Abstract by Xin Zhou

The increasing use of therapeutic proteins and peptides in the pharmaceutical field brings challenges to formulation designers, mainly due to the stability of the protein/peptide drug, low bioavailability and poor compliance. Especially for insulin delivery, avoiding many injections per day can release the patients from discomfort and pain. Therefore, it is essential to design and realize systems able to guarantee adequate sustained release of insulin to maintain the basal insulin level for a long time. Although protein amyloid aggregates are often associated with a number of diseases, they can also serve as functional biomaterials. Indeed, they can be considered as alternative formulations of specific protein drugs and, making use of the possibility for them to disassemble under specific conditions, they may be used as proteins or peptides depots for drug delivery applications.

In this project we explore the possibility to formulate human insulin into an amyloid spherical structure, named spherulites, and use the non-ionic surfactant polysorbate 80 (PS80) to modulate the aggregates features. The final aims of this PhD project are: 1) to unravel how PS80 alters the spherulite structure and stability and 2) to prove via in vivo experiments the possibility to use insulin spherulites as insulin delivery system.

Firstly, the effects of PS80 on insulin aggregation into spherulites was investigated (manuscript 1, Appendix І). We found that the addition of PS80 delayed and slowed down the amyloid aggregation kinetics. Furthermore, the PS80 varied both the morphology and structure of insulin amyloid spherulites in a PS80 concentration-dependent mode. More specifically, when the PS80 concentration was over its critical micelle concentration, insulin spherulites were characterized by a denser packing and higher β-sheets content. This translates into a higher stability of spherulites when exposed to alkaline solutions. In alkaline conditions, spherulites released insulin maintaining the native structure, with the insulin amount depending on the PS80 incorporated into the spheulites.

Secondly, we investigated the mechanism of growth of spherulites using a single-molecule microscopy approach (manuscript 2, Appendix Ⅱ). We highlighted a specific anisotropic pattern of the growth, which was characterized by two rate constants and affected by the addition of PS80.

Inspired by the in vitro release of native-like insulin from the spherulites, we then designed an in vivo study, which demonstrated that spherulites could release bioactive insulin in vivo (manuscript 3, Appendix Ⅲ). Similar to the in vitro study, the spherulites incubated with PS80 had higher stability in vivo and released a reduced amount of insulin with a different temporal profile. This indicates that PS80 may modulate insulin spherulites and via this approach, we may obtain tunable delivery systems to meet diverse demands from patients.