Radial glial cells as neuronal precursors: The next generation?

One of the major challenges in precursor cell biology of the central nervous system (CNS) has been to unambiguously identify different precursor cells so that we may better study them. Precursors have been largely characterized by the cell types they produce, compelling us to study them retrospectively without knowledge of which particular cell gave rise to specific progeny. Radial glial cells, easily identifiable cell types in the embryonic germinal zone, have been suggested recently to comprise a significant proportion of the neuronal precursor cell population of the developing brain (Malatesta et al., 2000; Miyata et al., 2001; Noctor et al., 2001, 2002). Radial glia can be identified by their characteristic bipolar morphology in which the cell soma resides in the ventricular zone (VZ) or subventricular zone (SVZ), bearing a long basal process that extends outwards toward the pial surface and a second, short apical process that contacts the ventricular wall. The existence of such cells has long been recognized and they have borne many names over the course of the field's history (Bentivoglio and Mazzarello, 1999). Pasko Rakic recognized that these cells express the astrocytic marker glial fibrillary acidic protein (GFAP) in the developing primate CNS and termed them “radial glia” (Rakic, 1972; Levitt and Rakic, 1980). In the mouse these cells do not begin to express GFAP until late in embryonic development, but can be identified with several other specific markers, including RC2 (Misson et al., 1988), brain lipid-binding protein (BLBP; Feng et al., 1994), vimentin (Dahl et al., 1981), nestin (Hockfield and McKay, 1985) and GLAST (Shibata et al., 1997). Rakic went on to show that radial glial cells act as a scaffold to support the migration of newly generated neurons from the embryonic germinal zone into the developing layers of the cortex, and until recently this was considered to be the major role for radial glial cells, before their postnatal transdifferentiation into astrocytes. The new role for radial glial cells as neuronal precursor cells in the embryo may dramatically change our understanding of CNS development, yet, at the same time, it raises several important and possibly controversial issues. In this mini-review, we critically examine our present understanding of radial glial cell regulation, specifically regarding the processes of induction, maintenance, and transdifferentiation. We also discuss our knowledge of cell lineage in the developing forebrain as it pertains to a putative role for radial glia as the major precursor population in vivo. Radial glial cells, defined by RC2 expression, first appear in the mouse forebrain at approximately embryonic day 9–10 (E9–10) and are the earliest cells to differentiate from the neuroepithelium (Misson et al., 1988). Little is known about the specification of radial glial cells, however, some studies suggest that certain classes of neurons may be responsible for their induction or differentiation. Soriano et al. (1997), have demonstrated that Cajal-Retzius cells transplanted into the adult cerebellum could induce a radial glial phenotype among Bergmann glia of the host brain, presumably through a secreted factor. In addition to Cajal-Retzius cells, transplantation of embryonic Purkinje cells into the adult cerebellum could also induce the reappearance of nestin positive radial glia (Sotelo et al., 1994), and the transplantation of E17 cortical neurons induced a radial glial phenotype among some astrocytes of the adult cortex (Leavitt et al., 1999). The mechanism by which immature neurons induce the radial glial phenotype is not well understood. It has been shown, however, that an unidentified 60 kDa protein secreted by cortical neurons, called RF60, is able to promote a radial glial phenotype in postnatal day 7 (P7) astrocytes in culture making this factor a strong candidate (Hunter and Hatten, 1995; Hunter-Schaedle, 1997). Glial growth factors (GGFs/neuregulins) have also been implicated in the promotion of the radial glial phenotype (Rio et al., 1997; Anton et al., 1997). GGF is expressed by migrating cortical neurons cells and induces the extension of the pial radial fiber by activating erbB receptors on radial glia, and perhaps by regulating BLBP expression. Further, erbB2 receptor deficient mice demonstrated disrupted arborization of radial glia cell endfeet (Anton et al., 1997), implicating an in vivo requirement for this signaling system in radial glial cell differentiation. A role for postmitotic neurons in the initial induction of the radial glial cell phenotype, however, has not been firmly established. The fact that the first postmitotic neurons appear at approximately E11 in the dorsal telencephalon (Caviness et al., 1995; Weiner and Chun, 1997) suggests that these cells may not play a role in the initial induction of radial glia as defined by RC2 expression, which occurs at approximately E9–10 in the forebrain (Misson et al., 1988), but rather in their differentiation. The number of RC2-expressing radial glial cells increases with increasing neurogenesis, and both radial glial cell number and neurogenesis peak at approximately E14–E15 (Caviness et al., 1995; Misson et al., 1988; Gadisseux et al., 1992) indicating that these two events are closely linked. Another mechanism by which radial glial cells may be specified and maintained could involve the cell contacted-mediated Notch signaling pathway (recently reviewed in Gaiano and Fishell, 2002). When retroviruses encoding activated Notch1 were injected intraventricularly into E9.5 embryos a radial glial fate was promoted among infected cells (Gaiano et al., 2000). A study from Campos and colleagues (2001) also supports a role for notch signaling in the induction of radial glia. This group observed that cells expressing Delta1 or 3, which are known ligands for the Notch1 receptor, were predominantly expressed by postmitotic cells that did not stain for the radial glial cell markers nestin and RC2. Additionally, they found that Delta1 expressing cells were in close opposition to radial glia and that decreased Delta1 expression correlated with the transdifferentiation of radial glia to astrocytes. The suggested role for Notch1 signaling in the specification of a radial glial cell fate, however, is not completely clear. The study by Gaiano et al. (2000) did not distinguish whether notch signaling was maintaining radial glia identity/morphology, specifying radial glial cell fate, or promoting a symmetric mode of cell division. Further, Campos et al. (2001) suggested that only postmitotic cells express Delta1 or 3, but the earliest postmitotic cells in the cortex are observed at approximately E11, as noted earlier, which is too late to be involved in the induction of RC2-expressing radial glial cells. Another report suggests that Delta1 expression in the forebrain begins at about E9 (Bettenhausen et al., 1995) that would be coincident with the appearance of the first radial glial cells. Therefore, a clear connection between the first appearance of radial glia, Delta1 expression and the first appearance of postmitotic cells has not been clearly established and requires further investigation. In any case, the role that notch plays in the regulation of the radial glial cell phenotype is likely restricted temporally. Cells infected with activated Notch1 were still able to transform into astrocytes postnatally (Chambers et al., 2001). Further, Notch1 and 3 were found to promote the differentiation of astroglia derived from multipotent adult hippocampal progenitors (Tanigaki et al., 2001). Therefore, the role Notch1 signaling plays in the regulation of radial glia is clearly context dependent. Understanding the mechanisms involved in changing the responsiveness of precursor cells to notch signaling will be another important avenue of study. In addition to the factors described above, several other genes and growth factors appear to be involved in the regulation of the radial glia cell phenotype. Mutations in the transcription factor Pax6 alter radial glia morphology, cell cycle, and number (Götz et al., 1998), as well, Pax 6 has been shown to regulate neurogenesis from glial cells (Heins et al., 2002). Mice deficient for Emx2, a homeobox gene expressed at high levels in Cajal-Retzius cells also demonstrate a disorganized radial glia scaffold, which is most likely due to loss of reelin expression by these cells at E13.5 (Mallamaci et al., 2000). Aberrant radial glia fibers were observed recently in the forebrains of mice where the gene lissencephaly (Lis1) was mutated to produce a truncated protein (Cahana et al., 2001). In addition signaling by transforming growth factor α can induce a bipolar radial-like morphology onto polygonal astrocytes in vitro. These cells upregulate the expression of GFAP and downregulate the expression of nestin inconsistent with the transdifferentiation of these cells into true radial glia (Zhou et al., 2001). Feng and colleagues (Feng et al., 1994; Feng and Heintz, 1995) have shown that signaling by BLBP, expressed primarily by radial glia, appears to be necessary for the induction of a bipolar radial glial morphology in postnatal cocultures of cerebellar granule and glial cells and the addition of a BLBP antibody to the cocultures substantially reduced the appearance of radial glial fascicles. BLBP expression appears to be turned on later than RC2 and in a subpopulation of radial glia (Hartfuss et al., 2000), suggesting it plays a role after the initial induction of radial glial cell identity. Interestingly, BLBP is not expressed in the vast majority of proliferating cells; rather its expression in radial glia is dynamically and temporally correlated with neural differentiation and migration (Feng et al., 1994). This observation also has implications for radial glia cells functioning as neuronal precursors as described below. In the mouse, radial glial cells begin to upregulate the astrocyte marker GFAP after the end of neurogenesis and begin to elaborate their ascending process, withdraw their descending process and ultimately transform into mature astrocytes (Misson et al., 1988; Voigt et al., 1989). Factors such as ciliary neurotrophic factor, leukemia inhibitory factor (Bonni et al., 1997; Murphy et al., 1997), EGF receptor signaling (Burrows et al., 1997), and members of the bone morphogenetic protein (BMP) family (Gross et al., 1996) have been demonstrated to play a role in astrocyte differentiation. The role they may play in the postnatal transdifferentiation of radial glia into astrocytes, however, is unknown. Other mechanisms that are likely to mediate this event include the completion of neuronal migration. When neurons are no longer actively migrating upon radial glial cells the signaling relationship between this two cell types would be broken, which might result in the initiation of the transdifferentiation process. Also, the disappearance of Cajal-Retzius cells, which occurs postnatally, appears to play a role in the disappearance of radial glia (Super et al., 2000). As described earlier these cells secrete factor(s) that can induce and support radial glial cell identity and in their absence the transdifferentiation of radial glia appears to be triggered. Finally, this transdifferentiation event is also likely to involve developmentally regulated epigenetic modifications such as changes in DNA methylation and chromatin structure (Takizawa et al., 2001). These studies and those recently reviewed by LePrince and Chanas-Sacré (2000) have begun to show some of the mechanisms that regulate the specification, maintenance and transdifferentiation of radial glia, but many questions remain to be answered. For example, little is known about how radial glial cell fate is initially specified and what directs the extension of the basal radial process toward the pial surface and regulates the polarity of radial glial cells. This issue becomes particularly important given the new findings of Miyata et al. (2001), which suggest that the radial process is continually lost and replaced during development. The result of in vivo studies suggested recently that radial glial cells comprise the vast majority of precursor cells that give rise to neurons in the cortical germinal zone (Hartfuss et al., 2001; Noctor et al., 2001, 2002; Götz et al., 2002). Hartfuss et al. (2001) used Ki-67, a marker of dividing cells, and nestin to label the precursor cell population in acutely dissociated cells derived from the cortex and ganglionic eminence of E12, E14, E16, and E18 mouse embryos. At E12, all precursor cells were found to express the radial glial marker RC2 and as development progressed the precursor cell population increasingly expressed the radial glial markers BLBP and GLAST. Further, this group reported that the application of DiI at the cortical ventricular surface of E15 mouse embryos resulted in 50% of labeled cells having processes that reached the pia, with the remainder bearing shorter processes. Noctor et al. (2001) first used a GFP retrovirus to label radial glial cells and their progeny in vivo and provided strong evidence for clones containing radial glia and their neuronal progeny. In their next study (Noctor et al., 2002) they used randomly obtained whole-cell recordings, which identified precursor cells at the ventricular surface of cortical slices from E12–E19 rat and then filled the recorded cells with the fluorescent dye Alexa-594-conjugated biocytin. This method showed that 100% of the recorded precursor cells had a radial glial morphology with a process ascending toward the pial surface at early as well as later stages. Using a second technique, they randomly labeled VZ cells of cortical slabs by delivering DiI-coated tungsten beads via a gene gun and quantified the percentage of VZ cells having a radial glial morphology at E12, E15, and E18. Radial glial morphology was defined as bipolar morphology with the cell body residing in the VZ and a radial process reaching either the marginal zone, cortical plate, or upper intermediate zone. At E12 virtually every labeled cell spanned the entire cerebral wall, however at E15 and E18 approximately 84% of labeled cells met this criteria, whereas approximately 15% had processes restricted to the VZ. The findings of these two groups raise several issues regarding the relationship of radial glial cells to the forebrain precursor cell population. Feng et al. (1994) reported that most cells that expressed the radial glial cell marker BLBP did not incorporate BrdU, and suggested that these cells played a role in supporting the migration of new neurons, but not in the expansion of the neuronal precursor population. While this contrasts with the results of Hartfuss et al. (2001), that suggested that subpopulations of proliferating precursor cells in the germinal zone express BLBP, it implies that not all radial glial cells are neuronal precursor cells. An idea supported by Gaiano et al. (2000) and Schmechel and Rakic (1979) who reported that not all radial glial cells are actively dividing during neurogenesis. A second issue that is raised relates to cell lineage. If the majority of cortical precursors are radial glial cells during neurogenesis and if radial glia transform into astrocytes postnatally then one would predict that a large proportion of clones examined in the postnatal mouse would contain neurons and at least one astrocyte. Such clones are rare, however, with the majority being either neuronal or astrocytic, additionally many neuronal clones are composed of single cells. The advent of ultrasound backscatter microscope guided injections into the developing neural tube has permitted the analysis of cell lineage beginning as early as E9.5. Analysis of E9.5 infected mice in the 3rd postnatal week showed that 34% of multicellular clones contained neurons only, 47% contained glia only, and 18% contained both glia and neurons (McCarthy et al., 2001). Neuron only clones tended to be smaller in size than glial or mixed clones (McCarthy et al., 2001). Interestingly, glial restricted clones were predominantly found ventrally suggesting that ventral radial glia might be more gliogenic than dorsal radial glia. Remarkably, studies beginning at this early stage also showed that neural precursor cells are already largely restricted to the generation of specific cell lineages (McCarthy et al., 2001). These results suggest that the decision of which radial glia will give rise to glia and which will give rise to neurons is made very early. Infections carried out at later stages of forebrain development (i.e., early to mid neurogenesis) showed similar results. Luskin et al. (1988) reported that the majority of cortical clones were also homogenous with regard to cell phenotype. Of 47 clones analyzed in postnatal animals, 39 were composed entirely of neurons, 38 of these were radially organized, six were composed of oligodendrocytes, and two clones were composed of astrocytes and only one of these contained a neuron. Other groups have reported similar cell type homogeneity within clustered clones (Price and Thurlow, 1988; Parnavalas et al., 1991; Grove et al., 1993; Luskin et al., 1993). The relatively rare existence of neuron and glia containing clones postnatally suggests that the radial glial population that transform into astrocytes postnatally play a minor role as neuronal precursor cells during development. Several explanations, however, may account for the apparent low number of bipotent astrocyte-neuron clones. In the cortex, dying astrocytes can be observed coincident with the transition of radial glia to astrocytes (Soriano et al., 1993), therefore some radial glia that would otherwise be present in the clone may be lost through cell death. Alternatively, some cells that appear to be radial glial cells may in fact be newly born migrating neurons (Nadarajah et al., 2001; Miyata et al., 2001). Miyata et al. (2001) have recently reported that a substantial population (approximately 56%) of radial glial cells divide asymmetrically to give rise to a daughter neuron that inherits the radial process and undergoes somal translocation to move into position in the cortical layers. The parent cell was then shown to grow a new process presumably in preparation for it's next division. In this situation it is possible that instead of elaborating a long ascending process into a multipolar form and transforming into an astrocyte postnatally, these cells may simply stop sprouting new processes. The fate of such cells may be something other than the astrocytic fate that has been showed for postnatal radial glia by previous analyses (i.e., cell death or transdifferentiation into a neuron). In these ways, a clone containing only neurons may be formed in the apparent absence of its radial glial parent cell. Further investigations into the postnatal fate of radial glia are required to address these issues. Widespread clones of the developing cortex have been shown to be composed of systematically spaced clustered cells, and contain multiple neuronal and glial phenotypes (Walsh and Cepko, 1993; Reid et al., 1997). Some precursor cells have been observed to migrate within the germinal zone (Fishell et al., 1993; Walsh and Cepko, 1993). It has been suggested that widespread clones are the result of multipotent neural stem cells migrating within the germinal zone, giving rise to progenitors that form homogenous clustered clones in a more defined region (Reid et al., 1995). Interestingly, Fishell et al. (1993) noted that these migrating cells had a characteristic bipolar morphology and expressed RC2 and nestin, but it was not noted how long the basal process was. Do radial glia, with processes spanning the entire cortex migrate within the germinal layer to give rise to widely dispersed clones, or perhaps more likely, could the population of shorter bipolar cells that exists in the VZ be capable of this migration? The more elongated radial glial cells may be more restricted progenitors, relegated to producing the single cell types that comprise clustered clones within the widespread clones. Another interesting possibility is that radial glial cells might divide to give rise to a neuron, which inherits the process, and the now shortened radial glial cell migrates within the VZ, occasionally crossing into new environments that promote the generation of different cell types. This would also ultimately result in the generation of multiple cell types within a single widespread clone. Further investigations regarding the motility of radial glia in the germinal zone is required to elucidate these and other possibilities, and is important for our understanding of the relationship between radial glia and the multipotent neural stem cell population of the developing CNS. Noctor et al. (2001, 2002) have convincingly demonstrated that radial glial cells proliferate to give rise to neurons having a pyramidal morphology in the cortex, however, pyramidal projection neurons are believed to be derived from a precursor cell population distinct from that which gives rise to interneurons in the telencephalon (reviewed in Parnavelas, 2000). In contrast to pyramidal neurons, the interneuron cell lineage in the telencephalon is primarily, if not exclusively generated in subpallial structures and these cells migrate tangentially, not radially, to their final positions (recently reviewed in Marin and Rubenstein, 2001). Halliday and Cepko (1992) reported that clones present in the striatal VZ, and some found in both the VZ and SVZ, were radially orientated, whereas clones in the SVZ only were nonradially orientated. It has been found that the lateral ganglionic eminence gives rise to neurons that migrate upon radial glia to populate both lateral and basolateral cortical regions and the primary olfactory cortex (De Carlos et al., 1996). Are radial glial cells of the ventral telencephalon able to give rise to both tangentially and radially migrating neurons? Or alternatively, is there a population of nonradial precursors that support the generation of tangentially migrating interneurons? If this was the case it would be likely that neural stem cells of the adult SVZ, which support the generation of interneurons destined for the olfactory bulb, would be descendants of this hypothetical short cell rather than a descendant of radial glia as has recently been proposed (Alvarez-Buylla et al., 2001). Further analysis of the population of short SVZ/VZ restricted cells in more ventral regions is required to elucidate this possibility. The ventral telencephalon has also been proposed to give rise to the majority of oligodendrocytes in the forebrain and some groups have suggested that radial glia may have a relationship to the oligodendrocyte lineage (Choi et al., 1983; Choi and Kim, 1985; Hirano and Goldman, 1988). Recent studies of the spinal cord have suggested that oligodendrocytes and astrocytes have distinct precursor cells in vivo (Zhou and Anderson, 2002; Lu et al., 2002) and these findings are supported by previous analyses of the oligodendroglial/astrocyte lineages (reviewed by Skoff and Knapp, 1991). Study of the radial glial/astrocyte/oligodendrocyte lineage in vivo has thus far been largely neglected. Furthermore, fluorescent dye lineage tracing experiments for radial glia have been restricted to studies of the cortical germinal zone (Voigt, 1989; Malatesta et al., 2000), which would not show the fate or potential of ventral radial glial cells. Interestingly, Notch1 signaling, which can promote the radial glial cell fate, appears to inhibit the generation of oligodendrocytes (Gaiano et al., 2000; Tanigaki et al., 2001), suggesting an antagonistic relationship between these lineages. To better understand the role radial glial cells play in other forebrain cell lineages, such as interneurons and oligodendrocytes, a greater emphasis must be placed on studying these cells in the ventral telencephalon. We have highlighted some of the issues surrounding the regulation of the radial glial cell phenotype and the role of radial glial cells as primary precursor cells of the developing CNS. Our goal was to expose areas in need of investigation in this rapidly developing field. In vivo lineage studies strongly suggest that radial glia are a very heterogeneous population of cells, some being restricted to the generation of neurons and others being restricted to glial lineages. As well, these subpopulations are likely established very early radial glia first from the Understanding these subpopulations and their relationship to the other cell lineages of the brain will be an important next in this field. Further, to the relationship between radial glial cells and cells with short processes in the VZ will be important to understand the precursor cell population of the developing CNS. regard to radial glial cell regulation, important areas for investigation include understanding the induction and maintenance of the radial glial cell as well as mechanisms that are involved in the disappearance of these cells of these and other questions may us to understand how radial glia and of radial glia to be This is necessary if we are to some of these processes in the mature brain for the of and CNS is a of a of is an for

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.