Fluorine-19 Magnetic Resonance Imaging as a Potential Tool in Preclinical Research --- With Focus on Diabetes

Diabetes remains one of the increasing threats to human health with over 422 million people suffering from the disease all over the world. Despite the advancements in understanding the disease and improvements in treatment, diagnosis and early prevention with effective monitoring of pre- and post- therapy of the disease is still urgently needed. Hereby, the function and mass of pancreatic islets (PIs) and in particular their beta cells play an important role for monitoring the progression of the disease. Efforts have been made to stimulate research efforts to track them in a quantitative way. Hereby, magnetic resonance imaging (MRI) is considered as a safe (no ionizing radiation) prime aid with excellent soft tissue contrast and high-resolution with detailed anatomical information. In order to further enhance the sensitivity of MRI for cellular imaging, contrast agents are often utilized to label the targeted cells either by in situ labeling after systemic injection or ex vivo pre-labeling with subsequent transplantation. Recently, due to its unique feature of being background free, fluorine based contrast agents together with 19F MRI techniques have attracted attention, with first successful applications in immune cell tracking like dendritic cells and T cells in immunotherapy. This might also open new perspectives for diabetes research. The main aim of this PhD thesis was to contribute to the development and validation of PI and beta cell labeling with different fluorine-based contrast agents to follow their fate in vivo using 19F MRI. Based on the properties of the contrast agents, both in situ and ex vivo labeling strategy have been explored.

In the context of in situ labeling, fluorine labeled D-mannoheptulose (19FMH) was tested for its high specificity towards GLUT-2 transporters, which are highly expressed by beta cells and hepatocytes. With the establishment of an in-house built, double-tuned radio-frequency coil, we could assess the distribution of 19FMH in vivo. Due to the quick clearance of the compound and relative low sensitivity of 19F MRI, limited in vivo signal were detected in the liver and potentially from the pancreas. Nevertheless, ex vivo 19F MR spectroscopy (MRS) confirmed the preferential uptake of 19FMH in tissue with high expression of the GLUT-2 transporter. These results suggest that synthesis of 18F-labeled FMH and detection by the highly sensitive PET could be a potential alternative solution for the future to overcome the sensitivity limitations of 19F MRI. Ex vivo labeling was applied to isolated PIs and INS-1E cells using liposomal particles containing a DiD fluorescent dye and perfluoro-5-crown-ether (PFCE) for both 19F MRI and fluorescence imaging (FLI) detection. In addition, lentiviral vector (LV) transduction of PIs and INS-1E cells was also performed for bioluminescent imaging (BLI). The PFCE labeling and LV transduction protocols were optimized for PIs and cells in vitro to be sufficient for in vivo imaging purposes without affecting the viability and functionality of PIs and INS-1E cells. Longitudinal multimodal in vivo imaging of transplanted PIs and cells at subcutaneous sites demonstrated the feasibility of longitudinal cell tracking for several months with PFCE labels. This presented imaging approach addresses a potential solution to overcome limitations in sensitivity, resolution and specificity of individual imaging techniques using a multimodal approach.

The inherent sensitivity limitation of 19F MRI was also tackled by implementation of data analysis tools like the compressed sensing (CS) technique with a focus on low signal-to-noise ratio (SNR) images of PFCE labeled PIs and INS-1E cells. Both, offline simulation and CS acquisition with different reconstruction algorithms showed a three- to fourfold increase of SNR per time. These preliminary results not only demonstrated again that the CS technique together with appropriate reconstruction algorithms is useful to reduce the scan time for 19F MRI application with low SNR, but also provides evidence that by using the CS techniques, extra signal of interest could be detected, which might be missed otherwise by using conventional 19F MRI acquisition schemes within the same scan time.

In this thesis, 19F MRI was also applied for a controlled release study from a pH-sensitive capsule. By combining 19F MRI and CT with fluoro-deoxyglucose (19FDG) and BaSO4 as contrast agents, respectively, information on early permeability of capsule material to water, the capsule’s integrity and its exact anatomical location of the site of release could be obtained both in vitro and in vivo in a hamster model. This novel imaging approach could serve as a valuable tool for understanding and evaluation of different models of controlled release mechanisms.

This PhD work has added knowledge to the further understanding, improvement and validation of 19F MRI as a preclinical imaging tool for cell tracking and controlled release. The future of 19F MRI remains bright due to its versatile applications in different biomedical fields. This work will hopefully contribute to a solid basis not only for its future applications and development in preclinical research, but also for the translation of the technique to the clinic.