Axonal remyelination by cord blood stem cells after spinal cord injury

Abstract

Human umbilical cord blood stem cells (hUCB) hold great promise for therapeutic repair after spinal cord injury (SCI). Here, we present our preliminary investigations on axonal remyelination of injured spinal cord by transplanted hUCB. Adult male rats were subjected to moderate SCI using NYU Impactor, and hUCB were grafted into the site of injury one week after SCI. Immunohistochemical data provides evidence of differentiation of hUCB into several neural phenotypes including neurons, oligodendrocytes and astrocytes. Ultrastructural analysis of axons reveals that hUCB form morphologically normal appearing myelin sheaths around axons in the injured areas of spinal cord. Colocalization studies prove that oligodendrocytes derived from hUCB secrete neurotrophic hormones neurotrophin-3 (NT3) and brain-derived neurotrophic factor (BDNF). Cord blood stem cells aid in the synthesis of myelin basic protein (MBP) and proteolipid protein (PLP) of myelin in the injured areas, thereby facilitating the process of remyelination. Elevated levels of mRNA expression were observed for NT3, BDNF, MBP and PLP in hUCB-treated rats as revealed by fluorescent in situ hybridization (FISH) analysis. Recovery of hind limb locomotor function was also significantly enhanced in the hUCB-treated rats based on Basso-Beattie-Bresnahan (BBB) scores assessed 14 days after transplantation. These findings demonstrate that hUCB, when transplanted into the spinal cord 7 days after weight-drop injury, survive for at least 2 weeks, differentiate into oligodendrocytes and neurons, and enable improved locomotor function. Therefore, hUCB facilitate functional recovery after moderate SCI and may prove to be a useful therapeutic strategy to repair the injured spinal cord.

Figure 1. Transdifferentiation of hUCB into neural phenotypes in vitro

RA-treated CD44⁺ cells were fixed, incubated with primary antibodies directed against NF-200 (A), GFAP (B) or APC (C), followed by incubation with fluorescein-conjugated or Texas-red conjugated secondary antibodies. After immunostaining, cells were counterstained with DAPI. Neural cell markers and phase-contrast images were merged. Extreme left panel shows bright field images and extreme right panel shows nuclei of the cells stained with DAPI. Scale bar = 50μm. (A) Cells expressing NF-200 and displaying neuron like morphology with long axonal projections. (B) Cells immunostained with the anti-GFAP antibody. Some of the cells are round and relatively small, whereas others contain long projections with immunoreactive filamentous structures that are visible in the cytoplasm. (C) APC-immunoreactive cells displaying morphology characteristic of oligodendrocytes, with flat cell body and short or long branched projections. Smaller, round immunoreactive cells are also occasionally present. All these cells exhibit CD44 markers specific for hUCB. A subpopulation of hUCB-derived cells growing in a monolayer before clone formation was found to be negative for all investigated antigens. The results are expressed as the mean ± SE of cell number from nine independent cultures (three parallel experiments from three separate cord blood preparations).

Figure 2. Survival and differentiation of hUCB in rat spinal cords

Differentiation of hUCB in injured spinal cords showing specific antigens: CD44 (hUCB marker) colocalized with (A) NF-200 (a neurofilament protein), (B) APC (oligodendrocyte marker) and (C) GFAP (an astrocyte marker). The differentiation of hUCB in vivo was observed after intraspinal grafting into injured spinal cords 7 days post-SCI. NF-200-, GFAP-, and APC-positive cells occurred in the vicinity of the injury site. Scale Bar = 100 μm. Results are from three independent sections 2 mm caudal from the injury epicenter (n ≥ 3).

Figure 3. Remyelination due to intraspinal grafting of hUCB

Section from injured rats showing loss of oligodendrocytes around the injury epicenter (A). The section is stained with Texas-Red conjugated APC antibody. (B) Greater preservation of oligodendrocytes in hUCB-treated sections. Merged image of section immunostained with Texas-Red conjugated APC antibody and FITC-conjugated CD44 antibody. For both A and B inset shows DAPI images. Transmission electron micrographs showing deformation of myelin sheath and axons in contused spinal cords (↑) (C & D). Myelin is thin and fragmented in many axons (*). In contrast, hUCB-treated sections showing normal myelin with several layers (E) (↑) indicates demyelinated axons undergoing remyelination. Scanning electron micrographs showing ruptured myelin (↑) in injured (F) and smooth myelin sheath (↑) in hUCB-treated (G) spinal cords. Magnification shown at 35000X for TEM and 15000X for SEM. (n ≥ 2).

Figure 4. Stem cell-mediated secretion of neurotrophic factors and synthesis of myelin proteins in treated rats

Immunohistochemical comparison of uninjured sham control, injured and hUCB-treated spinal cord sections was performed to analyze the secretion of neurotrophic hormones (NT3 and BDNF) and synthesis of myelin proteins (MBP and PLP). Paraffin sections from spinal cord blocks adjacent to the epicenter were probed with respective antibodies using DAB immunohistochemistry and counterstained with hematoxylin to stain the live nuclei and then photographed using bright-field microscope. Arrows indicate the stained portions with respective antibodies. Note the presence of demyelinated axons in the injured sections are indicated by *. Scale Bar = 200 μm. Results are from three independent sections caudal from the injury epicenter (n ≥ 3).

Figure 5. Confocal scanning microscope images demonstrate the secretion of neurotrophic hormones in spinal cords of rats

Cryosections from spinal cord blocks adjacent to the epicenter were processed for immunoflourescence studies as described in Materials and Methods. Sections were immunostained with FITC-conjugated CD44 and Texas-red conjugated APC antibodies. Further, they were DAB-stained with NT3 and BDNF antibodies. All the sections were stained with DAPI for showing nuclear localization. In hUCB-treated sections, remyelination was established by co-localization of APC with NT3 and BDNF (A & B). Arrows indicate NT3 and BDNF secreting hUCB-derived oligodendrocytes. Scale Bar = 200 μm. Quantitative estimation from sham control, injured and hUCB treated sections for NT3(C) and BDNF (E). Figs. D and F show quantitation of hUCB-derived oligodendrocytes secreting NT3 and BDNF respectively. Results are from three independent sections caudal from the injury epicenter (n ≥ 3). (Error bars indicate SEM. * Significant at p <0.05).

Figure 6. Confocal scanning microscope images demonstrate the synthesis of myelin proteins in spinal cords of rats

Cryosections from spinal cord blocks adjacent to the epicenter were processed for immunoflourescence studies as described in Methods. Sections were immunostained with FITC-conjugated CD44 and Texas-red conjugated APC antibodies. Further, they were DAB-stained with MBP and PLP antibodies. All the sections were stained with DAPI for showing nuclear localization. In hUCB-treated sections, remyelination was established by co-localization of APC with MBP and PLP (A & B). Arrows indicate MBP and PLP synthesizing hUCB-derived oligodendrocytes. Scale Bar = 200 μm. Quantitative estimation from sham control, injured and hUCB treated sections for MBP(C) and PLP (E). Figs. D and F show quantitation of hUCB-derived oligodendrocytes synthesizing MBP and PLP respectively. Results are from three independent sections caudal from the injury epicenter (n ≥ 3). (Error bars indicate SEM. * Significant at p <0.05).

Figure 7. Pattern of mRNA expression of neurotrophic factors and myelin proteins in spinal cords of rats

FISH analysis of neurotrophic factors and myelin proteins depicting sham control, injured, and hUCB-treated samples in the dorsal white matter region. Sequential serial sections hybridized with FITC-conjugated oligonucleotide antisense probes for NT3, BDNF, MBP and PLP were photographed using confocal microscope as described in Methods section. hUCB treated sections show colocalization (yellow) with Texas-red conjugated CD44 antibody, specific for hUCB. Inset shows representative Hoechst -33342 stained images. Scale Bar = 100 μm. Results are from three independent sections caudal from the injury epicenter (n ≥ 3).

Figure 8. Expression of neurotrophic factors and myelin proteins in injured and treated spinal cords of rats

RT-PCR analysis of neurotrophic factors and myelin proteins depicting control, sham control, injured, and hUCB-treated samples (A). House keeping gene GAPDH was used as loading control. Quantitative data showing pixel density of RT-PCR bands (n ≥ 2). (B). Western blot analysis of changes in the levels of neurotrophic factors and myelin proteins following spinal cord injury and hUCB treatment (C) and their corresponding quantitative analysis of bands using Image Pro software (D). There was no significant difference between control and sham controls. GAPDH was used as loading control. This figure shows representative gels and blots obtained in one experiment that was repeated three times with similar results (n ≥ 3). (Error bars indicate SEM. * Significant at p <0.05).

Figure 9. Hind limb functional recovery after spinal cord contusion

The BBB locomotor rating scale showed functional recovery of all hUCB-treated animals until day 21 post-SCI. The score indicating a gait characterized by no hind limb weight bearing and no coordinated hind limb movement is 0, whereas the score indicating a gait characterized by partial hind limb weight bearing and partial hind limb coordination is 13. The locomotor recovery scores averaged across hind limbs for weekly testing of rats with moderate contusion using NYU impactor (A). Each point represents the highest locomotor score achieved each day. Error bars indicate SEM ((n ≥ 5 per group). Video images of hind limb movements in rats (B). Spinal cord injury with NYU impactor resulted in paraplegia of hind limbs of rats when compared to normal rats. The hind limbs are internally rotated and the tail is not supporting the body weight. In hUCB-treated rats, hind limbs are recovered from the injury and are externally rotated. Pre = pre-operative; PO = post-operative. ↓ indicates transplantation point. (Error bars indicate SEM. * Significant at p <0.05).