Lipophilic dye-compatible brain clearing technique allowing correlative magnetic resonance/high-resolution fluorescence imaging in rat models of glioblastoma.

In this work we optimized a novel approach for combining in vivo MRI and ex vivo high-resolution fluorescence microscopy that involves: (i) a method for slicing rat brain tissue into sections with the same thickness and spatial orientation as in in vivo MRI, to better correlate in vivo MRI analyses with ex-vivo imaging via scanning confocal microscope and (ii) an improved clearing protocol compatible with lipophilic dyes that highlight the neurovascular network, to obtain high tissue transparency while preserving tissue staining and morphology with no significant tissue shrinkage or expansion. We applied this methodology in two rat models of glioblastoma (GBM; U87 human glioma cells and patient-derived human glioblastoma cancer stem cells) to demonstrate how vital the information retrieved from the correlation between MRI and confocal images is and to highlight how the increased invasiveness of xenografts derived from cancer stem cells may not be clearly detected by standard in vivo MRI approaches. The protocol studied in this work could be implemented in pre-clinical GBM research to further the development and validation of more predictive and translatable MR imaging protocols that can be used as critical diagnostic and prognostic tools. The development of this protocol is part of the quest for more efficacious treatment approaches for this devastating and still uncurable disease. In particular, this approach could be instrumental in validating novel MRI-based techniques to assess cellular infiltration beyond the macroscopic tumor margins and to quantify neo-angiogenesis.

A novel neuroferritinopathy mouse model (FTL 498InsTC) shows progressive brain iron dysregulation, morphological signs of early neurodegeneration and motor coordination deficits.

Neuroferritinopathy is a rare genetic disease with a dominant autosomal transmission caused by mutations of the ferritin light chain gene (FTL). It belongs to Neurodegeneration with Brain Iron Accumulation, a group of disorders where iron dysregulation is tightly associated with neurodegeneration. We studied the 498-499InsTC mutation which causes the substitution of the last 9 amino acids and an elongation of extra 16 amino acids at the C-terminus of L-ferritin peptide. An analysis with cyclic voltammetry on the purified protein showed that this structural modification severely reduces the ability of the protein to store iron. In order to analyze the impact of the mutation in vivo, we generated mouse models for the some pathogenic human FTL gene in FVB and C57BL/6J strains. Transgenic mice in the FVB background showed high accumulation of the mutated ferritin in brain where it correlated with increased iron deposition with age, as scored by magnetic resonance imaging. Notably, the accumulation of iron-ferritin bodies was accompanied by signs of oxidative damage. In the C57BL/6 background, both the expression of the mutant ferritin and the iron levels were lower than in the FVB strain. Nevertheless, also these mice showed oxidative alterations in the brain. Furthermore, post-natal hippocampal neurons obtained from these mice experienced a marked increased cell death in response to chronic iron overload and/or acute oxidative stress, in comparison to wild-type neurons. Ultrastructural analyses revealed an accumulation of lipofuscin granules associated with iron deposits, particularly enriched in the cerebellum and striatum of our transgenic mice. Finally, experimental subjects were tested throughout development and aging at 2-, 8- and 18-months for behavioral phenotype. Rotarod test revealed a progressive impaired motor coordination building up with age, FTL mutant old mice showing a shorter latency to fall from the apparatus, according to higher accumulation of iron aggregates in the striatum. Our data show that our 498-499InsTC mouse models recapitulate early pathological and clinical traits of the human neuroferritinopathy, thus providing a valuable model for the study of the disease. Finally, we propose a mechanistic model of lipofuscine formation that can account for the etiopathogenesis of human neuroferritinopathy.

Cellular magnetic resonance with iron oxide nanoparticles: long-term persistence of SPIO signal in the CNS after transplanted cell death.

AIM: To study the specificity of cellular MRI based on superparamagnetic iron oxide particles (SPIOs), especially within the CNS. MATERIALS & METHODS: A microglial cell line was engineered for the expression of a suicide gene, the receptor of diphtheria toxin (DT), and two reporter genes, green fluorescent protein and luciferase, in order to induce, in a controlled manner, cell death and test it through bioluminescence. SPIO-labeled DT-sensitive and control DT-insensitive cells were transplanted into the brains of mice, which underwent serial MRI and bioluminescence studies before and up to 90 days after DT-induced cell death. RESULTS: No variations in SPIO signal voids were detected along longitudinal monitoring in brain hemispheres transplanted with DT-sensitive cells. Ex vivo analyses showed persistence of iron nanoparticle deposits at transplantation sites. CONCLUSION: Due to the long-term persistence of signal after transplanted cell death, caution is advised when SPIOs are employed for cell tracking.

Characterization of biological features of a rat F98 GBM model: a PET-MRI study with [18F]FAZA and [18F]FDG.

INTRODUCTION: The prognosis of malignant gliomas remains largely unsatisfactory for the intrinsic characteristics of the pathology and for the delayed diagnosis. Multimodal imaging based on PET and MRI may assess the dynamics of disease onset and progression allowing the validation of preclinical models of glioblastoma multiforme (GBM). The aim of this study was the characterization of a syngeneic rat model of GBM using combined in vivo imaging and immunohistochemistry. METHODS: Four groups of Fischer rats were implanted in a subcortical region with increasing concentration of rat glioma F98 cells and weekly monitored with Gd-MR, [(18)F]FDG- and [(18)F]FAZA-PET starting one week after surgery. Different targets were evaluated on post mortem brain specimens using immunohistochemistry: VEGF, GFAP, HIF-1α, Ki-67 and nestin. RESULTS: Imaging results indicated that tumor onset but not progression was related to the number of F98 cells. Hypoxic regions identified with [(18)F]FAZA and high-glucose metabolism regions recognized with [(18)F]FDG were located respectively in the core and in external areas of the tumor, with partial overlap and remodeling during disease progression. Histological and immunohistochemical analysis confirmed PET/MRI results and revealed that our model resumes biological characteristics of human GBM. IHC and PET studies showed that necrotic regions, defined on the basis of [(18)F]FDG uptake reduction, may include hypoxic clusters of vital tumor tissue identified with [(18)F]FAZA. This last information is particularly relevant for the identification of the target volume during image-guided radiotherapy. CONCLUSIONS: In conclusion, the combined use of PET and MRI allows in vivo monitoring of the biological modification of F98 lesions during tumor progression.