Synthetic smart gel shows promise as artificial pancreas

smart gel artificial pancreas

Date: 22nd June 2020

Diabetes is one of the top ten causes of death worldwide and there are many approaches to diabetes management currently being employed.  From medication, insulin injections or pumps, to islet cell transplants or the use of artificial pancreata. However, each comes with its own strengths and weakness, and suitability to the user. Now scientists have developed a low-cost, electronic-free, glucose-responsive gel technology as an in vivo insulin delivery system.

For many, electronic-based “closed-loop” systems are the most suitable option for diabetes management as the device automatically and continuously controls blood glucose levels.  However, these type of electronic devices pose a number of technical issues such as high cost, burdensome sensor calibration, cumbersome wearing and risk of electronic failures.  An artificial pancreas without electronics would therefore be a huge advancement in treating diabetics.

A team of scientists led by Takayoshi Suganami, from Nagoya University, Japan, had previously started to develop such electronic-free technology, and they wanted build on this so-called smart gel technology by making optimisations that were directly relevant to clinical translation such as scalability and efficacy for glucose spikes  – daily blood glucose fluctuations. To build on this original work therefore the team needed to address these key challenges.

The original smart gel was designed as a protein-free, totally synthetic material-based device.  It exploited the sugar-responsive properties of boronic acid meaning the boronate gel-base was acutely glucose responsive.  This responsiveness manifested as a change in hydration properties of the gel in a glucose-dependant manner.  Under optimised conditions, a microscopically dehydrated layer or ‘skin layer’ formed, which effectively switched off the release or diffusion of the gel which was loaded with human insulin.  In contrast, high levels of glucose, caused the gel to become charged, causing the gel to hydrate, disappearance of the skin layer and subsequently allowing the release of insulin.  This delivery device, which was confined within a single catheter and implanted into mice, exhibited an artificial pancreas–like function in vivo for up to 3 weeks.

In order to address scale-up of the tech, the team optimised the system for use in the rat, which is 10 times the size of the mouse.  The device was diffusion dependant, therefore the team hypothesised by increasing the surface area they could effectively scale up the model.

To achieve this the team used a commercially available hollow fibre dialyser (dissembled into bundles of ~25 fibres) and coated the surface of the fibres with a thin coat of the gel.  The bundles were then further coated in a biocompatible polymer, and attached to an insulin reservoir, which was filled with recombinant human insulin.

In vivo characterisation

Initially the smart gel device was subcutaneously implanted into healthy rats and glucose tolerance tests (GTT) were performed in which glucose was injected following 6 hours of fasting on day 2 and 7.  Hypoglycemia did not occur in healthy rats implanted with the insulin-loaded device and it significantly attenuated a transient increase in blood glucose levels compared to control rats which had been implanted with a phosphate-buffered saline (PBS)-loaded device. In rats implanted with the smart gel device, the team also observed an increase in serum concentrations of human insulin that were concomitant with a decrease in endogenous concentrations of insulin – suggesting the device could reduce the burden of the endogenous insulin-secreting pancreatic β-cells.  The blood-glucose-lowering effect of the device was sustained for at least 7 days, showing durability of system.

To next evaluate the biosafety of the device, the team histologically examined the implant area at day 9.   They saw only mild inflammatory cell infiltration around the device and very little change in white blood cell numbers, suggesting the device was biocompatible.

Therapeutic efficacy for type 1 diabetic rats

To assess the therapeutic effect they implanted the device into a type 1 diabetic rat model, and demonstrated that the insulin-loaded device could effectively reduce blood glucose levels and that the device continuously released insulin up to day 7 (the end of the experiment).  In contrast, glucose levels remained high in rats implanted with the PBS-loaded device.

Whilst the device was effective at controlling glucose levels in a severe form of diabetes the team also wanted to tests its effectiveness for mild diabetes.

glucose fluctuations diabetes

Amelioration of glucose fluctuation for mild diabetes

A major complication for diabetic patients is athlerosclerosis, which is likely caused by daily glucose fluctuations such as hyperglycaemia. Mild diabetes was therefore induced in rats, and subcutaneous glucose levels were monitored with a continuous glucose monitoring system pre- and post-implantation of the insulin-loaded device.  The team found that, pre-implantation, the rats showed daily glucose fluctuations throughout the day and night.  However, in particular an increase in the subcutaneous glucose levels at night (hyperglycaemia) was seen whilst lower levels were seen during the day (although the levels during the day still remained higher than the control animals). As these are nocturnal animals their active phase was at night and was when most of the feeding occurred. Post-implantation a large decrease in the subcutaneous glucose levels at night were seen, whilst fluctuations during the day became almost stable.

Conclusion and future applications:

The increasing prevalence of diabetes means that there is a drive to develop a new generation of therapeutics and treatments for the disease.  Here, synthetic biology is playing a crucial and accelerating role in the development of these technologies.

We have recently seen, for example, light-activated, insulin-producing cells designed to combat diabetes).  Synthetic biology has also been applied to manufacturing processes and we recently reported the generation of nanomachines; engineered enzymes created in order to yield a cleaner and quicker method of producing anti-diabetic drugs.

Here the team have developed a synthetic smart gel device to function as an artificial pancreas.  Able to normalise glucose levels and ameliorate glucose fluctuations without inducing hypoglycaemia, this device may offer a lost-cost alternative to current electronic-based approaches.

It currently represents the only known chemical system that can achieves both durability and an acute response in a matter of seconds.

One of the key determinants for the translation into the clinic is the scalability of the system.  The team have demonstrated here proof-of-concept that this smart device could be scaled up from the mouse to the rats.  They have also developed a mathematical model, which indicated that the device could be further up-scaled by 10-fold.  One of the limiting factors may be the insulin reservoir, but the team suggest the use of a concentrated type of insulin, already in clinical use, may further scale up the effectiveness of the device.

It is still early days for this device and there are many hurdles to overcome before it can translate into a safe and efficient clinic-ready technology.  Many questions need to be addressed such as long-term durability and insulin release rate which would have to dramatically increase prior to clinical translation.  However, the attractiveness of the system lies in its synthetic electronic-free nature, not reliant on living cells, enzymes or gene-editing and it may offer an exciting alternative for diabetics in the future.


Matsumoto, A., H. Kuwata, S. Kimura, H. Matsumoto, K. Ochi, Y. Moro-oka, A. Watanabe, H. Yamada, H. Ishii, T. Miyazawa, S. Chen, T. Baba, H. Yoshida, T. Nakamura, H. Inoue, Y. Ogawa, M. Tanaka, Y. Miyahara and T. Suganami (2020). “Hollow fiber-combined glucose-responsive gel technology as an in vivo electronics-free insulin delivery system.” Communications Biology 3(1): 313.