Search

The role of the gaseous messengers NO and CO for β-cell function and survival is controversial. We examined this issue in the hyperglycemic-hyperinsulinemic ob/ob mouse, an animal model of type 2 obese diabetes, by studying islets from obese vs lean mice regarding glucose-stimulated insulin release in relation to islet NO and CO production and the influence of modulating peptide hormones. Glucose-stimulated increase in ncNOS-activity in incubated lean islets was converted to a decrease in ob/ob islets associated with markedly increased insulin release. Both types of islets displayed iNOS activity appearing after ~60 min in high-glucose. In ob/ob islets the insulinotropic peptides glucagon, GLP-1 and GIP suppressed NOS activities and amplified glucose-stimulated insulin release. The insulinostatic peptide leptin induced the opposite effects. Suppression of islet CO production inhibited, while stimulation amplified glucose-stimulated insulin release. Nonincubated isolated islets from young and adult obese mice displayed very low ncNOS and negligible iNOS activity. In contrast, production of CO, a NOS inhibitor, was impressively raised. Glucose injections induced strong activities of islet NOS isoforms in lean but not in obese mice and confocal microscopy revealed iNOS expression only in lean islets. Islets from ob/ob mice existing in a hyperglycemic in vivo milieu maintain elevated insulin secretion and protection from glucotoxicity through a general suppression of islet NOS activities achieved by leptin deficiency, high CO production and insulinotropic cyclic-AMP-generating hormones. Such a beneficial effect on islet function and survival might have its clinical counterpart in human leptin-resistant type 2 obese diabetes with hyperinsulinemia.

The Goto-Kakizaki (GK) rat displays a markedly reduced insulin response to glucose, a defect that is thought to be coupled to an impaired glucose signaling in the β-cell. We have examined whether carbon monoxide (CO), derived from β-cell heme oxygenase (HO), might be involved in the secretory dysfunction. Immunocytochemical labeling of constitutive HO (HO-2) showed no overt difference in fluorescence pattern in islets from GK vs. Wistar controls. However, isolated islets from GK rats displayed a markedly impaired HO activity measured as CO production (−50%), and immunoblotting revealed an approximately 50% reduction of HO-2 protein expression compared with Wistar controls. Furthermore, there was a prominent expression of inducible HO (HO-1) in GK islets. Incubation of isolated islets showed that the glucose-stimulated CO production and the glucose-stimulated insulin response were considerably reduced in GK islets compared with Wistar islets. Addition of the HO activator hemin or gaseous CO to the incubation media brought about a similar amplification of glucose-stimulated insulin release in GK and Wistar islets, suggesting that distal steps in the HO-CO signaling pathway were not appreciably affected. We conclude that the defective insulin response to glucose in the GK rat can be explained, at least in part, by a marked impairment of the glucose-HO-CO signaling pathway as manifested by a prominent decrease in glucose stimulation of islet CO production and a reduced expression of HO-2. A possible role of HO-1 expression as a compensatory mechanism in the GK islets is presently unclear.

We investigated, by a combined in vivo and in vitro approach, the temporal changes of islet nitric oxide synthase (NOS)-derived nitric oxide (NO) and heme oxygenase (HO)-derived carbon monoxide (CO) production in relation to insulin and glucagon secretion during acute endotoxemia induced by lipopolysaccharide (LPS) in mice. Basal plasma glucagon, islet cAMP and cGMP content after in vitro incubation, the insulin response to glucose in vivo and in vitro, and the insulin and glucagon responses to the adenylate cyclase activator forskolin were greatly increased after LPS. Immunoblots demonstrated expression of inducible NOS (iNOS), inducible HO (HO-1), and an increased expression of constitutive HO (HO-2) in islet tissue. Immunocytochemistry revealed a marked expression of iNOS in many β-cells, but only in single α-cells after LPS. Moreover, biochemical analysis showed a time dependent and markedly increased production of NO and CO in these islets. Addition of a NOS inhibitor to such islets evoked a marked potentiation of glucose-stimulated insulin release. Finally, after incubation in vitro, a marked suppression of NO production by both exogenous CO and glucagon was observed in control islets. This effect occurred independently of a concomitant inhibition of guanylyl cyclase. We suggest that the impairing effect of increased production of islet NO on insulin secretion during acute endotoxemia is antagonized by increased activities of the islet cAMP and HO-CO systems, constituting important compensatory mechanisms against the noxious and diabetogenic actions of NO in endocrine pancreas.

Islet production of nitric oxide (NO) and CO in relation to islet hormone secretion was investigated in mice given the NO synthase (NOS) inhibitor N G-nitro-l-arginine methyl ester (l-NAME) in their drinking water. In these mice, the total islet NO production was paradoxically increased, reflecting induction of inducible NOS (iNOS) in background of reduced activity and immunoreactivity of constitutive NOS (cNOS). Unexpectedly, normal mice fasted for 24 h also displayed iNOS activity, which was further increased in l-NAME-drinking mice. Glucose-stimulated insulin secretion in vitro and in vivo was increased in fasted but unaffected in fed mice after l-NAME drinking. Glucagon secretion was increased in vitro. Control islets incubated with different NOS inhibitors at 20 mM glucose displayed increased insulin release and decreased cNOS activity. These NOS inhibitors potentiated glucose-stimulated insulin release also from islets ofl-NAME-drinking mice. In contrast, glucagon release was suppressed. In islets from l-NAME-drinking mice, cyclic nucleotides were upregulated, and forskolin-stimulated hormone release, CO production, and heme oxygenase (HO)-2 expression increased. In conclusion, chronic NOS blockade evoked iNOS-derived NO production in pancreatic islets and elicited compensatory mechanisms against the inhibitory action of NO on glucose-stimulated insulin release by inducing upregulation of the islet cAMP and HO-CO systems.

We have studied, by a combined in vitro and in vivo approach, the relation between the inhibitory action of N(G)-nitro-l-arginine methyl ester (L-NAME), a selective inhibitor of nitric oxide synthase (NOS), on the activity of islet constitutive NOS (cNOS) and glucose regulation of islet hormone release in mice. The cNOS activity in islets incubated in vitro at 20 mM glucose was not appreciably affected by 0.05 or 0.5 mM L-NAME, but was greatly suppressed (-60%) by 5 mM L-NAME. Similarly, glucose-stimulated insulin release was unaffected by the lower concentrations of L-NAME but greatly enhanced in the presence of 5 mM of the NOS inhibitor. In incubated islets inhibition of cNOS activity resulted in a modestly enhanced insulin release in the absence of glucose, did not display any effect at physiological or subphysiological glucose concentrations, but resulted in a markedly potentiated insulin release at hyperglycaemic glucose concentrations. In the absence of glucose, glucagon secretion was suppressed by L-NAME. The dynamics of glucose-induced insulin release and (45)Ca(2+) efflux from perifused islets revealed that L-NAME caused an immediate potentiation of insulin release, and a slight increase in (45)Ca(2+) efflux. In islets depolarized with 30 mM K(+) in the presence of the K(+)(ATP) channel opener, diazoxide, L-NAME still greatly potentiated glucose-induced insulin release. Finally, an i.v. injection of glucose to mice pretreated with L-NAME was followed by a markedly potentiated insulin response, and an improved glucose tolerance. In accordance, islets isolated directly ex vivo after L-NAME injection displayed a markedly reduced cNOS activity. In conclusion, we have shown here, for the first time, that biochemically verified suppression of islet cNOS activity, induced by the NOS inhibitor L-NAME, is accompanied by a marked potentiation of glucose-stimulated insulin release both in vitro and in vivo. The major action of NO to inhibit glucose-induced insulin release is probably not primarily linked to changes in Ca(2+) fluxes and is exerted mainly independently of membrane depolarization events.

Recent observations suggest that carbon monoxide (CO) may serve as a neuroendocrine modulator in hypothalamus. Here we provide evidence, for the first time, that the islets of Langerhans contain the constitutive isoform of the CO-producing enzyme heme oxygenase (HO-2), the activity of which was found to modulate islet hormone release. Most insulin and glucagon cells in the rat endocrine pancreas expressed strong immunoreactivity for HO-2. In the exocrine parenchyma, scattered HO-2-positive ganglionic cell bodies were occasionally observed. Furthermore, Western blot analysis revealed the presence of HO-2 in isolated islets but not in acinar cells. Islet homogenates displayed a comparatively high HO-2 enzymatic activity measured as CO formation (approximately 600 pmol CO.min-1.mg islet protein-1). This HO-2 enzymatic activity was greatly suppressed by zincprotoporphyrin-IX (ZnPP-IX), a recognized inhibitor of HO activity. Neither ZnPP-IX nor the HO activator, hemin, influenced basal insulin release from isolated rat islets at low (1 mM) glucose. However, glucagon release at 1 mM glucose was increased by hemin and inhibited by ZnPP-IX. The hemin-induced increase in glucagon secretion was abolished by ZnPP-IX. Furthermore, a series of experiments at high glucose (16.7 mM) revealed that hemin induced a dose-dependent potentiation of glucose-stimulated insulin release. Moreover, glucose-induced insulin release was dose-dependently suppressed by ZnPP-IX but unaffected by protoporphyrin-IX, a compound known not to influence HO-2 activity in other tissues. Similarly, glucagon release at high glucose was dose-dependently increased by hemin and suppressed by ZnPP-IX. Finally, the hemin-induced increase in islet hormone release at high glucose was totally abolished by ZnPP-IX. The data strongly suggest that CO production positively modulates both glucagon and insulin secretion. We propose that CO may serve as a novel messenger molecule within the islets of Langerhans.

Production of nitric oxide through the action of nitric oxide synthase (NOS) has been detected in the islets of Langerhans. The inducible isoform of NOS (iNOS) is induced by cytokines and might contribute to the development of type-1 diabetes, while the constitutive isoform (cNOS) is thought to be implicated in the physiological regulation of insulin secretion. In the present study we have detected and quantified islet cNOS- and iNOS-derived NO production concomitant with measuring its influence on insulin secretion in the presence of different secretagogues: glucose, L-arginine, L-leucine and α-ketoisocaproic acid (KIC) both during fasting and freely fed conditions. In intact islets from freely fed mice both cNOS- and iNOS-activity was greatly increased by glucose (20 mmol/l). Fasting induced islet iNOS activity at both physiological (7 mmol/l) and high (20 mmol/l) glucose concentrations. NOS blockade increased insulin secretion both during freely fed conditions and after fasting. L-arginine stimulated islet cNOS activity and did not affect islet iNOS activity. l-leucine or KIC, known to enter the TCA cycle without affecting glycolysis, did not affect either islet cNOS- or iNOS activity. Accordingly, insulin secretion stimulated by L-leucine or KIC was unaffected by addition of L-NAME both during feeding and fasting. We conclude that both high glucose concentrations and fasting increase islet total NO production (mostly iNOS derived) which inhibit insulin secretion. The insulin secretagogues L-leucine and KIC, which do not affect glycolysis, do not interfere with the islet NO-NOS system.

We investigated implications of nitric oxide (NO) derived from islet neuronal constitutive NO synthase (ncNOS) and inducible NOS (iNOS) on insulin secretory mechanisms in the mildly diabetic GK rat. Islets from GK rats and Wistar controls were analysed for ncNOS and iNOS by HPLC, immunoblotting and immunocytochemistry in relation to insulin secretion stimulated by glucose or l-arginine in vitro and in vivo. No obvious difference in ncNOS fluorescence in GK vs control islets was seen but freshly isolated GK islets displayed a marked iNOS expression and activity. After incubation at low glucose GK islets showed an abnormal increase in both iNOS and ncNOS activities. At high glucose the impaired glucose-stimulated insulin release was associated with an increased iNOS expression and activity and NOS inhibition dose-dependently amplified insulin secretion in both GK and control islets. This effect by NOS inhibition was also evident in depolarized islets at low glucose, where forskolin had a further amplifying effect in GK but not in control islets. NOS inhibition increased basal insulin release in perfused GK pancreata and amplified insulin release after glucose stimulation in both GK and control pancreata, almost abrogating the nadir separating first and second phase in controls. A defective insulin response to l-arginine was seen in GK rats in vitro and in vivo, being partially restored by NOS inhibition. The results suggest that increased islet NOS activities might contribute to the defective insulin response to glucose and l-arginine in the GK rat. Excessive iNOS expression and activity might be deleterious for the beta-cells over time.

We have studied the influence of nitric oxide (NO) and carbon monoxide (CO), putative messenger molecules in the brain as well as in the islets of Langerhans, on glucose-stimulated insulin secretion and on the activities of the acid alpha-glucoside hydrolases, enzymes which we previously have shown to be implicated in the insulin release process. We have shown here that exogenous NO gas inhibits, while CO gas amplifies glucose-stimulated insulin secretion in intact mouse islets concomitant with a marked inhibition (NO) and a marked activation (CO) of the activities of the lysosomal/vacuolar enzymes acid glucan-1,4-alpha-glucosidase and acid alpha-glucosidase (acid alpha-glucoside hydrolases). Furthermore, CO dose-dependently potentiated glucose-stimulated insulin secretion in the range 0.1-1000 microM. In intact islets, the heme oxygenase substrate hemin markedly amplified glucose-stimulated insulin release, an effect which was accompanied by an increased activity of the acid alpha-glucoside hydrolases. These effects were partially suppressed by the guanylate cyclase inhibitor 1H-[1,2,4]oxadiazolo-[4,3-a]quinoxalin-1-one. Hemin also inhibited inducible NO synthase (iNOS)-derived NO production probably through a direct effect of CO on the NOS enzyme. Further, exogenous CO raised the content of both cGMP and cAMP in parallel with a marked amplification of glucose-stimulated insulin release, while exogenous NO suppressed insulin release and cAMP, leaving cGMP unaffected. Emiglitate, a selective inhibitor of alpha-glucoside hydrolase activities, was able to markedly inhibit the stimulatory effect of exogenous CO on both glucose-stimulated insulin secretion and the activityof acid glucan-1,4-alpha-glucosidase and acid alpha-glucosidase, while no appreciable effect on the activities of other lysosomal enzyme activities measured was found. We propose that CO and NO, both produced in significant quantities in the islets of Langerhans, have interacting regulatory roles on glucose-stimulated insulin secretion. This regulation is, at least in part, transduced through the activity of cGMP and the lysosomal/vacuolar system and the associated acid alpha-glucoside hydrolases, but probably also through a direct effect on the cAMP system.

A key question for understanding the mechanisms of pulsatile insulin release is how the underlying β-cell oscillations of the cytoplasmic Ca2+concentration ([Ca2+]i) are synchronized within and among the islets in the pancreas. Nitric oxide has been proposed to coordinate the activity of the β-cells by precipitating transients of [Ca2+]i. Comparing ob/ ob mice and lean controls, we have now studied the action of carbon monoxide (CO), another neurotransmitter with stimulatory effects on cGMP production. A strong immunoreactivity for the CO-producing constitutive heme oxygenase (HO-2) was found in ganglionic cells located in the periphery of the islets and in almost all islet endocrine cells. Islets from ob/ ob mice had sixfold higher generation of CO (1 nmol · min–1· mg protein–1) than the lean controls. This is 100-fold the rate for their constitutive production of NO. Moreover, islets from ob/ ob mice showed a threefold increase in HO-2 expression and expressed inducible HO (HO-1). The presence of an excessive islet production of CO in the ob/ ob mouse had its counterpart in a pronounced suppression of the glucose-stimulated insulin release from islets exposed to the HO inhibitor Zn-protoporhyrin (10 μM) and in a 16 times higher frequency of [Ca2+]itransients in their β-cells. Hemin (0.1 and 1.0 μM), the natural substrate for HO, promoted the appearance of [Ca2+]itransients, and 10 μM of the HO inhibitors Zn-protoporphyrin and Cr-mesoporphyrin had a suppressive action both on the firing of transients and their synchronization. It is concluded that the increased islet production of CO contributes to the hyperinsulinemia in ob/ ob mice. In addition to serving as a positive modulator of glucose-stimulated insulin release, CO acts as a messenger propagating Ca2+signals with coordinating effects on the β-cell rhythmicity.

Aims/hypothesis. To map the cellular location of inducible and constitutive nitric oxide synthase and haem oxygenase in rat islets to clarify the morphological background to putative nitric oxide and carbon monoxide pathways. Methods. Immunocytochemistry and confocal microscopy. Results. After treatment with endotoxin, immunoreactivity for inducible nitric oxide synthase was expressed in a large number of islet cells, most of which were insulin-immunoreactive beta cells and in single glucagon-immunoreactive and pancreatic polypeptide-immunoreactive cells. Somatostatin-immunoreactive cells lacked immunoreactivity for inducible nitric oxide synthase. In untreated rats, immunoreactivity for constitutive nitric oxide synthase occurred in the majority of insulin-immunoreactive and glucagon-immunoreactive cells, in most pancreatic polypeptide-immunoreactive and somatostatin-immunoreactive cells and in islet nerves. Similarly, immunoreactivity for constitutive haem oxygenase was detected in all four types of islet cells. Endotoxin treatment did not change the pattern of immunoreactivity for constitutive and inducible haem oxygenase. After treatment with alloxan, insulin-immunoreactivity was observed only in single islet cells, being almost devoid of immunoreactivity for constitutive nitric oxide synthase and haem oxygenase. Conclusion/interpretation. In vivo endotoxin-induced expression of inducible nitric oxide synthase in insulin-producing and in scattered glucagon-producing and pancreatic polypeptide-producing cells strengthens previous suggestions of a pathophysiological role for inducible nitric oxide synthase in the development of insulin-dependent diabetes mellitus. The presence of constitutive nitric oxide synthase and haem oxygenase in all four types of islet cells, together with recent functional data of ours support roles for nitric oxide and carbon monoxide as intracellular, paracrine or neurocrine modulators of islet hormone secretion. [Diabetologia (1999) 42: 978–986]

Carbon monoxide (CO) has been suggested as a novel messenger molecule in the brain. We now report on the cellular localization and hormone secretory function of a CO-producing constitutive heme oxygenase (HO-2) in mouse islets. Islet homogenates produced large amounts of CO which were suppressed dose-dependently by the HO inhibitor zincprotoporphyrin-IX (ZnPP-IX). We also show, for the first time, that glucose markedly stimulates the HO activity (CO production) in intact islets. A further potentiation was induced by the HO substrate hemin. Western blot showed that islet tissue expressed HO-2, and confocal microscopy revealed that HO-2 resided in insulin, glucagon, somatostatin, and pancreatic polypeptide cells. ZnPP-IX dose-dependently inhibited, whereas hemin enhanced, both insulin and glucagon secretion from glucose-stimulated islets. Stimulation or inhibition of CO production was accompanied by corresponding changes in islet cGMP levels. Exogenously applied CO stimulated insulin and glucagon release from isolated islets, whereas exogenous nitric oxide (NO) inhibited insulin and stimulated glucagon release. Islets stimulated by glucose or L-arginine displayed a marked increase in their NO-synthase (NOS) activity. Such an increase was suppressed by hemin, conceivably because NOS activity was inhibited by hemin-derived CO. Consequently, hemin enhanced L-arginine-induced insulin secretion. Insulin release stimulated by either hemin-derived CO or exogenous CO was strongly inhibited by the guanylate cyclase inhibitor ODQ, but it was unaffected by ZnPP-IX. Glucagon release induced by CO (but not by hemin) was inhibited by ODQ and partly inhibited by ZnPP-IX. We propose that the islets of Langerhans are equipped with a heme oxygenase-carbon monoxide pathway, which constitutes a novel regulatory system of physiological importance for the stimulation of insulin and glucagon release. This pathway is stimulated by glucose, is at least partly dependent on the cGMP system, and displays interaction with islet NOS activity.

Nitric oxide (NO) produced by islet constitutive NO synthase (cNOS) is a putative modulator of islet hormone secretion. We show here for the first time that the release of insulin induced byl-arginine orl-homoarginine is inhibited and that of glucagon stimulated in parallel with the rate of islet NO production. It was found thatl-homoarginine was ≈25–30% less potent thanl-arginine as an insulin secretagogue but equally potent as a glucagon secretagogue. Biochemical determination of islet cNOS activity revealed that the NO production with l-homoarginine as substrate was only ≈40% of that ofl-arginine. Selective inhibition of islet cNOS potentiated insulin release during amino acid stimulation. Moreover, inhibition of cNOS suppressed glucagon release, more so with l-arginine than with l-homoarginine as secretagogue, reflecting the relative rates of their NO production. In K+-depolarized islets, inhibition of cNOS enhanced the insulin response tol-arginine by 50% and that tol-homoarginine by 23%, largely corresponding to their relative NO production. The intracellular NO donor hydroxylamine dose dependently inhibited insulin but increased glucagon secretion in K+-depolarized and amino acid-stimulated islets. We conclude that both amino acids have a dual action on insulin release, since their stimulatory effects are reduced in parallel with the rates of their NO production. Furthermore, the greater NO production induced byl-arginine relative tol-homoarginine corresponds to NO-mediated increases in glucagon release. These NO effects are mainly exerted independently of membrane depolarization events.

Accumulated evidence links an important signal involved in glucose-stimulated insulin release to the activation of the islet lysosomal glycogenolytic enzyme acid glucan-1,4-alpha-glucosidase. We have analyzed the function of the lysosomal system/lysosomal enzyme activities in pancreatic islets of young (6-8 weeks), spontaneously diabetic, GK (Goto-Kakizaki) rats and Wistar control rats in relation to glucose-induced insulin release. The insulin secretory response to glucose was markedly impaired in the GK rat, but was restored by the adenylate cyclase activator forskolin. Islet activities of classical lysosomal enzymes, e.g.. acid phosphatase, N-acetyl-beta-D-glucosaminidase, beta-glucuronidase, and cathepsin D, were reduced by 20-35% in the GK rat compared with those in Wistar controls. In contrast, the activities of the lysosomal alpha-glucosidehydrolases, i.e.. acid glucan-1,4-alpha-glucosidase and acid alpha-glucosidase, were increased by 40-50%. Neutral alpha-glucosidase (endoplasmic reticulum) was unaffected. Comparative analysis of liver tissue showed that lysosomal enzyme activities were of the same magnitude in GK and Wistar rats. Notably, in Wistar rats, the activities of acid glucan-1,4-alpha-glucosidase and acid alpha-glucosidase were approximately 15-fold higher in islets than in liver. Other lysosomal enzymes did not display such a difference. Normalization of glycemia in GK rats by phlorizin administered for 9 days did not influence either the lysosomal alpha-glucosidehydrolase activities or other lysosomal enzyme activities in GK islets. Finally, the pseudotetrasaccharide acarbose, which accumulates in the lysosomal system, inhibited acid glucan-1,4-alpha-glucosidase activity in parallel with its inhibitory action on glucose-induced insulin release in intact Wistar islets, whereas no effect was recorded for either parameter in intact GK islets. In contrast, acarbose inhibited the enzyme activity equally in islet homogenates from both GK and Wistar rats, showing that the catalytic activity of the enzyme itself in disrupted cells was unaffected. We propose that dysfunction of the islet lysosomal/vacuolar system is an important defect impairing the transduction mechanisms for glucose-induced insulin release in the GK rat.

The expression and activities of constitutive nitric oxide synthase (cNOS) and inducible nitric oxide synthase (iNOS) in relation to insulin and glucagon secretory mechanisms were investigated in islets isolated from rats subjected to total parenteral nutrition (TPN) for 10 d. TPN is known to result in significantly increased levels of plasma lipids during the infusion time. In comparison with islets from freely fed control rats, islets taken from TPN rats at d 10 displayed a marked decrease in glucose-stimulated insulin release (4.65 +/- 0.45 ng/[islet x h] vs 10.25 +/- 0.65 for controls) (p0.001) accompanied by a strong iNOS activity (18.3 +/- 1.1 pmol of NO/[min x mg of protein]) and a modestly reduced cNOS activity (11.3 +/- 3.2 pmol of NO/[min x mg of protein] vs 17.7 +/- 1.7 for controls) (p0.01). Similarly, Western blots showed the expression of iNOS protein as well as a significant reduction in cNOS protein in islets from TPN-treated rats. The enhanced NO production, which is known to inhibit glucose-stimulated insulin release, was manifested as a strong increase in the cyclic guanosine 5'-monophosphate content in the islets of TPN-treated rats (1586 +/- 40 amol/islet vs 695 +/- 64 [p0.001] for controls). Moreover, the content of cyclic adenosine monophosphate (cAMP) was greatly increased in the TPN islets (80.4 +/- 2.1 fmol/islet vs 42.6 +/- 2.6 [p0.001] for controls). The decrease in glucose-stimulated insulin release was associated with an increase in the activity of the secretory pathway regulated by the cAMP system in the islets of TPN-treated rats, since the release of insulin stimulated by the phosphodiesterase inhibitor isobutylmethylxanthine was greatly increased both in vivo after iv injection and after in vitro incubation of isolated islets. By contrast, the release of glucagon was clearly reduced in islets taken from TPN-treated rats (33.5 +/- 1.5 pg/[islet x h] vs 45.5 +/- 2.2 for controls) (p0.01) when islets were incubated at low glucose (1.0 mmol/L). The data show that long-term TPN treatment in rats brings about impairment of glucose-stimulated insulin release, that might be explained by iNOS expression and a marked iNOS-derived NO production in the beta-cells. The release of glucagon, on the other hand, is probably decreased by a direct "nutrient effect" of the enhanced plasma lipids. The results also suggest that the islets of TPN-treated rats have developed compensatory insulin secretory mechanisms by increasing the activity of their beta-cell cAMP system.