Combinatory repair strategy to promote axon regeneration and functional recovery after chronic spinal cord injury

Abstract

Eight weeks post contusive spinal cord injury, we built a peripheral nerve graft bridge (PNG) through the cystic cavity and treated the graft/host interface with acidic fibroblast growth factor (aFGF) and chondroitinase ABC (ChABC). This combinatorial strategy remarkably enhanced integration between host astrocytes and graft Schwann cells, allowing for robust growth, especially of catecholaminergic axons, through the graft and back into the distal spinal cord. In the absence of aFGF+ChABC fewer catecholaminergic axons entered the graft, no axons exited, and Schwann cells and astrocytes failed to integrate. In sharp contrast with the acutely bridge-repaired cord, in the chronically repaired cord only low levels of serotonergic axons regenerated into the graft, with no evidence of re-entry back into the spinal cord. The failure of axons to regenerate was strongly correlated with a dramatic increase of SOCS3 expression. While regeneration was more limited overall than at acute stages, our combinatorial strategy in the chronically injured animals prevented a decline in locomotor behavior and bladder physiology outcomes associated with an invasive repair strategy. These results indicate that PNG+aFGF+ChABC treatment of the chronically contused spinal cord can provide a permissive substrate for the regeneration of certain neuronal populations that retain a growth potential over time, and lead to functional improvements.

Figure 1

Surgical intervention to bridge the gap after chronic contusive SCI. (a) Schematic diagram illustrating the repair strategy. At two months post-injury, a cystic cavity forms at the injury site (pink oval, top drawing). For repair surgery, the cavity is exposed via the dorsal cord surface, scar tissue lining the cavity is gently removed, and several autologous peripheral nerve segments (green dotted lines, bottom drawing) are longitudinally placed in the cavity to span the lesion. aFGF+ChABC (orange) is injected rostrally, caudally, and at the sides of the spinal cord adjacent to the graft. A glue made of aFGF, ChABC, and fibrin is used to stabilize the graft (blue oval, bottom drawing). (b) GFAP-stained spinal cord two months post-contusion. Note the large cavity in the epicenter of the injured site, as indicated by*.

Figure 2

TH+ nerve fibers regenerate through a PNG and re-enter the distal spinal cord following PNG+aFGF+ChABC treatment. (a–e) Images are from PNG+aFGF+ChABC-treated animals. The dotted line in each panel delineates the interface between the graft and spinal cord. (a) A representative low-magnification confocal image of a sagittal section with GFAP (green) and TH (red) staining displaying the overall anatomy of the PNG, spinal cord, and regenerating TH+ axons. Scale bar, 500 µm. (b) High-magnification confocal image of the rostral cord/PNG interface boxed in a. Large densities of TH+ axons (red) enter the PNG (GFAP-negative area) from the rostral end of the spinal cord (green). Scale bar, 100 µm. (c) High-magnification confocal image of caudal PNG/cord interface boxed in a. TH+ axons (red) grow through the PNG and cross the PNG-spinal cord interface back into the caudal spinal cord (green). Scale bar, 100 µm. (d) High-magnification confocal image showing NET+ axons (red) entering the PNG from the rostral spinal cord (green). Scale bar, 100 µm. (e) High-magnification confocal image showing NET+ axons (red) exiting the PNG back into the caudal spinal cord (green). Scale bar, 100 µm. (f–h) Images are from PNG-treated animals. The dotted line in each panel delineates the interface between the graft and spinal cord. (f) A representative low-magnification confocal image of a sagittal section with GFAP labeling (green) and TH staining (red) displaying the overall anatomy of the PNG, spinal cord, and regenerating TH+ axons. Scale bar, 500 µm. (g) High-magnification confocal image of the rostral cord/PNG interface boxed in f. Large densities of TH+ stained axons (red) enter the PNG (GFAP-negative area) from the rostral end of the spinal cord (green). Scale bar, 100 µm. (h) High-magnification confocal image of caudal PNG/cord interface boxed in f. TH+ axons (red) grow through the PNG, but fail to cross the PNG/spinal cord interface back into the caudal spinal cord (green). Scale bar, 100 µm. (i) Quantification of the density of TH-immunoreactive fibers found in the rostral spinal cord adjacent to the PNG, in the center of the PNG, and in the caudal spinal cord adjacent to the distal PNG. n = 6 per group. ***p < 0.001, Student’s t-test.

Figure 3

5-HT+ nerve fibers fail to regenerate in the chronically repaired cord. (a–d) Images are from PNG+aFGF+ChABC-treated animals. The dotted line in each panel delineates the interface between the graft and spinal cord. (a) A representative low-magnification confocal image of a sagittal section with GFAP labeling (green) and 5-HT staining (red) displaying the overall anatomy of the PNG, spinal cord, and regenerating 5-HT+ axons. Scale bar, 500 µm. (b) High-magnification confocal image of the rostral cord/PNG interface boxed in a. Note the small density of 5-HT+ stained axons (red) entering the PNG (GFAP-negative area) from the rostral end of the spinal cord (green). Scale bar, 100 µm. (c) High-magnification confocal image of the caudal PNG/cord interface boxed in a. Few 5-HT+ stained axons (red) reach the distal end of the graft, and no 5-HT axons exit the graft back into the caudal spinal cord (green). Scale bar, 100 µm. (d) Low-magnification confocal image of a sagittal section with GFAP labeling (green) and SERT staining (red) displaying the overall anatomy of the PNG, spinal cord, and lack of regenerating SERT+ axons. Scale bar, 500 µm. (e–g) Images are from PNG-treated animals. The dotted line in each panel delineates the interface between the graft and spinal cord. (e) A representative low-magnification confocal image of a sagittal section with GFAP labeling (green) and 5-HT staining (red) displaying the overall anatomy of the PNG, spinal cord, and regenerating 5-HT+ axons. Scale bar, 500 µm. (f) High-magnification confocal image of the rostral cord/PNG interface boxed in e. Note sparse 5-HT+ stained axons (red) entering the PNG (GFAP-negative area) from the rostral end of the spinal cord (green). Scale bar, 100 µm. (g) High-magnification confocal image of the caudal PNG/cord interface boxed in e. Note that no 5-HT+ stained axons (red) reach the distal end of the graft and no 5-HT+ axons exit the graft back into the caudal spinal cord (green). Scale bar, 100 µm. (h) Quantification of the density of 5-HT-immunoreactive fibers found in the rostral spinal cord adjacent to the PNG, in the center of the PNG, and in the caudal spinal cord adjacent to the distal PNG. n = 6 per group. *p < 0.05, Student’s t-test. Data represent mean ± S.E.M.

Figure 4

aFGF and ChABC promote astrocyte and Schwann cell integration at the interface of PNG and caudal spinal cord. Representative confocal images of sagittal sections showing PNG- (left column) or PNG+aFGF+ChABC- (right column) treated spinal cords 40 weeks post-injury. The top row shows low magnification of GFAP staining (white). Note the extensive migration of GFAP+ astrocytes throughout the PNG (GFAP-negative area) only following PNG+aFGF+ChABC treatment. Scale bar, 500 µm. Bottom row shows high magnification of PNG p75+ Schwann cells (red) and spinal cord GFAP+ astrocytes (green). Note the migration and integration of Schwann cells and astrocytes, including overlapping spatial domains (yellow), only following PNG+aFGF+ChABC treatment. Scale bar, 100 µm.

Figure 5

The raphe, but not the locus coeruleus, increases SOCS3 expression following chronic SCI. (a) Fluorescent image of the brainstem with TH staining illuminating the locus coeruleus. (b) Fluorescent image of the brainstem with 5-HT staining illuminating the raphe. (c) Fluorescent images of the locus coeruleus with TH (green) and SOCS3 (red) staining in naïve cord, eight weeks post SCI, or 40 weeks post SCI treated with PNG+aFGF+ChABC. (d) Fluorescent images of the raphe with 5-HT (green) and SOCS3 (red) staining in naïve cord, eight weeks post SCI, or 40 weeks post SCI treated with PNG+aFGF+ChABC. (e) Fluorescent images of the locus coeruleus with TH (green) and pS6 (red) staining in naïve cord, eight weeks post SCI, or 40 weeks post SCI treated with PNG+aFGF+ChABC. (f) Fluorescent images of the raphe with 5-HT (green) and pS6 (red) staining in naïve cord, eight weeks post SCI, or 40 weeks post SCI treated with PNG+aFGF+ChABC. Scale bar, 100 µm.

Figure 6

PNG+aFGF+ChABC increases lumbar TH+ axons around the anterior horn. (a) Representative confocal images of TH+ (top row) or 5-HT+ (bottom row) stained transverse sections in the ventral grey horn of the lumbar spinal cord 40 weeks post-injury. Scale bar, 100 µm. (b,c) Quantification of the density of TH+ (b) and 5-HT+ (c) immunoreactivity in the ventral grey horn of the lumbar spinal cord. n = 6 per group except naive n = 4. ****p < 0.0001. One-way ANOVA, Fisher’s Least Significant Difference post-hoc test. Data represent mean ± S.E.M.

Figure 7

PNG+aFGF+ChABC treatment maintains locomotor and improves micturition patterns. (a) Locomotor BBB score of rats prior to the repair surgery (Pre (8)) and following the repair surgery (weeks 9–40). n = 6 per group except at week 9 n = 5 for PNG and aFGF+ChABC. Between groups: ^&p = 0.0640, ^$p = 0.0931, ^%p = 0.0816, ^#p = 0.0794, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 Two-way repeated measures ANOVA Fisher’s Least Significant Difference post-hoc test. Within group ^δp = 0.0582, *p < 0.05, ***p < 0.001, ****p < 0.0001 One-way ANOVA, Fisher’s Least Significant Difference post-hoc test. (b) Representative smoothed metabolic cage traces 40 weeks after injury. (c,d) Metabolic cage quantification of average void frequency (c) and void volume (d). n = 6 per group. Between groups: **p < 0.01. Two-way repeated measures ANOVA, Fisher’s Least Significant Difference post-hoc test. Within group *p < 0.05 One-way ANOVA, Fisher’s Least Significant Difference post-hoc test. Data represent mean ± S.E.M.

Figure 8

PNG+aFGF+ChABC treatment improves urodynamic recovery. (a–c) Representative urodynamic recordings at 40 weeks after SCI. (a’-c’) Compressed view of bladder pressure recordings. Asterisks mark expanded void shown below. (a”-c”) Expanded time points from a’-c’. Grey box denotes the time between opening peak pressure (OPP) and closing peak pressure (CPP). Black dots denote EUS bursting. (d) Scatter plot of EUS bursting over three void cycles in relation to bladder contractions for individual animals. Grey box denotes the time interval between OPP and CPP (0–100%). (e) Percentage of animals displaying EUS bursting. (f) Average number of EUS bursts per void. (g) Percentage of EUS bursts between OPP and CPP of bladder contractions. (h) Average residual bladder volume following a void. (i) Bladder weight 40 weeks post-injury. (j) The bladder volume needed to initiate the first void. (k) The maximal pressure reached during a void. n = 6 per group. *p < 0.05, **p < 0.01. One-way ANOVA, Fisher’s Least Significant Difference post-hoc test. Data represent mean ± S.E.M.