Cell-internalized MRI contrast agents

Powerful contrast agents are essential for molecular imaging, a new and rapidly developing field that focuses on in vivo imaging of biological markers and processes. Chapter 1 provided an introduction on molecular imaging and gave an overview of research conducted in four major application areas: cancer, atherosclerosis, myocardial infarction and neurological disorders. Furthermore the parameters that play a critical role in molecular imaging were addressed, focusing on the challenges and opportunities of molecular imaging based on MRI. The aim of this thesis was to study the interaction of targeted MRI contrast agents with cultured mammalian cells, focusing on cellular location, relaxometry and quantification. Several lipid-based MRI contrast agents, including liposomes, quantum dots and emulsions, were incubated with human umbilical vein derived endothelial cells (HUVECs) as an in vitro model for assessing vital aspects of molecular MRI. In chapter 2 the in vitro model setup was presented that was used throughout this thesis work to investigate the consequences of uptake of high payload multimodal paramagnetic contrast agents on their T1- and T2-shortening efficacy. Cultured HUVECs were incubated with paramagnetic liposomes that were conjugated with a cyclic RGD-peptide to enable internalization via the 2343-integrin receptor. Non-targeted liposomes served as a control. Integrin targeting indeed strongly increased the uptake of paramagnetic liposomes, which were deposited in perinuclear vesicles. This amplification strategy, however, strongly reduced the longitudinal relaxivity of the internalized paramagnetic liposomes. Chapter 3 reported on the investigations into the longer-term fate of the cell-internalized liposomes by studying their relaxometric properties over 5 days, following an initial 24 hr loading period. Circa 25% of the Gd3+ content delivered to the cells via integrin-targeted liposomes was lost in the first 24 hr, which led to 65% and 77% reductions in R1 and R2, respectively, as compared to the original R1 and R2 enhancements. This implies that the remaining cell-associated gadolinium had relatively low effective r1 and r2 relaxivities. It was proposed that this is due to gradual release of Gd3+ from the chelate in the cell, followed by sequestration in an MR silent state. Most of the gadolinium internalized by cells following incubation with non-targeted liposomes was released in the 5-day follow-up period. In chapter 4 a mathematical model was presented that describes the effective longitudinal relaxation rate R1 for water protons in three cellular compartments (extracellular, cytoplasmic and vesicular subcellular spaces) as a function of compartment size, intercompartmental water exchange rates and local contrast agent concentration. The model was used to explain the effective R1 dependence on the overall concentration of cell-internalized Gd3+-containing liposomes that was measured in chapter 2. Relaxation parameters were simulated using a modified Bloch-McConnell equation including magnetization exchange between the three compartments. With the model several possible scenarios for internalized contrast agent distribution were evaluated. Relaxation parameters were calculated for contrast agent restricted to the cytoplasmic or vesicular compartments. The voxel contrast agent concentration dependencies of R1 can be used to qualitatively and quantitatively understand a number of different experimental observations reported in the literature. Most importantly the simulations reproduced the longitudinal relaxivity "quenching" for cell-internalized liposomes described in chapters 2 and 3. Chapter 5 described the relaxometric, optical and compositional properties of 2v43-integrin targeted- and non-targeted paramagnetic quantum dot micelles (pQDs) after incubation with HUVECs. pQDs are promising nanoparticles for bimodal molecular imaging purposes: their excellent optical properties allow for in vivo optical imaging in small animals, while the high Gd payload allows for detection using contrast-enhanced MRI. Cell-associated pQDs showed no concentration or time-dependent change in fluorescent intensities and cellular relaxivities, enabling accurate assessment of the their uptake by HUVECs using relaxometric and optical measurements. However, the molar ratio of Gd (originating from the lipid coating) to Cd (from the QD core material) in pellets containing pQD-incubated cells was significantly higher than the Gd/Cd molar ratio of the pQDs as prepared. It was proposed that this was due to non-specific lipid-transfer between the pQDs and the cellular membranes. These findings show that imaging read-outs from non-covalent contrast agent assemblies may become ambiguous due to reduced co-localization of the different imaging labels in biological environments. In vivo molecular imaging with targeted MRI contrast agents benefits from methods to quantify local contrast agent concentrations. In chapter 2-5, it was shown that quenching of the relaxivity induced by cellular internalization hampers contrast agent quantification based on changes in MRI contrast parameters. In the in vitro study described in chapter 6 we have investigated the quantification potential of a combination of 1H MRI, 19F MRI and 19F MRS, using a lipid-coated paramagnetic perfluorocarbon emulsion. HUVECs took up the 2343-integrin targeted emulsion to a higher extent than non-targeted emulsion. Association of the targeted emulsion with the cells resulted in a concentration-dependent proton R1 with different values for targeted and control nanoparticles, prohibiting unambiguous quantification of local contrast agent concentration, using 1H-MRI. Upon cellular association, the fluorine R1 remained constant with concentration, while the fluorine R2 showed a concentration-dependent increase. Even though the fluorine R2 was not constant, the 19F MRI and 19F MRS signals were linear and quantifiable as function of nanoparticle concentration. Detection limits, however, were considerably higher for 19F than for 1H-MR. Chapter 7 reported on the development of a multimodal contrast agent for combined SPECT/MRI imaging. The development of hybrid scanners in which two imaging modalities are combined in a single setup for simultaneous dual-modality scanning is a new and exciting development in medical imaging. Dual modality imaging greatly benefits from multimodal probes that can be detected by both techniques. The liposomal contrast agent described in this chapter, contained an 111Indium label for SPECT and gadolinium-chelates for MRI contrast, creating a platform for simultaneous quantification (based on SPECT) and precise spatial localization (based on MRI) using a single probe. To explore the utility of the liposomal probe, incubations were done with HUVECs and cell samples were analyzed using MRI, SPECT/CT, g-counting, confocal laser scanning microscopy (CLSM) and inductively coupled plasma mass spectrometry (ICP-MS). MRI measurements revealed quenching of the r1 relaxivities as reported in the previous chapters. Furthermore, it was found that the radiolabel allowed accurate quantification of cell-internalized probe using g-counting. In chapter 8, an overview was given of the results of incubations with the different contrast agents and the findings regarding cellular location, relaxometric properties and cellular uptake were discussed. Furthermore, the utility of gadolinium-based paramagnetic MRI contrast agents was reviewed, followed by a discussion on the perspectives of multimodality imaging