Caudal spinal cord of the teleost Sternarchus albifrons resembles regenerating cord

The morphology of spinal cord in the caudal-most spinal segments of normal adult Sternarchus albifrons is different from that of more rostral adult cord. The caudal segments are strikingly similar to the regenerating spinal cord observed after amputation of the tail in Sternarchus. In the caudal-most vertebral segment of normal spinal cord, ependymal cells are radially enlarged and are more numerous than in more rostral adult cord. Large processes of the ependymal cells extend into the central canal, which also contains a prominent Reissner's fiber. Invaginations of the outer surface of the spinal cord, with the associated basal lamina, are common. Lateral to the immediate ependymal layer, extracellular spaces contain longitudinally oriented neurites. Cell bodies and cell processes filled with dense-cored vesicles occur throughout the caudal-most segment of spinal cord, and are especially concentrated in the ventral half, interspersed with numerous capillaries. In all these respects the caudal-most segments of normal adult spinal cord in Sternarchus closely resemble regenerating spinal cord of Sternarchus. In both regions, at least some of the ependymal cells retain the ability to divide and generate new neurons and glial cells.     

Cellular organization of the central canal ependymal zone,a niche of latent neural stem cells in the adult mammalian spinal cord

A stem cell's microenvironment, or “niche,” is a critical regulator of its behaviour. In the adult mammalian spinal cord, central canal ependymal cells possess latent neural stem cell properties, but the ependymal cell niche has not yet been described. Here, we identify important similarities and differences between the central canal ependymal zone and the forebrain subventricular zone (SVZ), a well-characterized niche of neural stem cells. First, direct immunohistochemical comparison of the spinal cord ependymal zone and the forebrain SVZ revealed distinct patterns of neural precursor marker expression. In particular, ependymal cells in the spinal cord were found to be bordered by a previously uncharacterized sub-ependymal layer, which is relatively less elaborate than that of the SVZ and comprised of small numbers of astrocytes, oligodendrocyte progenitors and neurons. Cell proliferation surrounding the central canal occurs in close association with blood vessels, but unlike in the SVZ, involves mainly ependymal rather than sub-ependymal cells. These proliferating ependymal cells typically self-renew rather than produce transit-amplifying progenitors, as they generate doublets of progeny that remain within the ependymal layer and show no evidence of a lineage relationship to sub-ependymal cells. Interestingly, the dorsal pole of the central canal was found to possess a sub-population of tanycyte-like cells that express markers of both ependymal cells and neural precursors, and their presence correlates with higher numbers of dorsally proliferating ependymal cells. Together, these data identify key features of the spinal cord ependymal cell niche, and suggest that dorsal ependymal cells possess the potential for stem cell activity. This work provides a foundation for future studies aimed at understanding ependymal cell regulation under normal and pathological conditions.     

Proliferation, migration, and differentiation of endogenous ependymal region stem/progenitor cells following minimal spinal cord injury in the adult rat

Ependymal cells of the adult mammalian spinal cord exhibit stem/progenitor cell properties following injury. In the present study, we utilized intraventricular injection of 1,1'-dioctadecyl-6,6'-di(4-sulfophenyl)-3,3,3',3'-tetramethylindocarbocyanine (DiI) to label the ependyma lining the central canal to allow tracking of the migration of endogenous ependymal cells and their progeny after spinal cord injury (SCI). We developed a minimal injury model that preserved the integrity of the central canal and did not interfere with ependymal cell labeling. Three days following SCI, there was an 8.6-fold increase in the proliferative labeling index of the ependymal cells at the level of the needle track based on bromodeoxyuridine labeling, compared with 1 day post-injury. Terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) positive cells were not detected in the ependyma or surrounding gray matter, indicating that ependymal cells do not undergo apoptosis in response to minimal injury. Nestin was rapidly induced in the ependyma by 1 day and expression peaked by 7 days post-injury. We quantitated the number and distance of ependymal cell migration following minimal injury. The number of ependymal cells migrating from the region of the central canal increased by 3 days following minimal injury and DiI-labeled glial fibrillary acidic protein expressing cells were detected 14 days post-SCI, most of which migrated within 70 microm of the region of the central canal. These results show that a minimal SCI adjacent to the ependyma is sufficient to induce an endogenous ependymal cell response where ependymal stem/progenitor cells proliferate and migrate from the region of the central canal, differentiating primarily into astrocytes.     

Evidence for the c-ret protooncogene product (c-Ret) expression in the spinal tanycytes of adult rat

Expression of the protein product of c-ret (c-Ret) in the spinal cord of the adult rat was examined immunohistochemically at both electron and light microscopic levels. In the cervical, thoracic and lumbar segments of the spinal cord, a large number of c-Ret immunoreactive cells were found in both ependymal and subependymal layers of the central canal. These cells were ovoid or triangular in shape and had a well developed single cytoprocess which protruded into the central canal. None of the neuropeptides and neuronal markers examined, including substance P, CGRP, galanin, neuropeptide Y, tyrosine hydroxylase, methionine-enkephalin, choline acetyltransferase and glially fibrilally acidic protein, was present in these c-Ret immunoreactive cells in the spinal cord. Ultrastructurally, a desmosome-like structure was found between the apical part of the cytoprocess and the ependymal cell. These morphological observations indicated that c-Ret positive cells are spinal tanycytes. The present results suggest that spinal tanycytes in the rat express a trophic factor receptor and may respond to GDNF in the cerebrospinal fluid.     

The ependymal and glial configuration in the spinal cord of urodeles

Summary The structural organization of the ependymal and macroglial components of the central field of the spinal cord of postmetamorphic ribbed newts has been reinvestigated using elaborate fixation procedures for transmission electron microscopy. All along the central canal, the ependymal cells display ultrastructural features that strongly suggest a secretory activity. Infrequent mitotic images, occurring spontaneously among the ependymal cells, were observed. The tightly compacted periependymal stratum contains two types of glial cells: 1. oligodendrocytes, also observed outside this stratum as neuronal satellites, and 2. radial astrocytic cells, whose somata, exclusively located in the periependymal stratum, send their processes to the subpial lamina. The intercellular relationships between ependyma, oligodendrocytes and astrocytic cells are illustrated to show the continuity of the neuroepithelial configuration. Morphologic clues for identifying the cells of the central field of the urodele spinal cord are given. A gradient of differentiation of the oligodendroglial components could be postulated. In normal conditions, the astroglial differentiation is permanently arrested at the stage of radial glia. Some considerations concerning regeneration in the urodele spinal cord are submitted.Supported by grant INSERM C.R.L. 76.4.047.6 to Prof. J.C. Lacroix, Université Pierre et Marie Curie. Paris     

Development and differentiation of human ependymal cells transplanted to the eye of immunodeficient nude rats

Human ependymal cells were investigated before and after intraocular xenografting of fetal spinal cord to nude rats. Although most cells in the embryonic ependymal layers were undifferentiated, poorly differentiated cells lined the primitive central canal of 8-week-old fetal spinal cord. Mature ependymal cells were found in two different structures in transplanted spinal cord after 8 to 12 months, one a central canal-like structure and the other a disorganized structure composed of packed cells and microcysts. The central canal-like structure contained typical columnar ependymal cells with microvilli, cilia, and junctional complexes (zonulae adherentes). Tanycytes were also observed in this structure. The structure composed of packed cells consisted of abnormal ependymal cells with intracellular microvilli and cilia arranged chaotically and adhering to each other by means of irregular junctions. Evidence of pluripotentiality of ependymal cells was not found, even in the disorganized structures. Similarities and differences in ependymal cell differentiation between grafts and tumors are also discussed. The microenvironment appears to be a factor in the induction of cellular polarity and normal development of ependymal cells.     

Vasoactive intestinal polypeptide (VIP) immunoreactivity in the ependymal cells of the rat spinal cord

Vasoactive intestinal polypeptide (VIP) was demonstrated immunohistochemically in the entire ependymal and subependymal cells in all levels (cervical: C, thoracic: T, lumbar: L and sacral: S) of normal adult rat spinal cord. The VIP-immunoreactive basal processes from the apical ependymal cells coursed dorsally or ventrally along the median plane and reached the pia mater of the dorsal and ventral median septa. Many VIP-immunoreactive basal processes terminated on the blood vessels in the neuropil around the central canal. A few microvilli of the ependymal cells that project into the central canal also demonstrated intense VIP immunoreactivity. These observations suggest that ependymal cells may be involved in the modulation of VIP levels in the cerebrospinal fluid and regulation of vascular tone of the blood vessels in the spinal cord.     

Absence of the central canal and its obliterative process in the house shrew spinal cord (Suncus murinus)

Through studies of the spinal cord in a large series of the developmental stages of embryos, infant and adult house shrews (Suncus murinus), the central canal provided with ependymal cells was, using the light and electron microscopes, observed to disappear. The canal was obliterated completely from the level of the area postrema to the end of the spinal cord in the infant and adult animals and remained unreformed, the cause being entirely due to spontaneous cell death of ependymal cells lining the central canal during days 22 of gestation to 5 d after birth. Cell degeneration was marked by an increase in electron density of the cytoplasm, lysosomes prominent increase in number, and in later stages, by necrotic fragmentation of ependymal cells, which fragments were phagocytized and digested by macrophages, and microglias were observed in the empty spaces.     

Observations on the caudal end of the spinal cord

The development of the structural pattern of the lower sacral and coccygeal segments of the spinal cord in human, rabbit and monkey embryos and fetuses has been studied. The changes observed in serial sections from above downward are outlined, beginning with typical sections through the lower sacral cord. Among the changes, other than diminution in size of the spinal cord and reduction in size of the lower spinal nerves, there is a gradual disappearance of the posterior funiculus. As this occurs the gray matter appears to spread dorsally and the central canal widens. The gray matter becomes reduced in size and the lateral funiculus extends farther dorsally. A little lower down, the gray matter of the alar plate is reduced further in size and there is corresponding enlargement of the central canal. This enlargement constitutes the terminal ventricle. The spinal cord rapidly becomes smaller as both the fibers and the gray matter are diminished. In some specimens, fibers decussate dorsal to the lower end of the terminal ventricle. Little remains of the lower end of the spinal cord except the ependymal wall of the central canal and the surrounding fiber bundles. The shape and size of the lower end of the central canal is subject to variations. In the lower part of the spinal cord a longitudinal bundle on each side is formed by fiber contributions from the anterior horn cells in the basal plates. This bundle contributes fibers to the fifth sacral and the first and second coccygeal nerves. It is designated the sacrococcygeal fasciculus.     

Novel observations on the origin of ependymal cells in the ventricular zone of the rat spinal cord

Despite extensive investigations of gliogenesis, the time of origin of ependymal cells in the spinal cord has not yet been fully elucidated. Using a single dose of 5-bromo-2-deoxyuridine combined with various survival times we monitored: mitotic activity (short survival time), the presence of newly formed cells in the ventricular zone (intermediate survival time) and the formation of ependymal cells (long survival time) during the late embryonic and early postnatal development in the ventricular zone of the spinal cord of rats. In the period of study it was found that the ependymal cells populated this region in two waves. Most of the ependymal cells originated around embryonic day 18 and then between postnatal days 8 and 15. In addition, it was observed that in the ventricular zone of the spinal cord, proliferation and production of ependymal cells continues at the slower rate at least until day 36 of postnatal development. Elucidation of the relationship between progenitors in the embryonic ventricular zone and the relative quiescent ependymal lining of the central canal in adulthood could be important in the search for the adult neural stem cell niche.     

Development of the glycogen body through the whole length of the chick spinal cord

The development of the glycogen body was studied throughout the entire length of the chick spinal cord. The glycogen body cells first appeared at stage 31 on each side of the ependymal septum from the 26th to 28th segments of the spinal cord. By stage 34 the paired primordia of the glycogen body extended from the 25th to 29th segments. In the middle of the structure described classically as the glycogen body (i.e., the portion restricted to the level of the spinal nerves 26–29), these primordia were fused dorsally at stage 34 and had fused completely by stage 37. The paired primordia extended from the cervical enlargement to the lumbosacral portion of the spinal cord by stage 36 and extended to the upper cervical segments by stage 38. They were totally fused throughout the entire length of the spinal cord by stage 42. The glycogen containing cells, in the classical glycogen body level, appeared ventrolateral to the central canal at stage 35. Thereafter they increased in number and glycogen content, and extended rostrad and caudad from the classical glycogen body level. They fused to each other and then fused with the glycogen body. Therefore, the bilateral clusters of the glycogen-containing cells are considered as the ventral paired primordia of the glycogen body. The development of the glycogen body is essentially the same pattern as in the classical glycogen body throughout the entire length of the spinal cord.     

Dorsal border periaqueductal gray neurons project to the area directly adjacent to the central canal ependyma of the C4-T8 spinal cord in the cat

In a previous study horseradish peroxidase (HRP) injections in the upper thoracic and cervical spinal cord revealed some faintly labeled small neurons at the dorsal border of the periaqueductal gray (PAG). The present light microscopic and electronmicroscopic tracing study describes the precise location of these dorsal border PAG-spinal neurons and their terminal organization. Wheat germ agglutinin-conjugated HRP (WGA-HRP) injections into cervical and upper thoracic spinal segments resulted in several hundreds of small retrogradely labeled neurons at the dorsal border of the ipsilateral caudal PAG. These neurons were not found after injections in more caudal segments. WGA-HRP injections in the dorsal border PAG region surprisingly resulted in anterogradely labeled fibers terminating in the area dorsally and laterally adjoining the central canal ependyma of the C4-T8 spinal cord. No anterogradely labeled fibers were found more caudal in the spinal cord. The labeled fibers found in the upper cervical cord were not located in the area immediately adjoining the ependymal layer of the central canal, but in the lateral part of laminae VI, VII and VIII and in area X bilaterally. Electronmicroscopic results of one case show that the dorsal border PAG-spinal neurons terminate in the neuropil of the subependymal area and in the vicinity of the basal membranes of capillaries located laterally to the central canal. The terminal profiles contain electron-lucent and densecored vesicles, suggesting a heterogeneity of possible transmitters. A striking observation was the lack of synaptic contacts, suggesting nonsynaptic release from the profiles. The function of the dorsal border PAG-spinal projection is unknown, but considering the termination pattern of the dorsal border PAG neurons on the capillaries the intriguing similarity between this projection system and the hypothalamohypophysial system is discussed.     

Ascending propriospinal afferents to area X (substantia grisea centralis) of the spinal cord in the rat

Area X (the tenth area) of the spinal cord is a region surrounding the central canal and extending throughout the spinal cord length. Using anterograde and retrograde labeling techniques, ascending propriospinal projections to area X were examined in the rat. For anterograde tracing of axons, biotinylated dextran was injected into middle-thoracic, lumbar, or sacral-caudal segments. Unilateral injections resulted in bilateral labeling of terminals in area X of all segments rostral to the injections. The distribution of labeled terminals was conspicuous in regions dorsal and lateral to the central canal. The labeled axons were derived from the ventrolateral and the lateral cord. They coursed through lamina VII, giving off terminal axons. While giving off terminal axons in area X, they coursed further rostrally or caudally along the central canal or crossed over the central canal to terminate in the contralateral area X. Possible cells of origin of these ascending afferents were examined after injections of wheat germ agglutinin-horseradish peroxidase into regions surrounding the central canal (area X) at the cervical or thoracic level. Retrogradely labeled neurons were consistently seen in area X, and laminae VII and VIII of the thoracic and lumbar segments. The present study shows that ascending propriospinal axons project to area X of all spinal levels rostral to the cells of origin and suggests that some of these afferents may originate from neurons in area X and laminae VII and VIII. Based on previous data, it is surmised that area X functions, through these intricate interconnections, as a site for integration or modulation of somatic or nociceptive and visceroceptive sensation. Received: 3 July 1997 / Accepted: 8 October 1997     

Neuronal nitric oxide synthase immunoreactivity in ependymal cells during early postnatal development

Neuronal nitric oxide synthase (nNOS) immunoreactivity was observed in ependymal cell layer of the central canal of spinal cord of neonatal rats (2-20 days old). Neuronal nitric oxide synthase immunoreactivity was present in postnatal day 2 and this immunoreactivity gradually disappeared by postnatal day 16. The progressive decrease in nNOS staining with the increasing postnatal age may suggest that nNOS staining paralleled the maturation of the central canal and may also suggest that nNOS activity plays a role in the development of the ependymal cells.     

The pathogenesis of spinal cord involvement in dengue virus infection

To investigate the mechanisms of dengue (DEN) virus transmission within the spinal cord, severe combined immunodeficient mice were intracerebrally inoculated with DEN virus type 2. After inoculation, a high virus titer and antigens were detected in the brain and spinal cord. At early stages of the infection, ultrastructural examinations showed that a few virions were present in the cytoplasm of ependymal cells lining the central canal. As the infection progressed, virions were observed in the lumen of the rough endoplasmic reticulum (RER), RER-derived vesicles and the Golgi region of infected neurons. These data suggest that the inoculated DEN virus might spread to the neurons of the spinal cord via the cerebral spinal fluid and cause several neuronal pathological responses. Moreover, DEN virus was also observed in myelinated and unmyelinated nerve fibers and typical neuronal synapses. Some virion-containing vesicles appeared to be fused with the membrane of presynapses, indicating that neuron-to-neuron transport of DEN virus might occur in the spinal cord. Additionally, anterior, lateral and posterior horns of the spinal cord exhibited different numbers of the positive neurons and different staining intensities of the DEN antigen during the infection. This difference likely represents variation of susceptibility to the DEN virus among the neurons of the spinal cord.     

Sequential infection of glial cells by the murine hepatitis virus JHM strain (MHV-4) leads to a characteristic distribution of demyelination.

An antigenic variant of the neurotropic murine coronavirus JHMV, designated 2.2-V-1, causes marked demyelination in the relative absence of encephalitis. It is thus useful for the study of the pathogenesis of demyelinating lesions. To better understand the sequential events leading to demyelination, we have examined murine brain and spinal cord tissue at daily intervals after intracerebral inoculation, evaluating them for the distribution of viral antigen, leukocyte infiltration, and demyelination. Immunohistochemical staining indicated that virus established primary infection in the ependymal cells in both brain and spinal cord before spreading into nearby structures and along white matter tracts by cell-to-cell contact. Spread from brain to spinal cord appeared to occur via cerebrospinal fluid. Viral replication was focally cytocidal for ependymal cells, and essentially noncytocidal for other neural cells including glia. In brain, viral antigen and inflammation reached a peak at day 5 postinfection, and rapidly subsided by day 10 postinfection. In spinal cord, viral antigen was less abundant than in brain and was maximal between days 7 and 9 postinfection. The inflammatory response and demyelination, however, were more severe persisting from day 7 through day 19. In the spinal cord, demyelinating lesions developed initially in areas closer to the central canal and were detected most prominently in the anterior funiculi. This finding suggests that the permissiveness of the ependymal cell is crucial to viral entry and that sequential infection of glial cells leads to the characteristic distribution of demyelination.     

The pathogenesis of spinal cord involvement in the encephalomyelitis of mice caused by neuroadapted Sindbis virus infection

Weanling mice develop an acute encephalomyelitis with high mortality after intracerebral inoculation of neuroadapted Sindbis virus. The mice develop kyphoscoliosis and hindlimb paralysis. Immunohistochemical and in situ hybridization studies have demonstrated virus in the gray matter of the brain and spinal cord. Ventral horn cells are prominently infected, providing an anatomical basis for the clinical poliomyelitis. A novel route of spread of inoculated virus within the central nervous system has been found. The virus enters the ventricular system, and then travels caudally in the central canal of the spinal cord where ependymal cells are infected. The virus subsequently spreads into the gray matter. The distribution of virus in the spinal cord is likely dependent both on variations in the susceptibility of neural cells and on this route of entry and subsequent spread.     

Fine structure of regenerated ependyma and spinal cord in Sternarchus albifrons

The caudal-most regenerated spinal cord in Sternarchus albifrons consists solely of an ependymal tube. Ependymal cells are enlarged radially and are more numerous than in unregenerated cord. Projections of ependymal cell cytoplasm and Reissner's fiber fill most of the central canal. Small groups of neurites and cell processes filled with dense-cored vesicles lie between abluminal processes of ependymal cells. Rostral to this, additional cells appear dorsal and lateral to the inner ependymal layer. Some cell bodies contain numerous dense-cored vesicles. Larger bundles of neurites, some with synapses, are present. Invaginations of the peripheral edge of the cord create enclosed spaces lined with basal lamina. In the peripheral region, longitudinally oriented neurites extend through extracellular spaces or channels. The ventral portion at some levels of regenerated cord is completely filled with neurites, processes containing dense-cord vesicles, and capillaries. Similar masses of neurites and processes containing dense-cored vesicles lie outside the cord proper, in or near the meningeal layer. In rostral-most sections, the organization of regenerated spinal cord approaches that of normal cord, with the regenerated cord exhibiting groups of myelinated axons, differentiated fibrous astrocytes and oligodendroglia, cell bodies containing dense-cored vesicles, and differentiated electromotor neurons. These observations indicate a degree of pluripotency in some of the ependymal cells in adult Sternarchus. Moreover, they are consistent with a role of ependymal cells in the guidance of regenerating neurites.     

Ependymal cells of the filum terminale in fish (poecilia sphenops) adapted to freshwater and saltwater: Electron microscopic study

The ependymal lining of the central canal of the filum terminale and spinal cord in the vicinity of the caudal neurosecretory system in P. sphenops was examined in this study. Two general cell types based on shape and location were observed in the ependymal lining: cuboidal ependyma located in dorsal aspects of the filum terminal and columnar to pseudostratified ependymal cells found in ventrolateral and ventral aspects of the filum terminale. Comparison of the ependymal lining was made in animals adapted to saltwater and freshwater. In animals adapted to saltwater there was an increase in the basal infoldings of the cell membrane of the dorsal cuboidal ependyma. Infolding of the basal cell membrane is a phenomenon shared by cells known to participate in transport of electrolytes. Since a possible functional relationship between the ependyma of the third ventricle and median eminence has been shown, in future studies on the osmoregulatory function of the caudal neurosecretory system the ependymal lining of the central canal in this region should be considered.