Effects of PYY[3-36] in rodent models of diabetes and obesity.

BACKGROUND: Peptide YY (PYY) is a 36 amino-acid peptide secreted from ileal L cells following meals. The cleaved subpeptide PYY[3-36] is biologically active and may constitute the majority of circulating PYY-like immunoreactivity. The peptide family that includes PYY, pancreatic peptide and neuropeptide Y is noted for its orexigenic effect following intracerebroventricular administration. OBJECTIVE: To investigate the effects of peripheral (intraperitoneal and chronic subcutaneous) infusions of PYY[3-36] on food intake, body weight and glycemic indices. DESIGN/RESULTS: Food intake was measured in normal mice and in several rodent models of obesity and type II diabetes. In marked contrast to the reported central orexigenic effects, in the present study, PYY[3-36] acutely inhibited food intake by up to 45%, with an ED(50) of 12.5 microg/kg in fasted female NIH/Swiss mice. A 4-week infusion reduced weight gain in female ob/ob mice, without affecting the cumulative food intake. In diet-induced obese male mice, PYY[3-36] infusion reduced cumulative food intake, weight gain and epididymal fat weight (as a fraction of carcass) with similar ED(50)'s (466, 297 and 201 microg/kg/day, respectively) and prevented a diet-induced increase in HbA1c. Infusion at 100 microg/kg/day for 8 weeks in male fa/fa rats reduced the weight gain (288+/-11 vs 326+/-12 g in saline-infused controls; P<0.05), similar to effects in a pair-fed group. In female ob/ob and db/db mice, there was no acute effect of PYY[3-36] on plasma glucose concentrations. In male diabetic fatty Zucker rats, PYY[3-36] infused for 4 weeks reduced HbA1c and fructosamine (ED(50)'s 30 and 44 microg/kg/day). CONCLUSION: Peripheral PYY[3-36] administration reduced the food intake, body weight gain and glycemic indices in diverse rodent models of metabolic disease of both sexes. These findings justify further exploration of the potential physiologic and therapeutic roles of PYY[3-36].

Selective amylin inhibition of the glucagon response to arginine is extrinsic to the pancreas.

Amylin, a peptide hormone from pancreatic beta-cells, is reported to inhibit insulin secretion in vitro and in vivo and to inhibit nutrient-stimulated glucagon secretion in vivo. However, it has been reported not to affect arginine-stimulated glucagon secretion in vitro. To resolve if the latter resulted from inactive peptide (a problem in the early literature), those experiments were repeated here with well-characterized peptide and found to be valid. In isolated perfused rat pancreas preparations, coperfusion with 1 nM amylin had no effect on arginine-, carbachol-, or vasoactive intestinal peptide-stimulated glucagon secretion. Amylin also had no effect on glucagon output stimulated by decreasing glucose concentration from 11 to 3.2 mM or on glucagon suppression caused by increasing glucose from 3.2 to 7 mM. Amylin at 100 nM had no effect in isolated islets in which glucagon secretion was stimulated by exposure to 10 mM arginine, even though glucagon secretion in the same preparation was inhibited by somatostatin. In anesthetized rats, amylin coinfusion had no effect on glucagon secretion stimulated by insulin-induced hypoglycemia. To reconcile reports of glucagon inhibition with the absence of effect in the experiments just described, anesthetized rats coinfused with rat amylin or with saline were exposed sequentially to intravenous L-arginine (during a euglycemic clamp) and then to hypoglycemia. Amylin inhibited arginine-induced, but not hypoglycemia-induced, glucagon secretion in the same animal. In conclusion, we newly identify a selective glucagonostatic effect of amylin that appears to be extrinsic to the isolated pancreas and may be centrally mediated.

Lack of effect of calcitonin gene-related peptide and amylin on major markers of glucose metabolism in hepatocytes.

Effects of amylin and calcitonin gene-related peptide on several processes involved in carbohydrate metabolism were investigated in rat hepatocytes, non-parenchymal cells (Kupffer, Ito and endothelial cells) and alveolar macrophages. In hepatocytes, cAMP levels were increased 25-fold by glucagon (10 nM), less than 2-fold by calcitonin gene-related peptide (100 nM) and not at all by amylin (100 nM). In non-parenchymal cells and cultured alveolar macrophages, calcitonin gene-related peptide potently, and amylin weakly, stimulated cAMP levels. In hepatocytes neither amylin nor calcitonin gene-related peptide affected glycogen phosphorylase activity, glucose output, lactate uptake, glycogen synthesis, glycogen mass or tyrosine aminotransferase activity. The density of calcitonin gene-related peptide specific binding sites in parenchymal cells was 10-fold less then seen in non-parenchymal cells. We found no significant evidence of specific amylin binding sites. These results are consistent with the notion that amylin does not exert a direct effect in hepatocytes. However, we do not rule out that amylin may affect hepatic glucose output indirectly through Cori cycling of lactate derived from skeletal muscle or from interactions through non-parenchymal cells.

Different pharmacological characteristics in L6 and C2C12 muscle cells and intact rat skeletal muscle for amylin, CGRP and calcitonin.

1. We compared the ability of rat amylin, rat calcitonin gene-related peptide (CGRP) and rat and salmon calcitonins to elevate cyclic AMP levels and to inhibit [U-14C]-glucose incorporation into glycogen in insulin-stimulated intact rat soleus muscle and in two cell lines derived from rodent skeletal muscle, L6 and C2C12. 2. In intact soleus muscle, both amylin (EC50S of 0.7-6.1 nM) and salmon calcitonin (EC50S of 0.5-1.4 nM) were more potent than CGRP (EC50S of 5.6-15.8 nM) and were much more potent than rat calcitonin (EC50S of 50-137 nM) at stimulating cyclic AMP production, activating glycogen phosphorylase and inhibiting insulin-stimulated [14C]-glycogen formation. 3. In contrast, in both L6 and C2C12 cells, CGRP (EC50S of 0.042-0.12 nM) stimulated cyclic AMP formation and inhibited insulin-stimulated [U-14C]-glucose incorporation into glycogen approximately 1000 times more potently than amylin (EC50S 34-240 nM), while salmon calcitonin was without measurable effect. 4. There was a correlation between elevation of cyclic AMP and inhibition of insulin-stimulated [U-14C]-glucose incorporation into glycogen evoked by these peptides in both intact muscle (r2 = 0.69, P < 0.0004) and muscle cell lines (r2 = 0.96, P < 0.0001). 5. In conclusion, the effects of amylin, CGRP, and calcitonin on soleus muscle glycogen metabolism appear to be mediated by adenylyl cyclase-coupled receptors which show a pharmacological profile similar to high affinity amylin binding sites that have been previously reported in rat brain. In contrast, the effects of amylin and CGRP in L6 and C2C12 rodent muscle cell lines appear to be mediated by adenylyl cyclase-coupled receptors that behave like CGRP receptors.

Regulation of muscle glycogen metabolism by CGRP and amylin: CGRP receptors not involved.

The aim of the present study was to determine whether amylin and calcitonin gene-related peptide (CGRP) act through shared or distinct receptors to inhibit insulin-stimulated incorporation of [14C]-glucose into glycogen. Rat amylin was 3 fold more potent than either rat alpha CGRP or rat beta CGRP at reducing glycogen synthesis from [14C]-glucose in insulin-treated rat soleus muscle. This action was blocked by peptide antagonists, with the rank order of potency being AC187 > salmon calcitonin8-32 (sCT8-32) > h-alpha CGRP8-37 for antagonism of either amylin or CGRP. The antagonist potency order correlated with affinity for amylin receptors measured in rat nucleus accumbens but not CGRP receptors measured in rat L6 muscle cells. Inhibition of glucose incorporation into glycogen by amylin and CGRP appears to be mediated by shared receptors that have the pharmacological characteristics of amylin receptors, and are distinct from previously described CGRP receptors.

Differential antagonism of amylin's metabolic and vascular actions with amylin receptor antagonists.

High affinity amylin binding sites are present in the rat nucleus accumbens. These sites bind [125I]amylin with an affinity of 27 pM and have high affinity for salmon calcitonin (sCT) and moderately high affinity for calcitonin gene related peptide (CGRP). N-terminally truncated peptides were tested for their ability to compete for [125I]amylin binding to these sites and to antagonize the metabolic and vascular actions of amylin. CGRP(8-37), sCT(8-32), and ac-[Asn30,Tyr32]sCT(8-32) (AC187) inhibited [125I]amylin binding to rat nucleus accumbens. Order of potency at inhibiting amylin binding (AC187 > sCT(8-32) > CGRP(8-37)) differed from the order of potency at inhibiting [125I]CGRP binding to SK-N-MC neuroblastoma cells (CGRP(8-37) > AC187 > sCT(8-32)) . AC187 was the most potent antagonist of amylin's effects on isolated rat soleus muscle glycogen metabolism, and it was more effective than either sCT(8-32) or CGRP(8-37) at reducing amylin-stimulated hyperlactemia in rats. In contrast, CGRP(8-37) was the most potent peptide at antagonizing amylin-induced hypotension in rats. Amylin's hypotensive actions appear to be mediated by a weak action at CGRP receptors, while its metabolic actions are mediated by receptors with a distinct antagonist profile. AC187 is a potent antagonist of amylin binding sites in nucleus accumbens and of amylin's metabolic actions.

Amylin and epinephrine have no direct effect on glucose transport in isolated rat soleus muscle.

Amylin and epinephrine did not significantly affect insulin stimulated, or basal, 3-O-methylglucose transport in isolated rat soleus muscle, as measured by the release of 3-O-methylglucose from pre-loaded tissue. Both amylin and epinephrine inhibited insulin-stimulated 2-deoxyglucose uptake (by 25% and 38%, respectively) in soleus muscle from fed rats but not from fasted rats. The latter results are consistent with amylin and epinephrine stimulating glycogenolysis and inhibiting hexokinase activity by intracellular accumulation of glucose 6-phosphate. We conclude that amylin, like epinephrine, does not specifically inhibit glucose transporters in skeletal muscle.

Molecular physiology of amylin.

Amylin is a 37-amino acid peptide first isolated, purified, and characterized from the amyloid deposits in the pancrease of type 2 diabetics. It is synthesized and secreted primarily from pancreatic beta cells along with insulin. The ability of amylin to potently reduce insulin-stimulated incorporation of glucose into glycogen in skeletal muscle requires both an intact 2Cys-7Cys disulfide bond and a COOH-terminal amide. Amylin has structural and functional relationships to two other messenger proteins, calcitonin and CGRP. Amylin has relatively potent calcitonin-like activity on bone metabolism and weaker CGRP-like activity on the vasculature. CGRP is a slightly weaker agonist than amylin for metabolic responses. Although rat calcitonins are weak, teleost fish calcitonins are very potent agonists for amylin's metabolic effects. This group of peptides appears to act on a family of related G protein-coupled receptors; several variant calcitonin receptors have recently been cloned and expressed. These receptors appear to be coupled to adenylyl cyclase in many instances; recent evidence supports the view that amylin's effects on skeletal muscle occur, at least in large part, through activation of the cAMP pathway.

Shift from alpha- to beta-type adrenergic receptor-mediated responses in chronically endotoxemic rats.

Hepatocytes from chronically endotoxemic rats, or appropriate saline controls, were maintained in primary culture for 3 or 20 h. The ability of a variety of hormones to stimulate glycogen phosphorylase a was examined. At 3 h in culture, hepatocytes from endotoxemic rats had lower basal activities and exhibited impaired response to vasopressin, angiotensin II, and, to a lesser extent, norepinephrine and glucagon. The norepinephrine response was predominantly of the alpha-type in the saline rats but mixed alpha- and beta-type in the endotoxic cells. After 20 h in culture, vasopressin and angiotensin II responses were still impaired, while norepinephrine and glucagon responses were similar to those seen in the saline cells. The response to norepinephrine was predominantly of the beta-type in the endotoxic cells but still of the alpha-type in the saline cells. The results show that multiple mechanisms are involved in endotoxin-mediated inhibition of glycogen phosphorylase a activity and that alterations in intracellular calcium homeostasis play more of a significant role than adenosine 3',5'-cyclic monophosphate-mediated processes in diminished responsiveness of the liver seen in endotoxemia.

Steroid hormones inhibit induction of spontaneous nitric oxide production in cultured hepatocytes without changes in arginase activity or urea production.

The spontaneous formation of nitrites was examined in the medium of cultured rat hepatocytes and taken as a measure of nitric oxide generation. The rate of nitrite formation increased after 8-12 hr in culture which was blocked by the addition of dexamethasone, actinomycin D, or cycloheximide. Various glucocorticoids, mineralocorticoids, and sex steroids also inhibited nitrite formation by varying degrees, without affecting arginase activity or urea production. The inhibition of nitric oxide formation appears, therefore, not to be due to changes in the availability of arginine. The results suggest that nitric-oxide synthase is induced in hepatocytes in culture and show that anti-inflammatory glucocorticoids are not the only steroids that inhibit nitric oxide formation.

LPS inhibits PI-phospholipase C but not PC-phospholipase D or phosphorylase activation by vasopressin and norepinephrine.

Rats were infused with endotoxin (50 micrograms/100 g body wt) for 3 h, and the parenchymal cells of the liver were maintained in primary culture for 1-3 h. The effects of vasopressin, norepinephrine, and glucagon on the activation of phosphatidylinositol (PI)-phospholipase C, phosphatidylcholine (PC)-phospholipase D, and glycogen phosphorylase a were investigated. Activation of PI-phospholipase C was markedly reduced, particularly with norepinephrine. This confirms that one of the early metabolic impairments seen in acute endotoxin treatment is inhibition of PI-phospholipase C activity. However, the ability of vasopressin, norepinephrine, and glucagon to stimulate glycogen phosphorylase a and PC-phospholipase D was not affected by this endotoxin treatment. We conclude that activation of phosphorylase a by vasopressin and norepinephrine is not entirely dependent on the activation of PI-phospholipase C and inositol trisphosphate formation.

Effect of in vivo TNF administration on superoxide production and PKC activity of rat alveolar macrophages.

After the intravenous injection of recombinant human tumor necrosis factor (TNF)-alpha (6.0 x 10(5) U) into rats, phorbol 12-myristate 13-acetate (PMA)-stimulated superoxide anion (O2-) secretion was enhanced in suspensions of alveolar macrophages (AM phi) compared with saline-treated controls. No enhancement of spontaneous, A23187-stimulated, or opsonized zymosan (OPZ)-stimulated O2- release was observed. Intratracheal injection of TNF-alpha (6.0 x 10(5) U) did not result in enhancement of spontaneous or A23187-, OPZ-, or PMA-stimulated O2- release. Although no TNF-alpha was detected in the bronchoalveolar lavage fluid, small quantities of TNF-alpha and/or other mediators secreted by polymorphonuclear leukocytes present in the lung capillaries, veins, and arteries may have leaked into the alveolar compartment and primed AM phi for enhanced PMA-stimulated O2- release. The respiratory burst in macrophages and neutrophils appears to be dependent on the translocation of protein kinase C. We have demonstrated protein kinase C translocation in both TNF-alpha- and saline-treated AM phi on PMA stimulation, although no differences were observed due to TNF-alpha treatment.

Endotoxin and TNF alpha directly stimulate nitric oxide formation in cultured rat hepatocytes from chronically endotoxemic rats.

We examined the effects of endotoxin on nitric oxide formation in isolated rat hepatocytes in primary culture. Endotoxin was administered either in vivo, by continuous infusion for 30 or 3 h, or in vitro, on cultured cells. The spontaneous production of nitrites in hepatocytes from in vivo ET-infused rats was lower than equivalent saline controls in the absence of added stimuli. However in vitro addition of endotoxin in culture to hepatocytes from 30 h ET-infused rats greatly enhanced production relative to saline controls. This effect was mimicked by TNF alpha, and activators of protein kinase C (PMA and Ca2+ ionophore A23187). The effects of ET were blocked by NMMA, dexamethasone and protein synthesis inhibitors Actinomycin D and cycloheximide. No in vitro effect of ET was observed in the 3 h infusion model. The results show that chronic exposure to sub-lethal levels of ET primes liver parenchymal cells for the production of nitric oxide, when exposed in vitro to ET or TNF alpha.

Ethanol is a potent stimulator of phosphatidylcholine breakdown in cultured rat hepatocytes.

Addition of ethanol (17 to 340 mM) to cultured rat hepatocytes stimulated the breakdown of phosphatidylcholine phospholipases D and C as measured by an increase in the rate of release of choline and phosphocholine into the medium. The effects of ethanol were mimicked by propanol, dimethylsulfoxide and to a lesser extent methanol. The magnitude of the stimulation seen with ethanol was equivalent to and additive to that produced by glucagon vasopressin, norepinephrine, A23187 or PMA. In contrast, ethanol (340 mM) stimulated PI-specific phospholipase C activity by less than 20%. An equivalent stimulation of PC-specific phospholipase D and C was seen with as little as 20 mM ethanol and a 100% increase was seen with 340 mM ethanol. Ethanol did not significantly affect the ability of vasopressin, norepinephrine, ATP or A23187 to stimulate PI-specific phospholipase C. It is concluded that while ethanol is only a weak stimulator of PI-specific phospholipase C, it is a potent stimulator of phosphatidylcholine breakdown in rat hepatocytes.

Activation of membrane protein kinase C by glucagon and Ca(2+)-mobilizing hormones in cultured rat hepatocytes. Role of phosphatidylinositol and phosphatidylcholine hydrolysis.

We found that glucagon stimulated membrane protein kinase C (PKC) activity and phosphatidylcholine hydrolysis in 24 h-cultured rat hepatocytes. Phorbol myristate acetate, 8-bromo cyclic AMP, vasopressin, noradrenaline and the Ca2+ ionophore A23187 also stimulated membrane PKC activity. However, only vasopressin and noradrenaline stimulated inositol phosphate accumulation, whereas all agonists stimulated the rate of release of water-soluble choline metabolites into the medium. Choline, and to a much lesser extent phosphocholine, were released, suggesting predominantly phospholipase D activation. This was supported by the finding that the accumulation of phosphatidate and diacylglycerol was enhanced by the agents in [3H]myristate-labelled hepatocytes, as was [32P]phosphatidylethanol formation. Since the time courses for the release of choline into the medium and the accumulation of phosphatidate and diacylglycerol caused by vasopressin and glucagon were similar, the more rapid activation of PKC by vasopressin probably reflects diacylglycerol formation from phosphoinositide breakdown. The inability of glucagon to stimulate inositol phosphate production was not due to the prolonged culture, since similar results were obtained in 4 h cultures. We conclude that the stimulation of membrane PKC activity by glucagon correlates with accumulation of diacylglycerol and phosphatidate derived from the hydrolysis of phosphatidylcholine.

Effects of insulin on inositol phosphate production in cultured rat hepatocytes.

Addition of vasopressin (100 nM) to rat hepatocytes prelabelled with [3H]inositol stimulated the production of inositol phosphates in the presence of 20 mM Li+. Preincubation of hepatocytes with insulin (50 nM) or glucagon (10 nM) had no significant effect alone but enhanced the effects of vasopressin after a lag period of at least 1 min. The effects of insulin and glucagon appeared additive in this respect. Insulin also enhanced the norepinephrine-mediated stimulation of inositol phosphate accumulation. The enhancement by insulin of the effects of vasopressin required at least 0.5-5 nM insulin and did not involve changes in [3H]inositol lipid labelling or IP3 phosphatase activity. The effect of insulin appeared insensitive to prior treatment of hepatocytes with pertussis toxin (200 ng/ml for 18-24 h) or cholera toxin (100 ng/ml for 3-4 h). The glucagon enhancement of the effects of vasopressin was not affected by pertussis toxin but was mimicked by cholera toxin. The response of hepatocytes to vasopressin in the absence of Li+ was smaller and more transient. Under these conditions a 5 min prior incubation with insulin inhibited the stimulation by vasopressin of inositol phosphate accumulation. A similar inhibitory effect of prior insulin exposure on the transient activation by vasopressin of exogenous phosphatidylinositol 4,5-bisphosphate breakdown by hepatocyte homogenates was also seen. These data indicate that insulin, although having no effect on basal inositol phosphate accumulation, can either enhance or antagonise the effects of vasopressin in primary rat liver hepatocyte cultures depending on the experimental conditions.

Vasopressin and norepinephrine stimulation of inositol phosphate accumulation in rat hepatocytes are modified differently by protein f1nase C and protein kinase A.

Rat hepatocytes were maintained in primary monolayer culture for 24 h in the presence of serum. Treatment of hepatocytes with 1 microM 4 beta-phorbol 12 beta-myristate 13 alpha-acetate (PMA) for 5-15 min increased membrane-associated protein kinase C activity and concomitantly decreased soluble activity. Membrane protein kinase C activity returned to basal values within 1 h then decreased by more than 50% within 2 h. Prolonged (2-18 h) incubation with PMA did not further decrease protein kinase C activity. Pretreatment of hepatocytes with PMA for 5-15 min had little effect on the subsequent actions of 100 nM vasopressin but abolished the stimulation of inositol phosphate accumulation by 3 nM vasopressin and 20 microM norepinephrine. Long-term exposure (2-18 h) of hepatocytes to 1 microM PMA actually enhanced the effects of vasopressin and 20 microM norepinephrine. The stimulation by norepinephrine (20 microM) of inositol phosphate accumulation was abolished by the alpha 1-adrenergic antagonist prazosin (1 microM), whereas the beta-adrenergic antagonist propranolol (30 microM) had little effect. Addition of 8Br-cAMP (100 microM) or glucagon (10 nM) for 5 min or 8 h had no significant effect alone, but enhanced the subsequent vasopressin stimulation of inositol phosphate accumulation. There was no effect of 8Br-cAMP or glucagon on norepinephrine stimulation of phosphoinositide breakdown. These data indicate that the stimulation of phospholipase C activity in rat hepatocytes by 3 nM vasopressin is enhanced by cyclic AMP-dependent kinase but inhibited by protein kinase C. In contrast, down regulation of protein kinase C markedly enhanced the maximal phosphoinositide response due to both vasopressin and norepinephrine.

Vasopressin transiently stimulates phospholipase C activity in cultured rat hepatocytes.

Vasopressin stimulated phospholipase C activity in primary cultures of rat hepatocytes maintained for 18-24 h under serum free conditions. Soluble and membrane-associated phospholipase C activity was determined using exogenous [3H]phosphatidylinositol 4,5-bisphosphate ([3H]PIP2) in the presence of cholate, deoxycholate and NaCl. Exposure of hepatocytes for 5 s to vasopressin (100 nM) stimulated both membrane-associated and soluble phospholipase C activity by 30% and 40%, respectively. However, by 15 s this stimulation had disappeared. Addition of vasopressin to hepatocytes, previously labelled with [3H]inositol, stimulated inositol phosphate production within 5 s, but little further increase was seen over a 5-min incubation. These results indicate that vasopressin rapidly stimulates both soluble and membrane-associated phospholipase C activity.

Exposure of cultured hepatocytes to cyclic AMP enhances the vasopressin-mediated stimulation of inositol phosphate production.

Isolated rat hepatocytes in primary monolayer culture were maintained for 18-24 h in the presence of 10% (v/v) serum and [3H]inositol. Vasopressin (100 nM) stimulated the production of inositol mono-, bis- and tris-phosphates (IP1, IP2, and IP3). Prior exposure of hepatocytes to 8-bromo cyclic AMP (8Br-cAMP; 100 microM), but not 8-bromo cyclic GMP, enhanced the vasopressin-mediated stimulation of inositol phosphate accumulation, but had no significant effect on their formation in the absence of vasopressin. The effect of the cyclic AMP analogue was mimicked by glucagon (10 nM), and was seen whether cyclic AMP or glucagon was added 5 min or 12 h before the addition of vasopressin. An 8 h incubation with dexamethasone (100 nM) enhanced the accumulation of IP3, but not that of IP2 or IP1, in the presence of 8Br-cAMP and vasopressin. Cycloheximide or actinomycin D had little effect on the vasopressin stimulation of inositol phosphate accumulation, after an 8 h incubation in the presence or absence of 8Br-cAMP.