The usefulness of interfacial photopolymerization of poly(ethylene glycol) (PEG) diacrylate at a variety of concentrations and molecular weights to form hydrogel membranes for encapsulating porcine islets of Langerhans was investigated. The results from this study show in vitro and in vivo function of PEG-encapsulated porcine islets and the ability of PEG membranes to prevent immune rejection in a discordant xenograft model. Encapsulated islets demonstrated an average viability of 85% during the first week after encapsulation, slightly but significantly lower than unencapsulated controls. Encapsulated porcine islets were shown to be glucose responsive using static glucose stimulation and perifusion assays. Higher rates of insulin release were observed for porcine islets encapsulated in lower concentrations of PEG diacrylate (10-13%), not significantly reduced relative to unencapsulated controls, than were observed in islets encapsulated in higher concentrations (25%) of PEG diacrylate. Perifusion results showed biphasic insulin release from encapsulated islets in response to glucose stimulation. Streptozotocin-induced diabetic athymic mice maintained normoglycemia for up to 110 days after the implantation of 5,000-8,000 encapsulated porcine islet equivalents into the peritoneal cavity. Normoglycemia was also confirmed in these animals using glucose tolerance tests. PEG diacrylate-encapsulated porcine islets were shown to be viable and contain insulin after 30 days in the peritoneal cavity of Sprague-Dawley rats, a discordant xenograft model. From these studies, we conclude that PEG diacrylate encapsulation of porcine islets by interfacial photopolymerization shows promise for use as a method of xenoprotection toward a bioartificial endocrine pancreas.

Insulinopenic diabetes is known to produce endothelial dysfunction. This dysfunction could arise from either hyperglycemia or inadequate insulin. It is not known whether endothelial dysfunction occurs when hyperglycemia is present with elevated insulin levels. In this study, we utilized an experimental model of hyperglycemia with hyperinsulinemia to investigate latent endothelial dysfunction. Rats were continuously infused with glucose or saline for 72 h to achieve peak plasma glucose concentrations of approximately 25 mM. Plasma insulin rose by 12-fold in glucose-infused rats. No significant differences in serum electrolyte concentration were noted between control and glucose-infused rats after 72 h. Blood pressure was not altered by this intervention. Aortic rings taken from control rats relaxed to the endothelium-dependent vasodilators, acetylcholine and A-23187, and to the endothelium-independent vasodilator, nitroglycerin. Relaxation to acetylcholine but not to A-23187 or nitroglycerin was impaired in glucose-infused rat aortic rings. Incubation in vitro with either indomethacin or superoxide dismutase did not restore the impaired relaxation to acetylcholine in rings taken from glucose-infused rats. Thus hyperglycemia with hyperinsulinemia selectively impairs receptor-dependent, endothelium-dependent relaxation. These studies suggest that elevated glucose may be a common pathway leading to endothelial dysfunction in insulin-dependent diabetes mellitus and hyperglycemia-induced insulin resistance.

Insulin resistance is accentuated during periods of poor metabolic control in human non-insulin-dependent diabetes mellitus. The role of hyperglycemia in this suppression of insulin action is not clear. If glucose impairs insulin action, then the effect should be reproducible in vivo in tissues of normal intact rats. To test this possibility, normal rats were continuously administered 50% glucose in water (60-66 mg.kg-1.min-1) via an indwelling jugular catheter. After 72 h, these animals were hyperglycemic, hyperinsulinemic, and glucosuric compared with control rats infused for 72 h with normal saline (P less than 0.01). Basal glucose uptake in vivo was greater in muscle of glucose-infused rats. Insulin-stimulated glucose uptake in vivo and in vitro (by perfused hindquarters and isolated adipocytes) were suppressed in the glucose-infused group (P less than 0.01). Glycogen synthase activity was reduced 40% in extracts of muscle and adipose tissue of hyperglycemic rats. Basal and isoproterenol-stimulated lipolysis were increased, whereas insulin suppression of lipolysis was blunted in adipocytes from glucose-infused animals (P less than 0.01). Glucose infusion did not alter insulin binding by isolated adipocytes or solubilized skeletal muscle insulin receptors. These results suggest that a 72-h in vivo glucose infusion impaired insulin action in muscle and adipose tissue of normal rats by inducing postbinding defects similar to those observed in human diabetes mellitus during intervals of deteriorated metabolic control.

No abstract

No abstract