What have tissue culture studies told us about the development of oligodendrocytes?

SIRT2, tubulin deacetylation, and oligodendroglia differentiation

The mammalian silent information regulator 2 (SIRT2) is an NAD-dependent histone deacetylase with known roles in the regulation of the cell cycle. SIRT2 is also a tubulin deacetylase functioning as an early mitotic checkpoint, but its roles in regulating cytoplasmic microtubule dynamics were unknown. Novel findings now indicate that SIRT2 is specifically enriched in brain oligodendroglia, where it functions specifically in modulating the oligodendrocyte cytoskeleton during its differentiation and maturation.

The Lineages of Neuroglial Cells

The study of neuroglial cell lineages in the CNS identifies the time in development, when astrocytes and oligodendrocytes diverge from a common precursor cell. Recent studies using retroviral tracing show that the lineages for astrocytes and oligodendrocytes begin to diverge as early as embryonic day 13 (E13) in the cerebellum and as early as E15 in the forebrain. A very small percentage of glial precursor cells present in late pre- and postnatal development are pluripotential, but the vast majority of astrocytes and oligodendrocytes in the brain are derived from “committed” precursors. The precursors for these postmitotic astrocytes and oligodendrocytes are immature astrocytes and oligodendrocytes (progenitors) that express molecular properties unique to each of these cell types.

It is critical to distinguish glial lineage (thought of in terms of a glial cell's ancestry in normal development) from glial plasticity (the potential of a glial cell to alter its fate when its normal environment is changed). In tissue culture, a bipotential cell known as the O-2A cell generates oligodendrocytes under one set of culture conditions and retains plasticity to become a subtype of astrocyte under another set. Whether all cells phenotyped as O-2A cells in culture are bipotential or whether only a subset displays this capacity is still unclear.

Co-localization of NG2 proteoglycan and PDGF alpha-receptor on O2A progenitor cells in the developing rat brain

A detailed comparison in the developing rat central nervous system between the distribution of the NG2 proteoglycan and the alpha-receptor for platelet-derived growth factor (PDGF) shows that these two molecules are co-expressed by glial progenitor cells of the O2A lineage and can serve as reliable markers for identification of O2A cells in vivo. Our mapping experiments indicate that NG2-positive, PDGF alpha-receptor positive O2A cells are abundant throughout the developing central nervous system in both white and gray matter. The earliest cells immunoreactive for either of the two markers are found adjacent to the central canal of the embryonic day 15 (E15) spinal cord. These cells express only PDGF alpha-receptor and not NG2. By E17, process-bearing cells expressing both NG2 and PDGF alpha-receptor in a highly co-localized fashion are found throughout the central nervous system. The first postnatal week marks the peak in the number of NG2 and PDGF alpha-receptor immunoreactive cells, as well as the peak in the level of expression and the extent of co-localization of the two molecules. After the first week, the level of expression of both NG2 and PDGF alpha-receptor declines, although both molecules continue to be expressed in the adult brain. On O2A cells in the mature brain, NG2 and PDGF alpha-receptor are not as well co-localized at the subcellular level as they are on O2A cells in the younger brain. The functional consequences of co-localization and subsequent dissociation of NG2 and PDGF alpha-receptor on maturing O2A progenitors are investigated in the accompanying paper (Nishiyama et al.: J Neurosci Res 43:315-330, 1996).

The use of transplanted glial cells to reconstruct glial environments in the CNS

Transplantation studies have demonstrated that glia-depleted areas of the CNS can be reconstituted by the introduction of cultured cells. Thus, the influx of Schwann cells into glia-free areas of demyelination in the spinal cord can be prevented by the combined introduction of astrocytes and cells of the O-2A lineage. Although Schwann cell invasion of areas of demyelination is associated with destruction of astrocytes, the transplantation of rat tissue culture astrocytes ("type-1") alone cannot suppress this invasion, indicating a role for cells of the O-2A lineage in reconstruction of glial environments. By transplanting different glial cell preparations and manipulating lesions so as to prevent meningeal cell and Schwann cell proliferation it is possible to demonstrate that the behaviour of tissue culture astrocytes ("type-1") and astrocytes derived from O-2A progenitor cells ("type-2") is different. In the presence of meningeal cells, tissue culture astrocytes clump together to form cords of cells. In contrast, type-2 astrocytes spread throughout glia-free areas in a manner unaffected by the presence of meningeal cells or Schwann cells. Thus, progenitor-derived astrocytes show a greater ability to fill glia-free areas than tissue culture astrocytes. Similarly, when introduced into infarcted white matter in the spinal cord, progenitor-derived astrocytes fill the malacic area more effectively than tissue culture astrocytes, although axons do not regenerate into the reconstituted area.

Differentiation of the O-2A progenitor cell line CG-4 into oligodendrocytes and astrocytes following transplantation into glia-deficient areas of CNS white matter

The in vitro properties of the CG4 cell line have led to its increasing use as a cell line with which to study the behaviour of the O-2A progenitor cell. In this study we have examined the in vivo behaviour of the CG4 cell line following transplantation into areas of adult rat spinal cord white matter which have been permanently depleted of glial cells by the combination of local X-irradiation and direct injection of 0.1% ethidium bromide. Twenty-one days after transplantation, both myelin-forming oligodendrocytes and glial fibrillary acidic protein-positive astrocytes were identified within the lesion, indicating that the CG4 cell line has bipotential differentiation properties when introduced into a pathological environment consisting of demyelinated axons but devoid of oligodendrocytes or astrocytes. In this respect, the CG4 cell line resembles other glial progenitor cell lines that have been transplanted into similar lesions. In some areas of the lesion, remyelination was observed that was similar in extent to that achieved by growth factor-expanded populations of O-2A progenitor cells. The transplant origin of the cell types within the lesion was confirmed by retroviral incorporation of the lacZ marker gene, the expression of which allowed their identification by histochemistry. In conclusion, the in vivo properties of the CG4 cell line make it a highly suitable line with which to study the behaviour of O-2A progenitors following transplantation into normal and damaged CNS.

Glial cell lineage and development

Recent studies on the development of glial cells are making rapid progress in two major directions. While the factors that control gliogenesis during the perinatal period are becoming more clearly characterized, so too is the relationship between gliogenesis and the patterning of the embryonic central nervous system as a whole.

Divergent lineages for oligodendrocytes and astrocytes originating in the neonatal forebrain subventricular zone

Although previous studies have revealed that the prenatal rat ventricular zone contains separate progenitor cells for neurons, astrocytes, and oligodendrocytes during the development of the cerebral cortex as early as the beginning of neurogenesis (Luskin et al., 1993; Grove et al., 1993), it is still unclear whether there are bipotential progenitor cells in the neonatal telencephalic subventricular zone which give rise to both astrocytes and oligodendrocytes during the peak of gliogenesis. To investigate this possibility, discrete groups of clonally related cells, generated by infecting progenitor cells of the neonatal subventricular zone with a retroviral lineage tracer, were analyzed ultrastructurally. An intracerebral injection of retrovirus encoding the reporter gene E. coli beta-galactosidase (lacZ) was made into the subventricular zone of newborn rats. Two weeks later their brains were perfused, sectioned, and histochemically reacted with X-Gal to identify at the light microscopic level clones of lacZ-positive cells. The sections were processed for electron microscopy to enable the identity of clonally related cells to be assessed at the ultrastructural level. All of the clones analyzed contained cells of the same phenotype and could be divided into four distinct types: immature cell clones situated in the subependymal zone surrounding the lateral ventricle, oligodendrocytes clones, and white or gray matter astrocyte clones. Not all of the cells in every clone displayed ultrastructural features of a mature cell. Rather, in some glial clones the lacZ-positive cells appeared to be at different stages of differentiation. However, we never encountered clones which contained both macroglial subtypes or clones containing neurons. Although the existence of bipotential progenitor cells cannot be completely dismissed, our results indicate the absence of progenitor cells in vivo in the neonatal subventricular zone which divide and generate astrocytes and oligodendrocytes.