Astrocytes and guidance of outgrowing corticospinal tract axons in the rat. An immunocytochemical study using anti-vimentin and anti-glial fibrillary acidic protein

In the present investigation the role of astrocytes and their precursors in guidance of outgrowing corticospinal tract axons in the rat is studied. Antibodies against glial fibrillary acidic protein and vimentin are used to analyse immunogen expression of glial cells, whereas the postnatal outgrowth of corticospinal tract axons through the spinal cord was studied using anterogradely transported horseradish peroxidase. The first, leading corticospinal tract axons, being the objective of the present study, are characterized by dilatations at their distal ends, the growth cones. Growth cones of pioneer corticospinal tract axons are randomly distributed in the presumptive corticospinal tract area of the ventral most part of the dorsal funiculus. A dramatic change in glial cell labelling is found from the majority being vimentin immunoreactive and glial fibrillary acidic protein-negative at birth to almost all being the reverse at the end of the fourth postnatal week. From double labelling experiments it can be concluded that the vimentin-glial fibrillary acidic protein transition occurs within astrocyte precursor cells. The absence of glial fibrillary acidic protein-immunoreactive glial cells during the outgrowth period of pioneer corticospinal tract axons indicates that they cannot play a role in the guidance of outgrowing corticospinal tract pioneer axons. Vimentin-immunoreactive glial cells are present throughout the presumptive corticospinal tract area at the time of arrival of the leading corticospinal tract fibres. The vimentin-immunoreactive glial cells, which themselves are orientated perpendicular to the outgrowing corticospinal tract axons, are mainly arranged in longitudinal tiers parallel to the rostrocaudal axis. Electron microscopically, growth cones of pioneer corticospinal tract axons frequently exhibit protrusions into vimentin-immunoreactive glial cell processes, suggesting an adhesive type of contact. Therefore, in addition to a positional role, vimentin-immunoreactive glial cells probably play a chemical role in guidance of pioneer corticospinal tract axons. A prominent vimentin-immunoreactive glial septum was noted during corticospinal tract outgrowth in the midline raphe of the medulla oblongata and spinal cord whereas it is absent in the decussation area of corticospinal tract fibres. After the first postnatal week the major vimentin-immunoreactive glial barrier either completely disappears (medullary levels) or gradually reduces to a minor glial fibrillary acidic protein-immunoreactive one (spinal cord levels).(ABSTRACT TRUNCATED AT 400 WORDS)     

Postnatal development of the corticospinal tract in the rat

Summary Horseradish-peroxidase was used to anterogradely label and thus to trace the growth of corticospinal axons in rats ranging in age from one day to six months. Three to eight HRP-gels were implanted in the left cerebral hemisphere of the cortex. In each spinal cord three levels were studied, the cervical intumescence (C5), the mid-thoracic region (T5) and the lumbar enlargement (L3). The methodology employed for the electron microscopic visualization of HRP has been described previously (Joosten et al. 1987a).The outgrowth of labelled unmyelinated corticospinal tract axons in the rat spinal cord primarily occurs during the first ten postnatal days. The outgrowth of the main weve of these fibres is preceded by a number of pathfinding axons, characterized by dilatations at their distal ends, the growth cones. By contrast, later appearing unmyelinated axons, which presumably grow along the pathfinding axons, do not exhibit such growth cones. The first labelled pioneer axons can be observed in the cervical intumescence at postnatal day one (P1), in the mid-thoracic region at day three (P3) and in the lumbar enlargement at day five (P5).Prior to the entrance of the axons, the prospective corticospinal area or the pre-arrival zone is composed of fascicles consisting of unlabelled, unmyelinated fibres surrounded by lucent amorphous structures. During the outgrowth phase of the corticospinal fibres some myelinated axons could be observed within the outgrowth area even before day 14. These axons, however, were never labelled. These findings strongly suggest that the outgrowth area, which is generally denoted as the pyramidal tract, contains other axons besides the corticospinal fibres (and glial cells). The process of myelination of the labelled corticospinal tract axons in the rat spinal cord starts rostrally (C5) at about day 14 and progresses caudally during the third and fourth postnatal weeks. Although myelination seems to be largely complete at day 28 at all three spinal cord levels, some labelled unmyelinated axons are still present in the adult stage.     

Ipsilateral, ventral corticospinal tract of the adult rat: ultrastructure, myelination and synaptic connections

The corticospinal tract (CST) of the rat is a widely used model system in developmental, physiological, and regeneration studies. The CST of the rat consists of a main tract, that runs in the dorsomedial funiculus and several minor components. We have shown earlier that one of the minor components, the ipsilateral, ventral CST, projects all the way down the spinal cord in the adult rat and single fibers form large terminal arbors in their spinal target areas. Here we investigated its ultrastructure and compared it to that of CST fibers of the main tract. By the use of anterograde axonal tracing with biotin dextran-amine (BDA) and pre-embedding avidin-peroxidase histochemistry we investigated axon diameters and myelination using electron microscopy. Ipsilateral, ventral CST fibers were found to run in the ventral funiculus close to the midline. They were intermingled with heavily myelinated large diameter axons, presumably reticulospinal, vestibulospinal, or tectospinal fibers. Ipsilateral, ventral CST fibers were of small diameter (0.68 m, ±0.04) and about [frac34] of them were moderately myelinated (9.64 ± 0.7 layers of myelin). Co-localization of a rhodamine-dextrane anterograde tracer with the presynaptic marker synaptophysin using confocal microscopy and electron microscopy revealed varicosities on terminal arborisations to be presynaptic boutons and clearly demonstrated contacts to neurons in intermediate laminae of the spinal cord at lumbar spinal levels. This study extends our earlier work indicating that the ipsilateral, ventral CST component of the adult rat is a morphologically complete CST component and may perform similar functions to the main CST component.     

Axon guidance of outgrowing corticospinal fibres in the rat

This review is concerned with the development of the rat corticospinal tract (CST). The CST is a long descending central pathway, restricted to mammals, which is involved both in motor and sensory control. The rat CST is a very useful model in experimental research on the development of fibre systems in mammals because of its postnatal outgrowth throughout the spinal cord as well as its experimental accessibility. Hence mechanisms underlying axon outgrowth and subsequent target cell finding can be studied relatively easily. In this respect the corticospinal tract forms an important example and model system for the better understanding of central nervous system development in general.     

Alpha calcium/calmodulin-dependent protein kinase II immunoreactivity in corticospinal neurons: combination of axonal transport method and immunofluorescence

A combination of either retrograde or anterograde fluorescent tracer and immunofluorescence histochemistry using the monoclonal antibody specific for the alpha isoform of calcium/calmodulin-dependent protein kinase II (CaM kinase II) was employed to test whether CaM kinase II is expressed in somata of corticospinal neurons and their axons over their whole course. After the injection of carbocyanine dye DiI into the hindlimb area of the primary motor cortex of the rat, corticospinal axons and their terminal arbors were anterogradely labeled: DiI-labeled corticospinal fibers proceeded caudally in the ipsilateral internal capsule, cerebral peduncle and medullary pyramid, crossed at the pyramidal decussation and descended in the ventralmost area of the contralateral dorsal funiculus of the spinal cord. These DiI-labeled corticospinal axons expressed strong CaM kinase II immunoreactivity along their course. However, their terminal arbors within the gray matter of the lumbar cord were very weakly immunostained. With the injection of Fast Blue into the lumbar enlargement of the rat, somata of corticospinal neurons in layer V of the motor cortex were retrogradely labeled. The subsequent immunofluorescent histochemistry revealed that more than 80% of Fast Blue-labeled corticospinal neurons were immunostained with CaM kinase II antibody. The present immunohistochemical study demonstrated that CaM kinase II is strongly expressed in both somata and axons of a majority of corticospinal neurons, although we could not detect this enzyme in the corticospinal terminals in the spinal target areas.     

An observation on hitherto unknown corticospinal fibers that descend between the tractus corticospinalis lateralis and ventralis in the cat.

Following injection of wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP) into the pericruciate (sensorimotor) cortex on one side in the cat, we found labeled fibers surrounding the ventral horn of the spinal cord bilaterally, in addition to the crossed and the uncrossed lateral, ventral and dorsal corticospinal tracts (CSTs). These hitherto unknown fibers, which could be traced to the caudal end of the L2 segment, were predominantly contralateral to the WGA-HRP injection side. This study indicates the presence of newly recognized CST fibers in the cat.     

Anatomical evidence for the re-crossing of lateral corticospinal fibers via the posterior gray commissure in the cat spinal cord

This study demonstrated the projections of the corticospinal tract (CST) by using the anterograde transport of wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP). Following injection of WGA-HRP into the pre- and post-cruciate cortices (somatosensory area) on one side, labeled fibers were found in the spinal cord both in the lateral and in the anterior CSTs; both crossed, and uncrossed. The most conspicuous labeling was found in the crossed lateral CST. In addition, labeled lateral CST fibers were seen to recross via the posterior gray commissure. These re-crossed lateral CST fibers were most frequently observed in the cervical spinal segments.     

Immunocytochemical localization of cell adhesion molecule L1 in developing rat pyramidal tract

L1 is a representative of a family of carbohydrate neural cell adhesion molecules. The expression of L1 was studied during postnatal development of the rat pyramidal tract by immunohistology using polyclonal antibodies to L1 in spinal cord cervical intumescences. On postnatal day 1 (P1), L1 immunoreactivity was present in the entire dorsal funiculus, consisting of the ascending fasciculus gracilis and fasciculus cuneatus and the descending pyramidal tract. At that time the cervical pyramidal tract contains the first outgrowing corticospinal axons. At P4 both the fasciculus gracilis and the pyramidal tract are immunoreactive whereas the fasciculus cuneatus is negative. At P10 the pyramidal tract is intensely labelled whereas both ascending bundles are negatively stained. In the period between P4 and P10 the pyramidal tract is characterized by a massive outgrowth of corticospinal axons. During pyramidal tract myelination, between P10 and the end of the third postnatal week (P21), L1 immunoreactivity is progressively reduced. These observations suggest that L1 may play a prominent role in outgrowth, fasciculation and the onset of myelination of rat pyramidal tract axons. The differential L1 immunoreactivity of the pyramidal tract and the earlier developing ascending systems in rat dorsal funiculus indicate that this polyclonal antiserum is a useful differentiating marker for outgrowing fibre tracts.     

The axonal projections of the Hofmann nuclei in the spinal cord of the late stage chicken embryo

The projection pattern of the neurons of the paragriseal Hofmann nuclei was mapped in the chicken embryo using the lipophilic tracer DiI. This report focuses on the pattern of projection from the Hofmann nuclei major observed 1–4 days prior to hatching, at which time the projection appears to be substantially developed. (1) Each neuron extends a commissural axon through the ventral gray matter and across the midline in the ventral commissure. The axons originating from a single Hofmann nucleus cross within a stretch of the cord equivalent to about one spinal segment. There is a small overlap of the axon populations originating from adjacent Hofmann nuclei. After reaching the contralateral ventral columns the individual axons bifurcate and extend rostrally and caudally up to 5 spinal segments in each direction. The rostral and caudal trajectories differ; the rostral axons shift progressively more laterally while the caudal axons tend not to deviate from their initial course. (2) Throughout their longitudinal course the axons give rise to terminal collaterals that are concentrated in lamina 8. Rostrally and caudally the terminals decrease in density and become increasingly scattered. (3) Hofmann neurons are multipolar with 4–5 laterally directed primary dendrites whose arbors are restricted to the Hofmann nucleus major within which the neurons reside. (4) Hofmann neurons receive afferent terminals from a longitudinal column of commissural interneurons located contralaterally in close approximation to the central canal. Each Hofmann nucleus major is innervated by a rostrocaudally restricted subset of these presynaptic neurons. The axon trajectories of the presynaptic neurons are similar to those of the Hofmann neurons. (5) Paragriseal neurons that are not located within Hofmann nuclei major are also commissural intersegmental interneurons and tend to be clustered segmentally. The segmentation is clearest for the Hofmann nuclei minor, which are clusters of neurons iterated along the ventrolateral margin of the thoracic spinal cord but not organized in protruding lobes.     

The terminations of single, physiologically identified, somatosensory, corticospinal tract axons in the lumbar spinal cord of the cat.

1. Two hundred and twelve corticospinal axons were identified by stimulation in the hindlimb representation in area 3b of the somatosensory cortex and were recorded in the left dorsolateral funiculus of the spinal cord of the cat. The mean conduction velocity was 38 m/s, range 9-113 m/s. 2. Electrical stimulation of the receptive field evoked discharge in corticospinal axons with a mean latency of 36 ms (range 9-100 ms). 3. One hundred nine of the 212 recorded axons were successfully intra-axonally labeled by iontophoretic injection of horseradish peroxidase, with the mean length of labeled axon being 4.8 mm. Seventy-three of the labeled axons issued no collaterals, and 36 issued at least one labeled collateral into the spinal gray matter along the labeled portion of the parent axon. 4. Most labeled axons issued only one labeled collateral per spinal cord segment. Fourteen collaterals from 10 units were labeled well enough to permit reconstruction of their terminal arborizations. 5. Most terminal collaterals were oriented rostrocaudally and terminated in laminae V, VI, and VII. Most collaterals terminated within large mediolateral extents of the gray matter with no apparent topographic organization. 6. No collaterals terminated in laminae I or II or within the motoneuron pools, and no apparent correlation was found between their anatomic and physiological characteristics.     

Selective stimulation of dendrite outgrowth from identified corticospinal neurons by homotopic astrocytes.

Corticospinal neurons were identified in primary cultures of cortical neurons established from rats that had been injected with a fluorescent tracer to retrogradely label the corticospinal tract. We measured neurite outgrowth from corticospinal neurons after they had been co-cultured with astrocytes derived from either the cerebral cortex (homotopic region) or spinal cord (target region) of postnatal rats. The axon length of corticospinal neurons was increased when they were cultured on astroglial monolayers compared to a control monolayer (fibroblasts). However, no difference in axon length was noted on cortical versus spinal cord-derived astrocytes. On the other hand, total dendritic length was increased on cortical compared to spinal cord astrocytes. This increase in total dendrite length was not the result of differences in the length of primary dendrites, but primarily of a higher number of dendrites and increased branching on the cortical astroglia. If the corticospinal neurons were co-cultured without physical contact with the astrocytes, axonal and dendritic outgrowth were not stimulated when compared to the fibroblast control. The data indicate that dendritic growth from corticospinal neurons is preferentially promoted by astrocytes from the cerebral cortex, whereas axonal growth is not influenced by the anatomical origin of the astrocytes. The impact of these findings on our understanding of the role of astrocytes in the development and regeneration of the corticospinal tract is discussed.     

Growing corticospinal axons by-pass lesions of neonatal rat spinal cord

The anterograde transport of horseradish peroxidase was used to label newly growing corticospinal axons after they had entered lesioned regions of the neonatal rat spinal cord. Two types of lesions were made at thoracic and lumbar levels before the arrival of the first corticospinal axons. (1) Thermal lesions were produced by the brief application of a heated rod to the vertebral column and could destroy the normal growth path over several spinal segments. Corticospinal axons, when successful in growing distal to thermal lesions, did so at the same rate as in normal controls and retaining their normal relative positions and morphology, especially fasciculation. (2) Surgical lesions were produced by cutting the spinal cord and were limited to one segment but could result in a barrier in the normal growth path composed of a cyst or glial scar. Corticospinal axons that succeeded in growing distal to a surgical lesion did so by being deflected to unusual positions, became defasciculated, and sometimes their normal growth rate was slowed. That corticospinal axons could in many instances grow past the two types lf lesion suggests that a morphologically stereotyped glial scaffolding is not necessary for axon growth. The role of fasciculation in normal axon growth is highlighted by the disparate effects of the two types of lesion.     

Ryk-mediated Wnt repulsion regulates posterior-directed growth of corticospinal tract

Guidance cues along the longitudinal axis of the CNS are poorly understood. Wnt proteins attract ascending somatosensory axons to project from the spinal cord to the brain. Here we show that Wnt proteins repel corticospinal tract (CST) axons in the opposite direction. Several Wnt genes were found to be expressed in the mouse spinal cord gray matter, cupping the dorsal funiculus, in an anterior-to-posterior decreasing gradient along the cervical and thoracic cord. Wnts repelled CST axons in collagen gel assays through a conserved high-affinity receptor, Ryk, which is expressed in CST axons. Neonatal spinal cord secretes diffusible repellent(s) in an anterior-posterior graded fashion, with anterior cord being stronger, and the repulsive activity was blocked by antibodies to Ryk (anti-Ryk). Intrathecal injection of anti-Ryk blocked the posterior growth of CST axons. Therefore, Wnt proteins may have a general role in anterior-posterior guidance of multiple classes of axons.     

孕酮对脊髓损伤后神经纤维的保护作用

据报导孕酮可以加强髓鞘蛋白的表达量,稳定神经系统的白质。脊髓内移植背根节神经细胞后,其长出的神经纤维可长距离的在溃变神经纤维的支架上生长。为了研究脊髓损伤后孕酮对受损白质的溃变和再生神经纤维生长的影响,将C57小鼠脊髓胸8(Th8)节段皮质脊髓束横断后,将孕酮溶于芝麻油(10mg/ml)后皮下注射,隔日注射1次,持续2周采用丽春红2R-亮绿染色法显示皮质脊髓束在脊髓内的位置、形态。结果显示:皮下注射孕酮之后皮质脊髓束远端轴突更加完整,排列更加规则、致密,对照组中远端轴突碎裂严重,排列杂乱,纤维丢失。上述结果提示孕酮能够保护小鼠受损的皮质脊髓束远端轴突。     

The distribution of GAP-43 in normal rat spinal cord

Summary We have studied the distribution of the growth-associated protein GAP-43 in the spinal cord of adult rats by light and electron microscopy, using a new antiserum raised against GAP-43/-galactosidase fusion protein. We show that GAP-43 is present at all vertebral levels but is more concentrated in cervical and thoracic regions. In addition to heavy staining in the corticospinal tracts of the white matter, staining can be seen at the light microscopic level throughout the grey matter and is particularly heavy around the central canal and in the superficial dorsal horn. Electron microscopic examination revealed that GAP-43 immunostaining is confined to a subpopulation of axons and axon terminals. Staining occurs in small myelinated and unmyelinated fibres and in terminals which are mainly small and make single axodendritic or axosomatic synapses. Staining in such terminals occurs in the axoplasm but is heaviest immediately adjoining the axolema. Staining was not observed in dendrites, nor in large myelinated axons or large axon terminals. Our results indicate that GAP-43 is expressed in adult rat spinal cord in a subpopulation of small diameter fibres and axon terminals. The distribution and morphology of these terminals is consistent with several different possible origins including corticospinal projection neurons, small diameter primary afferent neurons, and descending raphe-spinal serotonin containing neurons.     

Collateralization of the cervical corticospinal tract in the rat

Anterograde staining with Phaseolus vulgaris leucoagglutinin (PHA-L) revealed the spinal arborization pattern of corticospinal tract (CST) fibers in the cervical enlargement of the rat. Within the confines of the pyramidal tract local nets of small fibers are present in addition to the rather large CST fibers with varicosities. CST termination is primarily located in lamina IV and extends into lamina V and VI. Extensive collateralization of CST axons was found interconnecting neurons located both in different horizontal laminae and in subsequent spinal cord segments. This complex pattern of CST collateralization is suggested to add a coordinative role in motor control to this tract both through serial axo-dendritic contacts in the spinal gray and through axo-axonal contacts in the white as well as the gray matter.     

Corticospinal tract fibers cross the ephrin-B3-negative part of the midline of the spinal cord after brain injury

The fibers of corticospinal tract (CST), which control fine motor function, predominantly project to the contralateral spinal cord, not recross to the ipsilateral side. Ephrin-B3, which is expressed in the midline of the spinal cord, and its receptor, EphA4, are crucial for preventing CST fibers from recrossing the midline in the developing spinal cord. However, these fibers can cross the midline to the denervated side after a unilateral CST or cortical injury. We determined the reason CST fibers can cross the midline after a cortical injury and the changes in ephrin-B3-EphA4 signaling associated with such a crossing. We first examined axonal sprouting from CST fibers after unilateral ablation of the motor cortex in postnatal and adult mice. CST fibers crossed the midline of the spinal cord after cortical ablation, especially when conducted during the early postnatal period. These fibers were well associated with functional recovery after the injury. We next assessed the mRNA expression of ephrin-B3 and EphA4 before and after the ablation. Surprisingly, no changes were detected in the expression patterns. We found, however, that ephrin-B3 expression in the ventral part of the midline disappeared after postnatal day 9 (P9), but was pronounced along the entire midline before P6. Most of the CST fibers crossed the midline through the ventral region, where ephrin-B3 expression was absent. Our results suggest that ephrin-B3 is not expressed along the entire midline of the spinal cord, and sprouting axons can cross the midline at ephrin-B3-negative areas.     

Two modes of corticospinal reinnervation occur close to spinal targets following unilateral lesion of the motor cortex in neonatal hamsters

Although it has been shown that unilateral neonatal cortical ablation induces bilateral corticospinal projections, the explanation for the pathways responsible for this bilateral innervation remains controversial. We hypothesized that such reinnervation may be supplied from newly formed fibers sprouting at the level rostral to, or at, or caudal to the pyramidal decussation. In order to test our hypothesis, we examined the brain and spinal cord of young hamsters which had a unilateral ablation of the right motor cortex at six days postnatally, and then received an injection of an anterograde neuronal lectin tracer, Phaseolus vulgaris-leucoagglutinin, into the hindlimb area of the left motor cortex at 21 days postnatally. For the identification of motoneurons in the lumbar spinal cord, some of these animals also received an injection of cholera toxin subunit B, a retrograde tracer, into the gastrocnemius muscle. A quantitative analysis in the left gray matter of the lumbar spinal cord indicated that the lectin labeling was two to eight times higher in cortically ablated animals than in intact animals. Immunohistochemical detection of the lectin revealed that innervation of the left spinal cord occurred close to targets at lower levels in the spinal cord. Two modes of reinnervation (types I and II) by the intact corticospinal tract were recognized. The type I fibers consisted of recrossing axon collaterals sprouted from the intact dorsal funiculus near their targets, while the type II fibers were recrossing parent axons which entered the intact, right gray matter several levels rostral to their targets, and then changed direction toward the targets. The recrossing at lower spinal levels yielded a large number of ipsilaterally labeled axons and their terminals in the gray matter of the denervated lumbar cord, with a distribution pattern similar to that seen on the intact side. The present results indicate that such ipsilateral innervation may play an important role in the sparing and recovery of function following neonatal hemicortical injury.