Prospective isolation of functionally distinct radial glial subtypes—Lineage and transcriptome analysis
Introduction
During development of the central nervous system (CNS) a multitude of diverse cell-types needs to be generated. Toward this end, subsets of progenitor cells appear to specialize in the production of one of the three major cell-types, neurons, oligodendrocytes or astrocytes. Such a lineage restriction also allows regulating the number of specific cell-types in an independent manner by amplifying the number of their immediate progenitors. That individual progenitor cells often generate only a single cell-type has been clearly demonstrated in many regions of the CNS (for recent review see Pinto and Götz, 2007) by genetically labelling single progenitor cells (Grove et al., 1993, Luskin et al., 1993, Walsh, 1993). For example, a considerable proportion of the early telencephalic progenitors labelled prior to the onset of neurogenesis generate exclusively either glial or neuronal progeny (McCarthy et al., 2001). While lineage analysis in vivo does not allow discriminating between the influence of extrinsic cues from the local environment or intrinsic fate determinants, this is possible using in vitro lineage analysis (Heins et al., 2001, Heins et al., 2002, Luskin et al., 1993, Malatesta et al., 2003, Qian et al., 1998, Williams and Price, 1995). Indeed, clonal analysis from single cells cultured in vitro (Shen et al., 2006) suggests that distinct fate decisions can occur in the absence of a specific local environment, implying that intrinsic molecular differences may govern fate decisions.
Considerable progress has been made in identifying transcriptional regulators of cell fate, such as members of the bHLH family of transcription factors (TFs) (Fode et al., 2000, Lu et al., 2002, Nieto et al., 2001, Schuurmans et al., 2004, Zhou and Anderson, 2002), or the homeobox transcription factors Pax6 and Dlx1/2 (Götz et al., 1998, Heins et al., 2002, Petryniak et al., 2007). Gain- and loss-of-function experiments demonstrated a potent role for these TFs in fate decisions between the neuronal and the oligodendroglial lineage. For example, deletion of Pax6, Dlx1/2, Mash1 and/or Neurogenin1 and 2 results in severe defects in neurogenesis in the developing dorsal telencephalon (Heins et al., 2002, Nieto et al., 2001, Petryniak et al., 2007, Schuurmans et al., 2004) and gain-of-function experiments confirmed the role of these TFs in promoting the neurogenic lineage respectively (Berninger et al., 2007a, Buffo et al., 2005, Heins et al., 2002, Mizuguchi et al., 2001, Sugimori et al., 2007). However, these gain-of-function experiments also reveal heterogeneity in the progenitor population as not all cells could be instructed towards a specific fate upon TF overexpression (see e.g., Heins et al., 2002, Sun et al., 2001), suggesting that other cues may be required to instruct a specific cell-type identity. Moreover, neurogenesis may be governed by different molecular mechanisms not only in different regions of the CNS, but also at different developmental stages within the same brain region. For example, recent data suggest that the mechanisms of neurogenesis are distinct for early neuronal progenitors generating deep layer neurons in the cortex and progenitors at later stages generating upper layer neurons (Schuurmans et al., 2004). While the former are Neurogenin-dependent, the later are Neurogenin-independent and instead appear to require Pax6 (Schuurmans et al., 2004). Intriguingly, it has been suggested that upper layer neurons derive largely from a special set of neurogenic progenitors (Nieto et al., 2004, Tarabykin et al., 2001), the basal progenitors that are located on top of the ventricular zone and form the embryonic subventricular zone (for review see Götz and Huttner, 2005). Basal progenitors are characterized by a lack of contact to either the apical, ventricular surface or the basal lamina located at the pial side (Götz and Huttner, 2005) and act as intermediate neuronal progenitors. Originating from radial glial cells, basal progenitors generate in most cases two postmitotic neurons (Haubensak et al., 2004, Miyata et al., 2004, Noctor et al., 2004, Wu et al., 2005). In contrast, radial glial cells that are attached at the basement membrane and the apical surface and possess an apico-basal polarity (Bentivoglio and Mazzarello, 1999, Cameron and Rakic, 1991, Cappello et al., 2006) (for review see, Götz and Huttner, 2005) self-renew while also generating neurons (Malatesta et al., 2003, Miyata et al., 2001, Noctor et al., 2001, Tamamaki et al., 2001). The different nature of these distinct sets of progenitors (radial glia versus basal progenitors) is consistent with the notion that the molecular determinants of upper and lower layer neuron generation are different. The notion that upper layer neurons are largely derived by indirect neurogenesis from basal progenitors is also consistent with the increase of basal progenitors in the number in cortical regions of primates that have considerably enlarged numbers of upper layer neurons (Dehay and Kennedy, 2007). However, despite the great relevance of these basal progenitors, it is not known which radial glial cells actually generate them.
Distinct lineages arising from radial glial cells have been described previously. Clonal analysis of the progeny of a single radial glial cell demonstrated that most of these generate neurons only, when isolated at mid-neurogenesis (Anthony et al., 2004, Malatesta et al., 2000, 2000). Other subsets of radial glia largely generate other radial glia or astroglial cells progeny (Anthony et al., 2004, Malatesta et al., 2000, 2000) and a further subset of radial glia that decreases during development is bi-potent generating neurons and glial progeny (Anthony et al., 2004, Malatesta et al., 2000, Malatesta et al., 2003). However, which set of radial glial cells may comprise the basal progenitors is not known. In fact, it is not even known whether basal progenitors may indeed arise from a specific subset of radial glial cells, or in a stochastic manner upon loss of the apical and basal attachment. Therefore we set out here to isolate subsets of radial glia that differ in their progeny, possibly including the generation of basal progenitors.
Indeed, besides the interest in the mechanisms leading to the specification of basal progenitors, the molecular determinants acting in basal progenitors or in radial glial cells appear to be profoundly different (for review see Pinto and Götz, 2007). While the transcription factor Pax6 is required for neurogenesis from radial glial cells in embryonic day (E) 14 cerebral cortex, its absence still allows neurogenesis from basal precursors (Heins et al., 2002). Thus, the molecular cues regulating neurogenesis from different sets of progenitors are strikingly different. To understand the molecular mechanisms of neurogenesis it is therefore important to separate these rather different types of progenitors. However, previous separation of neurogenic and non-neurogenic progenitors employing isolation of progenitor subtypes by cell surface molecules (Liepelt et al., 1990, Reinhardt-Maelicke et al., 1990, see however Maric and Barker, 2004, Rao and Mayer-Proschel, 1997, Strathmann et al., 2007) could not discriminate between radial glia or basal progenitors. Therefore, we aimed here to select subsets of radial glial cells generating distinct progeny in order to examine their molecular specification.
Towards this end, we took advantage of a mouse line in which the promoter elements of human glial fibrillary acidic protein (hGFAP) drives the expression of a strongly fluorescent version of GFP (Nolte et al., 2001). To attempt a selective enrichment of neurogenic versus non-neurogenic radial glia we chose to separate cells according to the difference in their levels of GFP expression, aiming to examine if the variation in GFP intensity was correlated with fate differences. This approach is based on the rationale that radial glia that are neuronal progenitors may down-regulate hGFAP-promoter driven activity and hence have less GFP and be less intense in their green fluorescence. Conversely, higher activity of this promoter may correlate with the generation of further radial glial cells. A fascinating question was whether or not the generation of basal progenitors may fall into any of these categories. In order to selectively isolate radial glial cells and avoid co-purification of the progeny of radial glial cells that may inherit GFP, we additionally labelled the GFP-positive (+) cells from the apical surface by staining for the apically localized pentaspan membrane protein prominin (CD133) (Götz and Huttner, 2005, Weigmann et al., 1997). Indeed, this approach allowed high and comparable levels of radial glial cells to be purified from the developing cortex. Strikingly, we found that radial glial cells with a strong activity of this promoter generate the basal progenitors, while radial glia with a weaker GFAP-promoter activity hardly gave rise to basal progenitors. This lineage difference further correlated with molecular differences between these functionally distinct subsets of radial glial cells. This analysis therefore provides novel insights into the lineage of radial glial cells and basal progenitors, for the first time separating radial glia that generates basal progenitors prospectively and describing their transcriptome.
Section snippets
Prospective isolation of distinct radial glial progenitor subtypes
In order to separate subsets of radial glia we combined two labelling methods, GFP expression under the human GFAP promoter and immunolabelling for the apically localized protein prominin (CD133, Weigmann et al., 1997). Prominin allows distinguishing cells that have an apical membrane domain, namely all ventricular zone (VZ) cells, from basal/SVZ precursors and postmitotic neurons, the known progeny of radial glial cells that inherit GFP from their ancestors (Götz and Huttner, 2005, Weigmann et
Novel insights into radial glial lineages in the developing cortex — direct versus indirect neurogenesis
Live time-lapse microscopy has revealed that basal progenitors typically generate two neurons in the developing cortex (Haubensak et al., 2004, Miyata et al., 2004, Noctor et al., 2004), even though a small proportion of these progenitors can also undergo further cell divisions (Wu et al., 2005). Moreover, these studies also observed that basal progenitors arise from radial glial cells (Miyata et al., 2001, Miyata et al., 2004, Noctor et al., 2001, Noctor et al., 2004). These observations
Animals
Cortices were isolated from time-mated Wistar rats (day of sperm detection was E1) and from hGFAP-eGFP transgenic mice (Nolte et al., 2001) that were maintained on a FVB/NCRL background (the day of vaginal plug was considered to be embryonic day (E) 0). Expression analysis e.g. in situ hybridization was performed in embryos from C57BL/6J mice.
Cell culture
Cortices were dissected from rat embryos at E15 but at E14 from hGFAP-eGFP mice as gestational lengths vary between rats and mice and the two ages are
Acknowledgments
We are particularly grateful to Leanne Godinho, Marie-Theres Schmid and Jovica Ninkovic for excellent comments on the manuscript. We are very grateful to Frank Kirchhoff for the hGFAP-eGFP mouse line, to Nat Heintz, Robert Hevner, Pierre Leprince and Jack Price for antibodies and Markus Möser, James Okano, Urban Lendhal, Carol Schuurmans, Dieter Meyer, Victor Tarabykin and Maike Sander for in situ probes. We thank M. Körbs, A. Steiner, A. Waiser, A. Bust, D. Franzen, C. Lach and M. Öcalan for
References (119)
- et al.
Stage-specific changes in gene expression in acutely isolated mouse CNS progenitor cells
Dev. Biol.
(2005) - et al.
Radial glia serve as neuronal progenitors in all regions of the central nervous system
Neuron
(2004) - et al.
Ets-2 and components of mammalian SWI/SNF form a repressor complex that negatively regulates the BRCA1 promoter
J. Biol. Chem.
(2003) - et al.
The history of radial glia
Brain Res. Bull.
(1999) - et al.
Isolation of Crb1, a mouse homologue of Drosophila crumbs, and analysis of its expression pattern in eye and brain
Mech. Dev.
(2002) - et al.
Neural progenitor genes. Germinal zone expression and analysis of genetic overlap in stem cell populations
Dev. Biol.
(2003) - et al.
Radial glial identity is promoted by Notch1 signaling in the murine forebrain
Neuron
(2000) - et al.
Control of dendritic field formation in Drosophila: the roles of flamingo and competition between homologous neurons
Neuron
(2000) - et al.
The orphan COUP-TF nuclear receptors are markers for neurogenesis from cnidarians to vertebrates
Dev. Biol.
(2004) - et al.
A genetic analysis of neural progenitor differentiation
Neuron
(2001)