Diabetes mellitus type 2 is a metabolic disorder that reaches epidemic dimensions and represents a growing burden for patients and society. Existing therapeutic possibilities have either unsatisfactory effects or unwanted side effects. The cost for development of a new medicine is estimated to be 2.6 B USD. Hence, for the benefit of people with or at risk of developing diabetes and diabetes complications it is important to invest in potential novel treatments with a high level of likelihood for success. I.e. there is great need for a translational medicine approach that can estimate what will happen in costly and time consuming clinical trials. The presented in vivo translational medicine validation platform (for short ‘in vivo platform’) in which rodent and human pancreatic islets are transplanted into the anterior chamber of the eye (ACE) of mice, with or without diabetes, has the opportunity to fulfill such a need. This in vivo platform will allow functional microscopic imaging and assessment of islet β-cell function and survival in the living organism non-invasively, longitudinally and at single-cell resolution.
The major objective of this ERC-PoC proposal was to establish a robust pharma-industry in vivo platform for validating novel diabetes treatments in early drug development. The major milestone of this ERC-PoC grant was the in vitro and in vivo validation of biosensors for β-cell function and survival in rodent and human pancreatic islets transplanted into the ACE of normal and diabetic mice to establish protocols for drug testing.
We focused on the validation of biosensors reporting on changes in cytoplasmic free calcium concentration ([Ca2+]i) and regeneration. [Ca2+]i is a well-respected readout of cell function and survival. Based on our experience gained within the ERC-AdG grant 338936, we concentrated on the GCaMP-family members GCaMP3 and GCaMP6. Within the activities of our ERC-AdG grant we have constructed several viral vectors encoding inducible and non-inducible Ca2+-biosensors as well as transgenic mice expressing Ca2+-biosensors in their β-cells (RIP-Cre:GCaMP3 and INS-Cre:GCaMP3). As a biosensor for β-cell regeneration we used the described Fucci-system (Sakaue-Sawano et al., Cell 2008).
With regard to using rodent islets as a screening system, biosensor-expressing islets from transgenic mice proved to be most robust. For transductions of human islets with viruses encoding biosensors, adenoviruses can be used as vehicles.
As immune-compromised recipients for human islets (to avoid rejection of the graft) we have tested nude mice, NSG mice as well as Rag-/- mice. All three mouse strains are suitable for engraftment and monitoring of human islets in vivo. Noteworthy, when testing NSG mice with different protocols for diet-induced obesity/diabetes we observed that these mice do not respond to dietary challenges as C57Bl/6 mice. Therefore, we are currently testing Rag-/- mice, which have a C57Bl/6 genetic background and do respond to diet intervention.
In order to study β-cell function in vivo in pancreatic islets transplanted into normal and diabetic mice, we tested different diet-intervention protocols in diabetes-prone male C57Bl/6J mice. When comparing high-fat-diet, Western-diet, and combination diets consisting of solid high-fat and drinking water supplemented with either 32% sucrose (HFHSD) or 32% fructose (HFHFrD), we observed differences in the severity and dynamics towards the progression of a diabetic phenotype. While all protocols led to obesity and whole-body insulin resistance, only the two combinatory diets HFHSD and HFHFrD led to the development of a diabetic phenotype within 4-8 weeks of treatment. Interestingly, while the HFHSD showed a more ‘β-cell-centered’ phenotype including β-cell insulin resistance and non-compensatory insulin release, the HFHFrD was more ‘liver-centered’ without developing β-cell insulin resistance but severe liver insulin resistance.
In conclusion, in this ERC-PoC we have successfully established robust protocols to be used by academia or pharma-industry allowing validation of lead-compounds for the treatment of diabetes in an early stage of development utilizing our in vivo imaging platform. This included validation of biosensors with major focus on Ca2+-biosensors, establishing ‘humanized mouse’ models as well as dietary-intervention protocols allowing the progression towards a diabetes phenotype within 4-8 weeks.
