Human neural stem cells differentiate and promote locomotor recovery in an early chronic spinal cord injury NOD-scid mouse model

Abstract

Background: Traumatic spinal cord injury (SCI) results in partial or complete paralysis and is characterized by a loss of neurons and oligodendrocytes, axonal injury, and demyelination/dysmyelination of spared axons. Approximately 1,250,000 individuals have chronic SCI in the U.S.; therefore treatment in the chronic stages is highly clinically relevant. Human neural stem cells (hCNS-SCns) were prospectively isolated based on fluorescence-activated cell sorting for a CD133(+) and CD24(-/lo) population from fetal brain, grown as neurospheres, and lineage restricted to generate neurons, oligodendrocytes and astrocytes. hCNS-SCns have recently been transplanted sub-acutely following spinal cord injury and found to promote improved locomotor recovery. We tested the ability of hCNS-SCns transplanted 30 days post SCI to survive, differentiate, migrate, and promote improved locomotor recovery.

Methods and findings: hCNS-SCns were transplanted into immunodeficient NOD-scid mice 30 days post spinal cord contusion injury. hCNS-SCns transplanted mice demonstrated significantly improved locomotor recovery compared to vehicle controls using open field locomotor testing and CatWalk gait analysis. Transplanted hCNS-SCns exhibited long-term engraftment, migration, limited proliferation, and differentiation predominantly to oligodendrocytes and neurons. Astrocytic differentiation was rare and mice did not exhibit mechanical allodynia. Furthermore, differentiated hCNS-SCns integrated with the host as demonstrated by co-localization of human cytoplasm with discrete staining for the paranodal marker contactin-associated protein.

Conclusions: The results suggest that hCNS-SCns are capable of surviving, differentiating, and promoting improved locomotor recovery when transplanted into an early chronic injury microenvironment. These data suggest that hCNS-SCns transplantation has efficacy in an early chronic SCI setting and thus expands the "window of opportunity" for intervention.

Figure 1. hCNS-SCns promote improved locomotor recovery on multiple tests.

(A) BMS locomotor performance is significantly improved in hCNS-SCns treated animals compared to vehicle controls (repeated measures ANOVA (p≤0.0022). A Bonferroni/Dunn post-hoc analysis at week 16 revealed a significant difference between hCNS-SCns and vehicle control (p≤0.02). There were no significant differences between hFbs and either hCNS-SCns or vehicle. (B) Recovery of coordination was significantly increased in hCNS-SCns treated animals compared to vehicle controls using chi square analysis for observed frequency (p≤0.047, Fisher's Exact Test). No statistically significant differences were found comparing hFbs with vehicle or hCNS-SCns transplanted animals. Error bars are not plotted as these bars represent the absolute percentage of animals reaching criteria. (C) CatWalk gait analysis showed that hCNS-SCns treated animals exhibited significantly increased swing speed compared to vehicle treated animals (p≤0.04, ANOVA, Fisher's PLSD). (D) von Frey analysis of mechanical allodynia showed no significant differences between any of the groups (p>0.05 ANOVA).

Figure 2. Transplanted hFbs survive and hCNS-SCns survive and migrate extensively.

Histology of spinal cords stained for human cytoplasm (SC121, brown and methyl green nuclear counterstain, blue) 16 weeks post-transplantation (A) reveal no human cells in vehicle treated animals, (B) few human cells in hFbs treated animals, and (C) many human cells surviving and migrating the length of the cord in hCNS-SCns treated animals. hCNS-SCns exhibit distinct morphologies in white vs. gray matter. (D) In white matter hCNS-SCns exhibit oligodendrocyte-like morphology. (E) In gray matter hCNS-SCns exhibit neuronal morphology Scale bar on left: 1000 µm. Scale bar on right: 50 µm. (F) 50,000 hFbs were initially transplanted (red line) and 11,701±3070 remain at the termination of the study. 75,000 hCNS-SCns were initially transplanted (blue line) and 215,711±48,978 were present at the termination of the study as estimated using unbiased stereological quantification. (G) Unbiased stereological quantification was used to assess the migration of transplanted cells in 1 mm blocks extending 8 mm rostral and caudal from the injury site including the spared tissue (s.t., vertical dashed line) surrounding the lesion. hCNS-SCns migrated up to 8 mm rostral and 5 mm caudal (left y-axis, blue). hFbs migrated 2 mm rostral and 2 mm caudal (right y-axis, red).

Figure 3. hCNS-SCns express Ki67 and nestin 16 weeks following transplantation.

(A–E) Human cytoplasm-positive cells, (SC121), red (B) were rarely associated with the cell cycle marker Ki67, green (C), DAPI counterstain, blue (A). Arrows indicate a double-labeled cell. Merged confocal image, showing rare hCNS-SCns expression of Ki67 indicating low mitotic activity (D). Orthogonal view of confocal image showing co-localization of Ki67 and SC121 (E). (F–J) Some SC121 positive human cells, red (G) expressed the immature neural marker nestin, green (H) DAPI counterstain, blue (F). Arrows indicate a double-labeled cell. Merged confocal image reveals many hCNS-SCns maintain immature phenotypes and nestin expression 16 weeks after transplantation (I). Orthogonal view of confocal image revealing co-localization of nestin and SC121 (J). Scale bars  = 20 µm and 10 µm in the bottom row.

Figure 4. hCNS-SCns mostly differentiate into oligodendrocytes and neurons, and few astrocytes.

(A–E) Several human nuclei positive cells, green (C), expressed Olig2 marker revealing immature oligodendrocytes, red (B), DAPI counterstain, blue (A). Arrows indicate a double-labeled cell. Merged confocal image, showing hCNS-SCns expression of Olig2 indicating differentiation to oligodendrocytes (D). Orthogonal view of confocal image showing co-localization of Olig2 and SC101 (E). (F–J) Some human nuclei-positive cells, SC101 green (H) also express the mature oligodendrocyte marker APC-CC1, red (G), DAPI counterstain, blue (F) Arrows indicate a double-labeled cell. Merged confocal image reveals some hCNS-SCns express APC-CC1 expression 16 weeks after transplantation. (I). Orthogonal view of confocal image revealing co-localization of APC-CC1 and human nuclei marker (J). (K–O) Human cytoplasm-positive cells SC121, green (M) also exhibit ß-tubulin III expression, red (L). DAPI counterstain, blue (K). Arrows indicate a double-labeled cell. Merged confocal image, revealing hCNS-SCns expression of ß-tubulin III (N). Orthogonal view of confocal image showing co-localization of ß-tubulin III and SC121 (O). (P–T) Few human cytoplasm cells SC121, red (Q) also expressed and the astrocyte marker GFAP, green (R). DAPI counterstain, blue (P) Arrowhead indicates a non-astrocytic human cell. Co-localization was rare indicating few hCNS-SCns exhibited astrocytic differentiation 16 weeks after transplantation (R). Orthogonal view of confocal image revealing co-localization of GFAP and SC121 (S). Scale bars  = 20 µm and 10 µm in the bottom row.

Figure 5. hCNS-SCns differentiation/fate quantification 16 weeks post-transplantation.

Bar graph revealing quantification of hCNS-SCns that expressed the proliferative marker Ki67, the immature neural marker nestin, immature oligodendrocyte marker Olig2, the mature oligodendrocyte marker APC-CC1, the neuronal marker ß-tubulin III and the astrocytic marker GFAP expressed as percentages.

Figure 6. Human cytoplasm co-localizes with paranodal protein CASPR.

(A) Orthogonal view of a confocal image of SC121 (red), CASPR (green) and DAPI counterstain (blue). The crosshair reveals co-localization of CASPR with SC121 and orthogonal projection. Arrows indicate additional SC121-positive axons exhibiting compact CASPR-positive paranodes. Arrowheads indicate diffusely distributed CASPR. (B–E) High-power images revealing examples of CASPR and SC121 co-localization. (B) High-power view of area in crosshair from (A). The two discrete CASPR-positive areas are ∼4 µm apart suggesting they are two paranodal regions of a single node. (C) High-power view of another co-localized axon revealing two discrete paranodal regions of a single node. (D, E) Additional high-power examples of SC121 co-localized with CASPR. Left scale bar  = 20 µm, right scale bars  = 1 µm.

Figure 7. hCNS-SCns transplantation does not alter lesion volume, spared tissue volume, or glial scar area.

(A) Representative spinal cord stained for GFAP to stereologically quantify lesion volume, indicated by the blue outline, spared tissue volume, quantified 500 µm rostrally and caudally from the lesion edges excluding the lesion, as depicted by the gray box, and the area of dense GFAP expression indicative of glial scarring excluding the lesion, outlined in black. (B) Staining of human GFAP (SC123), indicating rare astrocytic differentiation localized primarily near the injury site that did not exacerbate the glial scar. (C) Lesion volumes quantified using unbiased stereological probe Cavalieri Estimator show no significant difference for any of the three groups (p≥0.91 ANOVA). (D) Spared tissue volumes quantified using unbiased stereological probe Cavalieri Estimator show no significant difference for any of the three groups (p≥0.21 ANOVA). (E) Glial scar areas quantified using unbiased stereological probe Cavalieri Estimator show no significant difference for any of the three groups (p≥0.98 ANOVA). Scale bar  = 1000 µm.