Synthesis, characterisation and biological activity of FITC-insulin for the development of an artificial pancreas

In advanced drug development and delivery, fluorescence studies have clarified and improved aspectssuch as biodistribution, stability and metabolism with respect to the complexities imposed by thebiological systems. The ultimate aim of this study was to assess the delivery kinetics of a synthesisedfluorescently labelled insulin in an implantable artificial pancreas (INsmart device) developed by Tayloret al, 2016, which is currently being tested in-vivo. The first objective of this thesis was to produce a fluorescein isothiocyanate (FITC)-insulin conjugatewhich shows equivalent biological activity to native insulin under novel reaction conditions without theneed for using protecting groups and multi-step synthetic conditions. Secondly, the stability andsolubility profiles of the synthesised FITC-insulin conjugate in solution will be investigated for long-term storage and future applications. Thirdly, physiologically relevant glucose concentrations will beused to assess the performance of FITC-insulin delivery from an INsmart device. Lastly, other dyessuch as eosin isothiocyanate (EITC) and rhodamine B isothiocyanate (RBITC) will be assessed aslabelling candidates to produce other derivatised insulin conjugates. Mono-labelled FITC-insulin conjugate was successfully synthesised using a molar ratio of 2:1 (FITC:insulin) with short reaction times (up to 18h) at pH7 after studies were conducted to examine theeffects of reaction time, molar ratio and pH. The labelling position of this mono-labelled species wasidentified by MS-Orbitrap Fusion at the B1 residue (MonoB1). However, during synthesis,MonoB1conjugate always contains some trace amounts of unlabelled insulin that required furtherpurification by RP-HPLC using a gradient method. This HPLC method could identify four FITC-insulinconjugates including two mono-labelled species (labelled at the A1 or B1 position), di-labelled species(labelled at the A1 and B1) and tri-labelled species (labelled at the A1, B1 and B29). Further analysiswas performed using MALDI-MS to confirm the molecular weight of each conjugate produced. The biological activity of four FITC-insulin conjugates was assessed in human umbilical veinendothelial cells (HUVEC) and skeletal muscle cells (C2C12) via the insulin signalling pathway byexamining the levels of AKT phosphorylation (pAKT) and cell surface GLUT4. There was no significantdifference in pAKT and the GLUT4 cell surface levels observed for synthesised MonoB1 compared tonative insulin, highlighting that this conjugate was as biologically active as native insulin. The enhanced stability and solubility of FITC-insulin conjugate using diluting fluid containing m-cresol,glycerol and zinc oxide, which is typically contained in most commercial insulin formulations arebeneficial for setting up the in-vitro delivery study of the INsmart device. Because it allows aconcentrated depot of FITC-insulin in the device for release over an extended time. Improvement inthe smart gel formulation with diluting fluid and the use of correct membrane pore size showedpromising and reproducible results in the extended experiments where the INsmart device has beenset up and triggered multiple times with glucose and other dietary saccharides which act as controls toshow the specificity of the gel to glucose challenges. The results indicated that the device is capable ofdelivering basal insulin dose which can be boosted in response to multiple (11x) mealtime glucosesurges over a 5-day period. EITC and RBITC were used as fluorescent candidates to assess whether the same syntheticmethodology of fluorescently labelled insulin was applicable to different dye sizes. Mono-labelledinsulin conjugate with EITC was achieved using the same reaction conditions as in FITC-insulinsynthesis. Further development is still needed to remove unreacted RBITC during the synthesis ofRBITC-insulin conjugate.