NT-3 delivered by an adenoviral vector induces injured dorsal root axons to regenerate into the spinal cord of adult rats

Sensory axons interrupted in the dorsal roots of adult mammals are normally unable to regenerate into the spinal cord. We have investigated whether the introduction of a neurotrophin gene into the spinal cord might offer an approach to otherwise intractable spinal root injuries. The dorsal roots of the 4th, 5th, and 6th lumbar spinal nerves of adult rats were severed and reanastomosed. Fourteen to nineteen days later, adenoviral vectors containing either the LacZ or NT-3 genes were injected into the ventral horn of the lumbar spinal cord, resulting in strong expression of the transgenes in glial cells and motor neurons between 4 and 40 days after injection. When dorsal root axons were transganglionically labelled with HRP conjugated to cholera toxin subunit B, 16 to 37 days after dorsal root injury, large numbers of labelled axons could be seen to have regenerated into the cord, but only in those animals injected with vector carrying the NT-3 gene. The regenerated axons were found at the injection site, mainly in the grey matter, and had penetrated as deep as lamina V. Gene therapy with adenoviral vectors encoding a neurotrophin has therefore been shown to be capable of enhancing and directing the regeneration of a subpopulation of dorsal root axons (probably myelinated A fibres), into and through the CNS environment. J. Neurosci. Res. 54:554–562, 1998. © 1998 Wiley-Liss, Inc.     

Functional connections are established in the deafferented rat spinal cord by peripherally transplanted human embryonic sensory neurons

Functionally useful repair of the mature spinal cord following injury requires axon growth and the re-establishment of specific synaptic connections. We have shown previously that axons from peripherally grafted human embryonic dorsal root ganglion cells grow for long distances in adult host rat dorsal roots, traverse the interface between the peripheral and central nervous system, and enter the spinal cord to arborize in the dorsal horn. Here we show that these transplants mediate synaptic activity in the host spinal cord. Dorsal root ganglia from human embryonic donors were transplanted in place of native adult rat ganglia. Two to three months after transplantation the recipient rats were examined anatomically and physiologically. Human fibres labelled with a human-specific axon marker were distributed in superficial as well as deep laminae of the recipient rat spinal cord. About 36% of the grafted neurons were double labelled following injections of the fluorescent tracers MiniRuby into the sciatic and Fluoro-Gold into the lower lumbar spinal cord, indicating that some of the grafted neurons had grown processes into the spinal cord as well as towards the denervated peripheral targets. Electrophysiological recordings demonstrated that the transplanted human dorsal roots conducted impulses that evoked postsynaptic activity in dorsal horn neurons and polysynaptic reflexes in ipsilateral ventral roots. The time course of the synaptic activation indicated that the human fibres were non-myelinated or thinly myelinated. Our findings show that growing human sensory nerve fibres which enter the adult deafferentated rat spinal cord become anatomically and physiologically integrated into functional spinal circuits.     

Effects of sciatic nerve resection on L7 spinal roots and dorsal root ganglia in adult cats

The size, distribution, and number of nerve fibers and neuronal perikarya in the L7 spinal roots and ganglia of adult cats were examined 35, 90, and 190 days after ipsilateral sciatic nerve resection. With increasing survival time the size spectra of myelinated ventral root nerve fibers showed a progressive flattening of the alpha peak. In the dorsal roots the myelinated fiber size distribution exhibited a marked shift toward smaller sizes. The reduction in the proportion of large myelinated axons was particularly evident in the dorsal roots. Less clearcut changes were found in the size distribution of spinal ganglion neuronal perikarya. No significant loss of axons could be detected in ventral or dorsal roots. There was, however, a marked reduction in the number of dorsal root ganglion neurons. This discrepancy suggested the possibility that an initial loss of dorsal root axons was concealed by recurrent sprouting of axons from the proximal nerve stump. However, neuroma excision 90 days after nerve resection did not lead to any reduction in dorsal root axon numbers. Thus, any ingrowth of new axons to the dorsal root should occur from levels proximal to the neuroma. In comparison with previous findings in kittens, peripheral nerve resection in adult cats had significantly smaller effects on sizes and numbers of spinal root nerve fibers as well as of dorsal root ganglion neurons. Therefore, the potential for restitution of the peripheral innervation by axon regeneration appeared to be basically greater in mature than in immature animals.     

Carbonic anhydrase activity in the normal and injured peripheral nervous system of the rat

The carbonic anhydrase reactivity of primary neurons and axons of the L4 and L5 lumbar levels was studied in rats before and after various surgical procedures including transection of the spinal cord, removal of dorsal root ganglia, and transection of ventral or dorsal roots or spinal nerves. In normal animals, carbonic anhydrase reactivity was confined to large and medium size neurons of the dorsal root ganglia, and was also present in a sizeable percentage of cells scattered throughout the thoracolumbar sympathetic chain and in the celiac ganglion. At root level, enzymatic staining could be detected in 48.7% of the dorsal root myelinated axons of most sizes, whereas in ventral roots, it was restricted to small myelinated axons, in a proportion much higher at the L4 than in the L5 level. Spinal motoneurons remained unlabeled, despite procedures aimed at increasing the somal concentration of carbonic anhydrase, such as ventral root ligation and blocking of the fast or slow axoplasmic transport using colchicine or iminodiproprionitrile. However, it is likely that reactive ventral root axons originate from neurons situated segmentally in the spinal cord, and do not constitute aberrant sensory fibers, as carbonic anhydrase activity remained unchanged in the L4 and L5 ventral roots after removal of the corresponding spinal ganglia, whereas it disappeared after damage to the spinal cord at the lumbar level, or at a site distal to a ventral root section. Enzymatic staining of neurons of the dorsal root ganglia was not modified by a dorsal rhizotomy, but showed a marked decrease after transection of the spinal nerve.(ABSTRACT TRUNCATED AT 250 WORDS)     

Regeneration after spinal nerve root injury

Spinal nerve root avulsion has been considered as a central nervous type of injury and therefore not repaired surgically in man. The possibility for axonal regeneration after root avulsion or root lesion has been investigated in laboratory animals by means of up to date neurophysiological, morphological and tracing techniques. It is shown that, after ventral root avulsion and implantation into the spinal cord, alpha and probably also gamma motoneurons are able to regenerate within the spinal cord for a considerable distance before entering the implanted root and reinnervate previously denervated skeletal muscles. The regenerated neurons were found to respond to afferent activity with excitatory or inhibitory responses, and the regenerated axons could conduct action potentials that elicited muscle twitch responses. After dorsal root injury in the adult animal, regeneration into the spinal cord does not occur. However, regeneration of primary sensory neurons into appropriate locations of the spinal cord can be demonstrated in immature animals.     

Is CNS trauma a prerequisite for the elongation of CNS axons into denervated peripheral nerve?

The demonstration that some central nervous system (CNS) axons can regenerate when provided with a suitable environment raises the possibility of new treatments for CNS injury. However, at present the conditions for optimal regeneration are not well understood. For example, the methods used in previous studies have entailed CNS trauma as part of the research protocol (e.g. that resulting from the implantation of peripheral nerve grafts), and so the role of neuronal or axonal injury in the regrowth observed has been difficult to establish. To determine whether such injury is necessary for the central reinnervation of denervated peripheral nerve, the L5 dorsal root has been chronically denervated in rats by freeze-thawing its dorsal root ganglion (DRG), and the root has been left attached to either traumatized or non-traumatized spinal cord. The trauma induced was quite mild, and resulted from several vertical insertions of a fine needle. Two to 4 months later, retrogradely transported horseradish peroxidase (HRP) was used to label spinal neurons which sent axons into the denervated roots. HRP-labelled neurons were found in each of the spinal cords subjected to trauma, but no labelled neurons were observed in any of the non-traumatized cords. The number of HRP-labelled neurons in individual spinal cords was positively correlated with the degree of spinal cord trauma. We conclude first that the chronic and intimate presence of a denervated PNS tissue in continuity with the spinal cord is not, in itself, a sufficient stimulus to induce its reinnervation by CNS axons. Second, we conclude that under the conditions of this experiment CNS trauma is a prerequisite for the reinnervation of denervated peripheral nervous tissue by CNS axons.     

The distribution and morphology of parasympathetic preganglionic neurons in the cat sacral spinal cord as revealed by horseradish peroxidase applied to the sacral ventral roots

Parasympathetic preganglionic neurons in the sacral parasympathetic nucleus (SPN) of the cat were studied by applying horseradish peroxidase (HRP) to the sacral ventral roots. The results were compared to data from earlier experiments in which these same neurons were labelled by HRP applied to the pelvic nerve at a point much further from the spinal cord. The present experiments have shown. The total number of neurons in the SPN determined by ventral root labelling is equal to the number obtained by pelvic nerve labelling. This indicates that virtually all SPN neurons send their axons into the pelvic nerve. More extensive dendritic projections from SPN neurons were revealed than in the pelvic nerve experiments. In particular, a strong dendritic projection extended within the lateral marginal zone of the dorsal horn close to Lissauer's tract and horizontal dendrites projected well into the contralateral gray matter, some reaching the contralateral sacral parasympathetic nucleus. The neurons labelled via any one ventral root were all contained within a region equal in length to one spinal segment but shifted rostrally by a small amount. Thus, no evidence was obtained for a long intraspinal pathway in which axons of spinal cord neurons entered ventral roots many segments away from their somata.     

The Microtubular Pattern Changes at the Spinal Cord-Root Junction and Reverts at the Root-Peripheral Nerve Junction in Sensory and Motor Fibres of the Rat

In the rat, we studied the microtubular content of central nervous system (CNS) axons (pyramidal tract, dorsal funiculus, and intracord domain of motor axons), of radicular axons (ventral and dorsal roots), and of peripheral axons (sural and lateral gastrocnemius nerves). The microtubular density had an inverse relationship with the size of the axon. Within the CNS, values ranged from over 120 microtubules/microm2 for axons smaller than 0.1 microm2 of the pyramidal tract and dorsal funiculus to 24 for 3-microm motor axons (area, 7 microm2) in their spinal cord domain. Peripheral nerve and CNS axons of the same size had comparable microtubular densities. In contrast, the microtubular density of dorsal and ventral root axons was one half that of CNS or peripheral nerve axons of equal calibre. Considered along the axon, the microtubular density of motor and sensory fibres is high in the CNS domain, low in the root, and high again in the peripheral nerve domain. These observations are inconsistent with the notion that the cytoskeleton moves coherently away from the perikaryon. We conclude that the axonal microtubular content accords with the calibre of the fibre and with the anatomical region where it courses. We propose that axonal microtubules are regulated by local cues.     

Regeneration of adult dorsal root axons into transplants of embryonic spinal cord

Transplants of the embryonic rat spinal cord survive and differentiate in the spinal cords of adult and newborn host rats. Very little is known about the extent to which these homotopic transplants can provide an environment for regeneration of adult host axons that normally terminate in the spinal cord. We have used horseradish peroxidase injury filling and transganglionic transport methods to determine whether transected dorsal roots regenerate into fetal spinal cord tissue grafted into the spinal cords of adult rats. Additional transplants were examined for the presence of calcitonin gene-related peptide-like immunoreactivity, which in the normal dorsal horn is derived exclusively from primary afferent axons. Host animals had one side of the L4-5 spinal cord resected and replaced by a transplant of E14 or E15 spinal cord. Adjacent dorsal roots were sectioned and juxtaposed to the graft. The dorsal roots and their projections into the transplants were then labeled 2-9 months later. The tracing methods that used transport or diffusion of horseradish peroxidase demonstrated that severed host dorsal root axons had regenerated and grown into the transplants. In addition, some donor and host neurons had extended their axons into the periphery to at least the midthigh level as indicated by retrograde labeling following application of tracer to the sciatic nerve. Primary afferent axons immunoreactive for calcitonin gene-related peptide were among those that regenerated into transplants, and the projections shown by this immunocytochemical method exceeded those demonstrated by the horseradish peroxidase tracing techniques. Growth of the host dorsal roots into transplants indicates that fetal spinal cord tissue permits regeneration of adult axotomized neurons that would otherwise be aborted at the dorsal root/spinal cord junction. This transplantation model should therefore prove useful in studying the enhancement and specificity of the regrowth of axons that normally terminate in the spinal cord.     

Strategies for repair of the deafferented spinal cord

Deafferentation of the spinal cord by interruption of the sensory fibers in the dorsal roots highlights the problem of regeneration failure in the central nervous system. The injured dorsal root axons regenerate steadily, albeit slowly, in the peripheral compartment of the dorsal root, but abruptly cease to elongate when confronted with the interface between the peripheral and central nervous system, the dorsal root transitional zone (DRTZ). The glial cells of the CNS and their products together form this regeneration barrier. Recent years have witnessed several successful approaches to, at least in part, overcome this barrier. Particularly promising results have been obtained by (1) the replacement of adult non-regenerating dorsal root ganglion neurons with corresponding cells from embryonic or fetal donors, (2) the implantation of olfactory ensheathing cells at the DRTZ, and (3) immediate intrathecal infusion of growth factors to which dorsal root ganglion cells respond. In all these instances, growth of sensory axons into the adult spinal cord, as well as return of spinal cord connectivity, have been demonstrated. These findings suggest routes towards treatment strategies for plexus avulsion, and contribute to our understanding of possibilities to overcome regeneration failure in the spinal cord.     

Functional repair after dorsal root rhizotomy using nerve conduits and neurotrophic molecules

Functional recovery after large excision of dorsal roots is absent because of both the limited regeneration capacity of the transected root, and the inability of regenerating sensory fibers to traverse the dorsal root entry zone. In this study, bioresorbable guidance conduits were used to repair 6-mm dorsal root lesion gaps in rats, while neurotrophin-encoding adenoviruses were used to elicit regeneration into the spinal cord. Polyester conduits with or without microfilament bundles were implanted between the transected ends of lumbar dorsal roots. Four weeks later, adenoviruses encoding NGF or GFP were injected into the spinal cord along the entry zone of the damaged dorsal roots. Eight weeks after injury, nerve regeneration was observed through both types of implants, but those containing microfilaments supported more robust regeneration of calcitonin gene-related peptide (CGRP)-positive nociceptive axons. NGF overexpression induced extensive regeneration of CGRP(+) fibers into the spinal cord from implants showing nerve repair. Animals that received conduits containing microfilaments combined with spinal NGF virus injections showed the greatest recovery in nociceptive function, approaching a normal level by 7-8 weeks. This recovery was reversed by recutting the dorsal root through the centre of the conduit, demonstrating that regeneration through the implant, and not sprouting of intact spinal fibers, restored sensory function. This study demonstrates that a combination of PNS guidance conduits and CNS neurotrophin therapy can promote regeneration and restoration of sensory function after severe dorsal root injury.     

Structural plasticity of the adult CNS. Negative control by neurite growth inhibitory signals

The neuronal network of the adult central nervous system (CNS) retains a limited capacity for growth and structural change. This structural plasticity has been best studied in the context of lesion-induced growth and repair. More recently, structural changes underlying functional plasticity occurring under specific physiological conditions have also been documented, in particular in the cortex and the hippocampus. Areas known for their adult plastic potential retain high levels of the growth associated protein GAP-43, suggesting a persistence of important components of the intracellular growth machinery throughout life. Interestingly, a pronounced negative correlation exists between the levels of GAP-43 and myelination in the adult CNS. Because CNS myelin contains potent neurite growth inhibitory membrane proteins, neurite growth, sprouting and plasticity were investigated in the spinal cord and brain in areas where oligodendrocyte development and myelin formation was experimentally prevented, or in the presence of an inhibitor neutralizing antibody (mAB-IN-1). In all areas, lesion-induced or spontaneous sprouting was enhanced, in parallel with persistent high levels of GAP-43. Thus, spontaneous sprouting of side branches occurred from retinal axons in the optic nerve in the absence of myelin, and target-deprived retinal axons showed increased sprouting and innervation of the contralateral optic tectum in the presence of mAB IN-1. In experimentally myelin-free spinal cords collaterals from intact dorsal roots grew over long distances to innervate deafferented target regions following the section of three dorsal roots. Similarly, the corticospinal tract sprouted across the the midline and re-established a dense plexus of fibres on the contralateral side of the spinal cord following section of one corticospinal tract in juvenile rats. Following bilateral dorsal hemisection of the spinal cord including both corticospinal tracts in young and adult rats, long distance regeneration of corticospinal fibres leading to significant functional improvements of locomotion and certain reflexes was induced by the neurite growth inhibitor neutralizing antibody IN-1.     

Axonal regeneration from injured dorsal roots into the spinal cord of adult rats.

Injury to the central processes of primary sensory neurons produces less profound changes in the expression of growth-related molecules and less vigorous axonal regeneration than does injury to their peripheral processes. The left L4, L5, and L6 dorsal roots of deeply anaesthetized adult Sprague-Dawley rats were severed and reanastomosed, and in some animals, the ipsilateral sciatic nerve was crushed to increase the expression of growth-related molecules. After between 28 days and three months, the sciatic nerve of most animals was injected with transganglionic tracers and the animals were killed 2-3 days later. Other animals were perfused for electron microscopy. Very few regenerating axons entered the spinal cord of the rats without sciatic nerve injuries. Labelled axons, however, were always found in the spinal cord of rats with sciatic nerve injuries. They often entered the cord around blood vessels, ran rostrally within the superficial dorsal horn, and avoided the degenerating white matter. The animals with a conditioning sciatic nerve crush had many more myelinated axons around the dorsal root entry zone (DREZ) and on the surface of the cord. Thus, a conditioning lesion of their peripheral processes increased the ability of the central processes of myelinated A fibres to regenerate, including to sites (such as lamina II) they do not normally occupy. Astrocytes, oligodendrocytes, and meningeal fibroblasts in and around the DREZ may have inhibited regeneration in that region, but growth of the axons into the deep grey matter and degenerated dorsal column was also blocked.     

Effects of sciatic nerve crush on the L7 spinal roots and dorsal root ganglia in kittens

The number and size distribution of axons and neurons were examined in the L7 spinal roots and ganglia of kittens 14 to 220 days after early postnatal sciatic nerve crush. The results show that motoraxons in the ventral root as well as axons and perikarya of sensory neurons in the dorsal root remained growth-retarded throughout the examined period. This was most evident in the dorsal root. Both ventral and dorsal roots showed some loss of myelinated axons, but this was only half that previously observed after sciatic nerve resection. Whereas in the dorsal roots and dorsal root ganglia the loss seemed to be nonselective with respect to size, axons in the gamma range were primarily affected in the ventral roots. In the dorsal roots the proportion of unmyelinated axons was comparable with controls but in the ventral roots it was somewhat elevated. In most cases the loss of dorsal root ganglion neurons was relatively greater than the decrease of dorsal root axons.     

Immunosuppressants promote adult dorsal root regeneration into the spinal cord

Immunosuppressants promote neurite extension in culture and facilitate regeneration of peripheral nerves in vivo. However, their neurotrophic effects in the CNS have not been well studied. We utilized a rat dorsal root transaction model to examine the effects of cyclosporine A (CsA) and FK 506 on regeneration of the dorsal root into the spinal cord. After surgery, the rats received daily subcutaneous injections of CsA or FK 506. One month after surgery, dorsal root axons were immunohistochemically labeled to evaluate the extent of regeneration into the spinal cord. Dorsal root axons of CsA/FK 506-treated rats frequently entered into the spinal cord and arborized extensively. Administration of immunosuppressants markedly promoted regeneration of dorsal root axons into the spinal cord.     

Axonal regrowth through collagen tubes bridging the spinal cord to nerve roots

The capacity of central nervous system (CNS) axons to elongate from the spinal cord to the periphery throughout a tubular implant joining the ventral horn of the spinal cord to an avulsed root was investigated in a model of brachial plexus injury. The C5-C7 roots were avulsed by controlled traction and the C6 root was bridged to the spinal cord over a 3 mm gap by the use of a collagen cylinder containing or not containing an autologous nerve segment, or an autologous nerve graft. Nine months later, the functionnality and the quality of the axonal regrowth was evaluated by electrophysiology, retrograde labelling of neurons, and histological examination of the gap area. A normal electromyogram of the biceps was observed in all animals where the C6 root was bridged to the spinal cord. The mean average amplitude of the motor evoked potentials was comprised between 17.51 ± 12.03 μV in animals repaired with a collagen cylinder, and 27.83 ± 22.62 μV when a nerve segment was introduced in the tube. In nonrepaired animals spontaneous potentials reflecting a muscle denervation were observed at electromyography. Retrograde labelling indicated that a mean number of 58.88 ± 37.89 spinal cord neurons have reinnervated the biceps in animals repaired with a tube versus 78.38 ± 62.11 when a nerve segment was introduced in the channel, and 97.25 ± 56.23 in nerve grafting experiments. Analyses of the repair site showed the presence of numerous myelinated regenerating axons. In conclusion, our results indicate that spinal cord neurons can regenerate through tubular implants over a 3 mm gap, and that this axonal regrowth appeared as effective as in nerve grafting experiments. The combination of an implant and a nerve segment did not significantly increase the regeneration rate. J. Neurosci. Res. 49:425–432, 1997. © 1997 Wiley-Liss, Inc.     

Reinnervation of the gastrocnemius muscle by the contralateral S1 nerve root

Functional regeneration after transposition of a ventral nerve root was established in the adult cat. Reconstruction of the ventral root, using microsurgical methods, directed the right S1 ventral nerve root to innervate the left gastrocnemius muscle. Stimulus-induced unit responses were recorded from the left gastrocnemius muscle 5 to 8 months after the root cross, demonstrating the reestablishment of neuromuscular connections. The innervation of the left gastrocnemius muscle by neurons in the right ventral horn of the spinal cord was verified by injecting horseradish peroxidase into the muscle. Horseradish peroxidase reaction product was located in alpha and gamma motor neurons in the right S1 segment of the spinal cord. Computer-assisted determination of the soma area of the labeled neurons was compared with a normal S1 innervation of the gastrocnemius muscle. Analysis of the percentage of cells of a given soma area demonstrated an overall decrease in soma area in the operated animals. Because ventral root reconstruction can result in innervation of a foreign muscle, studies such as this may encourage repair or reconstruction of nerve roots to gain some functional recovery after spinal cord or nerve root injuries.     

Nerve growth factor receptor immunoreactivity in the neuronal perikarya of human sensory and sympathetic nerve ganglia

We examined immunohistochemically the dorsal root ganglia, sympathetic ganglia, spinal cord, ventral and dorsal roots, and sciatic nerves obtained at autopsy from adult humans, using a monoclonal antibody against the human nerve growth factor receptor. We observed labelling in a granular pattern in the neuronal perikarya of dorsal root and sympathetic nerve ganglia. Ventral horn cells and axons were not labelled.     

Nogo-A在成年大鼠脊髓和背根节的分布

目的研究Nogo—A在成年正常大鼠脊髓和背根节的分布。方法免疫组织化学方法(ABC法)和免疫荧光双标记法。结果正常成年大鼠的脊髓灰质分布有大量的Nogo—A免疫阳性的寡突胶质细胞、运动神经元和中间神经元，免疫阳性反应产物主要分布于细胞的胞体和部分突起中。Nogo—A广泛分布于穿行于脊髓白质的纤维包裹的髓鞘和轴突上。在脊髓前根、后根和坐骨神经的运动和感觉的有髓和无髓纤维也可观察到Nogo—A的表达。而背根神经节的神经元也大量表达Nogo—A，其强度由弱至强不等，广泛分布于大、中、小各类感觉神经元的胞质及突起中。结论Nogo—A在成年大鼠脊髓，背根神经节和外周神经纤维的广泛存在提示其在正常状态下的神经功能中可能起重要作用。