Understanding CNS remyelination: clues from developmental and regeneration biology

A guiding principle in remyelination research has been to seek clues to its nature in developmental studies on myelination. This "recapitulation hypothesis" argues that the regenerative response involves rerunning much the same programme as occurs during the developmental process. Here we examine the extent to which current evidence supports this hypothesis and whether this is a useful conceptual framework within which to study remyelination and suggest that an equally fruitful approach is to look to regenerative processes in other tissues.

Remyelination occurs as extensively but more slowly in old rats compared to young rats following gliotoxin-induced CNS demyelination

Age is one of the many factors that influence remyelination following CNS demyelination, although it is not clear whether it is the extent or rate of remyelination that is affected. To resolve this issue we have compared remyelination in young and old adult rat CNS following gliotoxin-induced demyelination. Remyelination of areas of ethidium bromide-induced demyelination in the caudal cerebellar peduncle reached completion by 4 weeks in young adult rats (2 months) but was not complete until 9 weeks in old adult rats (9-12 months). We have also shown that remyelination of lysolecithin-induced demyelination in the spinal white matter of old adult rats (18 months) can be extensive, with longer survival times (8 weeks) than have previously been examined. Thus, it is the rate rather than the extent of remyelination that changes in the ageing CNS. These results have important implications for understanding the mechanisms of remyelination, indicating that remyelination need not occur rapidly for it to be extensive. The capacity for the process of remyelination to continue over many weeks must also be borne in mind when assessing remyelination-enhancement strategies either by transplantation or promotion of endogenous mechanisms.

Myelination and behaviour of tenascin-C null transgenic mice

The extracellular matrix glycoprotein tenascin-C is widely expressed during development and repair, making it surprising that few abnormalities have been found in transgenic mice lacking this molecule. We have therefore re-examined the transgenic mice described by Saga et al. [Saga, Y., Yagi, T., Ikawa, Y., Sakakura, T. & Aizawa, S. (1992) Genes Dev., 6 1821-1831] in which tenascin-C was knocked-out by homologous recombination, focusing on two aspects of the nervous system likely to reveal any abnormalities that might follow the loss of tenascin-C. First, we have determined the pattern of myelin and distribution of oligodendrocyte precursor cells in those areas, such as the optic nerve and retina where local concentrations of tenascin-C have been proposed to act as barriers to oligodendrocyte precursor migration and so prevent inappropriate myelination. Secondly, we have examined the behaviour of the mice in a number of well-characterized tests, e.g. beam-walking, passive avoidance and the Morris water maze. We find no abnormalities of myelination or oligodendrocyte precursor distribution in adult mice, showing that local concentrations of tenascin-C are not the sole mechanism responsible for the pattern of myelination in these regions of CNS. However, we do find a number of behavioural abnormalities in these mice and show that hyperlocomotion and deficits in coordination during beam walking can be ascribed to tenascin-C deficiency. The effects on coordination are, however, not seen on a 129 genetic background. Taken together, these results significantly extend the phenotype associated with tenascin-C deficiency but argue against a role in myelination.

Distinctive patterns of PDGF-A, FGF-2, IGF-I, and TGF-beta1 gene expression during remyelination of experimentally-induced spinal cord demyelination

Although remyelination is a well-recognized regenerative process following both experimental and naturally occurring CNS demyelination, remarkably little is known about the molecules involved in its orchestration. In this study we have examined the mRNA expression of seven growth factors that influence oligodendrocyte lineage cells, during the remyelination of lysolecithin-induced demyelination in the rat spinal cord. These lesions involve rapid demyelination of axons, which undergo extensive remyelination between 10 and 28 days. The distribution and levels of expression of PDGF-A, IGF-I, CNTF, FGF-2, TGF-beta1, GGF-2, and NT-3 mRNAs were examined at 2, 5, 7, 10, 14, 21, and 28 days post-lesion induction, both within the lesion and within dorsal root ganglia whose axons transverse the lesion, by quantitative in situ hybridization using 35S-labeled oligonucleotide probes. large increases in IGF-I and TGF-beta1 mRNAs were evident within the spinal cord by 5 days. These levels peaked at 10 days at a time when new myelin sheaths appear and had declined by 28 days. Increases in FGF-2 and PDGF-A mRNAs were less intense and less widely distributed than those of IGF-I and TGF-1, but remained elevated for a longer duration. There were no changes in expression of CNTF, NT-3, or GGF-2 mRNAs within the lesioned cords; neither were ther changes in levels of expression of any growth factor mRNAs in the dorsal root ganglia. This work therefore indicates that some but not all members of the family of growth factors that affect the oligodendrocyte lineage are expressed during remyelination of demyelinated spinal cord axons and provides the data on which future studies on the specific roles of these factors in orchestrating this important regenerative process will be based.

The expression of myelin protein mRNAs during remyelination of lysolecithin-induced demyelination

To gain insights into the mechanisms of myelin repair in the CNS and to establish the extent to which this process resembles myelination in development we have examined the patterns of expression of transcripts of the major myelin proteins, myelin basic protein (MBP) and proteolipid protein (PLP) during remyelination of lysolecithin-induced demyelination in the adult rat spinal cord. Injection of 1 microliter 1% lysolecithin into the dorsal funiculus caused a dramatic decrease in levels of MBP exon 1 and MBP exon 2-containing transcripts and PLP/DM20 transcripts. Between 10 and 21 days post-lesion induction there was a gradual increase in levels of expression of all transcripts, which had returned to levels associated with normally myelinated spinal cord white matter at 21 days. These increases in levels of expression corresponded to the appearance of remyelinated axons, detected on toluidine blue-stained resin sections. Foci of high levels of expression occurred in regions of the lesion in which new myelin sheath formation was occurring, although the level of expression throughout the lesion never exceeded levels associated with myelin sheath maintenance in normal white matter due to the asynchronous pattern of remyelination. The changes in levels of expression of MBP exon 2 closely followed those of MBP exon 1. Our results indicate that (i) myelin protein gene expression associated with myelinogenesis during remyelination follows a similar pattern to that of myelinogenesis during development and that (ii) in rat models of demyelination changes of expression of MBP exon 1 and exon 2-containing transcripts are of equal value, an observation relevant to quantifying the effects of putative remyelination-enhancing strategies using the lysolecithin model.

Demyelination and remyelination of the caudal cerebellar peduncle of adult rats following stereotaxic injections of lysolecithin, ethidium bromide, and complement/anti-galactocerebroside: a comparative study

Experimentally induced demyelination due to the direct injection of gliotoxic agents has provided powerful models for studying the biology of remyelination. For the most part, these models have involved injection into white matter tracts of the spinal cord. However, the spinal cord has a number of limitations, such as the size of lesions that it is possible to make and its unsuitability for long-term direct cannulation for the delivery of putative remyelination-enhancing agents. In this study, we describe the natural history of three new models of demyelination/remyelination based on the stereotaxic injection of three gliotoxins: lysolecithin, ethidium bromide, and a combination of anti-galactocerebroside antibody and complement (GalC-ab/comp) into the caudal cerebellar peduncle of adult rats. All three agents produced large areas of demyelination with minimal axonal damage, which undergo extensive remyelination. Ethidium bromide- and GalC-ab/comp-induced lesions remyelinated more slowly than those induced by lysolecithin. The contribution to the remyelination of the lesion by Schwann cells reflects the degree of astrocyte damage incurred within the demyelinated area and is greatest for ethidium bromide-induced demyelination. These new models not only provide further insights into the mechanisms of CNS remyelination but also provide a valuable new resource for addressing a series of key issues relevant to current efforts to promote CNS remyelination either by the enhancement of intrinsic processes or by the transplantation of myelinogenic cells.

Trapidil-mediated inhibition of CNS remyelination results from reduced numbers and impaired differentiation of oligodendrocytes

In a previous study, we described the inhibitory effects of the growth factor-antagonist, trapidil, on spontaneously occurring oligodendrocyte remyelination in the rat spinal cord following lysolecithin-induced demyelination [30]. The objective of the present study was to further investigate the mechanisms of trapidil-mediated impairment of remyelination and thus obtain greater insight into the steps at which growth factors may be involved in remyelination. To this end, an ultrastructural analysis of the cellular composition of lesions from control and trapidil-treated animals was undertaken. Demyelination was created in the dorsal funiculus of 6-week-old female rats by the injection of 1.0 microliter of 1% lysolecithin. The animals received daily intraperitoneal injections of trapidil (80 mg/kg) or saline for 21 days, beginning on the day of lesion induction. Quantitative electron microscopic examination of lesions from both groups of animals showed that trapidil-treated lesions had reduced numbers of oligodendrocytes (P = 0.02) with a higher relative proportion of immature phenotypes, but increased numbers of microglia (P = 0.0009) and dystrophic axons (P0.02). In addition, the numbers of myelin lamellae around remyelinated axons were fewer in trapidil-treated animals. These results suggest that trapidil-mediated impairment of CNS remyelination is due to a blockage of growth factor-mediated proliferation and/or recruitment of remyelinating cells. Furthermore, the presence of oligodendrocytes with a more immature phenotype and the decreased thickness of the myelin sheaths of remyelination in the trapidil-treated animals indicate an impairment of growth factor-mediated differentiation.

Paralysis in hedgehogs (Erinaceus europaeus) associated with demyelination

Paraplegia affected 14 hedgehogs (Erinaceus europaeus) in a wildlife rescue hospital over a period of six months. Postmortem examination revealed demyelination in the brain and spinal cord and an inflammatory response in the meninges, choroid plexus and CNS. The peripheral nervous system was not affected. In the spleen, lungs and liver there was an accumulation of megakaryocytes and other evidence of extramedullary haemopoiesis, but there was no haematological evidence of anaemia. The pattern of disease incidence and the nature of the changes in the CNS suggest they were of viral origin, but no causal agent was isolated and the possibility of a neurotoxin cause cannot be ruled out.

Transplanting myelin-forming cells into the central nervous system: principles and practice

Although transplantation of myelin-forming cells into the central nervous system (CNS) has recently attracted much attention as a potential therapy for repairing persistent demyelination found in the demyelinating diseases such as multiple sclerosis and the leukodystrophies, it is worth remembering that the technique was originally conceived of as an experimental technique for manipulating in vivo environments to study interactions between different cell types in either repair or development. It is in this capacity that the technique is still predominantly used. Nevertheless, information, both technical and biological, that the continued use of the technique yields also often provides material for assessing the feasibility of glial cell transplantation as a therapeutic procedure. In this article, we describe some of the guiding principles of transplantation of myelinogenic cells into the mammalian CNS, focusing initially on the recipient environment and then considering the donor material. The division of the discussion into recipient and donor is one of convenience since in reality the interactions between the two cannot be considered in isolation.

The expression of myelin basic protein exon 1 and exon 2 containing transcripts during myelination of the neonatal rat spinal cord--an in situ hybridization study

During myelination, myelin proteins are expressed in a highly coordinated sequence. Of the four major isoforms of myelin basic protein, the proportion of the 21.5kDa and 17 kDa isoforms (which contain sequences encoded by myelin basic protein [MBP] exon 2) is enriched during active myelination in the mouse, suggesting that the alternative splicing of MBP transcripts containing exon 2 information is developmentally regulated. In this study, we compare the expression of MBP exon 1 and MBP exon 2 mRNAs in the neonatal rat spinal cord to establish whether developmental regulation of exon 2 mRNAs occurs in the rat in a manner similar to that previously described in the mouse. The expression of MBP mRNAs, together with that of proteolipid protein (PLP/DM-20) mRNA, in the developing white matter tracts increased dramatically between P7 and P10, corresponding to an increase in the extent of myelination. High levels of expression of each mRNA species examined were maintained between P10 and P14 as myelination proceeded. At P21, the expression of MBP exons 1 and 2 was reduced in the ventrolateral funiculi but was maintained at high levels in the dorsal funiculus. At P45, a further downregulation of both MBP mRNAs was apparent. By contrast, high levels of PLP/DM-20 expression were maintained from P10 onwards. In the grey matter, expression of MBP and PLP/DM-20 mRNAs increased more gradually and peak expression occurred later than in the white matter tracts. In this study, we therefore provide a description of myelin protein gene expression during post-natal development of the rat spinal cord. We have also shown that in the rat spinal cord, changes in the levels of MBP exon 2 expression associated with myelination reflect the changes of all MBP transcripts.

Peptidomimetic aminomethylene ketone inhibitors of interleukin-1 beta-converting enzyme (ICE)

Pyridone-based peptidomimetic inhibitors of recombinant human Interleukin-1 beta-converting enzyme (ICE, caspase-1) with an aminomethylene ketone activating group in the P1' position are described. Several analogues with sub-nanomolar Ki's versus ICE and improved aqueous solubility are reported.

Do olfactory glia have advantages over Schwann cells for CNS repair?

To a large extent the success of axon regeneration and sustained remyelination which distinguishes the PNS from the CNS is attributable to differences in their respective glial environments. For this reason, many have been attracted to the idea that repair of the CNS might be achieved by transplanting Schwann cells into areas of CNS pathology. Schwann cells will not only promote regeneration but will also myelinate axons thereby making them an appropriate cell type to mediate repair of lesions characterised by demyelination as well as axotomy. The recent discovery that olfactory glia are capable of forming myelin sheaths, together with their well-documented ability to support axon regeneration, means that these cells have a range of repair properties similar to that of Schwann cells. It is not clear at present which of these two alternatives, the Schwann cells or the olfactory glial cell, would be of greater benefit for achieving regeneration of axons or remyelination of persistent demyelination following transplantation into the CNS. In this article we review the repair properties of olfactory glia and identify the areas in which their use for repairing the CNS may have advantages over Schwann cells.

Local recruitment of remyelinating cells in the repair of demyelination in the central nervous system

The distance over which remyelinating cells within surrounding intact tissue are stimulated to respond to a demyelinating lesion and migrate toward it is unknown. To address this issue we have conducted a series of experiments in which the generation of remyelinating cells in tissue surrounding a spontaneously repairing area of demyelination induced in the adult rat spinal cord is suppressed by exposure to X-irradiation. By regulating the area of X-irradiation relative to the length of the demyelinating lesion within dorsal white matter we have shown that remyelinating cells are not recruited over distances greater than 2 mm into areas of demyelination, implying that most of the remyelinating cells are locally generated. This result indicates that there is only a narrow rim of normal tissue surrounding an area of demyelination from which remyelinating cells can be recruited. The depletion of cells within this rim may account for the poor remyelination associated with large areas of demyelination and following repeated episodes of demyelination. We have also shown that, in contrast to Schwann cells, oligodendrocyte lineage cells recruited into lesions have a limited ability to rapidly repopulate large areas of demyelination. Attempts to enhance remyelination in situations where it fails should therefore focus on increasing the size of the surrounding area from which remyelinating cells can be recruited by augmenting the level of recruitment signal, and preventing premature differentiation of oligodendrocytes so as to maximize their migratory and proliferative potential.

Remyelination in demyelinating disease

In multiple sclerosis, partial remyelination is conspicuous in many lesions, and is thought to contribute significantly to lasting recovery from acute relapse. However, myelin repair ultimately fails during progression of the disease, as disability and handicap accumulate. In this chapter we explore the biological background to myelin repair in CNS demyelinating disease, and the reasons underlying the failure of more widespread and lasting remyelination in multiple sclerosis. Experimental studies provide clear evidence that therapies promoting myelin repair can be highly successful in the CNS, and we discuss the clinical approaches which might allow the translation of these laboratory studies to neurological practice, together with some of the potential hazards and pitfalls likely to arise.

To what extent is oligodendrocyte progenitor migration a limiting factor in the remyelination of multiple sclerosis lesions?

In this article we describe a series of experimental approaches, involving the use of gliotoxin-induced demyelination, X-irradiation and glial cell transplantation, which examine the size of the area around demyelinating lesions from which new remyelinating cells are generated, and the distance over which they are able to migrate. Taken together, these studies suggest that the recruitment of remyelinating cells takes place over a very limited area and that long distance migration of remyelinating cells is not a feature of remyelination. The implications of these findings for spontaneous remyelination of multiple sclerosis plaques, and the development of strategies for enhancing remyelination are discussed.

Growth factors and remyelination in the CNS

It is now well established that there is an inherent capacity within the central nervous system (CNS) to remyelinate areas of white matter that have undergone demyelination. However this repair process is not universally consistent or sustained, and persistent demyelination occurs in a number of situations, most notably in the chronic multiple sclerosis (MS) plaque. Thus there is a need to investigate ways in which myelin deficits within the CNS may be restored. One approach to this problem is to investigate ways in which the inherent remyelinating capacity of the CNS may be stimulated to remyelinate areas of long-term demyelination. The expression of growth factors, which are known to be involved in developmental myelinogenesis, in areas of demyelination strongly suggests that they are involved in spontaneous remyelination. Therefore delivery of exogenous growth factors into areas of persistent demyelination is a potential therapeutic strategy for stimulating remyelination. This review will discuss the evidence that growth factors may have a role in promoting CNS remyelination by enhancing the survival and stimulating the proliferation and recruitment of remyelinating oligodendrocytes.

The effects of the growth factor-antagonist, trapidil, on remyelination in the CNS

In this study we describe the effects of trapidil, a putative platelet-derived growth factor-antagonist, on spontaneously occurring remyelination in rat spinal cord. Demyelination was created in the dorsal funiculus of 6- and 11-week-old female rats by the direct injection of 1.0 microliter of 1% lysolecithin. The animals received daily intra-peritoneal injections of either trapidil or saline for 21 days, commencing on the day of lesion induction. The 11-week-old rats receiving trapidil (60 mg/kg) showed a significant decrease in the extent of oligodendrocyte remyelination. Moreover, those axons that were remyelinated by oligodendrocytes tended to have thinner myelin sheaths than axons remyelinated by oligodendrocytes in the control group. In the 6-week-old group, the dose of trapidil which inhibited oligodendrocyte remyelination in the 11-week-old animals had a minimal effect on the extent of oligodendrocyte remyelination and no effect on the quality of myelin sheath formation. A higher dose of trapidil (80 mg/kg) was required before significant impairment of oligodendrocyte remyelination was achieved in the younger age group, implying an age-dependent effect of growth factor-inhibition of CNS remyelination. These results indicate an important role for growth factors, and in particular PDGF, in the orchestration of oligodendrocyte remyelination in the rodent CNS.

Contrasting effects of mitogenic growth factors on oligodendrocyte precursor cell migration

We have examined the effects of the mitogenic growth factors platelet derived growth factor (PDGF), basic fibroblast growth factor (bFGF) and glial growth factor-2 (GGF-2) on oligodendrocyte precursor migration. In an agarose drop migration assay PDGF and bFGF stimulated migration while GGF-2 had no effect. The migration-enhancing effect of bFGF cannot be blocked by neutralising antibodies against PDGF, confirming that this effect is direct and not mediated via upregulation of PDGF receptors. Based on our results, we propose a model in which the differing effects of PDGF and GGF-2 ensure appropriate numbers of oligodendrocyte precursor cells in the vicinity of axons to be myelinated during development.

Transplanting oligodendrocyte progenitors into the adult CNS

This review covers a number of aspects of the behaviour of oligodendrocyte progenitors following transplantation into the adult CNS. First, an account is given of the ability of transplanted oligodendrocyte progenitors, grown in tissue culture in the presence of PDGF and bFGF, to extensively remyelinate focal areas of persistent demyelination. Secondly, we describe how transplanted clonal cell lines of oligodendrocyte progenitors will differentiate into astrocytes as well oligodendrocytes following transplantation into pathological environments in which both oligodendrocytes and astrocytes are absent, thereby manifesting the bipotentially demonstrable in vitro but not during development. Finally, a series of studies examining the migratory behaviour of transplanted oligodendrocyte progenitors (modelled using the oliodendrocyte progenitor cell line CG4) are described. These show that CG4 cells do not survive (or migrate) when transplanted into the normal adult CNS. However, if they are transplanted into CNS tissue that has previously been exposed to 40 Gy of x-irradiation then transplanted CG4 cells survive, divide and migrate over large distances. Moreover, within an x-irradiated environment, migrating transplanted CG4 cells are able to enter remotely located foci of demyelination and contribute to the remyelination of the demyelinated axons within. These studies demonstrate that although the normal adult CNS does not appear to support survival and migration of the CG4 cell line, it is possible to manipulate the environment in such a way that these nonpermissive properties of the environment can be overcome.

Schwann cell-like myelination following transplantation of an olfactory bulb-ensheathing cell line into areas of demyelination in the adult CNS

In this study we have transplanted a clonal olfactory bulb-ensheathing cell line into focal areas of the rat spinal cord which contain demyelinated axons but neither oligodendrocytes nor astrocytes. The cell line was created by retroviral incorporation of the temperature-sensitive Tag gene into FACS-sorted 04+ cells from 7-day-old rat pup olfactory bulb. The spinal cord lesions were obtained by injecting small volumes of ethidium bromide into the dorsal white matter of spinal cord previously exposed to 40 Grays of X-irradiation. Many of the axons were remyelinated by PO+ myelin sheaths 21 days after transplantation. Light and electron microscopy revealed cells engaging and myelinating axons in a manner highly reminiscent of Schwann cells within similar lesions. GFAP+ cells were also present within the lesion. This study provides the first in vivo evidence that olfactory bulb-ensheathing cells are able to produce peripheral-type myelin sheaths around axons of the appropriate diameter.

Remyelination in the CNS of the hypothyroid rat

A number of studies have provided good evidence to indicate a role for thyroid hormone in myelination. Since myelination and remyelination have many shared objectives, and may therefore involve similar mechanisms, we examined whether thyroid hormone may also have a role in remyelination by both Schwann cells and oligodendrocytes of spinal cord axons that had been demyelinated by the injection of ethidium bromide in thyroidectomized rats. Neither the extent of oligodendrocyte remyelination nor the thickness of myelin sheath formed by remyelinating oligodendrocytes was affected by hypothyroidism. However, the extent of remyelination carried out by Schwann cells was decreased in hypothyroid rats compared with normal control animals.

Transplanted CG4 cells (an oligodendrocyte progenitor cell line) survive, migrate, and contribute to repair of areas of demyelination in X-irradiated and damaged spinal cord but not in normal spinal cord

In this study, we have examined the behavior of a lac-Z-transfected O- 2A progenitor cell line, CG4, following transplantation into normal and X-irradiated adult rat spinal cord, and we have also addressed the issue of whether CG4 cells transplanted remotely from ethidium bromide-induced demyelinating lesions in both X-irradiated and nonirradiated spinal cord are able to contribute to their repair. Following transplantation into X-irradiated spinal cord, CG4 cells survive, divide, and migrate extensively. The migration occurs mainly within the parenchymal tissue of the cord without preference for white or gray matter. Moreover, CG4 cells migrating away from their point of introduction are able to enter areas of demyelination and remyelinate the demyelinated axons therein. In contrast, when CG4 cells are transplanted into nonirradiated spinal cord, their survival is limited to areas of damage created by the injection procedure. The CG4 cells do not survive in undamaged, nonirradiated spinal cord. When transplanted remotely from areas of demyelination they are unable to traverse intervening areas of normal white matter, although they may enter lesions if transplanted into their close vicinity. These results have important implications for the development of potential therapeutic strategies for the treatment of multifocal demyelinating disorders that are based on glial cell transplantation.