quantitated both basal and magnesium-stimulated NDP kinase activity in soluble and particulate fractions of retina derived from STZ-diabetic rats to examine whether abnormalities in G-protein function in diabetes are attributable to alterations in retinal NDP kinase. There was no effect of diabetes either on the basal or the magnesium-activated retinal NDP kinase activity. This study represents the first characterization of NDP kinase activity in rat retina, and suggests that in diabetes, this enzyme may not be rate-limiting and/or causal for the observed alterations in retinal G-protein functions.

Purine nucleotide- and sugar phosphate-induced inhibition of the carboxyl methylation and catalysis of protein phosphatase-2A in insulin-secreting cells: protection by divalent cations

Recently, we demonstrated that the 36 kDa catalytic subunit of protein phosphatase 2A (PP2Ac) undergoes methylation at its C-terminal leucine in normal rat islets, human islets and isolated beta cells; this modification increases the catalytic activity of PP2A [Kowluru et al. Endocrinology. 137:2315-2323, 1996]. Previous studies have suggested that adenine and guanine nucleotides or glycolytic intermediates [which are critical mediators in beta cell function] also modulate phosphatase activity in the pancreatic beta cell. Therefore, we examined whether these phosphorylated molecules specifically regulate the carboxyl methylation and the catalytic activity of PP2A in beta cells. Micromolar concentrations of ATP, ADP, GTP or GDP each inhibited the carboxyl methylation of PP2Ac and, to a lesser degree, the catalytic activity of PP2A. Likewise, the carboxyl methylation of PP2Ac and its catalytic activity were inhibited by [mono- or di-] phosphates of glucose or fructose. Additionally, however, the carboxyl methylation of PP2Ac was significantly stimulated by divalent metal ions (Mn2+ > Mg2+ > Ca2+ > control). The nucleotide or sugar phosphate-mediated inhibition of carboxyl methylation of PP2Ac and the catalytic activity of PP2A were completely prevented by Mn2+ or Mg2+. These data indicate that divalent metal ions protect against the inhibition by purine nucleotides or sugar phosphates of the carboxyl methylation of PP2Ac perhaps permitting PP2A to function under physiologic conditions. Therefore, these data warrant caution in interpretation of extant data on the regulation of phosphatase function by purine nucleotides.

Ceramide-activated protein phosphatase-2A activity in insulin-secreting cells

Okadaic acid (OKA)-sensitive phosphatase (PP2A) activity may modulate nutrient-induced insulin secretion from pancreatic beta cells [Kowluru et al., Endocrinology 137 (1996) 2315-2323]. Ceramides, a new class of lipid second messengers may regulate PP2A [Dobrowsky and Hannun, J. Biol. Chem. (1992) 267, 5048-5051], and might play a role in cytokine-mediated apoptosis in beta cells [Sjöholm, FEBS Lett. 367 (1995) 283-286]. Therefore, we investigated the regulation of PP2A-like activity by ceramides in isolated beta (HIT-T15 or INS-1) cells. Cell-permeable (C2, C6 or C18) ceramides stimulated OKA-sensitive (but not -insensitive) phosphatase activity in a concentration-dependent manner (0-12.5 microM), with maximal stimulation (+50-100%) at < 12.5 microM. C2-dihydroceramide (a biologically inactive analog of C2 ceramide) failed to augment PP2A-like activity. Stimulatory effects of ceramides do not appear to be mediated via activation of the carboxyl methylation of the catalytic subunit of protein phosphatase 2A, since no effects of ceramides (up to 25 microM) were demonstrable on this parameter. These data identify a ceramide-activated protein phosphatase as a possible locus at which ceramides might exert their effects on beta cells leading to altered insulin secretion, and decreased cell viability followed by apoptotic cell demise.

Evidence for differential roles of the Rho subfamily of GTP-binding proteins in glucose- and calcium-induced insulin secretion from pancreatic beta cells

We utilized clostridial toxins (with known specificities for inhibition of GTPases) to ascertain the contribution of candidate GTPases in physiologic insulin secretion from beta cells. Exposure of normal rat islets or isolated beta (HIT-T15) cells to Clostridium difficile toxins A and B catalyzed the glucosylation (and thereby the inactivation) of Rac, Cdc42, and Rho endogenous to beta cells; concomitantly, either toxin reduced glucose- or potassium-induced insulin secretion from rat islets and HIT cells. Treatment of beta cells with Clostridium sordellii lethal toxin (LT; which modified only Ras, Rap, and Rac) also reduced glucose- or potassium-induced secretion. However, clostridial toxin C3-exoenzyme (which ADP-ribosylates and inactivates only Rho) was without any effect on either glucose- or potassium-induced insulin secretion. These data suggest that Cdc42, Rac, Ras, and/or Rap (but not Rho) may be needed for glucose- or potassium-mediated secretion. The effects of these toxins appear to be specific on stimulus-secretion coupling, since no difference in metabolic viability (assessed colorimetrically by quantitating the conversion of the tetrazolium salt into a formazan in a reduction reaction driven by nutrient metabolism) was demonstrable between control and toxin (A or LT)-treated beta cells. Toxin (A or LT) treatment also did not alter glucose- or potassium-mediated rises in cytosolic free calcium concentrations ([Ca2+]i), suggesting that these GTPases are involved in steps distal to elevations in [Ca2+]i. Recent findings indicate that the carboxyl methylation of Cdc42 is stimulated by only glucose, whereas that of Rap (Kowluru et al., J Clin Invest 98: 540-555, 1996) and Rac (present study) are regulated by glucose or potassium. Together, these findings provide direct evidence, for the first time, that the Rho subfamily of GTPases plays a key regulatory role(s) in insulin secretion, and they suggest that Cdc42 may be required for early steps in glucose stimulation of insulin release, whereas Rap and/or Rac may be required for a later step(s) in the stimulus-secretion coupling cascade (i.e. Ca2+-induced exocytosis of insulin).

Glucose activates the carboxyl methylation of gamma subunits of trimeric GTP-binding proteins in pancreatic beta cells. Modulation in vivo by calcium, GTP, and pertussis toxin

The gamma subunits of trimeric G-proteins (gamma1, gamma2, gamma5, and gamma7 isoforms) were found to be methylated at their carboxyl termini in normal rat islets, human islets and pure beta [HIT-T15] cells. Of these, GTPgammaS significantly stimulated the carboxyl methylation selectively of gamma2 and gamma5 isoforms. Exposure of intact HIT cells to either of two receptor-independent agonists--a stimulatory concentration of glucose or a depolarizing concentration of K+--resulted in a rapid (within 30 s) and sustained (at least up to 60 min) stimulation of gamma subunit carboxyl methylation. Mastoparan, which directly activates G-proteins (and insulin secretion from beta cells), also stimulated the carboxyl methylation of gamma subunits in intact HIT cells. Stimulatory effects of glucose or K+ were not demonstrable after removal of extracellular Ca2+ or depletion of intracellular GTP, implying regulatory roles for calcium fluxes and GTP; however, the methyl transferase itself was not directly activated by either. The stimulatory effects of mastoparan were resistant to removal of extracellular Ca2+, implying a mechanism of action that is different from glucose or K+ but also suggesting that dissociation of the alphabetagamma trimer is conducive to gamma subunit carboxyl methylation. Indeed, pertussis toxin also markedly attenuated the stimulatory effects of glucose, K+ or mastoparan without altering the rise in intracellular calcium induced by glucose or K+. Glucose-induced carboxyl methylation of gamma2 and gamma5 isoforms was vitiated by coprovision of any of three structurally different cyclooxygenase inhibitors. Conversely, exogenous PGE2, which activates Gi and Go in HIT cells and which thereby would dissociate alpha from beta(gamma), stimulated the carboxyl methylation of gamma2 and gamma5 isoforms and reversed the inhibition of glucose-stimulated carboxyl methylation of gamma subunits elicited by cyclooxygenase inhibitors. These data indicate that gamma subunits of trimeric G-proteins undergo a glucose- and calcium-regulated methylation-demethylation cycle in insulin-secreting cells, findings that may imply an important role in beta cell function. Furthermore, this is the first example of the regulation of the posttranslational modification of G-protein gamma subunits via nonreceptor-mediated activation mechanisms, which are apparently dependent on calcium influx and the consequent activation of phospholipases releasing arachidonic acid.

Glucose- and GTP-dependent stimulation of the carboxyl methylation of CDC42 in rodent and human pancreatic islets and pure beta cells. Evidence for an essential role of GTP-binding proteins in nutrient-induced insulin secretion

Several GTP-binding proteins (G-proteins) undergo post-translational modifications (isoprenylation and carboxyl methylation) in pancreatic beta cells. Herein, two of these were identified as CDC42 and rap 1, using Western blotting and immunoprecipitation. Confocal microscopic data indicated that CDC42 is localized only in islet endocrine cells but not in acinar cells of the pancreas. CDC42 undergoes a guanine nucleotide-specific membrane association and carboxyl methylation in normal rat islets, human islets, and pure beta (HIT or INS-1) cells. GTPgammaS-dependent carboxyl methylation of a 23-kD protein was also demonstrable in secretory granule fractions from normal islets or beta cells. AFC (a specific inhibitor of prenyl-cysteine carboxyl methyl transferases) blocked the carboxyl methylation of CDC42 in five types of insulin-secreting cells, without blocking GTPgammaS-induced translocation, implying that methylation is a consequence (not a cause) of transfer to membrane sites. High glucose (but not a depolarizing concentration of K+) induced the carboxyl methylation of CDC42 in intact cells, as assessed after specific immunoprecipitation. This effect was abrogated by GTP depletion using mycophenolic acid and was restored upon GTP repletion by coprovision of guanosine. In contrast, although rap 1 was also carboxyl methylated, it was not translocated to the particulate fraction by GTPgammaS; furthermore, its methylation was also stimulated by 40 mM K+ (suggesting a role which is not specific to nutrient stimulation). AFC also impeded nutrient-induced (but not K+-induced) insulin secretion from islets and beta cells under static or perifusion conditions, whereas an inactive structural analogue of AFC failed to inhibit insulin release. These effects were reproduced not only by S-adenosylhomocysteine (another methylation inhibitor), but also by GTP depletion. Thus, the glucose- and GTP-dependent carboxyl methylation of G-proteins such as CDC42 is an obligate step in the stimulus-secretion coupling of nutrient-induced insulin secretion, but not in the exocytotic event itself. Furthermore, AFC blocked glucose-activated phosphoinositide turnover, which may provide a partial biochemical explanation for its effect on secretion, and implies that certain G-proteins must be carboxyl methylated for their interaction with signaling effector molecules, a step which can be regulated by intracellular availability of GTP.

Carboxylmethylation of the catalytic subunit of protein phosphatase 2A in insulin-secreting cells: evidence for functional consequences on enzyme activity and insulin secretion

We report the carboxylmethylation of a 36-kDa protein in intact normal rat islets and clonal beta (INS-1) cells. This protein was predominantly cytosolic. Its carboxylmethylation, as assessed by vapor phase equilibration assay, was resistant to inhibition by N-acetyl-S-trans, trans-farnesyl-L-cysteine, a competitive substrate for cysteine methyl transferases. These data suggest that the methylated C-terminal amino acid is not cysteine. The methylated protein was identified as the catalytic subunit of protein phosphatase 2A (PP2Ac) by immunoblotting. The carboxylmethylation of the PP2Ac increased its catalytic activity, suggesting a key role in the functional regulation of PP2A. Therefore, we studied okadaic acid, a selective inhibitor of PP2A that acts by an unknown mechanism. Okadaic acid (but not 1-nor-okadaone, its inactive analog) inhibited (Ki = 10 nM) the carboxylmethylation of PP2Ac and phosphatase activity in the cytosolic fraction (from normal rat islets and clonal beta-cells) as well as in intact rat islets. Furthermore, methylated PP2Ac underwent rapid demethylation (t 1/2 = 40 min) catalyzed by a methyl esterase localized in islet homogenates. Ebelactone, a purported inhibitor of methyl esterases, significantly delayed (> 200 min) the demethylation of PP2Ac. Furthermore, ebelactone reversibly inhibited glucose- and ketoisocaproate-induced insulin secretion from normal rat islets. These data identify, for the first time, a methylation-demethylation cycle for PP2Ac in the beta-cell and suggest a key functional relationship between PP2A activity and the carboxylmethylation of its catalytic subunit. These findings thus suggest a negative modulatory role for PP2A in nutrient-induced insulin exocytosis.

Characterization of prenylcysteine methyltransferase in insulin-secreting cells

Prenylcysteine carboxymethyltransferase, an enzyme involved in the post-translational modification of many signalling proteins, was characterized in insulin-secreting INS-1 cells and normal rat pancreatic islets. The activity of this enzyme was monitored by the methylation of an artificial substrate (a prenylated cysteine analogue) with S-adenosy1[methyl-3H]methionine as methyl donor. More than 95% of the methyltransferase activity was associated with the membranes, and high-salt treatment only partially extracted the enzyme from the membranes. The highest specific activity was in the insulin-granule-enriched 25000 g pellet obtained by differential centrifugation. However, a highly purified insulin-enriched fraction obtained by density centrifugation in Percoll did not exhibit methyltransferase activity. The analyses of marker enzymes for cellular organelles revealed that the methyltransferase was co-localized, with the plasma membrane and probably the endoplasmic reticulum, but not with the mitochondria or lysosomes. Guanosine 5'-[gamma-thio]-triphosphate failed to increase methyltransferase activity directly, although it promotes the methylation of GTP-binding proteins. Mastoparan, Ca2+, cAMP and the protein kinase C activator phorbol 12-myristate 13-acetate did not alter enzyme activity. In addition, methyltransferase activity was not stably modified by stimulation of intact cells using glucose or other agents. However, the carboxymethylation of certain low-molecular-mass G-proteins is increased by glucose stimulation; conversely, treatment of cells with N-acetyl-S-trans,trans-farnesyl-L-cysteine inhibited glucose- and forskolin-induced insulin secretion. These results suggest that the membrane-associated prenylcysteine carboxymethyltransferase may be constitutively active and that the methylation of target proteins in vivo is regulated by the access of these proteins to the methyltransferase, as well as by their active (GTP-liganded) configuration.

A novel regulatory mechanism for trimeric GTP-binding proteins in the membrane and secretory granule fractions of human and rodent beta cells

Recently we described roles for heterotrimeric and low-molecular-mass GTP-binding proteins in insulin release from normal rat islets. During these studies, we observed that a protein with an apparent molecular mass (37 kDa) similar to that of the beta subunit of trimeric GTP-binding proteins underwent phosphorylation in each of five classes of insulin-secreting cells. Incubation of the beta cell total membrane fraction or the isolated secretory granule fraction (but not the cytosolic fraction) with [gamma-32P]ATP or [gamma-32P]GTP resulted in the phosphorylation of this protein, which was selectively immunoprecipitated by an anti-serum directed against the common beta subunit of trimeric G-proteins. Disruption of the alpha beta gamma trimer (by pretreatment with either fluoroaluminate or guanosine 5'(-)[gamma-thio]triphosphate) prevented beta subunit phosphorylation. Based on differential sensitivities to pH, heat and the histidine-selective reagent diethyl pyrocarbonate (and reversal of the latter by hydroxylamine), the phosphorylated amino acid was presumptively identified as histidine. Incubation of pure beta subunit alone or in combination with the exogenous purified alpha subunit of transducin did not result in the phosphorylation of the beta subunit, but addition of the islet cell membrane fraction did support this event, suggesting that membrane localization (or a membrane-associated factor) is required for beta subunit phosphorylation. Incubation of phosphorylated beta subunit with G alpha.GDP accelerated the dephosphorylation of the beta subunit, accompanied by the formation of G alpha-GTP. Immunoblotting detected multiple alpha subunits (of Gi, G(o) and Gq) and at least one beta subunit in the secretory granule fraction of normal rat islets and insulinoma cells. These data describe a potential alternative mechanism for the activation of GTP-binding proteins in beta cells which contrasts with the classical receptor-agonist mechanism: G beta undergoes transient phosphorylation at a histidine residue by a GTP-specific protein kinase; this phosphate, in turn, may be transferred via a classical Ping-Pong mechanism to G alpha.GDP (inactive), yielding the active configuration G alpha.GTP in secretory granules (a strategic location to modulate exocytosis).

Non-specific stimulatory effects of mastoparan on pancreatic islet nucleoside diphosphokinase activity: dissociation from insulin secretion

We examined whether mastoparan (MAS)-induced insulin secretion might involve the activation of nucleoside diphosphokinase (NDP kinase), which catalyzes the conversion of GDP to GTP, a known permissive factor for insulin secretion. MAS and MAS 7 (which activate GTP-binding proteins), but not MAS 17 (an inactive analog), stimulated insulin secretion from normal rat islets. In contrast to their specific effects on insulin secretion, MAS, MAS 7 and MAS 17 each stimulated formation of the phosphoenzyme-intermediate of NDP kinase, as well as its catalytic activity. These effects were mimicked by several cationic drugs. Thus, caution is indicated in using MAS to study cellular regulation, since some of its effects appear to be non-specific, and may be due, in part, to its amphiphilic, cationic nature.

Characterization of nucleoside diphosphokinase activity in human and rodent pancreatic beta cells: evidence for its role in the formation of guanosine triphosphate, a permissive factor for nutrient-induced insulin secretion

We have recently demonstrated a permissive role for GTP in nutrient-induced insulin secretion. One of the possible loci at which GTP might exert its regulatory effects include one (or more) of the GTP-binding proteins which we have identified in subcellular fractions (including secretory granules) of pancreatic islets. Herein, we characterize nucleoside diphosphokinase (NDP kinase) activity, which catalyzes the transphosphorylation of nucleotide diphosphate (e.g., GDP) to nucleotide triphosphates (e.g., GTP) in insulin-secreting cells. The presence of NDP kinase activity in normal rat and human islets, and pure beta (RIN and HIT) cells, was verified by three distinct approaches: first, its catalytic activity (formation of GTP or GTP gamma S from GDP and ATP or ATP gamma S); secondly, by immunologic detection; and third, by quantitating the phosphoenzyme intermediate of NDP kinase, which is involved in a ping-pong phosphotransfer mechanism. Subcellularly, NDP kinase is predominantly cytosolic (with a tetrameric molecular mass of 85-90 kDa) and requires divalent metal ions and thiols for its activity. UDP, which forms an abortive complex with the enzyme, inhibited its activity in a concentration-dependent manner (Ki = 2 mM). The phosphorylated intermediate of NDP kinase was differentially sensitive to heat, acidic pH, and a histidine-selective reagent, diethyl pyrocarbonate, suggesting that (one of) the phosphoamino acid(s) may be histidine. These data demonstrate that in beta cells NDP kinase undergoes transient phosphorylation and suggest that this phosphate, in turn, is transferred to GDP. If the GTP which is formed thereby is bound to, or channelled to, relevant GTP-binding proteins, it would facilitate the formation of active form of these proteins.(ABSTRACT TRUNCATED AT 250 WORDS)

Subcellular localization and kinetic characterization of guanine nucleotide binding proteins in normal rat and human pancreatic islets and transformed beta cells

The subcellular localization and the kinetics of the GTPase activities of monomeric and heterotrimeric GTP-binding proteins were investigated in normal rat and human pancreatic islets and were compared to those obtained using a transformed hamster beta cell line (HIT cells). The [alpha-32P]GTP overlay technique revealed the presence of at least four low-molecular-mass proteins (approx. 20-27 kDa) in normal rat islets, which were enriched in the secretory granule fraction compared to the membrane fraction (with little abundance of these proteins in the cytosolic fraction). In contrast, in HIT cells, these proteins (at least six) were predominantly cytosolic. Three of these proteins were immunologically identified as rab3A, rac2, and CDC42Hs in islets as well as in HIT cells. In addition, pertussis toxin augmented the ribosylation of at least one heterotrimeric G-protein of about 39 kDa (probably G(i) and/or G(o)) in the membrane and secretory granule fractions of normal rat and human islets, whereas at least three such Ptx substrates (36-39 kDa) were found in HIT cell membranes. Kinetic activities revealed the presence of at least three such activities (Km for GTP of 372 nM, 2.2 microM, and 724 microM) in islet homogenates which were differentially distributed in various subcellular fractions; similar activities were also demonstrable in HIT cell homogenates. Thus, these studies demonstrate the presence of both monomeric G-proteins intrinsic to the secretory granules of normal rat islets which can be ascribed to beta cells; since these G-proteins are regulated by insulinotropic lipids (as described in the accompanying article), such proteins may couple the activation of phospholipases (endogenous to islets) to the exocytotic secretion of insulin. These findings also suggest that caution is necessary in extrapolating data concerning G-proteins from cultured, transformed beta cell lines to the physiology of normal islets, in view of both qualitative and quantitative differences between the two preparations.

Regulation of guanine-nucleotide binding proteins in islet subcellular fractions by phospholipase-derived lipid mediators of insulin secretion

In the accompanying article (Kowluru, A., Rabaglia, M.E., Mose, K.E. and Metz, S.A. (1994) Biochim. Biophys. Acta 1222, 348-359) we identified three specific GTPase activities in islet subcellular fractions; most notably, two of these were enriched in the secretory granules. In the present study, we describe the regulation of GTPase activity in subcellular fractions of normal rat and human islets by insulinotropic lipids with a similar rank order as their insulin-releasing capacity. Arachidonic acid (AA), lysophosphatidylcholine (LPC), or phosphatidic acid (PA) inhibited the GTPase activities significantly (by 60-80%) in islet homogenates; each also selectively inhibited certain GTPases in specific individual fractions. Less insulinotropic fatty acids, such as linoleic acid and oleic acid, inhibited GTPase to a lesser degree, whereas lysophosphatidic acid (LPA), phosphatidylcholine (PC) or palmitic acid, which do not acutely promote secretion, were ineffective. Similar inhibitory effects of these lipids were also demonstrable in fractions of human islets as well as those of transformed beta-cells (HIT cells). The effects of lipids were not attributable to their detergent properties (since several detergents failed to mimic lipid effects) or to inhibition of GTP binding (since they actually increased GTP gamma S binding modestly, and moreover, in reconstituted fractions, they potentiated GDP/GTP exchange activity up to 2-fold). These data indicate that the insulinotropic nature of the lipids might be due, in part, to their ability to maintain G-proteins in their GTP-bound (active) configuration by increasing GTP binding and decreasing its hydrolysis. These studies comprise the first evidence for the regulation by biologically active lipids of endocrine cell G-proteins at a locus distal to plasma membrane events (i.e., on endocrine secretory granules), and provide thereby a possible novel mechanism whereby the activation of islet endogenous phospholipases might culminate in insulin exocytosis.

Stimulation by prostaglandin E2 of a high-affinity GTPase in the secretory granules of normal rat and human pancreatic islets

Recent reports of a pertussis-toxin (Ptx)-sensitive inhibition of glucose-induced insulin release by prostaglandin E2 (PGE2) in transformed beta-cells prompted us to look for the presence of prostaglandin-regulatable GTP-binding proteins (G-proteins) on the secretory granules of normal pancreatic islets. PGE2 (but not PGF2 alpha, PGA2, PGB2 or PGD2) stimulated in a concentration-dependent manner a high-affinity GTPase activity in the secretory-granule-enriched fractions of both normal rat and human islets. Similar results were found after sucrose-density-gradient-centrifugation-based isolation of secretory granules to those after a differential-centrifugation procedure. Half-maximal stimulation occurred at 800 nM PGE2, a concentration known to inhibit both phases of glucose-induced insulin secretion from pure beta-cell lines. The GTPase stimulatory effect of PGE2 was blocked virtually totally by Ptx pretreatment; it was not due to an effect on substrate binding since no measurable effect of PGE2 on binding of guanosine 5'-[gamma-[35S]thio]triphosphate was observed in cognate fractions. Other Ptx-sensitive inhibitors of insulin secretion (such as adrenaline or clonidine) also stimulated GTPase activity, suggesting that one (or more) inhibitory exocytotic G-proteins (i.e. a putative GEi) is located on the secretory granules. These studies demonstrate, for the first time in an endocrine gland, the presence of a regulatable G-protein, strategically located on the secretory granules where it might regulate the exocytotic cascade distal to both plasma-membrane events and the generation of soluble mediators of insulin secretion.

Modulation of insulin secretion from normal rat islets by inhibitors of the post-translational modifications of GTP-binding proteins

Many GTP-binding proteins (GBPs) are modified by mevalonic acid (MVA)-dependent isoprenylation, carboxyl methylation or palmitoylation. The effects of inhibitors of these processes on insulin release were studied. Intact pancreatic islets were shown to synthesize and metabolize MVA and to prenylate several candidate proteins. Culture with lovastatin (to inhibit synthesis of endogenous MVA) caused the accumulation in the cytosol of low-M(r) GBPs (labelled by the [alpha-32P]GTP overlay technique), suggesting a disturbance of membrane association. Concomitantly, lovastatin pretreatment reduced glucose-induced insulin release by about 50%; co-provision of 100-200 microM MVA totally prevented this effect. Perillic acid, a purported inhibitor of the prenylation of small GBPs, also markedly reduced glucose-induced insulin secretion. Furthermore, both N-acetyl-S-trans,trans-farnesyl-L-cysteine (AFC), which inhibited the base-labile carboxyl methylation of GBPs in islets or in transformed beta-cells, and cerulenic acid, an inhibitor of protein palmitoylation, also reduced nutrient-induced secretion; an inactive analogue of AFC (which did not inhibit carboxyl methylation in islets) had no effect on secretion. In contrast with nutrients, the effects of agonists that induce secretion by directly activating distal components in signal transduction (such as a phorbol ester or mastoparan) were either unaffected or enhanced by lovastatin or AFC. These data are compatible with the hypothesis that post-translational modifications are required for one or more stimulatory GBPs to promote proximal step(s) in fuel-induced insulin secretion, whereas one or more inhibitory GBPs might reduce secretion at a more distal locus.

Small elevations of glucose concentration redirect and amplify the synthesis of guanosine 5'-triphosphate in rat islets

Recent studies suggest a permissive requirement for guanosine 5'-triphosphate (GTP) in insulin release, based on the use of GTP synthesis inhibitors (such as myocophenolic acid) acting at inosine monophosphate (IMP) dehydrogenase; herein, we examine the glucose dependency of GTP synthesis. Mycophenolic acid inhibited insulin secretion equally well after islet culture at 7.8 or 11.1 mM glucose (51% inhibition) but its effect was dramatically attenuated when provided at < or = 6.4 mM glucose (13% inhibition; P < 0.001). These observations were explicable by a stimulation of islet GTP synthesis derived from IMP since, at high glucose: (a) total GTP content was augmented; (b) a greater decrement in GTP (1.75 vs. 1.05 pmol/islet) was induced by mycophenolic acid; and (c) a smaller "pool" of residual GTP persisted after drug treatment. Glucose also accelerated GTP synthesis from exogenous guanine ("salvage" pathway) and increased content of a pyrimidine, uridine 5'-triphosphate (UTP), suggesting that glucose augments production of a common regulatory intermediate (probably 5-phosphoribosyl-1-pyrophosphate). Pathway-specific radiolabeling studies confirmed that glucose tripled both salvage and de novo synthesis of nucleotides. We conclude that steep changes in the biosynthesis of cytosolic pools of GTP occur at modest changes in glucose concentrations, a finding which may have relevance to the adaptive (patho) physiologic responses of islets to changes in ambient glucose levels.

Functional alterations of G-proteins in diabetic rat retina: a possible explanation for the early visual abnormalities in diabetes mellitus

We examined changes in guanosine triphosphate-dependent signal transduction mechanisms in the retina from the early stages of the streptozotocin-diabetic rat, a model for Type 1 (insulin-dependent) diabetes mellitus. Guanosine triphosphate binding, guanosine triphosphatase activity, and binding of (azido) guanosine triphosphate decreased significantly in the retina as early as 2 weeks after the induction of diabetes. The ability of guanosine triphosphate to inhibit forskolin-stimulatable adenyl cyclase was also abolished. These data suggest functional deterioration of G-proteins, especially Gi, in diabetic retina. Further studies using retinal rod outer segments revealed deterioration in light-sensitive, guanosine triphosphate-dependent functions of transducin in diabetic rats. Pertussis toxin-catalysed ADP ribosylation of the alpha subunit of transducin, a heterotrimeric G-protein of rod outer segments, was also reduced in diabetes. No functional effects were seen in purified subunits of transducin subjected to non-enzymatic glycation in vitro. On the other hand, incubation of non-diabetic rod outer segments with (12-0-tetradeconyl) phorbol-13-acetate, a protein kinase C agonist, in the presence of magnesium and adenosine triphosphate resulted in the reduction of guanosine triphosphate-binding and hydrolysis, thus indicating that protein kinase C may be involved in the regulation of these activities. The significance of these observations in the early visual abnormalities associated with diabetes is discussed.

Phospholipid N-methylation in diabetic erythrocytes: effects on membrane Na+, K+ ATPase activity

Phospholipid methylation was quantified in non-diabetic and streptozotocin diabetic rat erythrocytes. While the total mass of methylated lipids remained the same in both groups, the relative abundance of individual methylated lipid species differed significantly in diabetic erythrocytes. Moreover, incubation of erythrocytes membranes with S-adenosyl methionine, a substrate for methyl transferases, not only increased membrane lipid methylation but also decreased Na+, K+ ATPase activity significantly. These results suggest that phospholipid methylation may cause the observed depression of erythrocyte Na+, K+ ATPase activity in diabetes and could contribute to the altered rheology of erythrocytes in diabetes.

Preferential excretion of glycated albumin in C57BL-Ks-J mice: effects of diabetes

Urinary excretion of glycated albumin was quantitated in genetically hyperglycemic mice (C57BL-Ks-J, db/db mice), a model for non-insulin-dependent diabetes mellitus, and compared with their non-diabetic littermates. The data indicated a preferential excretion of glycated albumin in non-diabetic mice. This phenomenon of 'editing' of glycated albumin is decreased significantly in diabetic mice. Quantitative measurements of overall excretion of glycated albumin suggested that the loss of editing in diabetic mice is due to the dilution of glycated albumin by the unmodified albumin which is excreted in large amounts in diabetic mice. Therefore, the loss of editing observed in this model resembled the one we characterized in insulin-dependent diabetic humans and a streptozotocin-diabetic rat model.

Erythrocyte sodium-potassium ATPase activity and thiol metabolism in genetically hyperglycemic mice

Erythrocyte sodium pump activity, osmotic fragility, and thiol status were measured in genetically hyperglycemic (db/db) mice and compared with their nondiabetic littermates (db/m). The data showed no major differences in these parameters. However, erythrocytes from streptozotocin (Stz)-induced diabetic rats had significantly lower activity of sodium pump and thiols with an almost fourfold increase in osmotic fragility as compared with erythrocytes from nondiabetic rats. Sorbinil (an aldose reductase inhibitor) treatment of Stz-diabetic rats normalized all these lesions, suggesting a key role for polyol pathway. However, sorbitol levels in erythrocytes from db/db and db/m mice were undetectable. The data suggest that in db/db mice, the relative lack of polyol pathway, a potential consumer of NADPH, may provide erythrocytes with optimal NADPH for glutathione reductase system, thus maintaining normal GSH levels even at the height of hyperglycemia. Thus, the genetically hyperglycemic mice may serve as a useful model to study diabetes related complications without involving polyol pathway.

Protein glycation: effects upon protein recognition by the proximal tubule

Reabsorption of 125I labeled unmodified and glycated albumin by the proximal tubule was studied using micropuncture technique in wistar rats of different ages. The data show a significant reabsorption of unmodified albumin while the glycated albumin was virtually excluded from reabsorption process. Among different explanations considered, we think molecular charge of the protein seems to be a key factor for the altered recognition of glycated albumin by the proximal tubule.

Reversible sodium pump defect and swelling in the diabetic rat erythrocyte: effects on filterability and implications for microangiopathy

We have found a defect in the ouabain-sensitive Na+, K+-ATPase (Na+ pump, EC 3.6.1.37) of erythrocytes from streptozocin diabetic rats. This defect was accompanied by an increase in cell volume and osmotic fragility and a decrease in the cytosolic K+/Na+ ratio. There was also a doubling in the time needed for diabetic erythrocytes to pass through 4.7-micron channels in a polycarbonate filter. Our data are consistent with a primary defect in the erythrocyte Na+ pump and secondary changes in cell volume, osmotic fragility, K+/Na+ ratio, and cell filterability. All were reversed or prevented in vivo by insulin or the aldose reductase inhibitor Sorbinil. Protein kinase C agonists (phorbol ester and diacylglycerol) and agonist precursor (myoinositol) reversed the Na+ pump lesion, suggesting that protein kinase C-dependent phosphorylation of the 100-kDa subunit regulates Na+ pump activity and that insulin can influence erythrocyte protein kinase C activity. Ouabain inhibition of the erythrocyte Na+ pump also produced increases in cell size and reductions in rates of filtration. Theoretical treatment of the volume changes also predicts reduction in filterability as a consequence of cell swelling. We suggest that enlarged erythrocytes could play a role in the evolution of the microvascular changes of diabetes mellitus.

Glycation of calmodulin: chemistry and structural and functional consequences

In the presence of Ca2+ and glucose, calmodulin incorporates 2.5 mol of glucose/mol of protein. In the absence of Ca2+, only 1.5 mol of glucose is incorporated per mole of calmodulin. Glycation of calmodulin is associated with variable reductions in its capacity to activate three Ca2+/calmodulin-dependent brain target enzyme systems, including adenylyl cyclase, phosphodiesterase, and protein kinase. In addition, glycated calmodulin exhibits a 54% reduction in its Ca2+ binding capacity. Isolated CNBr cleavage fragments of glycated calmodulin suggest that glycation follows a nonspecific pattern in that each of seven available lysines is susceptible to modification. A limit observed on the extent of glycation appears related to the accompanying increase in negative charge on the protein. Glycation results in minimal structural rearrangements in calmodulin, and the Ca2+-induced increase in alpha-helix content and radius of gyration is the same for glycated and unmodified calmodulin. Since glycated calmodulin's Ca2+ binding capacity is reduced, this implies that the Ca2+-induced conformational changes in calmodulin do not require all four Ca2+ binding sites to be occupied. Examination of the lysine positions in calmodulin suggests that Ca2+ binding to domains II and IV is sufficient to induce these changes. The functional consequences of calmodulin glycation therefore cannot be attributed to inhibition of these conformational changes. An alternative explanation is that the inhibition arises from interference at the target enzyme binding site by bound glucose. While glycation shows minimal structural effects, a large pH dependence is observed for the alpha-helix content of unmodified calmodulin.(ABSTRACT TRUNCATED AT 250 WORDS)

Suggested mechanism for the selective excretion of glucosylated albumin. The effects of diabetes mellitus and aging on this process and the origins of diabetic microalbuminuria

In previous studies in the Sprague-Dawley rat, Williams and coworkers reported the phenomenon of selective urinary excretion of glucosylated albumin (editing, i.e., the percent glucosylation of urinary albumin is more than that of plasma albumin) by the mammalian kidney. Ghiggeri and coworkers subsequently found that the extent of editing is reduced in human diabetics. Moreover, the reduction in editing in diabetes correlates inversely with levels of microalbuminuria. We also find reduction in the extent of editing in diabetic humans. We find a striking inverse correlation not only with the magnitude of microalbuminuria but also with the extent of plasma albumin glucosylation. In contrast, we found little correlation between the reduction in editing and the duration of diabetes in human subjects. Stz induced diabetes in the Sprague-Dawley rat is associated with a striking and rapid reduction in editing which develops virtually with the same kinetics exhibited by the appearance of hyperglycemia. This loss of editing is rapidly reversed by daily administration of insulin but not by aldose reductase inhibitors. Mannitol infusion in anesthetized Wistar rats resulted in an increase in urine volume, GFR, and microalbuminuria, and was also accompanied by a marked reduction in editing. This reduction was rapidly reversed by a cessation of mannitol infusion. We propose here that glucosylated albumin (in contrast to unmodified albumin) is not reabsorbed by the proximal tubule, and thus, is preferentially excreted in the urine. We postulate that the increase in GFR which emerges as a consequence of increased plasma osmolality in diabetes mellitus delivers more albumin to the proximal tubule than can be reabsorbed. This results in a dilution of excreted glucosylated albumin molecules by excreted unmodified albumin, which appears as the early microscopic albuminuria of diabetes. Paradoxically, the fall in apparent editing is accompanied by an absolute increase in the total quantity of glucosylated albumin excreted. In contrast, we found that editing of glucosylated albumin by the normal kidney is found to gradually decline as a function of age without the appearance of microalbuminuria. This suggests that a different mechanism operates to produce the loss of editing seen with aging in man, and as clearly (but in a shorter absolute time intervals) in the Fischer-344 rat.

Secretagogue-responsive and -unresponsive pools of phosphatidylinositol in pancreatic islets

The effect of glucose on phosphatidylinositol turnover was studied. Phosphatidylinositol of rat pancreatic islets was labeled with myo[2-3H]inositol in the presence of various secretagogues (16.7 mM D-glucose, 22 mM D-mannose, 20 mM D-glyceraldehyde) and nonsecretagogues (3.3 mM D-glucose, 20 mM pyruvate, 16.7 mM D-galactose, 16.7 mM L-glucose). Upon subsequent stimulation with 16.7 mM D-glucose, only the islets that were labeled in the presence of secretagogues showed a loss of radioactivity from phosphatidylinositol. No loss of radioactivity from phosphatidylinositol occurred in the presence of 3.3 mM D-glucose even after labeling in the presence of secretagogues. A comparison of the subcellular distribution of labeled phosphatidylinositol in islets before and after stimulation with insulinotropic glucose revealed a loss of radioactivity from the plasma membrane fraction as judged by subcellular fractionation with a sucrose gradient. These results support a hypothesis advanced previously that pancreatic islets contain a unique pool of phosphatidylinositol that undergoes rapid turnover only in the presence of insulinotropic concentrations of D-glucose or other secretagogues [R. S. Rana, R. J. Mertz, A. Kowlura, J. F. Dixon, L. E. Hokin, and M. J. MacDonald (1985) J. Biol. Chem. 260, 7861-7867]. On the basis of the subcellular fractionation studies reported here, the secretagogue-responsive phosphatidylinositol pool appears to be located primarily in the plasma membrane of pancreatic islets.

Enzymes of phospholipid metabolism in rat pancreatic islets: subcellular distribution and the effect of glucose and calcium

The effect of glucose and calcium on the activities of the phosphatidylinositol cycle enzymes, CDP-diglyceride inositol transferase, diacylglycerokinase, and lysophosphatidylcholine 2-acyltransferase in rat pancreatic islets was studied. Calcium inhibited the activity of CDP-diglyceride inositol transferase but had no effect on lysophosphatidylcholine 2-acyltransferase and diacylglycerokinase activities. Upon preincubation of islets in a concentration of glucose known to stimulate insulin release, the activity of lysophosphatidylcholine 2-acyltransferase, but not that of diacylglycerokinase or the CDP-diglyceride inositol transferase, was stimulated. Subcellular fractionation of pancreatic islets showed that secretory granule membranes were enriched in CDP-diglyceride inositol transferase, whereas lysophosphatidylcholine 2-acyltransferase activity was highest in the microsomal membranes. The activation of 2-acyltransferase by incubating islets in insulinotropic glucose, and the calcium sensitivity of CDP-diglyceride inositol transferase, suggest that these enzymes may have roles in regulation of insulin secretion.

Evidence for calcium enhanced phosphorylation of pyruvate kinase by pancreatic islets

Pancreatic islet cytosol contains a calcium-calmodulin dependent protein kinase that can mediate the phosphorylation of an endogenous protein that has an Mr of 57 000, as well as exogenous muscle pyruvate kinase (subunit Mr, 57 000). EGTA and trifluoperazine decreased the phosphorylation. Alkaline inactivation of pyruvate kinase made it a better substrate for the kinase. As in rat islet cytosol, rabbit islet cytosol catalyzed the phosphorylation of a 57 000 Mr protein in the presence of calcium and calmodulin. This phosphoprotein was immunoprecipitated with anti-pyruvate kinase antibody. This is consistent with the idea that the 57 000 Mr phosphoprotein in islet cytosol is the subunit of pyruvate kinase. The paper following this paper shows that the kinetic and immunologic properties of the islet pyruvate kinase indicate it is the M2 isoenzyme and that its phosphorylation does not affect its catalytic activity.

Phospholipid methyltransferase activity in pancreatic islets: activation by calcium

Pancreatic islet homogenates contain a Mg2+-requiring phospholipid methyltransferase activity, the activity of which was doubled by calcium (K0.5 less than 5 microM). Other divalent metal ions stimulated the activity from 11 to 35%, but zinc and strontium were inhibitory. Cyclic AMP had no effect on the enzyme activity and cyclic GMP inhibited it slightly. Calcium increased the Vmax of the enzyme without affecting its Km with respect to S-adenosylmethionine (6 microM). Chlorpromazine, trifluoperazine, and dibucaine inhibited the calcium-stimulatable activity without affecting the activity in the absence of calcium. Phosphatidylserine stimulated, and arachidonic acid and palmitic acid inhibited, the basal enzyme activity. The methylated products were found to be primarily mono- and dimethylphosphatidylethanolamine (30%) and phosphatidylcholine (43%) and an, as yet unidentified, nonpolar lipid fraction (27%), as judged by thin-layer chromatography. In the presence of calcium, incorporation of methyl groups into phosphatidylcholine, mono- and dimethylphosphatidylethanolamine, and nonpolar lipids was increased by 131, 60, and 46%, respectively. Based on the localization of the enzyme activity in the insulin secretory granule fraction, it is proposed that phospholipid methylation plays a role in coupling the stimulus to the initial events in insulin secretion, leading to the exocytosis of insulin.

Evidence for glucose-responsive and -unresponsive pools of phospholipid in pancreatic islets

The effect of glucose on the metabolism of phospholipids in pancreatic islets was studied with three radioactive phospholipid precursors, [32P]orthophosphate, [3H]myoinositol, and [3H]arachidonic acid, to determine the conditions necessary for studying the breakdown of prelabeled phospholipids. Islets were incubated in the presence of a radioactive precursor for 60 or 90 min and in the presence of either 3.3 or 16.7 mM glucose to prelabel phospholipids. To study the breakdown of prelabeled phospholipid, the unincorporated precursor was removed and the islets were reincubated for 15 or 20 min under conditions that either did or did not stimulate insulin release. Prelabeling in the presence of a noninsulinotropic concentration of glucose (3.3 mM) supported the incorporation of precursors into almost all islet phospholipids studied. Prelabeling in an insulinotropic concentration of glucose (16.7 mM) increased the incorporation of precursors into a number of phospholipids even more; and reincubation in 16.7 mM glucose caused a rapid loss of radioactivity from specific phospholipids (phosphatidylinositol and/or phosphatidylcholine, depending on the precursor). This breakdown was observed only when islets had been prelabeled in 16.7 mM glucose. The amount of radioactivity lost from phospholipid corresponded roughly to the additional amount incorporated during the prelabeling in the high concentration of glucose. Radioactivity in phospholipids in islets prelabeled in 3.3 mM glucose or in nonsecretagogue metabolic fuels, such as malate plus pyruvate, did not decrease when the islets were subsequently exposed to 16.7 mM glucose, nor did it decrease in 3.3 mM glucose when these islets had been prelabeled in 16.7 mM glucose. Glyceraldehyde, an insulin secretagogue, but not galactose or L-glucose which are not insulin secretagogues, stimulated phospholipid breakdown in islets that had been prelabeled in 16.7 mM glucose. Depriving islets of extracellular calcium, a condition that inhibits insulin release, inhibited phospholipid breakdown. The results suggest that pancreatic islets contain a glucose-responsive and a glucose-unresponsive phospholipid pool. The glucose-responsive pool becomes labeled and undergoes rapid turnover only under stimulatory conditions and may play a role in the stimulus-secretion coupling of insulin release.

Evidence for phosphorylation of pancreatic islet pyruvate kinase

Rabbit pancreatic islet cytosol catalyzes the calcium-activated phosphorylation by [gamma 32P]ATP of a protein with a molecular weight of 57,000 that is precipitated with antipyruvate kinase antibodies. We were unable to demonstrate that phosphorylation in the presence of calcium or cAMP had any immediate effect on rat pancreatic islet pyruvate kinase activity. This finding is consistent with our inability to confirm the finding of others that pancreatic islets contain phosphoenolpyruvate carboxykinase activity (Diabetes, 34:246, 1985). Since the carboxykinase catalyzes phosphoenolpyruvate formation and pyruvate kinase catalyzes essentially the opposite reaction, if the carboxykinase were present in the beta cell, pyruvate kinase would need to be inhibited to prevent recycling of phosphoenolpyruvate.

Activation of pancreatic islet myosin ATPase by ATP and actin

Previous work from our laboratory indicated that pancreatic islets contain myosin light chain kinase, a calcium- and calmodulin-activated enzyme. This enzyme catalyzes phosphorylation of myosin which, in tissues containing smooth muscle, is believed to permit the ATPase of myosin to be activated by actin. The current report shows that incubating islet cytosol with ATP under conditions that should permit phosphorylation of myosin markedly enhances islet myosin ATPase activity in the presence of actin. It has been suggested that contractile proteins power insulin granule movements in the beta cell. Phosphorylation of myosin may be one of the means of coupling stimuli to insulin secretion.

Stimulation of phospholipid methylation by glucose in pancreatic islets

A two fold stimulation in the incorporation of [3H-methyl] groups from [3H-methyl] methionine into phospholipids was seen in intact pancreatic islets within six minutes of exposure to a glucose concentration that stimulates insulin release. Nonstimulatory sugars, L-glucose and D-galactose, as well as dibutyryl cAMP, did not affect phospholipid methylation in islet cells. A calcium channel blocker, verapamil, inhibited methylation. These studies suggest that the signal for glucose-induced insulin release could involve phospholipid methylation.

Glucose-induced, calcium-mediated protein phosphorylation in intact pancreatic islets

Conditions for studying protein phosphorylation in intact pancreatic islets were developed in order to study the effects of glucose and other effectors. Islets were incubated in Krebs-Ringer bicarbonate buffer containing 5 mM malate and 5 mM pyruvate (metabolic fuels that are not insulin secretagogues) for 150 min to permit incorporation of 32Pi into islet phosphate pools. Glucose or other effectors were then added, and the incubation was terminated after 10 to 30 min. Glucose increased phosphorylation of four islet peptides with molecular weights of 20,000, 33,000, 43,000 and 57,000. The calcium channel blockers, verapamil and D-600, inhibited phosphorylation of each of the four proteins, and trifluoperazine inhibited phosphorylation of the proteins with molecular weights of 20,000 and 57,000. The results indicate that glucose-induced insulin release may be mediated in part by protein phosphorylation, and that calcium may act as an intracellular messenger in coupling the glucose stimulus to the secretory process.

Protein phosphorylation in pancreatic islets: evidence for separate Ca2+ and cAMP-enhanced phosphorylation of two 57,000 Mr proteins

There is a phosphopeptide that has an Mr of 53,000 to 60,000 in insulin-secreting tissues and there is general agreement that this peptide can be phosphorylated in a calcium-dependent manner. The present report shows that there are at least two phosphoproteins with Mr's near 57,000 in rat pancreatic islet cytosol. One peptide has an Mr of 57,000, a pl of 7.5 - 8 and is phosphorylated in a Ca2+-enhanced manner, and the other has an Mr of 54,000, a pl of 5 - 5.5 and is phosphorylated in a cAMP-enhanced manner, as judged by two-dimensional polyacrylamide gel electrophoresis. Sepharose 4B chromatography indicated that the former polypeptide resides in a native protein complex that has an Mr of about 500,000 and the latter in a complex that has an Mr of about 180,000. Tritiated azido cyclic AMP binds to an islet polypeptide that has an Mr of 54,000. The results suggest that Ca2+ and cAMP could regulate stimulus-secretion coupling in pancreatic islets via protein phosphorylation.