Antisense Vimentin CDNA Combined With ChABC Reduces Glial Scar Formation And Promotes Axon Regeneration After Spinal Cord Injury In Rats

Background and Objectives:The exact mechanism for a lack of spontaneous recovery following spinal cord injury (SCI) remains unclear. Reactive gliosis occurs after central nervous system (CNS) injury, and results in the formation of glial scar and cystic cavities. Chondroitin sulphate proteoglycans (CSPGs), the extracellular matrix (ECM) components of glial scar, are regarded as important inhibitory molecules for axon regeneration. CSPGs consist of a core protein and a variable number of glycosaminoglycan (GAG) side chains. Chondroitinase ABC (ChABC), an enzyme that catalyzes hydrolysis of GAG chains, helps to improve motor, sensory and autonomic functions following spinal cord lesions. However, CNS regeneration is still an issue to be solved.As a basic cytoskeletal protein, intermediate filaments (IF) underpin cellular architecture and take part in cell differentiation and proliferation. The major hallmark of glial scar formation is the accumulation of hypertrophic astrocytes with increased expression of IF glial fibrillary acidic protein (GFAP) and embryonic IF vimentin. Vimentin, as a cytoskeleton-associated protein, seems necessary for the formation of GFAP networks and reactive gliosis. Infection with retroviruses carrying full-length vimentin antisense cDNA may reduce vimentin and GFAP expression in reactive astrocytes in vitro and in vivo.Glial scar is, in essence, the response of glial cells, especially astrocytes, to an injury. Reactive astrocytes profusely reexpress vimentin, and persistently secrete CSPGs after injury. The meshwork of tightly interwoven astrocytic processes and the inhibitory CSPGs molecules in ECM form a mechanical and biochemical barrier for axon regrowth. Both CSPGs and CSPGs-producing cells are the key elements for glial scar formation. Treatment targeting at both intra- and extra-cellular components may modify the physical structure and biochemical composition of glial scar, thus establishing a favorable microenvironment for axon regeneration and regrowth of residual fibres. In this study, we administered both full-length antisense vimentin cDNA and ChABC to the hemisected rat spinal cord to investigate the effect of the combined therapy on the formation of glial scar and cystic cavity, and axon regeneration.Materials and MethodsPartⅠ:Recombinant retroviruses carrying antisense vimentin cDNA were previously constructed, and adult female Sprague- Dawley rats weighing 250-300 g were used. After anesthetization, lateral hemisection was made at T9-T10 level by creating a 2-mm-long longitudinal incision along the midline of the cord with a scalpel under a dissecting microscope, followed by lateral incision at the rostral and caudal ends and removal of the tissues by aspiration. A catheter was fed rostrally to lay immediately caudal to the laminectomy, with one end externalized behind the neck for access during dosing. ChABC (10 U/ml; Sigma, St. Louis, MO) was used to degrade CSPGs, and retroviruses containing antisense vimentin cDNA were concentrated 1000-fold to inhibit vimentin expression. The dose used for either single or combined treatments was 6μl. Rats of the control group were administered normal saline or retroviruses without carrying vimentin cDNA. Under inhalation anesthesia, the animal was slowly injected with the agents every other day for 2 weeks (totally eight times) via the catheter using a mechanical syringe driver over 10 min.The spinal tissue 3 mm rostral and caudal to the lesion (totally 8 mm long) or spinal tissue at the equivalent site in normal rats was obtained surgically 2 weeks after injury. Total RNA or protein was obtained, GFAP, vimentin, and CSPGs expression were examined using RT-PCR and Western blot. The spinal cords were also examined by immunohistochemistry, and change of glial scar and cystic cavity after combined treatments was evaluated.PartⅡ:Dorsal spinal cord hemisection model was made by hemisecting spinal cord at T9-T10 level in rats. After treatment with antisense vimentin cDNA and ChABC, axon regeneration of wound rats were examined by anterograde tracing with BDA and retrograde tracing with HRP, then hind limb placing response and Basso Beattie and Bresnahan locomotor rating score (BBB) of rats were also determined.Results1. Vimentin mRNA and protein expressed at the surface of white matter, some special glial cells, and the ependymal cells of the central cannal in normal spinal cord. GFAP mRNA and protein expressed in both white matter and gray matter of normal rats, especially at the surface of white matter and gray matter around the central canal. NG2 (one of the most important CSPGs molecule) was not detected by RT-PCR or Western blotting with 2B6 monoclone antibody (an antibody which recognize the core protein of CSPG, but not combine with the whole CSPG molecule), while detection with CS-56 monoclone antibody by immunohistochemistry or immunoflurescense showed positive staining at the surface of white matter and around the neurons of anterior horn, especially at the pia matter of the normal spinal cord.2. Spinal cord became slimmed 2 weeks after SCI, with up-regulation of vimentin, GFAP and NG2 mRNA. Vimentin and GFAP profusely expressed near the injury site, and amounts of vimentin and GFAP positive astrocytes accumulated and proliferated at the wound site, with hypertrophied cell bodies and interwoved prosesses. CSPGs highly expressed at the injury epicenter, distributed more focused than vimentin and GFAP. The cell bodies and processes of vimentin and GFAP positive astrocytes interwoved with each other and constituted the dense glial scar at injury site, with highly expressed IF proteins 8 weeks following SCI. Vimentin expression were more focused at the scar site, compared with GFAP, which indicates that vimentin may be a better biochemical marker. CSPGs expression was down-regulated 8 weeks following SCI than 2 weeks, and focused on the region of glial scar.3. Vimentin and GFAP mRNA and protein expression in antisense vimentin group and combined treatment group were decreased 2 weeks following SCI, compared with the normal saline group, while the ChABC group and the vimentin control group showed no difference. The NG2 mRNA expression showed no difference in each injury group, while CS-56 expression decreased in ChABC group and combined treatment group, compared with the normal saline group, and showed no difference in antisense viemntin group and vimentin control group. Western blotting showed positive 2B6 antibody (an antibody which recognize the core protein of CSPG, but not combine with the whole CSPG molecule) immunostaining in ChABC group and combined group with no difference, and negative staining in normal saline group, antisense vimentin group and antisense vimentin control group. While after digestion with ChABC in vitro, the homogenate of normal saline group displayed increased staining compared with ChABC group and combined group. Immunohistochemistry showed that 8 weeks following SCI, vimentin, GFAP expression down-regulated in antisense vimentin group and combined group, and did not change in antisense vimentin control group and ChABC group, compared with normal saline group. CS-56 immunostaining decreased in ChABC group and combined group, and did not change in antisense vimentin control group and antisense vimentin group, compared with normal saline group. Cystic cavity appeared in injured spinal cord 2 week after SCI, and matured and stabilized 8 weeks later, with dense glial scar tissue forming the cavity wall. The size of cavity significantly decreased in each treated group than control group, and there was no difference between the treated groups.4. Immunoflurescence showed NF200 positive neuronal processes were in close relation with astrocytic GFAP positive processes. 8 weeks after SCI, NF200 positive neuronal fibres near the cystic cavity did not penetrate into the cavity wall, while did after reducing glial scar with combined treatment.5. Anteriograde tracing with BDA showed that normal corticospinal tract (CST) located at the dorsal column, just symmetrically posterior to the central canal, and extended straightly along the spinal cord. 8 weeks following SCI, CST arrested near the glial scar just rostral to the injury site with characteristic endbulbs, and a few fibres presented regenerative morphology, but no fibres penetrated into the scar. After combined treatment, fewer degenerative endbulbs arrested rostral to the injury site, and more BDA positive fibres winded distortedly across or into glial scar area, and projected to the area caudal to the injury site. A few fibres arose from the injury site and penetrated into the gray matter. Distribution of BDA positive axon seemed to be in close relation to the distribution of CSPGs molecules 8 weeks after SCI. More BDA positive fibres were distributed in CSPGs low expression area, and some fibres winded along the border of CSPGs expression, with characteristic growth cone morphology. A few distorted fibres penetrated into CSPGs high expression area. 6. Retrograde tracing with HRP showed that 8 week following SCI, no HRP positive fibres penetrated into the injured area of glial scar in the rats of normal saline group, while some fibres did in the rats of treated group, especially the combined treated group.7. Hindlimb BBB scores results showed that no significant difference existed between either groups in the first three weeks, while 4 week later, the treated rats showed better results than the controlled animals, while there was no difference between the combined group and the single treated group.8. Hindlimb placing response disappeared 2 weeks after injury in all 6 rats, with the maximal digit spreading distance (MDSD) decreased. 8 weeks later, 1 of 6 rats restored hindlimb response with improved MDSD. 4, 4 of 6 rats restored hindlimb response in antisense vimentin cDNA or ChABC group respectively, with significant improved MDSD. 5 of 6 rats restored in combined group with increased MDSD, compared with rats of antisense vimentin cDNA or ChABC group.Conclusion1. Characteristic reactive gliosis appeared, and glial scar and cystic cavity formed following SCI in rats. 2 weeks after SCI, vimentin, GFAP and CSPGs mRNA and protein expression increased dramatically, with the appearance of glial scar and cystic cavity. 8 weeks after SCI, vimentin and GFAP expression up-regulated more significantly, and CSPGs still expressed profusely, with the formation of mature and stabilized glial scar and cavity. The hypertrophied processes and cell bodies of reactive astrocytes and the ECM constituted the dense glial scar, which became a physical and biochemical barrier to the tiny residue or sprouted fibres. Regeneration of CST arrested rostrally to the glial scar, and could not penetrated into scar areas or project to the caudal area.2. Vimentin and GFAP expression near glial scar decreased after antisense vimentin cDNA treatment, and CSPGs molecules in scar tissue were partly digested with ChABC administration. Antisense vimentin cDNA combined with ChABC reduced glial scar and cavity formation 8 weeks after SCI.3. CST arrested just rostral to the glial scar and could not penetrate into the scar tissue or projected to the caudal area 8 weeks following SCI. Antisense vimentin cDNA combined with ChABC might promote CST regeneration, help regenerative fibres penetrate into the scar area, even project to the caudal spinal cord. 4. Antisense vimentin cDNA combined with ChABC can improve hindlimb spreading, increase the maximal digit spreading distance in hindlimb placing response experiment, and increase the hindlimb BBB score, thus promote locomotion restoration of SCI rats.