Article21 October 2004free access Mash1 specifies neurons and oligodendrocytes in the postnatal brain Carlos M Parras Carlos M Parras Institut de Génétique et de Biologie Cellulaire et Moléculaire, Illkirch, France Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK Search for more papers by this author Rossella Galli Rossella Galli Stem Cell Research Institute, HS Raffaele, Milan, Italy Search for more papers by this author Olivier Britz Olivier Britz Institut de Génétique et de Biologie Cellulaire et Moléculaire, Illkirch, France Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK Search for more papers by this author Sylvia Soares Sylvia Soares UMR7501, CNRS-UPC, Université P&M Curie, Paris, France Search for more papers by this author Christophe Galichet Christophe Galichet Institut de Génétique et de Biologie Cellulaire et Moléculaire, Illkirch, France Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK Search for more papers by this author James Battiste James Battiste Center for Basic Neuroscience, UT Southwestern Medical Center, Dallas, TX, USA Search for more papers by this author Jane E Johnson Jane E Johnson Center for Basic Neuroscience, UT Southwestern Medical Center, Dallas, TX, USA Search for more papers by this author Masato Nakafuku Masato Nakafuku Division of Developmental Biology, Cincinnati Children's Hospital Research Foundation, Cincinnati, OH, USA Search for more papers by this author Angelo Vescovi Angelo Vescovi Stem Cell Research Institute, HS Raffaele, Milan, Italy Search for more papers by this author François Guillemot Corresponding Author François Guillemot Institut de Génétique et de Biologie Cellulaire et Moléculaire, Illkirch, France Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK Search for more papers by this author Carlos M Parras Carlos M Parras Institut de Génétique et de Biologie Cellulaire et Moléculaire, Illkirch, France Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK Search for more papers by this author Rossella Galli Rossella Galli Stem Cell Research Institute, HS Raffaele, Milan, Italy Search for more papers by this author Olivier Britz Olivier Britz Institut de Génétique et de Biologie Cellulaire et Moléculaire, Illkirch, France Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK Search for more papers by this author Sylvia Soares Sylvia Soares UMR7501, CNRS-UPC, Université P&M Curie, Paris, France Search for more papers by this author Christophe Galichet Christophe Galichet Institut de Génétique et de Biologie Cellulaire et Moléculaire, Illkirch, France Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK Search for more papers by this author James Battiste James Battiste Center for Basic Neuroscience, UT Southwestern Medical Center, Dallas, TX, USA Search for more papers by this author Jane E Johnson Jane E Johnson Center for Basic Neuroscience, UT Southwestern Medical Center, Dallas, TX, USA Search for more papers by this author Masato Nakafuku Masato Nakafuku Division of Developmental Biology, Cincinnati Children's Hospital Research Foundation, Cincinnati, OH, USA Search for more papers by this author Angelo Vescovi Angelo Vescovi Stem Cell Research Institute, HS Raffaele, Milan, Italy Search for more papers by this author François Guillemot Corresponding Author François Guillemot Institut de Génétique et de Biologie Cellulaire et Moléculaire, Illkirch, France Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK Search for more papers by this author Author Information Carlos M Parras1,2, Rossella Galli3, Olivier Britz1,2, Sylvia Soares4, Christophe Galichet1,2, James Battiste5, Jane E Johnson5, Masato Nakafuku6, Angelo Vescovi3 and François Guillemot 1,2 1Institut de Génétique et de Biologie Cellulaire et Moléculaire, Illkirch, France 2Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK 3Stem Cell Research Institute, HS Raffaele, Milan, Italy 4UMR7501, CNRS-UPC, Université P&M Curie, Paris, France 5Center for Basic Neuroscience, UT Southwestern Medical Center, Dallas, TX, USA 6Division of Developmental Biology, Cincinnati Children's Hospital Research Foundation, Cincinnati, OH, USA *Corresponding author. Division of Molecular Neurobiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, UK. Tel.: +44 208 816 2740; Fax: +44 208 816 2109; E-mail: [email protected] The EMBO Journal (2004)23:4495-4505https://doi.org/10.1038/sj.emboj.7600447 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Progenitors in the telencephalic subventricular zone (SVZ) remain mitotically active throughout life, and produce different cell types at embryonic, postnatal and adult stages. Here we show that Mash1, an important proneural gene in the embryonic telencephalon, is broadly expressed in the postnatal SVZ, in progenitors for both neuronal and oligodendrocyte lineages. Moreover, Mash1 is required at birth for the generation of a large fraction of neuronal and oligodendrocyte precursors from the olfactory bulb. Clonal analysis in culture and transplantation experiments in postnatal brain demonstrate that this phenotype reflects a cell-autonomous function of Mash1 in specification of these two lineages. The conservation of Mash1 function in the postnatal SVZ suggests that the same transcription mechanisms operate throughout life to specify cell fates in this structure, and that the profound changes in the cell types produced reflect changes in the signalling environment of the SVZ. Introduction Progenitor cells in the telencephalon, which generate the vast array of neurons and glia found in the adult cerebral cortex, basal ganglia and olfactory bulb, have unique characteristics. While in most regions of the central nervous system (CNS), neurogenesis and gliogenesis take place during embryonic development and cease before or soon after birth, new neurons, oligodendrocytes and astrocytes are generated in the telencephalon well into postnatal life. Indeed, stem cells in the telencephalon continuously proliferate and differentiate from embryonic stages to adulthood, although their cellular output changes dramatically over time. At embryonic stages, multipotent progenitors located in the telencephalic ventricular and subventricular zones produce a multitude of neuronal subtypes and a first wave of oligodendrocyte and astrocyte precursors. In particular, progenitors in the ventral telencephalon have been shown to produce both GABAergic interneurons and oligodendrocytes (He et al, 2001; Yung et al, 2002), while some progenitors in the dorsal telencephalon generate both cortical neurons and astrocytes (Williams and Price, 1995). Embryonic ventricular zone progenitors then give rise, soon before birth, to a distinct postnatal subventricular zone (SVZ) with unique features (Marshall et al, 2003; Pencea and Luskin, 2003; Tramontin et al, 2003). A distinct characteristic of the postnatal SVZ is that it produces a second wave of astrocytic and oligodendrocytic precursors, destined to colonise the corpus callosum and cerebral cortex. Production of astrocytes and oligodendrocytes from SVZ precursors peaks during the first 2 weeks of postnatal life. Another important characteristic of the SVZ is that it contains specialised neuronal precursors, which migrate through the rostral migratory stream (RMS) into the olfactory bulb, where they differentiate into olfactory interneurons (Luskin, 1993). Olfactory interneurons are continuously produced by the SVZ from late embryogenesis onwards, and the adult SVZ is dedicated to the production of this cell type (Doetsch et al, 1999; Alvarez-Buylla and Garcia-Verdugo, 2002; Pencea and Luskin, 2003). Important differences therefore exist between progenitor cells in the embryonic and postnatal telencephalon, in that the former are largely multipotent and produce many types of neurons, while the latter are mostly lineage-restricted and produce essentially one neuronal subtype, olfactory interneurons. Mechanisms controlling cell fate specification have been extensively investigated in the embryonic CNS, where basic helix–loop–helix (bHLH) transcription factors have a central role (Bertrand et al, 2002). In the embryonic ventral telencephalon, the proneural protein Mash1 is essential for the production of neuronal precursor cells (Casarosa et al, 1999; Horton et al, 1999). In the dorsal telencephalon, together with other proneural proteins of the Neurogenin family, Mash1 promotes the neuronal commitment of multipotent progenitors, while inhibiting their astrocytic differentiation (Nieto et al, 2001). In addition, Mash1 expression in embryonic telencephalic progenitors can activate a GABAergic subcortical differentiation programme that involves induction of the Dlx homeodomain (HD) protein family (Fode et al, 2000). Although GABAergic neurons and oliodendrocytes appear to originate from a common lineage in the telencephalon (He et al, 2001; Yung et al, 2002), a different family of bHLH proteins, Olig1 and Olig2, has been implicated in specification of telencephalic oligodendrocyte precursors (Lu et al, 2002). Whether specification of neuronal and glial progenitors in the adult brain relies on the same transcription mechanisms as in the embryo is unknown. The limited differentiation potential of postnatal progenitors could reflect differences in the regulation of proneural factors operating in both embryonic and adult brain, or alternatively the recruitment of a different set of cell fate determinants to control neurogenesis and gliogenesis at postnatal stages (Kintner, 2002). In this paper, we have addressed the role of Mash1 in the postnatal SVZ, and we show that this gene is required at this stage for the specification of both neurons and oligodendrocytes. Thus, a common mechanism underlies the generation of two of the three cell types produced in the postnatal brain. Our results also indicate that, despite their limited differentiation potential, fate specification of progenitors in the postnatal brain relies on the same intrinsic mechanisms as in the embryo. Results Mash1 expression in the neonatal brain We began investigating the mechanisms underlying cell fate specification in the postnatal SVZ by asking whether the proneural gene Mash1, which is essential for neurogenesis in the embryonic ventral telencephalon (Casarosa et al, 1999; Horton et al, 1999), remains expressed in the SVZ after birth. In sections of brains harvested at birth (P0), numerous cells expressing Mash1 RNA and protein were found in the SVZ, as well as in the RMS, that extends from the lateral ventricles to the olfactory bulb (Figure 1A and data not shown). The neonatal SVZ contains numerous neuronal and glial precursors, but in double-labelling experiments, only a fraction of Mash1-positive cells expressed either the neuronal precursor marker, βIII-tubulin (14%; Figure 1B), or the oligodendrocyte precursor cell (OPC) markers NG2 (4%; Figure 1C), PDGFRα (4%; data not shown) and Olig2 (40%; Figure 1D). Most Mash1+ cells were not labelled with lineage-specific markers, possibly because they were too immature. To examine the fate of Mash1+ progenitors, we used a mouse transgenic line in which the LacZ reporter gene is under the control of Mash1 regulatory sequences (Mash1∷LacZ mice; Verma-Kurvari et al, 1996). At P0, the LacZ transgene was expressed, like Mash1, in the SVZ and the RMS, but also in the bulb where Mash1 RNA and protein is not detected, indicating that the βgal protein is stable enough to trace the fate of Mash1+ progenitors (Figure 1E). Double-labelling experiments in Mash1∷LacZ mice showed that the vast majority of βIII-tubulin+ neuronal precursors in the SVZ and RMS are βgal+ (Figure 1F), while a smaller number βgal+ cells express the OPC marker O4 (Figure 1G). Mash1∷LacZ mice were also analysed at P7, when the white matter is more differentiated than at birth. A large fraction of NG2+ OPCs expressed βgal in the corpus callosum (77%; Figure 1H), fimbria (75%), anterior commissure (67%) and olfactory bulb (45%). Thus, Mash1 is expressed in both olfactory interneuron precursors and oligodendrocyte precursors in the early postnatal SVZ. Figure 1.Mash1-expressing cells in the neonatal SVZ belong to two cell lineages. (A) Mash1 transcripts (purple) in a sagittal section through the rostral telencephalon of a P0 mouse are found in the dorsal and lateral parts of the SVZ (dSVZ and lSVZ, respectively), along the rostral migratory stream (RMS) and in the centre of the olfactory bulb. (B–D) Double antibody labelling of sagittal sections of a P0 brain showing Mash1+ cells (green) coexpressing the neuronal marker βIII-tubulin (B, red) and the oligodendrocyte precursor markers NG2 (C, red) and Olig2 (D, red). Double-labelled cells are marked by arrows. (E) X-gal staining of a sagittal brain section from a P3 Mash1∷LacZ transgenic mouse, showing the distribution of Mash1-βgal+ cells. (F, G) Double labelling of a sagittal brain section from a P0 transgenic mouse for βgal (red) and βIII-tubulin (F, green) or the OPC marker O4 (G, green, arrow). (H) Double labelling of a sagittal brain section from a P7 transgenic mouse for βgal (blue, X-gal precipitate) and NG2 (brown). The inset is an enlargement of the area outlined in the corpus callosum (CC). Scale bars: 20 μm. Download figure Download PowerPoint Mash1 expression in the adult brain To better define the stage in the olfactory interneuron lineage when Mash1 is expressed, we turned to the adult brain where the different types of progenitors in the lineage can be identified with specific antibodies (Alvarez-Buylla and Garcia-Verdugo, 2002). Mash1 expression in the adult rostral telencephalon was overall very similar to that observed at birth, with labelled cells located near the lateral ventricles and in the RMS and expressing proliferation markers (Figure 2A; Supplementary Figure 1A). The majority of Mash1+ cells in the SVZ had an antibody labelling profile characteristic of transit amplifying progenitors of the olfactory interneuron lineage, that is, positive for the HD proteins Dlx and negative for the neuroblast markers PSA-NCAM and mCD24 (56% of Mash1+ cells; Figure 2B; Doetsch et al, 2002). A smaller fraction of Mash1+ cells had characteristics of migrating neuroblasts (Dlx+, PSA-NCAM+, mCD24+; 26%; Figure 2B; Supplementary Figure 1B), while very few Mash1+ cells expressed GFAP, which marks SVZ astrocytes and stem cells (Supplementary Figure 1C; Doetsch et al, 1999). In adult Mash1∷LacZ transgenic mice (Supplementary Figure 1E), there was a excellent match between the expression of the LacZ transgene and the neuroblast markers PSA-NCAM and mCD24 in the SVZ and RMS, thus confirming that Mash1 is transiently expressed in all progenitors of the olfactory interneuron lineage (Figure 2C; Supplementary Figure 1F). βgal expression was also detected in oligodendrocyte precursors in the telencephalic white matter and grey matter (data not shown), but we have not yet attempted to define the exact stage of Mash1 expression in this lineage. Figure 2.Mash1+ cells in the adult SVZ are progenitors of the olfactory interneuron lineage. (A) Double labelling of a sagittal section through the SVZ of an 8-week-old brain, for Mash1 (red) and BrdU (green) after 2 h of BrdU incorporation. Double-labelled cells are indicated by arrows. (B) Triple labelling for Mash1 (green), the progenitor and neuroblast marker Dlx (red) and the neuroblast marker mCD24 (blue). Mash1+ transit amplifying progenitors (Dlx−, mCD24− and Dlx+, mCD24−) are marked by arrows and Mash1+ neuroblasts (Dlx+, mCD24+) by an arrowhead. (C) Double labelling of a sagittal section through the RMS of an 8-week-old Mash1-LacZ transgenic mouse, for βgal (red) and the neuroblast marker PSA-NCAM (green). Most neuroblasts in the RMS are double labelled. Scale bars: 20 μm. Download figure Download PowerPoint To confirm the position of Mash1+ cells in the adult neuronal lineage, we examined the time course of Mash1 expression during regeneration of the adult SVZ. Infusion of the antimitotic drug AraC into the adult brain eliminates rapidly dividing cells, including transit amplifying progenitors and neuroblasts, while leaving slow dividing stem cells unaffected (Doetsch et al, 1999). Almost all Mash1+ cells were eliminated by a 6-day-long AraC treatment (1.4% Mash1+ cell remaining 12 h after the end of the treatment, compared with control brains treated with vehicle only) and their number remained low after 24 h (3.2%). The number of Mash1+ cells reached control level 48 h after the end of the treatment (94.3%), while the number of PSA-NCAM+ neuroblasts remained low (1.8%). This time course of Mash1 expression in the regenerating SVZ closely matches that reported for transit amplifying progenitors or C cells (Doetsch et al, 1999). Altogether, our results demonstrate that the majority of Mash1+ cells in the adult SVZ have characteristics of undifferentiated transit amplifying progenitors of the olfactory neuron lineage, and that Mash1 expression is only transiently maintained in differentiating neuroblasts. Mash1 mutant phenotype at birth The above data suggested that Mash1 might have a function in the postnatal SVZ. Since mice carrying a null mutation in Mash1 die soon after birth, they could be used to address the role of Mash1 in the generation of oligodendrocytes and olfactory interneurons by SVZ progenitors, which has already started at this stage (Spassky et al, 2001; Ivanova et al, 2003; Pencea and Luskin, 2003). We first compared the number and distribution of all dividing progenitors, marked by BrdU incorporation, in the telencephalon of Mash1 mutant and control newborns (Figure 3). The number of BrdU+ cells was significantly reduced in the lateral and dorsal SVZ and the RMS of Mash1 mutants (Figure 3D). There was no evidence of increased apoptosis (data not shown) and no accumulation of progenitor cells in other parts of the Mash1 mutant telencephalon (Figure 3D), suggesting that the reduction in the number of BrdU+ cells in the SVZ and RMS is due to a defect in the generation of progenitor cells rather than in their subsequent migration or survival. Figure 3.Less dividing progenitors are found in the SVZ and RMS of Mash1 mutants at birth. (A) Sagittal section of P0 brain showing the different zones used to quantify the defects in BrdU incorporation reported in (A). (B, B′, C, C′) Distribution of BrdU+ cells in the SVZ (B, B′) and proximal RMS (C, C′) of P0 control (B, C) and Mash1 mutants (B′, C′), after 30 min of BrdU incorporation. (D) Quantification of BrdU+ cells in different brain regions, as defined in (A), in P0 control (blue bars) and Mash1 mutants (purple bars). White numbers in purple bars are ratios of BrdU+ cell numbers in Mash1 mutants over controls. *P<0.05, **P<0.01, ***P<0.001, Student's t-test. Download figure Download PowerPoint We then examined which types of cells were affected in Mash1 mutants. The olfactory bulb at birth contains neuronal precursors that have migrated along the RMS (Pencea and Luskin, 2003), and is also an active site of production of oligodendrocytes, generated locally from the rostral extension of the ventricular wall (Spassky et al, 2001). The overall size of the bulb was reduced by one-third in Mash1 mutant newborns (Figure 4C, C′ and F). There was a further reduction in the number of olfactory bulb neurons, marked by βIII-tubulin (Figure 4G), with granule cells more severely affected than periglomerular neurons (Figure 4H). To determine whether this neuronal deficit was due to a defect in neurogenesis at birth, rather than in interneuron production at earlier stages, we measured the fraction of neuronal precursors, marked by coexpression of βIII-tubulin and the mitotic marker Ki67, in acutely dissociated olfactory bulb cells (Figure 4A and A′). The percentage of double-labelled neuronal precursors was reduced by about two-thirds in mutant bulbs compared with controls (Figure 4G), indicating that neonatal neurogenesis is severely compromised in the absence of Mash1 function. Figure 4.Reduced number of neuronal and oligodendrocyte precursors in the olfactory bulb of Mash1 mutants at birth. (A, A′) Double labelling for βIII-tubulin (green) and Ki67 (red) and counterstaining with DAPI (blue), to identify double-labelled neuronal precursors (arrows) in acutely dissociated olfactory bulb cells from P0 control (A) and Mash1 mutant (A′). (B–E, B′–E′) Sagittal sections of olfactory bulb from P0 control (B–E) and Mash1 mutant (B′–E′), labelled for the GABAergic neuron marker GAD65 (B, B′) and the OPC markers NG2 (C, C′, D, D′) and Olig2 (E, E′). (F, G) Quantification of all cells (F), and of neurons and neuronal precursors (G) in acutely dissociated control (blue bars) and Mash1 mutant (purple bars) bulb. (H, I) Quantification of GAD65+ neurons in the granular and periglomerular layers and of TH+ dopaminergic neurons (H), NG2+, O4+ and Olig2+ OPCs (I) in sagittal sections of control and Mash1 mutant bulb. White numbers in purple bars are ratios of numbers of labelled cells in Mash1 mutants over controls. *P<0.05, **P<0.01, Student's t-test. Scale bars: 20 μm. Download figure Download PowerPoint As Mash1 is also expressed in the oligodendrocyte lineage at birth (Figure 1), we examined whether OPCs and oligodendrocytes were affected in Mash1 mutant newborns. As is the case for neuronal precursors, the number of NG2+, O4+ and Olig2+ OPCs was severely reduced in mutant bulbs (Figure 4C–E, C′–E′ and I). Oligodendrogenesis appeared to be less affected in other parts of Mash1 mutant brains, and we are currently examining whether particular subsets of oligodendrocyte precursors are missing in different brain regions in the absence of Mash1. Together, these data indicate that Mash1 has a major role in the neonatal SVZ, where it is required for the generation of both neuronal and oligodendrocyte precursors. However, given the complexity of this structure, this study could not rule out that multiple defects contribute to the mutant phenotype in the postnatal SVZ, such as a defect in migration of precursors from other parts of the brain, or a defect in cell production at embryonic stages that would indirectly affect the postnatal SVZ. To better define the cellular function of Mash1 in the postnatal brain, we turned to a simpler model of differentiation of SVZ progenitors in culture. Mash1 mutant phenotype in progenitor cultures Neural progenitors can be propagated as ‘neurospheres’ in cultures containing the mitogens EGF and FGF, and then differentiated into neurons, oligodendrocytes and astrocytes by switching cultures to serum-containing medium (Reynolds and Weiss, 1992; Morshead et al, 1994; Gritti et al, 1999). We first confirmed that Mash1 has a similar expression profile in neurosphere cultures established from the lateral ventricular walls of newborn mice (see Materials and methods and Supplementary Figure 2) as in vivo, and therefore that this culture system could be used to address the role of Mash1 in postnatal progenitors. Mash1 was expressed in a large fraction of neurosphere cells cultivated with mitogens (55–70%; Supplementary Figure 2), and double-labelling experiments showed that these cells did not express neuronal- or glial-specific markers. The fraction of cells expressing Mash1 decreased rapidly as cultures were switched to a differentiation medium, and many of these cells transiently expressed NG2 or βIII-tubulin (Supplementary Figure 2G). These results indicate that Mash1 is expressed in SVZ-derived cultures in a manner very similar to that observed in the postnatal SVZ in vivo, that is, in immature, lineage marker-negative progenitors, and then transiently in differentiating neuronal and oligodendrocyte precursors. We then established neurosphere cultures from the lateral ventricular walls of Mash1 mutants and control littermates at birth. Mash1 mutant and control neurospheres were propagated for up to four passages with similar efficiency (data not shown), suggesting that Mash1 does not have a significant role in the self-renewal of neural progenitors in culture. After 2–4 passages, neurospheres were dissociated and cultivated in differentiating conditions. Progenitors from control cultures produced large numbers of βIII-tubulin+ neurons, O4+ oligodendrocytes and GFAP+ astrocytes after 7 days in culture (Figure 5A, C and D). In striking contrast, Mash1 mutant cultures contained much fewer neurons and oligodendrocytes, while the number of astrocytes was increased (Figure 5A′, C and D). The number of nestin+ precursors was similar in mutant and control cultures throughout the course of the culture (Figure 5D), indicating that the reduction in the number of neurons and oligodendrocytes in mutant cultures is not due to a delay in precursor differentiation. Figure 5.Defect in specification of neurons and oligodendrocytes in Mash1 mutant progenitor cultures. (A, A′) Triple labelling for βIII-tubulin (green), O4 (red) and GFAP (blue) and counterstaining with DAPI (white) in control (A) and mutant (A′) dissociated neurosphere cultures after 7 days. (B, B′) Double labelling for GFP (green) and O4 (red) in clonal cultures of GFP+ control (B) or mutant (B′) progenitors cocultivated with an excess of GFP-negative wild-type progenitors. The arrows mark GFP+ oligodendrocytes in the wild-type clone (B), which are absent in the mutant clone (B′). (C) Quantification of neurons (βIII-tubulin+), oligodendrocytes (O4+) and astrocytes (GFAP+) in control and mutant cultures after 7 days. White numbers in striped bars are ratios of numbers of labelled cells in mutants over control cultures. (D) Time course of the generation of neurons, oligodendrocytes and astrocytes in control and mutant progenitor cultures. (E) Quantification of the different types of clones generated in control and mutant progenitor cultures. Each pair of bars represents the frequency, in control and mutant cultures, of a particular type of clone, whose cell type composition is indicated by a colour code and letter code under the bars (N: neurons in green; O: oligodendrocytes in red; A: astrocytes in blue). The average size of each type of clone is indicated under the corresponding bar. *P<0.05, **P<0.01, Student's t-test. Scale bars: 10 μm. Download figure Download PowerPoint The reduced production of neurons and oligodendrocytes in mutant cultures could reflect a defect in the specification of neuronal and oligodendrocytic precursors in these cultures, or alternatively the correct specification of precursors with reduced proliferation capacities. Moreover, the reduction in the number of OPCs may reflect an indirect role of Mash1-dependent neurons, rather than a cell autonomous role of Mash1, in the induction of oligodendrocytes. To distinguish between these hypotheses, we performed a clonal analysis of Mash1 mutant progenitors in culture, by cocultivating mutant or control neurosphere cells marked with green fluorescent protein (GFP) with a large excess of GFP-negative, wild-type neurosphere cells (Figure 5B and B′; see Materials and methods). In control experiments, GFP+ cells gave rise, after 7 days in culture, to a variety of clones of different size and cellular composition, including some clones containing neurons, oligodendrocytes and astrocytes and others containing only two of the three cell types (Figure 5E). When GFP+ Mash1 mutant progenitors were cultivated instead, the overall number of clones formed after 7 days was not significantly changed (data not shown), but their composition was dramatically different. There was a drastic reduction in the frequency of all types of clones containing neurons, and of most types of clones containing oligodendrocytes, and a parallel increase in the frequency of astrocytic clones (Figure 5E). Importantly, average clone size was not significantly different in mutant and control cultures after 3 days (wild-type cultures: 9.4±4.2, n=2; mutant cultures: 5.6±2.3, n=2), and clones of similar composition, including neurons- and oligodendrocytes-containing clones, were also of the same size in control and mutant cultures after 7 days (Figure 5E). Moreover, the number of apoptotic cells was not significantly increased in mutant cultures (data not shown). Together, these results ruled out that the loss of neurons and oligodendrocytes in mutant cultures could be due to a defect in precursor proliferation or survival. Rather, these experiments demonstrate that Mash1 is required for the specification of both neuronal and oligodendrocyte precursors, which in the absence of Mash1 function adopt instead an astrocytic fate. Only a subset of oligodendrocytes was missing in Mash1 mutant clonal cultures, as already observed in nonclonal cultures (Figure 5C and D). Interestingly, oligodendrocytes present in mixed astros–oligos clones were spared by the Mash1 mutation, whereas oligodendrocytes found in other types of wild-type clones (i.e. neurons/astros/oligos clones, neurons/oligos clones and oligos-only clones) were mostly missing in