Syntenin‐a promotes spinal cord regeneration following injury in adult zebrafish

In contrast to mammals, adult zebrafish recover locomotor function after spinal cord injury, in part due to the capacity of the central nervous system to repair severed connections. To identify molecular cues that underlie regeneration, we conducted mRNA expression profiling and found that syntenin‐a expression is upregulated in the adult zebrafish spinal cord caudal to the lesion site after injury. Syntenin is a scaffolding protein involved in mammalian cell adhesion and movement, axonal outgrowth, establishment of cell polarity, and protein trafficking. It could thus be expected to be involved in supporting regeneration in fish. Syntenin‐a mRNA and protein are expressed in neurons, glia and newly generated neural cells, and upregulated caudal to the lesion site on days 6 and 11 following spinal cord injury. Treatment of spinal cord‐injured fish with two different antisense morpholinos to knock down syntenin‐a expression resulted in significant inhibition of locomotor recovery at 5 and 6 weeks after injury, when compared to control morpholino‐treated fish. Knock‐down of syntenin‐a reduced regrowth of descending axons from brainstem neurons into the spinal cord caudal to the lesion site. These observations indicate that syntenin‐a is involved in regeneration after traumatic insult to the central nervous system of adult zebrafish, potentially leading to novel insights into the cellular and molecular mechanisms that require activation in the regeneration‐deficient mammalian central nervous system.     

Spinal cord transplants permit the growth of serotonergic axons across the site of neonatal spinal cord transection

These experiments were designed to determine whether transplants of fetal spinal cord tissue into lesioned spinal cord in newborn rats provide a terrain that supports the growth of serotonergic (5-HT) axons across the site of the lesion. Although descending serotonergic axons can regenerate after chemical lesions in adult animals, they show little regrowth after surgical lesions. In newborn animals, 5-HT axons do not regrow after either chemical or mechanical lesions since the axotomized raphe-spinal neurons die. After partial spinal cord lesions made in developing animals, immature axons can take an aberrant route around the site of the lesion to reach normal target areas. Even these robust, late-growing, uninjured axons, however, are unable to grow through the site of the spinal cord lesion. Immunocytochemical labeling was used to determine if descending serotonergic axons grow into fetal spinal cord transplants, and whether these axons cross the transplant to reach spinal cord levels caudal to the lesion. Spinal cord transection at a mid-thoracic spinal cord level on the day of birth resulted in a dramatic decrease in 5-HT immunoreactivity caudal to the lesion by one day postoperative. 5-HT immunoreactivity caudal to the lesion was abolished by 5 days postoperative and did not return after acute or chronic (6 months) survival periods. When a transplant of fetal spinal cord tissue was placed into the lesion site, 5-HT axons were identified throughout the transplant. At spinal cord levels caudal to the transection and transplant, the serotonergic axons were identified in the host spinal cord in both the white and gray matter. This 5-HT innervation was not confined to spinal cord segments adjacent to the lesion site but extended to spinal cord segments as far as lower lumbar levels. The reinnervation of the host spinal cord caudal to the transection was far less than that seen in unlesioned adult rat spinal cord. Horseradish peroxidase (HRP) injected caudal to the transection and transplant, retrogradely labeled neurons within the medullary raphe nuclei. The HRP and 5-HT results both depended on apposition of the transplant with the rostral and caudal stumps of the host spinal cord; without such apposition, labeling was abolished. These results indicate that the presence of a transplant at the site of the neonatal lesion modifies the environment at the lesion site in such a manner as to support the elongation of identified axons across the site of the lesion and into the host cord caudal to the lesion.     

Collateral branching in axonal projections to spinal cord from paramedian reticular nucleus neurons

Experiments were done in cats to identify neurons in the paramedian reticular nucleus (PRN) sending collateral axons to the region of the intermediolateral nucleus (IML) at different levels of the thoracic cord by using lectin-conjugated horseradish peroxidase (HRP) and double-labeling fluorochrome histochemistry to retrogradely label PRN neurons. Injections of Fast blue (FB) into the spinal cord at the T2 level centered in the region of the IML were coupled with injections of Nuclear yellow (NY) into the ipsilateral cord at either the T4 or T7 levels centered in the region of the IML. Neurons in the PRN retrogradely labeled after diffusion of HRP into the region of the IML at the T2 level were observed throughout the rostrocaudal extent of the ventral PRN. In addition, a few labeled neurons were noted in the ventral portion of the dorsal PRN. About 40% of the neurons in the PRN which were labeled with FB after an injection at the T2 level were also labeled with NY injected into the cord in further caudal segments. These data suggest that the PRN may exert its influence on the cardiovascular system partly through collateral axonal branches to widely separated populations of sympathetic preganglionic neurons in different spinal segmental levels.     

Demonstration of the potential for chronically injured neurons to regenerate axons into intraspinal peripheral nerve grafts

Experiments were carried out to determine if neurons damaged by injury to the spinal cord retain the ability to regenerate their axonal process for a prolonged period of time after the initial response to injury and if peripheral nerve (PN) grafts could support the regrowth of these processes. True blue (TB) was injected into one side of the adult rat lumbar spinal cord to label neurons with axons coursing through this region. Seven days later spinal cord tissue surrounding the injection sites was removed by aspiration to create a hemisection cavity 3-4 mm in length. Four weeks later scar tissue lining the lesion cavity was removed prior to grafting 1 cm segments of autologous tibial nerve to the rostral and the caudal surfaces of the cavity wall. The distal end of each graft was ligated and left unapposed to spinal cord tissue. Four weeks later the distal end of each PN graft was exposed to nuclear yellow (NY) to retrogradely label neurons that had grown an axon into the graft. Neurons containing both TB and NY were deemed capable of axonal regeneration while in a chronically injured state. Double-labeled (TB/NY) neurons were found in the ipsilateral spinal cord in laminae IV through X, excluding IX, and in Laminae VI and VII contralateral to the lesion. Most neurons were located within 10 mm of the lesion, with the majority caudal to the lesion. Nearly 50% (range 24-74%) of lumbar dorsal root ganglion neurons containing TB also were labeled with NY.(ABSTRACT TRUNCATED AT 250 WORDS)     

Axonal regeneration of Clarke's neurons beyond the spinal cord injury scar after treatment with chondroitinase ABC

We have previously demonstrated that enzymatic digestion of chondroitin sulfate proteoglycan (CSPG) at the scar promotes the axonal regrowth of Clarke's nucleus (CN) neurons into an implanted peripheral nerve graft after hemisection of the spinal cord. The present study examined whether degradation of CSPG using chondroitinase ABC promoted the regeneration of CN neurons through the scar into the rostral spinal cord in neonatal and adult rats. Following hemisection of the spinal cord at T11, either vehicle or chondroitinase ABC was applied onto the lesion site. The postoperative survival periods were 2 and 4 weeks. The regenerated CN neurons were retrogradely labeled by Fluoro-Gold injected at spinal cord level C7. In the sham group, there was no regeneration of injured CN neurons in both neonatal and adult rats. Treatment with 2.5 unit/ml chondroitinase ABC in neonates resulted in 11.8 and 8.3% of the injured CN neurons regenerated into the rostral spinal cord at 2 and 4 weeks, respectively. In adults, 9.4 and 12.3%, at 2 and 4 weeks, respectively, of the injured CN neurons regenerated their axons to the rostral spinal cord. The immunoreactivity for the carbohydrate epitope of CSPG was dramatically decreased around the lesion site after treatment with chondroitinase ABC compared to sham control in both neonatal and adult animals. Our results show that axonal regeneration in the spinal cord can be promoted by degradation of CSPG with chondroitinase ABC. This result further suggests that CSPG is inhibitory to the regeneration of neurons in the spinal cord after traumatic injury.     

GDNF and NGF released by synthetic guidance channels support sciatic nerve regeneration across a long gap

The present work was performed to determine the ability of neurotrophic factors to allow axonal regeneration across a 15-mm-long gap in the rat sciatic nerve. Synthetic nerve guidance channels slowly releasing NGF and GDNF were fabricated and sutured to the cut ends of the nerve to bridge the gap. After 7 weeks, nerve cables had formed in nine out of ten channels in both the NGF and GDNF groups, while no neuronal cables were present in the control group. The average number of myelinated axons at the midpoint of the regenerated nerves was significantly greater in the presence of GDNF than NGF (4942 +/-1627 vs. 1199 +/-431, P < or = 0.04). A significantly greater number of neuronal cells in the GDNF group, when compared to the NGF group, retrogradely transported FluoroGold injected distal to the injury site before explantation. The total number of labelled motoneurons observed in the ventral horn of the spinal cord was 98.1 +/-23.4 vs. 20.0 +/-8.5 (P < or = 0.001) in the presence of GDNF and NGF, respectively. In the dorsal root ganglia, 22.7% +/- 4.9% vs. 3.2% +/-1.9% (P +/-0.005) of sensory neurons were labelled retrogradely in the GDNF and NGF treatment groups, respectively. The present study demonstrates that, sustained delivery of GDNF and NGF to the injury site, by synthetic nerve guidance channels, allows regeneration of both sensory and motor axons over long gaps; GDNF leads to better overall regeneration in the sciatic nerve.     

Neurotrophism without neurotropism: BDNF promotes survival but not growth of lesioned corticospinal neurons.

Neurotrophic factors exert many effects on the intact and lesioned adult central nervous system (CNS). Among these effects are prevention of neuronal death (neurotrophism) and promotion of axonal growth (neurotropism) after injury. To date, however, it has not been established whether survival and axonal growth functions of neurotrophins can be independently modulated in injured adult neurons in vivo. To address this question, the ability of brain-derived neurotrophic factor (BDNF) to influence corticospinal motor neuronal survival and axonal growth was examined in two injury paradigms. In the first paradigm, a survival assay, adult Fischer 344 rats underwent subcortical lesions followed by grafts to the lesion cavity of syngenic fibroblasts genetically modified to secrete high amounts BDNF or, in control subjects, the reporter gene green fluorescent protein. In control subjects, only 36.2 +/- 7.0% of the retrogradely labeled corticospinal neurons survived the lesion, whereas 89.8 +/- 5.9% (P < 0.001) of the corticospinal neurons survived in animals that received BDNF-secreting grafts. However, in an axonal growth assay, BDNF-secreting cell grafts that were placed into either subcortical lesion sites or sites of thoracic spinal cord injury failed to elicit corticospinal axonal growth. Despite this lack of a neurotropic effect on lesioned corticospinal axons, BDNF-secreting cell grafts placed in the injured spinal cord significantly augmented the growth of other types of axons, including local motor, sensory, and coerulospinal axons. Immunolabeling for tyrosine kinase B (trkB) demonstrated that BDNF receptors were present on corticospinal neuronal somata and apical dendrites but were not detected on their projecting axons. Thus, single classes of neurons in the adult CNS appear to exhibit disparate survival and growth sensitivity to neurotrophic factors, potentially attributable at least in part to differential trafficking of neurotrophin receptors. The possibility of tropic/trophic divergence must be considered when designing strategies to promote CNS recovery from injury.     

Serotonin-containing projections to the thalamus in the rat revealed by a horseradish peroxidase and peroxidase antiperoxidase double-staining technique

The retrograde transport of horseradish peroxidase (HRP) has been used in combination with peroxidase antiperoxidase (PAP) immunocytochemistry in order to investigate serotonin-containing projections to the thalamus of the rat. Sections were histochemically stained to reveal retrogradely transported HRP and then PAP immunostained using a monoclonal anti-serotonin (5-HT) antibody. Following HRP injections into the ventral thalamus, retrogradely labelled cells were observed in a number of sites in the brainstem and including areas known to be rich in 5-HT-containing neurons. At rostral levels of the dorsal raphe nucleus, retrogradely labelled cells were observed both on the midline and in a distinct lateral group extending diffusely into the periaqueductal gray (PAG). In both of these areas many 5-HT-immunoreactive HRP retrogradely labelled neurons were observed. However, except for the most rostral levels of the dorsal raphe nucleus, such double-labelled cells represented only a small proportion of the total population of 5-HT-immunoreactive neurons. In the lateral group, the retrograde labelling was mainly unilateral to the injection site but some contralateral labelling was also seen. At caudal levels of the dorsal raphe nucleus, retrogradely labelled cells were observed predominantly in the lateral group. At the level of the dorsolateral tegmental nucleus, few 5-HT of 5-HT/HRP labelled cells were observed in the lateral group, although HRP retrogradely labelled neurons were present. Double-stained cells were detected also in the medial raphe nucleus (corresponding to the B8 cell group according to the nomenclature of Dahlström and Fuxe¹³), among the fibres of the medial lemniscus (B9), and in nucleus raphe pontis (B5).     

Grafts of BDNF-producing fibroblasts rescue axotomized rubrospinal neurons and prevent their atrophy

We have reported that intraspinal transplants of fibroblasts genetically modified to express brain-derived neurotrophic factor (BDNF) promote rubrospinal axon regeneration and functional recovery following subtotal cervical hemisection that completely ablated the rubrospinal tract. In the present study we examined whether these transplants could prevent cell loss and/or atrophy of axotomized Red nucleus neurons. Adult rats received a subtotal spinal cord cervical hemisection followed by a graft of unmodified fibroblasts or fibroblasts producing BDNF into the lesion cavity. One or 2 months later, fluorogold was injected several segments caudal to the lesion-transplant site to retrogradely label those Red nucleus neurons whose axons have regenerated. Unmodified fibroblasts failed to protect against either cell loss or atrophy. Neuron counts and soma-size measurements in Nissl-stained preparations showed a 45% loss of recognizable neurons and 40% atrophy of the surviving neurons in the injured Red nucleus. Grafts of BDNF-producing fibroblasts reduced neuron loss to less than 15% and surviving neurons showed only a 20% decrease in mean soma size. Soma size analysis of fluorogold-labeled Red nucleus neurons indicated that the Red nucleus neurons whose axons regenerated caudal to the graft did not atrophy. We conclude that fibroblasts engineered ex vivo to secrete BDNF and grafted into a partial cervical hemisection promote axon regeneration while reducing cell loss and atrophy of neurons in the Red nucleus. These results suggest that transplants of genetically engineered cells could be an important tool for delivery of therapeutic factors that contribute to the repair of spinal cord injury.     

Rubrospinal neurons fail to respond to brain-derived neurotrophic factor applied to the spinal cord injury site 2 months after cervical axotomy

Numerous experimental therapies to promote axonal regeneration have shown promise in animal models of acute spinal cord injury, but their effectiveness is often found to diminish with a delay in administration. We evaluated whether brain-derived neurotrophic factor (BDNF) application to the spinal cord injury site 2 months after cervical axotomy could promote a regenerative response in chronically axotomized rubrospinal neurons. BDNF was applied to the spinal cord in three different concentrations 2 months after cervical axotomy of the rubrospinal tract. The red nucleus was examined for reversal of neuronal atrophy, GAP43 and Talpha1 tubulin mRNA expression, and trkB receptor immunoreactivity. A peripheral nerve transplant paradigm was used to measure axonal regeneration into peripheral nerve transplants. Rubrospinal axons were anterogradely traced and trkB receptor immunohistochemistry performed on the injured spinal cord. We found that BDNF treatment did not reverse rubrospinal neuronal atrophy, nor promote GAP-43 and Talpha1 tubulin mRNA expression, nor promote axonal regeneration into peripheral nerve transplants. TrkB receptor immunohistochemistry demonstrated immunoreactivity on the neuronal cell bodies, but not on anterogradely labeled rubrospinal axons at the injury site. These findings suggest that the poor response of rubrospinal neurons to BDNF applied to the spinal cord injury site 2 months after cervical axotomy is not related to the dose of BDNF administered, but rather to the loss of trkB receptors on the injured axons over time. Such obstacles to axonal regeneration will be important to identify in the development of therapeutic strategies for chronically injured individuals.