Neural precursors exhibit distinctly different patterns of cell migration upon transplantation during either the acute or chronic phase of EAE: a serial MR imaging study

Abstract

As the complex pathogenesis of multiple sclerosis contributes to spatiotemporal variations in the trophic micromilieu of the central nervous system, the optimal intervention period for cell-replacement therapy must be systematically defined. We applied serial, 3D high-resolution magnetic resonance imaging to transplanted neural precursor cells (NPCs) labeled with superparamagnetic iron oxide nanoparticles and 5-bromo-2-deoxyuridine, and compared the migration pattern of NPCs in acute inflamed (n = 10) versus chronic demyelinated (n = 9) brains of mice induced with experimental allergic encephalomyelitis (EAE). Serial in vivo and ex-vivo 3D magnetic resonance imaging revealed that NPCs migrated 2.5 ± 1.3 mm along the corpus callosum in acute EAE. In chronic EAE, cell migration was slightly reduced (2.3 ± 1.3 mm) and only occurred in the lateral side of transplantation. Surprisingly, in 6/10 acute EAE brains, NPCs were found to migrate in a radial pattern along RECA-1(+) cortical blood vessels, in a pattern hitherto only reported for migrating glioblastoma cells. This striking radial biodistribution pattern was not detected in either chronic EAE or disease-free control brains. In both acute and chronic EAE brain, Iba1(+) microglia/macrophage number was significantly higher in central nervous system regions containing migrating NPCs. The existence of differential NPC migration patterns is an important consideration for implementing future translational studies in multiple sclerosis patients with variable disease.

FIG. 1

Clinical progression of EAE following ICV delivery of NPCs and MR imaging of the distribution of SPIO-labeled NPCs on day 1 after ICV delivery in EAE. a: EAE was induced by subcutaneous injection of MOG_35–55 peptide and mice were monitored daily for signs of neurologic disability. SPIO-labeled NPCs were delivered ICV after either 14 days (NPC acute EAE) or 28 days (NPC chronic EAE) and mice were clinically monitored for an additional 8 days. b: On day 1 after ICV delivery in acute and chronic EAE, SPIO-labeled NPCs were detected by 3D RARE MR imaging along the ventricular system as far as 2.0 mm caudal from bregma (the site of NPC injection). NPC distribution in disease-free control mice was similar. In all mice, NPCs were present predominantly along portions of the ventricular system of the right cerebral hemisphere; ipsilateral to the site of delivery.

FIG. 2

Serial in vivo 3D RARE and ex vivo 3D FLASH MR imaging of SPIO-labeled NPCs following ICV delivery in EAE. SPIO-labeled NPCs were noninvasively imaged using 3D RARE on days 1, 4, and 7 after ICV transplantation. Representative serial in vivo 3D RARE MR image slices (250 µm thick, 148 µm in plane resolution) of SPIO-labeled NPCs following ICV transplantation at the chronic degenerative phase of EAE (a)–(c), at the peak of acute disability (f)–(h) and, in disease-free control mice (l)–(m) are shown. High-resolution ex vivo 3D FLASH MR images (day 8 post-ICV delivery) from corresponding brain regions are included to highlight the biodistribution of SPIO-labeled NPCs in chronic (d), acute (i) and control (n) brain. Regions within insets (white rectangle in d, i, and n) are shown in greater detail for chronic (e), acute (j) and control (o) brains. Radial distribution of SPIO-labeled NPCs along vascular structures in acute brain is indicated by white arrows on day 7 (h) and presented in greater detail in (j). SPIO-labeled NPCs migrate along the corpus callosum after delivery during the chronic neurodegenerative phase of EAE with signs of migration evident on day 1 (a). NPCs continue to migrate medially along the corpus callosum on day 4 (b) and day 7 (c). For the acute phase of the EAE time course (at peak of disease), radial spreading of hypointensities is present in the brain parenchyma on day 4 (g) and day 7 (h) which is absent in both chronic and control brain. Hypointense MR signal from SPIO-labeled NPCs is restricted to the ventricular system in disease-free control mice (k)–(0).

FIG. 3

Calculation of NPC migration distance within acute and chronic EAE brain. The maximum distance of NPC migration was measured along the corpus callosum beginning from the edge of the lateral ventricle to the most distant hypointense pixel. a: A representative example of a NPC migration path (indicated by a white dashed line) following transplantation during the chronic phase of EAE. b: Mean maximum NPC migration distance following transplantation in the acute phase was not significantly different from NPCs transplanted during the chronic phase of EAE (P > 0.05).

FIG. 4

Maximum extent of NPC migration as detected by serial in vivo 3D RARE and ex vivo 3D MGE MR imaging of SPIO-labeled NPCs following ICV delivery in EAE. SPIO-labeled NPCs were noninvasively imaged on s 1, 4, and 7 after ICV transplantation. Serial in vivo 3D RARE MR image slices (250 µm thick, 148 µm in plane resolution) representative of the peak extent of NPC migration following ICV transplantation at the chronic degenerative phase (a)–(c) and at the peak of acute disability of EAE (f)–(h) are shown. In chronic EAE, NPCs migrated robustly within rostral regions of the corpus callosum whereas, in acute EAE, extensive NPC migration was frequently evident within caudal levels of the brain. For the acute phase of the EAE time course (at peak of disease), radial hypointensities are present in the brain parenchyma on day 1 (f), day 4 (g) and day 7 (h). In chronic EAE, NPCs were detected within the rostral corpus callosum on day 1 (a) and these cells continued to migrate laterally along the corpus callosum on day 4 (b) and day 7 (c). High-resolution ex vivo 3D MGE MR images (day 21 post-ICV delivery) from corresponding brain regions are included to highlight the biodistribution of SPIO-labeled NPCs in chronic (d) and acute (i) brain. Regions within insets (white rectangle in d, i) are shown in greater detail for chronic (e) and acute (j) brains. Radial distribution of SPIO-labeled NPCs along vascular structures in acute brain is indicated by white arrows (j).

FIG. 5

Statistical difference mapping of the in vivo distribution of SPIO-labeled NPCs 1 week after ICV delivery during the acute (b,d,f, and h, n = 10) and chronic (c,e,g, and i, n = 9) phase of EAE. After anatomical MR image registration and signal intensity normalization, average pixel intensity values were calculated for each grayscale MR image slice in each experimental group. Coronal image slices of the average grayscale pixel intensity values for disease free control (a) acute (b) and chronic (c) EAE brains are shown for the region of the mouse CNS exhibiting the highest cell migratory activity. After subtraction of the average grayscale pixel intensity from corresponding coronal MR images of ICV-transplanted disease-free control brains (n = 4), the probability that residual pixels are detected 1 week after ICV delivery of NPCs during acute (d) or chronic (e) EAE phase was computed. Probability values ranging from 100% to a 70% cutoff above chance (e.g., 50%) were assigned to a pseudo-color scale and pixels within the MR image were colored accordingly. For example, green pixels indicate that a hypointense voxel was specifically identified in 90% of diseased brains (e.g., present in acute or chronic and absent in disease-free brain). The process of average pixel subtraction was repeated to generate sagittal probability maps of hypointensity distribution within a 4 mm rostro-caudal expanse of the CNS (32 contiguous 250 µm slices) on day 7 after ICV delivery in acute EAE (f and h) and chronic EAE (g and i). Hypointensities were detected in the cortex with high probability following ICV delivery of NPCs in acute EAE (f, arrow) but not in chronic EAE (g). In a sagittal MR image slice near the midline of the CNS, NPCs were detected along the corpus callosum with high probability in chronic EAE (i, arrow) but not in acute EAE (H).

FIG. 6

DAB-enhanced Prussian blue histological detection of SPIO-labeled NPCs. In vivo 3D RARE (a) and ex vivo 3D MGE (b) MR imaging of chronic EAE brain demonstrates spreading of hypointense signal along the corpus callosum caudal to the ICV injection site. DAB-enhanced Prussian blue staining of the inset (white rectangle) indicated in A and B confirmed the presence of SPIO-labeled cells migrating along the corpus callosum (c). In acute EAE brain, in vivo 3D RARE (d) and ex vivo 3D MGE (e) MR imaging revealed extensive radial spreading of hypointense signal within the brain parenchyma; particularly in regions of mouse somatosensory cortex. The distribution of hypointensities detected by in vivo and ex vivo MR imaging (white rectangle in d and e) correlated highly with the distribution of SPIO-labeled cells along structures resembling cortical blood vessels as detected histologically using Perl’s stain (f). Scale bar = 500 µm in c and 250 µm in f.

FIG. 7

NPCs distribute along RECA-1⁺ blood vessels following ICV delivery during the acute phase of EAE. DAB-enhanced Prussian blue histochemistry for iron oxide reveals numerous NPCs distributed along structures resembling cortical blood vessels in acute EAE brain (a). NPCs are found in the cortex associated with blood vessels surrounded by inflammatory infiltrates, as shown by hematoxylin and eosin (purple) and DAB-enhanced Prussian blue (brown) staining (b). SPIO-labeled NPCs were detected along the periphery of a vascular structure (c). Immunohistochemistry for the vascular endothelial cell marker RECA-1 (green) and BrdU⁺ NPCs (red) confirmed the presence of transplanted NPCs in close proximity to cortical blood vessels in mice that received SPIO-labeled NPCs at the peak (acute EAE phase) period of clinical disability (d). Scale bar = 100 µm in panels a and b and 50 µm in panels c and d.

FIG. 8

NPC migration occurs within regions of Iba1⁺ macrophage activity. a: Iba1⁺ macrophages (red) were detected in close proximity to migratory EGFP⁺ NPCs within regions of the corpus callosum adjacent to lateral ventricular space (LV). b: The number of Iba1⁺ cells was counted in regions of the corpus callosum that contained EGFP⁺ migratory NPCs and rostral or caudal regions of the corpus callosum from the same animal where NPCs were retained within the lateral ventricle. Scale bar = 100 µm.