Evaluation of cell transplant-mediated attenuation of diffuse injury in experimental autoimmune encephalomyelitis using onVDMP CEST MRI.

The development and translation of cell therapies have been hindered by an inability to predict and evaluate their efficacy after transplantation. Using an experimental autoimmune encephalomyelitis (EAE) mouse model of multiple sclerosis (MS), we studied attenuation of the diffuse injury characteristic of EAE and MS by transplanted glial-restricted precursor cells (GRPs). We assessed the potential of on-resonance variable delay multiple pulse (onVDMP) chemical exchange saturation transfer (CEST) MRI to visualize this attenuation. Allogeneic GRPs transplanted in the motor cortex or lateral ventricles attenuated paralysis in EAE mice and attenuated differences compared to naïve mice in onVDMP CEST signal 5 days after transplantation near the transplantation site. Histological analysis revealed that transplanted GRPs co-localized with attenuated astrogliosis. Hence, diffuse injury-sensitive onVDMP CEST MRI may complement conventional MRI to locate and monitor tissue regions responsive to GRP therapy.

Quantitative "Hot Spot" Imaging of Transplanted Stem Cells using Superparamagnetic Tracers and Magnetic Particle Imaging (MPI).

Magnetic labeling of stem cells enables their non-invasive detection by magnetic resonance imaging (MRI). Practically, most MRI studies have been limited to visualization of local engraftment as other sources of endogenous hypointense contrast complicate the interpretation of systemic (whole body) cell distribution. In addition, MRI cell tracking is inherently non-quantitative in nature. We report here on the potential of magnetic particle imaging (MPI) as a novel tomographic technique for non-invasive hot spot imaging and quantification of stem cells using superparamagnetic iron oxide (SPIO) tracers. Neural and mesenchymal stem cells, representing small and larger cell bodies, were labeled with three different SPIO tracer formulations, including two preparations that have previously been used in clinical MRI cell tracking studies (Feridex® and Resovist®). Magnetic particle spectroscopy (MPS) measurements demonstrated a linear correlation between MPI signal and iron content, for both homogeneous solutions of free particles in solution and for internalized and aggregated particles in labeled cells over a wide range of concentrations. The overall MP signal ranged from 1×10-3 - 3×10-4 Am2/g Fe, which was equivalent to 2×10-14 - 1×10-15 Am2 per cell, indicating that cell numbers can be quantified with MPI analogous to the use of radiotracers in nuclear medicine or fluorine tracers in 19F MRI. When SPIO-labeled cells were transplanted in mouse brain, they could be readily detected by MPI at a detection threshold of about 5×104 cells, with MPI/MRI overlays showing an excellent agreement between the hypointense MRI areas and MPI hot spots. The calculated tissue MPI signal ratio for 100,000 vs. 50,000 implanted cells was 2.08. Hence, MPI has potential to be further developed for quantitative and easy-to-interpret, tracer-based non-invasive imaging of cells, preferably with MRI as an adjunct anatomical imaging modality.

Survival of neural progenitors allografted into the CNS of immunocompetent recipients is highly dependent on transplantation site.

Allografts continue to be used in clinical neurotransplantation studies; hence, it is crucial to understand the mechanisms that govern allograft tolerance. We investigated the impact of transplantation site within the brain on graft survival. Mouse [Friend leukemia virus, strain B (FVB)] glial precursors, transfected with luciferase, were injected (3 × 10(5)) into the forceps minor (FM) or striatum (STR). Immunodeficient rag2(-/-) and immunocompetent BALB/c mice were used as recipients. Magnetic resonance imaging (MRI) confirmed that cells were precisely deposited at the selected coordinates. The graft viability was assessed noninvasively with bioluminescent imaging (BLI) for a period of 16 days. Regardless of implantation site, all grafts (n = 10) deposited in immunodeficient animals revealed excellent survival. In contrast, immunocompetent animals only accepted grafts at the STR site (n = 10), whereas all the FM grafts were rejected (n = 10). To investigate the factors that led to rejection of FM grafts, with acceptance of STR grafts, another group of animals (n = 19) was sacrificed during the prerejection period, on day 5. Near-infrared fluorescence imaging with IRDye 800CW-polyethylene glycol probe displayed similar blood-brain barrier disruption at both graft locations. The morphological distribution of FM grafts was cylindrical, parallel to the needle track, whereas cells transplanted into the STR accumulated along the border between the STR and the corpus callosum. There was significantly less infiltration by both innate and adaptive immune cells in the STR grafts, especially along the calloso-striatal border. With allograft survival being dependent on the transplantation site, the anatomical coordinates of the graft target should always be taken into account as it may determine the success or failure of therapy.

Imaging of cellular therapies.

Cellular therapy promises to revolutionize medicine, by restoring tissue and organ function, and combating key disorders including cancer. As with all major developments, new tools must be introduced to allow optimization. For cell therapy, the key tool is in vivo imaging for real time assessment of parameters such as cell localization, numbers and viability. Such data is critical to modulate and tailor the therapy for each patient. In this review, we discuss recent work in the field of imaging cell therapies in the clinic, including preclinical work where clinical examples are not yet available. Clinical trials in which transferred cells were imaged using magnetic resonance imaging (MRI), nuclear scintigraphy, single photon emission computed tomography (SPECT), and positron emission tomography (PET) are evaluated from an imaging perspective. Preclinical cell tracking studies that focus on fluorescence and bioluminescence imaging are excluded, as these modalities are generally not applicable to clinical cell tracking. In this review, we assess the advantages and drawbacks of the various imaging techniques available, focusing on immune cells, particularly dendritic cells. Both strategies of prelabeling cells before transplant and the use of an injectable label to target cells in situ are covered. Finally, we discuss future developments, including the emergence of multimodal imaging technology for cell tracking from the preclinical to the clinical realm.

Basic research and clinical applications of non-hematopoietic stem cells, 4-5 April 2008, Tubingen, Germany.

From 4 to 5 April 2008, international experts met for the second time in Tubingen, Germany, to present and discuss the latest proceedings in research on non-hematopoietic stem cells (NHSC). This report presents issues of basic research including characterization, isolation, good manufacturing practice (GMP)-like production and imaging as well as clinical applications focusing on the regenerative and immunomodulatory capacities of NHSC.

Applicability and limitations of MR tracking of neural stem cells with asymmetric cell division and rapid turnover: the case of the shiverer dysmyelinated mouse brain.

LacZ-transfected C17.2 neural stem cells (NSCs) were labeled with the superparamagnetic iron oxide formulation Feridex prior to ICV injection in shi/shi neonates. Feridex labeling did not alter cell differentiation in vitro and in vivo. Initially, MR images obtained at 11.7T correlated closely to NSC distribution as assessed with anti-dextran and anti-beta-galactosidase double-fluorescent immunostaining. However, at 6 days postgrafting there was already a pronounced mismatch between the hypointense MR signal and the histologically determined cell distribution, with a surprisingly sharp cutoff rather than a gradual decrease of signal. Positive in vivo BrdU labeling of NSCs showed that significant cell replication occurred post-transplantation, causing rapid dilution of Feridex particles between mother and daughter cells toward undetectable levels. Neural differentiation experiments demonstrated asymmetric cell division, explaining the observed sharp cutoff. At later time points (2 weeks), the mismatch further increased by the presence of non-cell-associated Feridex particles resulting from active excretion or cell death. These results are a first demonstration of the inability of MRI to track rapidly dividing and self-renewing, asymmetrically dividing SCs. Therefore, MR cell tracking should only be applied for nonproliferating cells or short-term monitoring of highly-proliferative cells, with mitotic symmetry or asymmetry being important for determining its applicability.

Radiopaque alginate microcapsules for X-ray visualization and immunoprotection of cellular therapeutics.

Alginate-poly-L-lysine-alginate (APA) microcapsules have been explored as vehicles for therapeutic drug and cell delivery. The permselectivity of these capsules provides a unique means of controlled drug release and immunoisolation of encapsulated cells. Immunoisolation is especially attractive as it abrogates the need for chronic immunosuppressive therapy and opens up the possibility for the delivery of numerous cell sources including xenogeneic grafts. APA microcapsules containing cellular therapeutics have proven effective in the short-term treatment of a wide range of diseases requiring enzyme or endocrine replacement therapy, including type I diabetes. If these microcapsules could be noninvasively monitored with X-ray imaging modalities (i.e., fluoroscopy, CT, and digital subtraction angiography), questions such as the ideal transplantation site, the best means of delivery, and the long-term survival of grafts could be better addressed. We have developed two novel alginate-based radiopaque microcapsule formulations containing either barium sulfate (Ba X-Caps) or bismuth sulfate (Bi X-Caps). As compared to conventional, nonradiopaque APA capsules, Ba X-Caps and Bi X-Caps containing human cadaveric islets resulted in a decrease in cellular viability of less than 5% up to 14 days after encapsulation. Both radiopaque capsules were found to be permeable to lectins < or =75 kDa, but were impermeable to lectins > or =120 kDa, thus ensuring the blockage of the penetration of antibodies while allowing free diffusion of insulin and nutrients. The glucose-responsive insulin secretion of the radiopaque encapsulated human islets was found to be unaltered compared to that of unlabeled controls, with human C-peptide levels ranging from 3.21 to 2.87 (Ba X-Caps) and 3.23 to 2.87 (Bi X-Caps) ng/islet at 7 and 14 days postencapsulation, respectively. Using fluoroscopy, both Ba X-Caps and Bi X-Caps could be readily visualized as single radiopaque entities in vitro. Furthermore, following transplantation in vivo in mice and rabbits, single capsules could be identified with no significant change in contrast for at least 2 weeks. This study represents the first attempt at making radiopaque microcapsules for X-ray guided delivery and imaging of cellular therapeutics. While human cadaveric islets were used as a proof-of-principle, these radiopaque capsules may have wide ranging therapeutic applications for a variety of cell types.

MR imaging of lineage-restricted neural precursors following transplantation into the adult spinal cord.

Neural precursor cell (NPC) transplantation is a promising strategy for treatment of CNS injuries and neurodegenerative disorders because of potential for cell replacement. An important element of future clinical applications is development of a non-invasive procedure to follow NPC fate. We show that neuronal-restricted precursors (NRPs) and glial-restricted precursors (GRPs), NPCs with lineage restrictions for neurons and glia, respectively, can be labeled in vitro with the superparamagnetic iron oxide contrast agent Feridex. Following engraftment into intact adult spinal cord, labeled cells robustly survived in white and gray matter and migrated selectively along white matter tracts up to 5 mm. Localization of cells was reliably established using ex vivo magnetic resonance imaging of spinal cords. Imaging coincided with histological detection of iron and the human alkaline phosphatase transgene in most grafting sites, including the stream of migrating cells. Following transplantation, magnetically labeled cells exhibited mature morphologies and differentiated into neurons, astrocytes, and oligodendrocytes, similar to grafts of unlabeled NRPs and GRPs. Interestingly, Feridex-labeled cells, but not unlabeled cells, induced influx of ED1-positive macrophages/microglia. Small numbers of these phagocytic cells took up iron from grafted cells, while the majority of Feridex label was found in transplanted cells. We conclude that Feridex labeling does not inhibit NPC differentiation and can be used to reliably localize NPCs by MRI following engraftment into adult CNS, with the possible exception of areas of rapidly proliferating cells. The present results are relevant for MR-guided clinical application of transplantation strategies in treatment of spinal cord injury and other CNS pathologies.

Instant MR labeling of stem cells using magnetoelectroporation.

For cellular MR imaging, conventional approaches to intracellular magnetic labeling of nonphagocytic cells rely on the use of secondary compounds such as transfection agents and prolonged incubation of cells. Magnetoelectroporation (MEP) was investigated as an alternative method to achieve instant (<1 s) endosomal labeling with the FDA-approved formulation Feridex, without the need for adjunct agents or initiating cell cultures. While MEP was harmful at higher voltages or pulse durations, the procedure could be properly calibrated using a pulse of 130 V and 17 ms. Labeling was demonstrated for stem cells from mice, rats, and humans; the uptake of iron was in the picogram range and comparable to values obtained using transfection agents. MEP-labeled stem cells exhibited an unaltered viability, proliferation, and mitochondrial metabolic rate. Labeled mesenchymal stem cells (MSCs) and neural stem cells (NSCs) differentiated into adipogenic, osteogenic, and neural lineages in an identical fashion as unlabeled cells, while containing Feridex particles as demonstrated by double immunofluorescent staining. MEP-labeled NSCs proliferated normally following intrastriatal transplantation and could be readily detected by MR imaging in vivo. As MEP circumvents the use of secondary agents, obviating the need for clinical approval, MEP labeling may be ideally suitable for bedside implementation.

In vivo and ex vivo MRI detection of localized and disseminated neural stem cell grafts in the mouse brain.

The application of stem cells as delivery vehicles opens up the opportunity for targeting therapeutic proteins to the damaged or degenerating central nervous system. Neural stem cell (NSC) lines have been shown to engraft, differentiate and correct certain central nervous system diseases. The present study was performed to test the ability of magnetic resonance imaging (MRI) in detecting transplanted NSCs under conditions of limited migration in the normal adult mouse brain versus widespread migration when the cells are transplanted neonatally. The C17.2 NSC line was labeled in vitro with superparamagnetic iron oxide (SPIO) particles and the labeled cells were implanted intracranially. Serial in vivo gradient echo MR imaging was performed using a 4.7 T horizontal bore magnet. High resolution ex vivo images of the isolated brains were performed at 9.4 T, and the presence of iron was correlated with Prussian blue staining in histological sections. Adult animals injected with SPIO-labeled stem cells exhibited hypointense regions near the injection site that were observed up to 32 days after injection. In neonatally transplanted animals, MR signal intensity from transplanted NSCs was not apparent in in vivo imaging but ex vivo MR images revealed small hypointense regions throughout the brain including the olfactory bulbs, cortex and the cerebellum, reflecting the wide distribution of the engrafted cells. These regions were correlated with Prussian blue staining, which confirmed the presence of SPIO particles inside the engrafted cells. We have shown that MRI is capable of differentiating localized and widespread engraftment of C17.2 stem cells in the central nervous system.

MR-trackable intramyocardial injection catheter.

There is growing interest in delivering cellular agents to infarcted myocardium to prevent postinfarction left ventricular remodeling. MRI can be effectively used to differentiate infarcted from healthy myocardium. MR-guided delivery of cellular agents/therapeutics is appealing because the therapeutics can be precisely targeted to the desired location within the infarct. In this study, a steerable intramyocardial injection catheter that can be actively tracked under MRI was developed and tested. The components of the catheter were arranged to form a loopless RF antenna receiver coil that enabled active tracking. Feasibility studies were performed in canine and porcine myocardial infarction models. Myocardial delayed-enhancement (MDE) imaging identified the infarcted myocardium, and real-time MRI was used to guide left ventricular catheterization from a carotid artery approach. The distal 35 cm of the catheter was seen under MRI with a bright signal at the distal tip of the catheter. The catheter was steered into position, the distal tip was apposed against the infarct, the needle was advanced, and a bolus of MR contrast agent and tissue marker dye was injected intramyocardially, as confirmed by imaging and postmortem histology. A pilot study involving intramyocardial delivery of magnetically labeled stem cells demonstrated the utility of the active injection catheter system.

111In oxine labelled mesenchymal stem cell SPECT after intravenous administration in myocardial infarction.

Mesenchymal stem cells (MSCs) have shown therapeutic potential if successfully delivered to the intended site of myocardial infarction. The purpose of this pilot study was to test the feasibility of 111In oxine labelling of MSCs and single photon emission computed tomography (SPECT) imaging after intravenous administration in a porcine model of myocardial infarction. Adult farm pigs (n=2) were subjected to closed chest experimental myocardial infarction. 111In oxine labelled MSCs (1 x 10(7) to 2 x 10(7) cells) were infused intravenously, and SPECT imaging was performed initially and on days 1, 2, 7 and 14. High quality SPECT images were obtained through 2 weeks of imaging. High initial MSC localization occurred in the lungs and slow progressive accumulation occurred in the liver, spleen and bone marrow. Renal activity was mild and persistent throughout imaging. No appreciable accumulation occurred in the myocardium. It is concluded that 111In oxine radiolabelling of MSCs is feasible, and in vivo imaging with SPECT provides a non-invasive method for sequentially monitoring cell trafficking with good spatial resolution. Because intravenous administration of MSCs results in significant lung activity that obscures the assessment of myocardial cell trafficking, alternative routes of administration should be investigated for this application.

Combination of transfection agents and magnetic resonance contrast agents for cellular imaging: relationship between relaxivities, electrostatic forces, and chemical composition.

The purpose of this study was to investigate the changes in electrostatic and magnetic resonance (MR) properties observed when MR contrast agents (CAs) (Feridex, MION-46L, or G5-dendrimer-DOTA-Gd) are combined with transfection agents (TAs) under various conditions for use as a CA-TA complex basis for cellular labeling and MRI. CAs were incubated with various classes of TAs for 0-48 hr in solutions of varying concentrations and pH values. NMR relaxation rates (1/T(1), 1/T(2)), MRI and zeta potential (ZP) of CA-TA solutions were measured. TAs decreased the 1/T(1) and 1/T(2) of G5-DOTA-Gd, Feridex, and MION-46L by 0-95%. Altering the pH of G5-DOTA-Gd-TA decreased the T(1)-weighted signal intensity (SI) on MRI from 0 to 78%. Measured ZP values for G5-DOTA-Gd, Feridex, and MION-46L were -51, -41, and -2.0 mV, respectively. The TA LV had a negative ZP, while the other TAs had ZPs ranging from +20 to +65 mV. The alteration of the ZP and NMR relaxivities of the MR CAs, Feridex, MION-46L, and G5-DOTA-Gd by TAs has been demonstrated. These results enhance our understanding of the relationship between electrostatic and MR properties.

Magnetodendrimers allow endosomal magnetic labeling and in vivo tracking of stem cells.

Magnetic resonance (MR) tracking of magnetically labeled stem and progenitor cells is an emerging technology, leading to an urgent need for magnetic probes that can make cells highly magnetic during their normal expansion in culture. We have developed magnetodendrimers as a versatile class of magnetic tags that can efficiently label mammalian cells, including human neural stem cells (NSCs) and mesenchymal stem cells (MSCs), through a nonspecific membrane adsorption process with subsequent intracellular (non-nuclear) localization in endosomes. The superparamagnetic iron oxide nanocomposites have been optimized to exhibit superior magnetic properties and to induce sufficient MR cell contrast at incubated doses as low as 1 microg iron/ml culture medium. When containing between 9 and 14 pg iron/cell, labeled cells exhibit an ex vivo nuclear magnetic resonance (NMR) relaxation rate (1/T2) as high as 24-39 s-1/mM iron. Labeled cells are unaffected in their viability and proliferating capacity, and labeled human NSCs differentiate normally into neurons. Furthermore, we show here that NSC-derived (and LacZ-transfected), magnetically labeled oligodendroglial progenitors can be readily detected in vivo at least as long as six weeks after transplantation, with an excellent correlation between the obtained MR contrast and staining for beta-galactosidase expression. The availability of magnetodendrimers opens up the possibility of MR tracking of a wide variety of (stem) cell transplants.

Molecular factors that determine Curie spin relaxation in dysprosium complexes.

Dysprosium complexes can serve as transverse relaxation (T(2)) agents for water protons through chemical exchange and the Curie spin relaxation mechanism. Using a pair of matched dysprosium(III) complexes, Dy-L1 (contains one inner-sphere water) and Dy-L2 (no inner-sphere water), it is shown that the transverse relaxation of bulk water is predominantly an inner-sphere effect. The kinetics of water exchange at Dy-L1 were determined by (17)O NMR. Proton transverse relaxation by Dy-L1 at high fields is governed primarily through a large chemical shift difference between free and bound water. Dy-L1 forms a noncovalent adduct with human serum albumin which dramatically lengthens the rotational correlation time, tau(R), causing the dipole-dipole component of the Curie spin mechanism to become significant and transverse relaxivity to increase by 3-8 times that of the unbound chelate. These findings aid in the design of new molecular species as efficient r(2) agents.

Differentiation between hemosiderin- and ferritin-bound brain iron using nuclear magnetic resonance and magnetic resonance imaging.

MRI is an optimal clinical (research) tool to provide information on brain morphology and pathology and to detect metal ions that possess intrinsic magnetic properties. Non-heme iron is abundantly present in the brain in three different forms: "low molecular weight" complexes, iron bound to "medium molecular weight complexes" metalloproteins such as transferrin, and "high molecular weight" complexes as ferritin and hemosiderin. The total amount and form of iron may differ in health and disease, and MRI can possibly quantify and monitor such changes. Ferritin-bound iron is the main storage form of iron and is present predominantly in the extrapyramidal nuclei where its amounts normally increase as a function of age. Ferritin is water soluble and shortens both, T1 and T2 relaxation, with as result a signal change on the MR images. Hemosiderin, a degradation product of ferritin, is water-insoluble with a stronger T2 shortening effect than ferritin. The larger cluster size of hemosiderin and its water-insolubility also explain a lack of significant T1-shortening effect on T1-weighted images. Using both in vitro specimens and intact brain tissue in vivo we demonstrate here that MRI may be able to distinguish between ferritin- and hemosiderin-bound iron.

Dynamic relaxometry: application to iron uptake by ferritin.

We introduce dynamic relaxometry as a novel technique for studying biochemical reactions, such as those leading to mineral formation (biomineralization). This technique was applied to follow the time course of iron oxidation and hydrolysis by the protein ferritin. Horse spleen apoferritin was loaded with single additions of 4, 10, 20, 40, and 100 ferrous ions per protein, and with multiple additions of 4, 10, 20, and 100 ferrous ions. The NMR T2 relaxation time was then measured sequentially and continuously for up to 24 h. At low loading factors of 4-10 Fe atoms/molecule, the iron is rapidly bound and oxidized by the protein on a time scale of approximately 15 s to several minutes. At intermediate loading factors (10-40), rapid initial oxidation was observed, followed by the formation of antiferromagnetic clusters. This process occurred at a much slower rate and continued for up to several hours, but was inhibited at lower pH values. At higher loading factors (40-1000), iron oxidation may occur directly on the core, and this process may continue for up to 24 h following the initial loading. Dynamic relaxometry appears to be a potentially powerful technique that may well have applications beyond the study of iron uptake by the ferritin protein.