The remyelinating potential and in vitro differentiation of MOG-expressing oligodendrocyte precursors isolated from the adult rat CNS.

There is a long-standing controversy as to whether oligodendrocytes may be capable of cell division and thus contribute to remyelination. We recently published evidence that a subpopulation of myelin oligodendrocyte glycoprotein (MOG)-expressing cells in the adult rat spinal cord co-expressed molecules previously considered to be restricted to oligodendrocyte progenitors [G. Li et al. (2002) Brain Pathol., 12, 463-471]. To further investigate the properties of MOG-expressing cells, anti-MOG-immunosorted cells were grown in culture and transplanted into acute demyelinating lesions. The immunosorting protocol yielded a cell preparation in which over 98% of the viable cells showed anti-MOG- and O1-immunoreactivity; 12-15% of the anti-MOG-immunosorted cells co-expressed platelet-derived growth factor alpha receptor (PDGFRalpha) or the A2B5-epitope. When cultured in serum-free medium containing EGF and FGF-2, 15-18% of the anti-MOG-immunosorted cells lost anti-MOG- and O1-immunoreactivity and underwent cell division. On removal of these growth factors, cells differentiated into oligodendrocytes, or astrocytes and Schwann cells when the differentiation medium contained BMPs. Transplantation of anti-MOG-immunosorted cells into areas of acute demyelination immediately after isolation resulted in the generation of remyelinating oligodendrocytes and Schwann cells. Our studies indicate that the adult rat CNS contains a significant number of oligodendrocyte precursors that express MOG and galactocerebroside, molecules previously considered restricted to mature oligodendrocytes. This may explain why myelin-bearing oligodendrocytes were considered capable of generating remyelinating cells. Our study also provides evidence that the adult oligodendrocyte progenitor can be considered as a source of the Schwann cells that remyelinate demyelinated CNS axons following concurrent destruction of oligodendrocytes and astrocytes.

Decline in rate of colonization of oligodendrocyte progenitor cell (OPC)-depleted tissue by adult OPCs with age.

Rates of remyelination decline with age and this has been attributed to slower recruitment of oligodendrocyte progenitor cells (OPCs) into areas of demyelination and slower differentiation of OPCs into remyelinating oligodendrocytes. When considering causes for reduced recruitment rates, intrinsic causes (alterations in biological properties of OPCs) need to be separated from extrinsic causes (age-related differences in the lesion environment). Using 40 Gy of X-irradiation to deplete tissue of its endogenous OPC-population, we examined the effects of age on the rate at which adult rat OPCs colonize OPC-depleted tissue. We found a significant reduction in the rate of colonization between 2 and 10 months of age (0.6 mm/week versus 0.38 mm/week). To determine if this represented an intrinsic property of OPCs or was due to changes in the environment that the cells were recolonizing, OPCs from 10-month-old animals were transplanted into 2-month-old hosts and OPCs from 2-month-old animals were transplanted into 10-month-old hosts. These experiments showed that the transplanted OPCs retained their age-related rate of colonization, indicating that the decline in colonizing rates of OPCs with age reflects an intrinsic property of OPCs. This age-related decline in the ability of OPCs to repopulate OPC-depleted tissue has implications for understanding remyelination failure in multiple sclerosis (MS) and developing therapies for remyelination failure.

Modelling large areas of demyelination in the rat reveals the potential and possible limitations of transplanted glial cells for remyelination in the CNS.

Transplantation of myelin-forming glial cells may provide a means of achieving remyelination in situations in which endogenous remyelination fails. For this type of cell therapy to be successful, cells will have to migrate long distances in normal tissue and within areas of demyelination. In this study, 40 Gy of X-irradiation was used to deplete tissue of endogenous oligodendrocyte progenitors (OPCs). By transplanting neonatal OPCs into OPC-depleted tissue, we were able to examine the speed with which neonatal OPCs repopulate OPC-depleted tissue. Using antibodies to NG-2 proteoglycan and in situ hybridisation to detect platelet-derived growth factor alpha-receptor Ralpha (PDGFRalpha) mRNA to visualise OPCs, we were able to show that neonatal OPCs repopulate OPC-depleted normal tissue 3-5 times more rapidly than endogenous OPCs. Transplanted neonatal OPCs restore OPC densities to near-normal values and when demyelinating lesions were made in tissue into which transplanted OPCs had been incorporated 1 month previously, we were able to show that the transplanted cells retain a robust ability to remyelinate axons after their integration into host tissue. In order to model the situation that would exist in a large OPC-depleted area of demyelination such as may occur in humans; we depleted tissue of its endogenous OPC population and placed focal demyelinating lesions at a distance (< or =1 cm) from a source of neonatal OPCs. In this situation, cells would have to repopulate depleted tissue in order to reach the area of demyelination. As the repopulation process would take time, this model allowed us to examine the consequences of delaying the interaction between OPCs and demyelinated axons on remyelination. Using this approach, we have obtained data that suggest that delaying the time of the interaction between OPCs and demyelinated axons restricts the expression of the remyelinating potential of transplanted OPCs.

Remyelination of demyelinated CNS axons by transplanted human schwann cells: the deleterious effect of contaminating fibroblasts.

Areas of demyelination can be remyelinated by transplanting myelin-forming cells. Schwann cells are the naturally remyelinating cells of the peripheral nervous system and have a number of features that may make them attractive for cell implantation therapies in multiple sclerosis, in which spontaneous but limited Schwann cell remyelination has been well documented. Schwann cells can be expanded in vitro, potentially affording the opportunity of autologous transplantation; and they might also be spared the demyelinating process in multiple sclerosis. Although rat, cat, and monkey Schwann cells have been transplanted into rodent demyelinating lesions, the behavior of transplanted human Schwann cells has not been evaluated. In this study we examined the consequences of injecting human Schwann cells into areas of acute demyelination in the spinal cords of adult rats. We found that transplants containing significant fibroblast contamination resulted in deposition of large amounts of collagen and extensive axonal degeneration. However, Schwann cell preparations that had been purified by positive immunoselection using antibodies to human low-affinity nerve growth factor receptor containing less than 10% fibroblasts were associated with remyelination. This result indicates that fibroblast contamination of human Schwann cells represents a greater problem than would have been appreciated from previous studies.

Remyelination of Demyelinated CNS Axons by Transplanted Human Schwann Cells: The Deleterious Effect of Contaminating Fibroblasts.

Areas of demyelination can be remyelinated by transplanting myelin-forming cells. Schwann cells are the naturally remyelinating cells of the peripheral nervous system and have a number of features that may make them attractive for cell implantation therapies in multiple sclerosis, in which spontaneous but limited Schwann cell remyelination has been well documented. Schwann cells can be expanded in vitro, potentially affording the opportunity of autologous transplantation; and they might also be spared the demyelinating process in multiple sclerosis. Although rat, cat, and monkey Schwann cells have been transplanted into rodent demyelinating lesions, the behavior of transplanted human Schwann cells has not been evaluated. In this study we examined the consequences of injecting human Schwann cells into areas of acute demyelination in the spinal cords of adult rats. We found that transplants containing significant fibroblast contamination resulted in deposition of large amounts of collagen and extensive axonal degeneration. However, Schwann cell preparations that had been purified by positive immunoselection using antibodies to human low-affinity nerve growth factor receptor containing less than 10% fibroblasts were associated with remyelination. This result indicates that fibroblast contamination of human Schwann cells represents a greater problem than would have been appreciated from previous studies.

Schwann cells transplanted into normal and X-irradiated adult white matter do not migrate extensively and show poor long-term survival.

Although Schwann cells are able to enter the central nervous system (CNS) when the integrity of the glia limitans is disrupted, their ability to migrate through intact CNS remains unclear. We have addressed this issue by transplanting lacZ-labeled Schwann cells into normal adult spinal cord white matter, and into X-irradiated spinal cord (an environment that, unlike normal spinal cord, permits the migration of transplanted oligodendrocyte progenitors). Schwann cell cultures, obtained from neonatal rat sciatic nerve and expanded using bovine pituitary extract and forskolin, were transfected by repeated exposure to retroviral vectors encoding the Escherichia coli lacZ gene. The normal behavior of the transduced cells was confirmed by transplantation into a nonrepairing area of demyelination in the spinal cord, where they formed myelin sheaths around demyelinated axons. A single microliter containing 4 x 10(4) cells was then transplanted into unlesioned normal and X-irradiated white matter of the spinal cord of adult syngeneic rats. One hour after injection, blue cells were observed as a discrete mass within the dorsal funiculus with a longitudinal distribution of 2-3 mm, indicating the extent of passive spread of the injected cells. At subsequent survival times (1, 2, and 4 weeks posttransplantation) blue cells had a distribution that was no more extensive than that seen 1 h after transplantation. However, the number of Schwann cells declined with time following transplantation such that at 4 weeks there were few surviving Schwann cells in both X-irradiated and nonirradiated spinal cord. These results indicate that transplanted Schwann cells do not migrate extensively and show poor long-term survival when introduced into a normal CNS environment.

Transplanted glial cells migrate over a greater distance and remyelinate demyelinated lesions more rapidly than endogenous remyelinating cells.

Glial cell transplantation offers a means of remyelinating areas of demyelination in situations where endogenous remyelination fails. How effective such a strategy would be if undertaken in human demyelinating disease is not yet clear since it is very difficult to create large areas of demyelination in adult rodents that would mimic the situation found in a human disease such as multiple sclerosis. When CNS tissue is subjected to 40 Grays of X-irradiation, remyelination is suppressed in the X-irradiated area unless cells migrate into, or are introduced into the X-irradiated area. In the present experiments, by appropriate positioning of lead shielding we have created a "starting gate" from which oligodendrocyte progenitors must depart in order to colonise areas of demyelination. When the starting gate is located at one end of the area of demyelination, endogenous cells fail to colonise throughout an area of demyelination over the ensuing month. In contrast, when transplanted oligodendrocyte precursors are faced with the same situation, the whole area of demyelination is remyelinated over the same period. To determine how far transplanted cells can migrate to areas of demyelination and also to study how quickly the cells can colonise areas of demyelination we injected cells at some distance from areas of demyelination made in X-irradiated tissue. In these experiments, we found that transplanted cells could repopulate up to 9 mm in 2 months compared to 4 mm recorded for endogenous cells (Franklin et al. [1997] J. Neurosci. Res. 50:337-344). These experiments demonstrate that transplanted cells have a far greater ability to colonise areas of demyelination than endogenous cells.

Redistribution of bisbenzimide Hoechst 33342 from transplanted cells to host cells.

Identification of transplanted cells within host tissue is an important component of many transplantation experiments. In this study, Schwann cells labelled with the fluorochrome bisbenzimide (Hoechst 33342, H33342) and transduced with the lac-Z gene were introduced into normal white matter and their distribution was examined 2 h, 24 h and 4 weeks after transplantation. At 2 and 24 h following transplantation, H33342-labelled cells were more widely distributed than lac-Z-labelled cells in both longitudinal and transverse directions. By 4 weeks following transplantation, no lac-Z-labelled cells could be found. However, H33342-labelled cells were observed in and around the glial scar. Therefore, labelling of host cells by transfer of H33342 dye from transplanted cells has to be considered whenever this dye is used as a transplant marker.

Behavioural consequences of oligodendrocyte progenitor cell transplantation into experimental demyelinating lesions in the rat spinal cord.

Glial cell transplantation has the potential to be developed into a clinical treatment for human demyelinating diseases because of its demonstrated efficacy in remyelinating experimentally demyelinated axons. As a step towards clinical application it is necessary to demonstrate that the procedure is safe and efficacious in promoting behavioural recovery. In this study we transplanted glial cell progenitors into demyelinating lesions induced by intraspinal injection of ethidium bromide in the rat. Locomotor function after transplantation was assessed using a beam-walking test that has previously been shown able to detect deficits associated with demyelination in the dorsal funiculus of the rat spinal cord. Two groups of animals with transplants were examined. In one group, spontaneous remyelination was prevented by exposure of the lesion to 40 Gy of X-irradiation; in the other, male glial cells were transplanted into nonirradiated female recipients, permitting their identification by use of a probe specific to the rat Y chromosome. Following transplantation, there was severe axon loss in a large proportion of the irradiated animals and those affected did not recover normal behavioural function. In contrast, both the small proportion of the irradiated group that sustained only mild axon loss and the nonirradiated recipients of transplants recovered normal function on our behavioural test. We conclude that glial cell transplantation is able to reverse the functional deficits associated with demyelination, provided axonal loss is minimal.

The demonstration by transplantation of the very restricted remyelinating potential of post-mitotic oligodendrocytes.

To examine the remyelinating ability of post-mitotic oligodendrocytes, we subjected cell preparations derived from neonatal and adult rats to 40 Grays of X-irradiation to remove mitotically active cells and injected them into areas of demyelination in which the inherent ability to generate remyelinating cells had been inhibited. The extensive remyelination seen following implantation of non-irradiated neonatal and adult cells was almost completely abolished when the transplanted cell suspension was exposed to 40 Grays of X-irradiation, demonstrating that effective remyelination requires the generation of cells by mitosis. Radiation-resistant and therefore non-dividing oligodendrocytes were detected in areas of demyelination following transplantation of neonatal cultures and oligodendrocyte preparations derived from the adult nervous system. However, the pattern of myelin formation associated with the radiation-resistant oligodendrocytes from the two sources was different. Following implantation of X-irradiated neonatal cultures, a small number of oligodendrocytes could be found within the area of demyelination, and although these cells formed sheets of myelin membrane, they did not form myelin sheaths. After implantation of X-irradiated adult cells, in addition to the aberrant myelin formation seen with the neonatal cells, some myelin sheaths were observed. Our findings confirm that effective remyelination requires cell division and suggest that there may be diverse populations of radiation-resistant oligodendrocytes in the adult nervous system, some of which can form myelin sheaths and others of which can only make myelin sheets. Important for the interpretation of our previous studies is the demonstration here that 40 Grays of X-irradiation per se does not inhibit oligodendrocytes from remyelinating axons.

Attempts to produce astrocyte cultures devoid of oligodendrocyte generating potential by the use of antimitotic treatment reveal the presence of quiescent oligodendrocyte precursors.

The presence of oligodendrocyte precursor cells which cannot be removed from primary cultures by antibody-dependent techniques complicates the interpretation of transplantation experiments designed to examine the potential of astrocytes to influence remyelination (Blakemore et al.; Glia 13:79-91, 1995). In the present series of experiments we have investigated the use of the antimitotic cytosine arabinoside to eliminate oligodendrocyte precursors from mixed glial cell cultures following immunolytic removal of both oligodendrocytes and oligodendrocyte progenitors using A2B5 and O4 monoclonal antibodies. Our results indicate that not all oligodendrocyte precursors are involved during the subsequent regeneration of oligodendrocytes since a population of precursors survive 3-day and 12-day exposure to cytosine arabinoside. Maintaining immunolysed cultures in serum-free medium containing cytosine arabinoside for 23 days, removed the potential to generate large clones of oligodendrocytes both in vitro and following transplantation. However, a small number of oligodendrocyte precursors survived this treatment and generated single oligodendrocytes in vitro and isolated clusters of oligodendrocyte remyelination following transplantation. Overall, these results indicate that oligodendrocyte precursors have considerable potential to generate oligodendrocytes, but, since they can also survive for long periods in a quiescent state, their complete elimination from immunolysed astrocyte cultures by the use of an antimitotic is unreliable, if not impossible.

Glial cell transplants that are subsequently rejected can be used to influence regeneration of glial cell environments in the CNS.

Transplantation of glial cells into demyelinating lesions in CNS offers an experimental approach which allows investigation of the complex interactions that occur between CNS glia, Schwann cells, and axons during remyelination and repair. Earlier studies have shown that 1) transplanted astrocytes are able to prevent Schwann cells from participating in CNS remyelination, but that they are only able to do so with the cooperation of cells of the oligodendrocyte lineage, and 2) transplanted mouse oligodendrocytes can remyelinate rat axons provided their rejection is controlled by immunosuppression. On the basis of these observations, we have been able to prevent the Schwann cell remyelination that normally follows ethidium bromide demyelination in the rat spinal cord by co-transplanting isogeneic astrocytes with a potentially rejectable population of mouse oligodendrocyte lineage cells. Since male mouse cells were used it was possible to demonstrate their presence in immunosuppressed recipients using a mouse Y-chromosome probe by in situ hydridisation. When myelinating mouse cells were rejected by removal of immunosuppression, the demyelinated axons were remyelinated by host oligodendrocytes rather than Schwann cells, whose entry was prevented by the persistence of the transplanted isogeneic astrocytes. The oligodendrocyte remyelination was extensive and rapid, indicating that the inflammation associated with cell rejection did not impede repair. If this host oligodendrocyte remyelination was prevented by local X-irradiation, the lesion consisted of demyelinated axons surrounded by processes from the transplanted astrocytes. By this approach, it was possible to create an environment which resembled the chronic plaques of multiple sclerosis. Thus, these experiments demonstrate that in appropriate circumstances the temporary presence of a population of glial cells can alter the outcome of damage to the CNS.

Repair of demyelinated lesions by glial cell transplantation.

Transplantation of oligodendrocyte lineage cells results in myelination of naked axons. The most extensive remyelination is achieved using A2B5 positive progenitor cells and these observations encourage the view that glial transplantation may be a useful treatment in human demyelinating diseases. However, several issues need to be resolved before glial cell transplantation can be applied clinically; this review focuses on the choice of cells, their source, whether chronically demyelinated axons can be repaired by transplantation, and whether the principles of glial transplantation established in small rodents are applicable to other mammalian species and to man.

In vitro and in vivo characterisation of glial cells immortalised with a temperature sensitive SV40 T antigen-containing retrovirus.

An oncogene-carrying replication-defective retrovirus was used to establish immortalised lines of murine glial cells. Primary cultures of early postnatal cerebellar cells were infected with a retrovirus based on the Murine Moloney Leukemia Virus containing a temperature-sensitive mutant of the Simian Virus 40 large T antigen (SV40 T) oncogene and a gene coding for resistance to the antibiotic G418. Infected cells were selected in G418 and after several in vitro passages cells expressing the O4 antigen were established as a cell line. At a later time point O4-positive single-cell clones were established. Two different types of clones were obtained: 1) "plastic" clones consisting of cells which initially had a morphological and antigenic phenotype of young glial precursor cells but which gradually lost these features, and 2) "stable" cell clones including a clone with the immunological and electrophysiological characteristics of Schwann cells. Culture of the latter cells in the presence of 1 mM dibutyryl cyclic adenosine monophosphate for a period of at least 10 days induced a change in shape and a shift in antigen expression towards a more "differentiated" maturation stage. When the SV40 T O4-positive immortalised cell line isolated on the cell sorter was transplanted into demyelinated lesions in adult rats, cells were observed ensheathing axons and forming limited amounts of PNS-type myelin. Glial cells immortalised with a temperature-sensitive mutant of the SV40 T oncogene thus retain many physiological properties of their primary culture counterparts and can be induced to undergo limited differentiation in vitro and in vivo. These cell lines, which represent immature CNS glia or Schwann cells, are providing useful tools for investigating the role of cell surface antigens involved in neuron-glial interactions.

Lines of glial precursor cells immortalised with a temperature-sensitive oncogene give rise to astrocytes and oligodendrocytes following transplantation into demyelinated lesions in the central nervous system.

Immortalised lines of murine glial precursor cells expressing the neomycin resistance gene and a temperature-sensitive mutation of the SV 40 T oncogene were established from cultures of oligodendrocytes and precursor cells infected with a replication-incompetent, helper-free retrovirus. At the permissive temperature (33 degrees C), they could be continually propagated in vitro and cells were present expressing the 04 antigen specific for glial precursor cells and oligodendrocytes. At 38 degrees C, where the expression of the T antigen is down regulated, cell division largely ceased. During early passage in vitro, limited differentiation to a more mature phenotype, as evidenced by expression of GFAP and the oligodendrocyte marker 01 was observed at both 33 degrees C and 38 degrees C. When transplanted into demyelinating lesions in the spinal cords of adult rats early passages of the lines yielded myelin-forming oligodendrocytes and astrocytes. Cells from later passages of the lines although failing to synthesise myelin still associated specifically with the demyelinated axons. These experiments demonstrate the retention of physiological properties of these oncogene-carrying glial cells when transplanted in vivo and suggest that such immortalised populations can be used for the isolation of molecules regulating glial cell function.

The reconstruction of an astrocytic environment in glia-deficient areas of white matter.

Injection of ethidium bromide into X-irradiated spinal cord white matter produces a lesion in which demyelinated axons reside in an environment that is permanently depleted of glial cells. By transplanting defined populations of glial cells into this lesion it is possible to recreate normal or novel glial environments. In this study we have transplanted cultures of astrocytes into the X-irradiated ethidium bromide lesion in order to (1) assess the ability of these cells to relate to components within the lesion environment and thereby contribute to tissue reconstruction and (2) establish an astrocytic environment around demyelinated axons that resembles pathological states such as the chronic demyelinated plaques of multiple sclerosis. In order to focus attention on the interactions between astrocytes and demyelinated axons we developed a protocol for depleting astrocyte cultures of oligodendrocyte lineage cells and Schwann cells based on complement-mediated immunocytolysis and in vitro X-irradiation. In addition to establishing the ability of transplanted astrocytes to form an astrocytic matrix around demyelinated axons, this study has also revealed the diversity of cell types present within neonatal forebrain cultures.

Repair of demyelinated lesions by transplantation of purified O-2A progenitor cells.

The transplantation of well defined populations of precursor cells offers a means of repairing damaged tissue and of delivering therapeutic compounds to sites of injury or degeneration. For example, a functional immune system can be reconstituted by transplantation of purified haematopoietic stem cells, and transplanted skeletal myoblasts and keratinocytes can participate in the formation of normal tissue in host animals. Cell transplantation in the central nervous system (CNS) has been proposed as a means of correcting neuronal dysfunction in diseases associated with neuronal loss; it might also rectify glial cell dysfunction, with transplanted oligodendrocyte precursor cells eventually allowing repair of demyelinating damage in the CNS. Here we use co-operating growth factors to expand purified populations of oligodendrocyte type-2 astrocyte (O-2A) progenitor cells for several weeks in vitro. When injected into demyelinating lesions in spinal cords of adult rats, created in such a way as to preclude host-mediated remyelination, these expanded populations are capable of producing extensive remyelination. In addition, transplantation of O-2A progenitor cells genetically modified to express the bacterial beta-galactosidase gene gives rise to beta-galactosidase-positive oligodendrocytes which remyelinate demyelinated axons within the lesion. These results offer a viable strategy for the manipulation of neural precursor cells which is compatible with attempts to repair damaged CNS tissue by precursor transplantation.

The differentiation of glial cell progenitor populations following transplantation into non-repairing central nervous system glial lesions in adult animals.

The non-repairing nature of the locally x-irradiated ethidium bromide (EB)-induced demyelinating white matter lesion has been further validated by showing that injections of two cultures which promote host remyelination of EB lesions in normal tissue do not do so in x-irradiated lesions. The behaviour of an oncogene-immortalized glial cell line and a growth-factor-expanded glial progenitor population have been examined following transplantation into the non-repairing EB lesion. Our studies indicate that the selected glial cell populations were each capable of establishing glial environments around demyelinated axons. Extensive oligodendrocyte remyelination with little astrocytic presence was observed in lesions transplanted with growth-factor-expanded optic nerve progenitors, while less extensive oligodendrocyte remyelination with the establishment of astrocyte-like cells was found in lesions transplanted with ts A58-SV40T immortalized glial cells. Prolonged expansion of both populations resulted in a loss of differentiation to normal glial phenotypes.

The behaviour of meningeal cells following glial cell transplantation into chemically-induced areas of demyelination in the CNS.

Following transplantation of cultures of CNS glia containing leptomeningeal cells into ethidium bromide-induced demyelinating lesions, meningeal cells adopt either compacted or diffuse arrangements. The compacted arrangements involved no interactions with other cellular components of the remyelinating environment, and were particularly prominent following transplantation of cultures containing a high proportion of fibronectin-positive meningeal cells. The diffuse arrangements involved interactions with astrocytes, Schwann cells and endothelial cells, and contributed to a fragmented appearance of the lesion. Such observations indicate that meningeal cell contamination should be avoided when attempts are being made to repopulate glial-deficient lesions in the CNS by transplanting central glial cells, since their effect is to partition the glial environment.