Recently, a new multifunctional, bio‐inorganic nanocomposite membrane with the ability to self‐regulate the release of insulin in response to blood glucose (BG) levels was reported. Herein, the application of this material as part of a small, implantable, closed‐loop insulin delivery device designed to continuously monitor BG concentrations and regulate insulin release is proposed. The insulin delivery device consists of a nanocomposite glucose‐responsive plug covalently bound to an insulin reservoir made of surface‐modified silicone. The plug is prepared with crosslinked bovine serum albumin (BSA) and enzymes (glucose oxidase (GOx) and catalase (CAT)), pH‐responsive hydrogel nanoparticles, and multifunctional MnO2 nanoparticles. The plug functions both as a glucose sensor and controlled delivery unit to release higher rates of insulin from the reservoir in response to hyperglycemic BG levels and basal insulin rates at normal BG concentration. The surfaces of the device are modified by silanization followed by PEGylation to ensure its safety and biocompatibility and the stability of encased insulin. Our results show that insulin release can be modulated in vitro in response to glucose concentrations. In vivo experiments show that the glycemia of diabetic rats can be controlled with implantation of the prototype device. The glucose‐responsiveness of the device is also demonstrated by rapid drop in BG level after challenging diabetic rats with bolus injection of glucose solution. In addition, it is demonstrated that surface PEGylation of the device is necessary for reducing the immune response of the host to the implanted foreign object and maintaining insulin stability and bioactivity. With this molecular architecture and the bio‐inorganic nanocomposite plug, the device has the ability to maintain normal BG levels in diabetic rats.

We have developed glucose-responsive implantable microdevices for closed-loop delivery of insulin and conducted in vivo testing of these devices in diabetic rats. The microdevices consist of an albumin-based bioinorganic membrane that utilizes glucose oxidase (GOx), catalase (CAT) and manganese dioxide (MnO(2)) nanoparticles to convert a change in the environmental glucose level to a pH stimulus, which regulates the volume of pH-sensitive hydrogel nanoparticles and thereby the permeability of the membrane. The membrane is integrated with microfabricated PDMS (polydimethylsiloxane) structures to form compact, stand-alone microdevices, which do not require tethering wires or tubes. During in vitro testing, the microdevices showed glucose-responsive insulin release over multiple cycles at clinically relevant glucose concentrations. In vivo, the microdevices were able to counter hyperglycemia in diabetic rats over a one-week period. The in vitro and in vivo testing results demonstrated the efficacy of closed-loop biosensing and rapid response of the 'smart' insulin delivery devices.

It has been argued whether insulin accelerates or prevents atherosclerosis. Although results from in vitro studies have been conflicting, recent in vivo mice studies demonstrated antiatherogenic effects of insulin. Insulin is a known activator of endothelial nitric oxide synthase (NOS), leading to increased production of NO, which has potent antiatherogenic effects. We aimed to examine the role of NOS in the protective effects of insulin against atherosclerosis. Male apolipoprotein E-null mice (8 wk old) fed a high-cholesterol diet (1.25% cholesterol) were assigned to the following 12-wk treatments: control, insulin (0.05 U/day via subcutaneous pellet), N(ω)-nitro-l-arginine methyl ester hydrochloride (l-NAME, via drinking water at 100 mg/l), and insulin plus l-NAME. Insulin reduced atherosclerotic plaque burden in the descending aorta by 42% compared with control (plaque area/aorta lumen area: control, 16.5 ± 1.9%; insulin, 9.6 ± 1.3%, P < 0.05). Although insulin did not decrease plaque burden in the aortic sinus, macrophage accumulation in the plaque was decreased by insulin. Furthermore, insulin increased smooth muscle actin and collagen content and decreased plaque necrosis, consistent with increased plaque stability. In addition, insulin treatment increased plasma NO levels, decreased inducible NOS staining, and tended to increase phosphorylated vasodilator-stimulated phosphoprotein staining in the plaques of the aortic sinus. All these effects of insulin were abolished by coadministration of l-NAME, whereas l-NAME alone showed no effect. Insulin also tended to increase phosphorylated endothelial NOS and total neuronal NOS staining, effects not modified by l-NAME. In conclusion, we demonstrate that insulin treatment decreases atherosclerotic plaque burden and increases plaque stability through NOS-dependent mechanisms.

Anti-mitogenic agents currently used to prevent restenosis in drug-eluting stents delay re-endothelialization. Delayed re-endothelialization is now considered as the main cause of late stent thrombosis with drug-eluting stents, which emphasizes the need for new treatments. We have shown that systemic insulin treatment decreases neointimal growth and accelerates re-endothelialization after arterial injury in a rat model of restenosis. However, systemic insulin treatment cannot be given to non-diabetic individuals because of the risk of hypoglycemia. Thus, we investigated whether local insulin treatment is also effective in reducing neointimal growth after arterial injury. Rats were given local vehicle or local insulin delivered via Pluronic gel applied around the carotid artery immediately following balloon injury. Plasma glucose and systemic insulin levels were not affected by local insulin treatment. Insulin decreased intimal area at 28 days (P < 0.05) and also inhibited vascular smooth muscle cell migration by 60% at 4 days (P < 0.05). NPH (a longer-lasting insulin) also decreased neointimal area. These results indicate that local insulin treatment can lead to decreased restenosis, suggesting a protective vascular effect of insulin in vivo and that local insulin treatment, possibly via insulin-eluting stents, may be clinically relevant.