ObjectiveTo evaluate the effect of the combination of collagen scaffold and brain-derived neurotrophic factor (BDNF) on the repair of transected spinal cord injury in rats.MethodsThirty-two Sprague-Dawley rats were randomly divided into 4 groups: group A (sham operation group), T9, T10 segments of the spinal cord was only exposed; group B, 4-mm T9, T10 segments of the spinal cord were resected; group C, 4-mm T9, T10 segments of the spinal cord were resected and linear ordered collagen scaffolds (LOCS) with corresponding length was transplanted into lesion site; group D, 4-mm T9, T10 segments of the spinal cord were resected and LOCS with collagen binding domain (CBD)-BDNF was transplanted into lesion site. During 3 months after operation, Basso-Beattie-Bresnahan (BBB) locomotor score assessment was performed for each rat once a week. At 3 months after operation, electrophysiological test of motor evoked potential (MEP) was performed for rats in each group. Subsequently, retrograde tracing was performed for each rat by injection of fluorogold (FG) at the L2 spinal cord below the injury level. One week later, brains and spinal cord tissues of rats were collected. Morphological observation was performed to spinal cord tissues after dehydration. The thoracic spinal cords including lesion area were collected and sliced horizontally. Thoracic spinal cords 1 cm above lesion area and lumbar spinal cords 1 cm below lesion area were collected and sliced coronally. Coronal spinal cord tissue sections were observed by the laser confocal scanning microscope and calculated the integral absorbance (IA) value of FG-positive cells. Horizontal tissue sections of thoracic spinal cord underwent immunofluorescence staining to observe the building of transected spinal cord injury model, axonal regeneration in damaged area, and synapse formation of regenerated axons.ResultsDuring 3 months after operation, the BBB scores of groups B, C, and D were significantly lower than those of group A (P<0.05). The BBB scores of group D at 2-12 weeks after operation were significantly higher than those of groups B and C (P<0.05). Electrophysiological tests revealed that there was no MEP in group B; the latencies of MEP in groups C and D were significantly longer than that in group A (P<0.05), and in group C than in group D (P<0.05). Morphological observation of spinal cord tissues showed that the injured area of the spinal cord in group B extended to both two ends, and the lesion site was severely damaged. The morphologies of spinal cord tissues in groups C and D recovered well, and the morphology in group D was closer to normal tissue. Results of retrograde tracing showed that the gray matters of lumbar spinal cords below the lesion area in each group were filled with FG-positive cells; in thoracic spinal cords above lesion sites, theIA value of FG-positive cells in coronal section of spinal cord in group A was significantly larger than those in groups B, C, and D (P<0.05), and in groups C and D than in group B (P<0.05), but no significant difference was found between groups C and D (P>0.05). Immunofluorescence staining results of spinal cord tissue sections selected from dorsal to ventral spinal cord showed transected injured areas of spinal cords which were significantly different from normal tissues. The numbers of NF-positive axons in lesion center of group A were significantly larger than those of groups B, C, and D (P<0.05), and in groups C and D than in group B (P<0.05), and in group D than in group C (P<0.05).ConclusionThe combined therapeutic approach containing LOCS and CBD-BDNF can promote axonal regeneration and recovery of hind limb motor function after transected spinal cord injury in rats.