Regeneration of Propriospinal Axons in Rat Transected Spinal Cord Injury through a Growth-Promoting Pathway Constructed by Schwann Cells Overexpressing GDNF

Abstract

Unsuccessful axonal regeneration in transected spinal cord injury (SCI) is mainly attributed to shortage of growth factors, inhibitory glial scar, and low intrinsic regenerating capacity of severely injured neurons. Previously, we constructed an axonal growth permissive pathway in a thoracic hemisected injury by transplantation of Schwann cells overexpressing glial-cell-derived neurotrophic factor (SCs-GDNF) into the lesion gap as well as the caudal cord and proved that this novel permissive bridge promoted the regeneration of descending propriospinal tract (dPST) axons across and beyond the lesion. In the current study, we subjected rats to complete thoracic (T11) spinal cord transections and examined whether these combinatorial treatments can support dPST axons' regeneration beyond the transected injury. The results indicated that GDNF significantly improved graft-host interface by promoting integration between SCs and astrocytes, especially the migration of reactive astrocyte into SCs-GDNF territory. The glial response in the caudal graft area has been significantly attenuated. The astrocytes inside the grafted area were morphologically characterized by elongated and slim process and bipolar orientation accompanied by dramatically reduced expression of glial fibrillary acidic protein. Tremendous dPST axons have been found to regenerate across the lesion and back to the caudal spinal cord which were otherwise difficult to see in control groups. The caudal synaptic connections were formed, and regenerated axons were remyelinated. The hindlimb locomotor function has been improved.

Keywords: Schwann cell; descending propriospinal axon; functional recovery; glial response; glial scar; glial-cell-derived neurotrophic factor; regeneration; spinal cord injury; spinal transection; transplantation.

Figure 1

(A) Timeline of experimental design. (B) Schematic description of a continuous axonal growth-promoting pathway. Transplantation of Schwann cells overexpressing glial-cell-line-derived neurotrophic factor inside both the lesion and caudal host cord forming an axonal growth pathway extending from the axonal cut ends to the site of innervation in the distal spinal cord.

Figure 2

Transected descending propriospinal tract (dPST) axons regenerate into and beyond T11 transection lesion site after treatment with SCs-GDNF graft in the lesion site and caudal beyond the lesion. (A) A montage image for BDA-labeled dPST axons (red), host astrocytes (GFAP, green) in a sagittal section at low magnification of a case receiving SCs-GDNF transplantation into the lesion gap at T11 and injection of the SCs-GDNF into the distal host spinal cord. BDA injection site was located 3 mm rostral to the lesion. Regions at different sites were boxed and examined at higher magnification at rostral graft–host interface (B,C), within the graft (D,E), caudal graft–host interface (F,G), and cord spinal cord proper (H,I) in the subsequent images. (B,C) At the rostral host–graft interface, regenerated dPST axons (white arrows in (C,E,G,I)) crossed the host–graft boundary (white dash line) and grew into the graft. (D,E) Within the SCs-GDNF graft, continued growth of regenerated dPST axons was found (white arrows in (E)). (F,G) At the distal graft–host interface (yellow dash line in (F,G)), regenerated axons (white arrows in (G)) penetrated into the caudal host cord. (H,I) Within the distal host spinal cord, regenerated dPST axons (white arrows in (E)) were clearly seen. (J) Percentage of regenerated axons appearing in the caudal host cord over the regenerated axon appearing in the graft epicenter at different distances from caudal graft–host boundary among three groups. *** p < 0.001, * p < 0.05, the distal SCs-GDNF group versus the distal DMEM group; ### p < 0.001, distal SCs-GFP group versus distal DMEM group. $$$ p < 0.001, $$ p < 0.01, $ p < 0.05, distal SCs-GDNF group versus distal SCs-GFP group. Scale bars: (A), 1000 µm; (B–I), 50 µm. Animal number: caudal injection of DMEM group n = 12; caudal injection of SCs-GFP group n = 14; caudal injection of SCs-GDNF group n = 16. Bar heights represent means ± SD.

Figure 3

Regenerated dPST axons formed synapse on host neurons in the distal host spinal cord. (A–C) BDA-labeled dPST axons (red) colocalize with the presynaptic marker synaptophysin (SYN) (green). (D,E,F) Higher-magnification images from dash square area in (C) showed the colocalization (yellow arrows) between bouton structure (red) and synaptophysin-stained structure (green). (G) Immuno-electron microscopy showed that a regenerated dPST presynaptic terminal, containing dark reaction product of BDA (red arrow) formed a synapse with a host dendritic spine (green arrow). Scale bars: (A–C) 50 µm; (D–F) 10 µm; (G) 500 nm.

Figure 4

SCs overexpressing GDNF promoted the myelination within the transplant. (A–C) P0-positive myelination in SC expressing GFP (SC-GFP) in control group. (D–F) Many more myelinations found in SC expressing GDNF graft. (G–I) Higher-magnification images from (D–F) showed the relationship of GFP SCs with P0 positive myelin. (J) Quantitative analysis showed P0 intensity was increased in the graft of SC-GDNF compared with in the graft of SC-GFP, ** p < 0.01; (K), Immuno-electron microscopy confirms that regenerating BDA-containing dPST axons (red arrow) were being myelinated (green arrow). Scale bar: (A–F) = 50 μm; (G–I) = 5 μm. (K) = 500 nm.

Figure 5

SCs-GDNF grafted within and caudal to a spinal transection modified astroglial responses at the caudal graft–host interface. Representative photomicrographs of GFAP expression in the caudal graft–host interface in a case that received DMEM (A–C) or a case that received SCs-GDNF (D–F) injections into the caudal host spinal cord. In the DMEM-injected case, increased expression of GFAP (B,C) was found at the distal graft–host interface. In contrast, in the SCs-GDNF-injected case, the expression of GFAP was considerably reduced (E,F). Such reduction was correlated with significant regeneration of dPST axons into the caudal host spinal cord (D,F). High magnification of boxed area in (F) was further appreciated in (G) colocalization of regenerated axons (BDA, red). (H) Quantitative analyses show that the difference in GFAP expressions in the three groups was statistically significant (***: p < 0.001). Scale bar: (A–F) 500 μm; (G) 150 μm.

Figure 6

GDNF, expressed by SCs-GFP, promoted bidirectional migration of astrocytes into the graft and SCs into the host. (A–C) Low magnification shows the bidirectional migration of astrocytes and SCs towards each other at the rostral graft–host interface. (D–F) High magnification of boxed area in (A–C) shows the intermingling of astrocyte and SCs, creating a blurring boundary. (G–I) In the absence of GDNF, grafted GFP-expressing Schwann cells (SCs-GFP, green) and host astrocytes (GFAP-IR, red) were separated by a sharp boundary. Scale bar: (A–C) 500 μm; (D–I) 50 μm.

Figure 7

GDNF-induced parallel alignment between migratory astrocytes and regenerated axons. (A–D) Low magnification images showed migration of host astrocytes (GFAP IR red) into the SCs-GDNF graft. (E–G) At high magnification, a significant amount of regenerated descending propriospinal tract (dPST) axons (F, BDA-labeled, green) were shown in close association with the host astrocytes migrated into the graft region (E, GFAP-IR, red) in the XY, XZ, and YZ planes. Scale bar: (A) 1000 μm; (B–D) 100 μm; (E–G) 50 μm.

Figure 8

Partial recovery of hindlimb motor function after axonal regeneration through and beyond a continuous SCs-GDNF growth-promoting pathway established after a complete spinal transection. (A) Improved BBB locomotor recovery was found in the group that received intraspinal transplantation and caudal injection of SCs-GDNF (green), compared with caudal injection of either SCs-GFP (blue) or DMEM (red) (** p < 0.01, * p < 0.05, SCs-GDNF vs. DMEM; ## p < 0.01, # p < 0.05, SCs-GDNF vs. SCs-GFP). (B,C) Scatter plot showing BBB score of different groups in 9th week (B) and 10th week (C) after spinal transection. (D) Correlation between number of regenerating axons across the caudal boundary and the open field locomotive score at 10th week. Bar heights represent means ± SD.