Nature 449, 351-355 (20 September 2007) | doi:10.1038/nature06090; Received 28 May 2007; Accepted 12 July 2007; Published online 26 August 2007

Differential Notch signalling distinguishes neural stem cells from intermediate progenitors

Ken-ichi Mizutani1,2,5, Keejung Yoon1,2,5,6, Louis Dang1,3, Akinori Tokunaga1,2 & Nicholas Gaiano1,2,3,4

  1. Institute for Cell Engineering,
  2. Department of Neurology,
  3. Department of Neuroscience, and,
  4. Department of Oncology, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205, USA
  5. These authors contributed equally to this work.
  6. Present address: Division of Brain Diseases, Center for Biomedical Sciences, National Institute of Health, Tongil-Lo 194, Eunpyung-Gu, Seoul, 122-701, Korea.

Correspondence to: Nicholas Gaiano1,2,3,4 Correspondence and requests for materials should be addressed to N.G. (Email: gaiano@jhmi.edu).


During brain development, neurons and glia are generated from a germinal zone containing both neural stem cells (NSCs) and more limited intermediate neural progenitors (INPs)1, 2, 3. The signalling events that distinguish between these two proliferative neural cell types remain poorly understood. The Notch signalling pathway is known to maintain NSC character and to inhibit neurogenesis, although little is known about the role of Notch signalling in INPs. Here we show that both NSCs and INPs respond to Notch receptor activation, but that NSCs signal through the canonical Notch effector C-promoter binding factor 1 (CBF1), whereas INPs have attenuated CBF1 signalling. Furthermore, whereas knockdown of CBF1 promotes the conversion of NSCs to INPs, activation of CBF1 is insufficient to convert INPs back to NSCs. Using both transgenic and transient in vivo reporter assays we show that NSCs and INPs coexist in the telencephalic ventricular zone and that they can be prospectively separated on the basis of CBF1 activity. Furthermore, using in vivo transplantation we show that whereas NSCs generate neurons, astrocytes and oligodendrocytes at similar frequencies, INPs are predominantly neurogenic. Together with previous work on haematopoietic stem cells4, this study suggests that the use or blockade of the CBF1 cascade downstream of Notch is a general feature distinguishing stem cells from more limited progenitors in a variety of tissues.

The Notch signalling pathway is of fundamental importance to a wide variety of processes during embryonic development and in the adult5, 6. Of particular interest, is the fact that Notch signalling regulates stem cells in many different settings, including the nervous system, haematopoietic system, skin and gut7. The Notch receptors are transmembrane proteins activated by Delta and Jagged ligands6. On activation, the intracellular domain of Notch (NICD) is cleaved by γ-secretase and translocates into the nucleus to interact with the transcriptional regulator CBF1. In the developing nervous system, the NICD–CBF1 complex activates target genes such as Hes1 and Hes5 (ref. 8), which antagonize proneural genes and neuronal differentiation. Little is known about the specificity of Notch function in different proliferative neural cell types, although genetic evidence suggests that Notch signalling is not used uniformly in all telencephalic progenitors9, 10.

To characterize endogenous Notch signalling during brain development we generated a transgenic mouse line (TNR, for transgenic Notch reporter) expressing enhanced green fluorescent protein (EGFP) in cells with pathway activation. The transgene includes a CBF1-responsive element (CBFRE, four CBF1-binding sites and the basal simian virus 40 (SV40) promoter11) upstream of EGFP (Fig. 1a). Proof-of-principle experiments (Supplementary Fig. 2), including short hairpin RNA (shRNA)-mediated knockdown of CBF1, and previous reports4, 12, 13, indicate that the TNR transgene faithfully reports CBF1 activity. In the telencephalic ventricular zone (VZ), EGFP is expressed together with the neural progenitor markers Nestin and CD133 (ref. 14) (Fig. 1b, c, and Supplementary Fig. 2i–k). On embryonic day (E) 10.5 most telencephalic VZ cells express EGFP (Supplementary Fig. 2h), whereas at E14.5 cell clusters express either high (EGFPhi) or low (EGFPlo/neg) levels of EGFP (Fig. 1c–f). Immunostaining revealed that both EGFPhi and EGFPlo/neg cells contain cleaved (activated) Notch1 (clN1)15 (Fig. 1f), raising the possibility that not all cells activate CBF1 in response to Notch activation.

Figure 1: CBF1 signalling heterogeneity in the telencephalic VZ.
Figure 1 : CBF1 signalling heterogeneity in the telencephalic VZ. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

a, Strategy to detect CBF1 activation in the TNR line. b, c, Nestin (b) is expressed in the telencephalic (telen.) VZ, as is EGFP (c), in TNR embryos. d, e, EGFPhi cells in the E14.5 neocortex (NCX, d) and lateral ganglionic eminence (LGE, e) are interspersed with EGFPlo/neg cells. DAPI, 4,6-diamidino-2-phenylindole. f, Double labelling reveals that cleaved (activated) Notch1 (red) is present in cells expressing high and low levels of EGFP. g, CBFRE–EGFP-CAG–DsRed2 and plasmids expressing NICD1, wild-type or DNA-binding mutant (DBM) forms of CBF1–VP16, together with nucleus-localized β-galactosidase (nls-lacZ). CP/IZ, cortical plate/intermediate zone. h, Quantification of cell position in the E15.5 neocortex after in utero electroporation at E12.5. i, NICD1 retains cells in the proliferative zone (VZ/SVZ) (Supplementary Fig. 4) but does not upregulate EGFP in all cells. j, CBF1–VP16 also retains cells in the proliferative zone but drives widespread EGFP expression. k, CBF1DBM–VP16 does not retain cells in the proliferative zone or upregulate EGFP. Scale bars,200μm (b, c), 25μm (df) and 50μm (ik). In h, P<0.01 for A–B, A–C, B–C, G–H, G–I and H–I comparisons, and P<0.03 for E–F comparison; n = 3 for each.

High resolution image and legend (202K)

To characterize CBF1 activity in VZ cells, we performed DNA electroporation in utero. First, at E12.5 we electroporated a plasmid containing CBFRE–EGFP and also DsRed2 driven by the CAG promoter16 to permit the detection of all electroporated cells (CBFRE–EGFP-CAG–DsRed2; Fig. 1g). Consistent with the mosaic EGFP expression in the TNR VZ, only a subset of DsRed2+ VZ cells was EGFPhi at E15.5 (Supplementary Fig. 3). Next, CBFRE–EGFP-CAG–DsRed2 was electroporated together with plasmids expressing NICD1 (constitutively active), or an activated CBF1 (CBF1 fused to the VP16 transactivation domain; CBF1–VP16 (ref. 17)) (Fig. 1g). At E15.5 most cells expressing NICD1 or CBF1–VP16 were in the VZ and subventricular zone (SVZ), which is consistent with the ability of Notch signalling to inhibit neurogenesis (Fig. 1h and Supplementary Fig. 4). However, whereas most VZ cells expressing CBF1–VP16 were EGFP+ (Fig. 1j), many cells expressing NICD1 were not (Fig. 1i). These data suggest that VZ cells are heterogeneous in their response to Notch activation, and that in a subset of cells NICD1 inhibits neurogenesis in a CBF1-independent manner.

Recent work has found that the neocortical VZ is heterogeneous2, with radial glial NSCs driving gene expression from the glutamate–aspartate transporter promoter (GLASTp), and neurogenic INPs driving expression from the tubulin α1 promoter (Tα1p)2, 18. To characterize CBF1 activity in those cell types, we electroporated CBFRE–EGFP, or a plasmid driving enhanced yellow fluorescent protein (EYFP) from the Hes5 promoter19, together with GLASTp–DsRed2 or Tα1p–DsRed2 (Fig. 2a). Cells with CBF1 activity (EGFP+ or EYFP+) were largely GLASTp–DsRed2+ and Tα1p–DsRed2 (Fig. 2b, c, and Supplementary Fig. 5a). Similarly, in TNR embryos EGFPhi VZ cells were largely GLASTp–DsRed2+ and Tα1p–DsRed2- (Supplementary Fig. 5b, c). CBF1 activity is therefore present in NSCs but not in INPs.

Figure 2: Analysis of CBF1 function in NSCs and INPs.
Figure 2 : Analysis of CBF1 function in NSCs and INPs. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

a, Plasmids used to identify cells with CBF1 activity, or with radial glial NSC or INP character. Hes5p, Hes5 promoter. b, c, After E12.5 co-electroporations, VZ cells with CBF1 activity (green) at E15.5, reported as expression from CBFRE–EGFP (b) or from Hes5p–EYFP (c), are predominantly GLASTp–DsRed2+ and Tα1p–DsRed2-. df, VZ cells with shRNA-mediated knockdown of CBF1 from E13.5 to E15.5 are predominantly Tα1p–DsRed2+ (d), indicating INP character, whereas controls are either Tα1p–DsRed2+ or GLASTp–DsRed2+ (e). f, Quantification of these data; the asterisk indicates P<0.01; n = 3. g, Many Tα1p–DsRed2+ cells in the VZ were proliferating (BrdU+, arrows) after CBF1 knockdown. CP/IZ, cortical plate/intermediate zone. h, shRNA knockdown of CBF1 promotes migration to the CP (third panel), which can be inhibited by NICD1 (fourth panel; see also Supplementary Fig. 6). shRNA-Luc to knockdown luciferase was used as a negative control (first and second panels). i, Quantification of the data in h and also Supplementary Fig. 6. j, Activation of CBF1 targets by using CBF1–VP16 was ineffective at shifting INPs towards an astroglial fate more consistent with NSC character (see also Supplementary Fig. 11). Asterisk, P<0.05; n = 3. Scale bars, 50μm (c; also applies to be, g) and 100μm (h). In i, P<0.03 for A–B, E–F, G–H, I–J, K–L, A–G, C–I, E–K, B–H and D–J comparisons; n = 4 for each. See Supplementary Fig. 13 for the individual channels. All error bars indicate s.d.

High resolution image and legend (243K)

On the basis of the difference in CBF1 activity between NSCs and INPs, we tested whether a loss of CBF1 signalling could convert NSCs into INPs. After shRNA-mediated knockdown of CBF1 in vivo at E13.5, many electroporated cells migrated from the VZ to the SVZ and cortical plate, indicating that CBF1 knockdown promoted neurogenesis (Fig. 2h, i, and Supplementary Fig. 6a). Those cells remaining in the VZ after knockdown were largely Tα1p–DsRed2+ and incorporated bromodeoxyuridine (BrdU) (Fig. 2d–g), suggesting that blockade of CBF1 signalling promotes neurogenesis by converting NSCs into INPs.

Our findings that clN1 is present in EGFPlo/neg cells in the VZ (Fig. 1f), and that NICD1 inhibits the differentiation of INPs without activating CBF1 (Fig. 1h,i), suggested that INPs use CBF1-independent Notch signalling. Consistent with this idea was our observation that expression of NICD1 in cells with shRNA-mediated knockdown of CBF1 inhibited the exit of Tα1p–DsRed2+ cells (INPs) from the VZ or SVZ (Fig. 2h, i, and Supplementary Fig. 6). NICD1 did not inhibit cell migration from the VZ to the SVZ, which is consistent with the notion that INPs in the VZ give rise to those in the SVZ3. These data indicate that Notch signalling inhibits both NSC and INP differentiation. However, whereas CBF1-dependent Notch signalling promotes NSC character in the VZ, CBF1-independent Notch signalling promotes INP character in the VZ and SVZ. Experiments with a γ-secretase inhibitor to block ligand-dependent Notch processing have also provided evidence that Notch has a function in EGFPlo/neg cells regardless of their CBF1 signalling status (Supplementary Fig. 7).

Although our in vivo data suggested that EGFPhi cells are NSCs and that EGFPlo/neg cells are neurogenic INPs, we tested this presumption directly. First, using fluorescence-activated cell sorting (FACS) we found that the neural progenitor markers CD133 (ref. 14) (Fig. 3a) and Nestin (Supplementary Fig. 8b) were expressed in both EGFPhi and EGFPlo/neg telencephalic cells from E14.5 TNR embryos. This finding was consistent with our observation that both EGFPhi and EGFPlo/neg cells are present in the telencephalic proliferative zone in vivo. Because CD133 is a cell surface marker, we were able to isolate CD133+/EGFPhi cells (presumptive NSCs) and CD133+/EGFPlo/neg cells (presumptive INPs) by FACS (Fig. 3a,b). Quantitative reverse-transcriptase-mediated polymerase chain reaction (RT–PCR) revealed that the former expressed higher levels of Notch receptors and targets than the latter (Fig. 3c), although no differences were detected in the expression of Rbpsuh (which encodes CBF1) or of the pathway inhibitors Numb and Numblike6 (not shown). We next used the neurosphere assay to examine the stem cell character of CD133+/EGFPhi and CD133+/EGFPlo/neg cells. In the presence of fibroblast growth factor 2 (FGF2), EGFPhi cells generated more numerous and larger primary and secondary neurospheres than EGFPlo/neg cells (Fig. 3d–f), suggesting that the former are indeed NSCs and the latter are INPs. Consistent with this notion was our observation that CD133+/EGFPlo/neg cells generated neurospheres similar to those derived from CD133+/Tα1p–EGFP+ cells (compare Supplementary Fig. 9 and Fig. 3f). We used quantitative RT–PCR to compare gene expression in neurospheres generated from EGFPhi and EGFPlo/neg cells. Notably, the former continued to express higher levels of EGFP, Hes1 and Hes5, whereas the latter expressed higher levels of Mash1, a gene antagonized by the Hes genes (Fig. 3d, e, g)8. The heritability of attenuated CBF1 signalling in EGFPlo/neg cells (presumptive INPs) would explain why those cells are often found clustered in the VZ in vivo (Fig. 1c–f).

Figure 3: In vitro analysis of EGFPhi and EGFPlo/neg cells.
Figure 3 : In vitro analysis of EGFPhi and EGFPlo/neg cells. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

a, FACS plot of E14.5 TNR ganglionic eminence cells. b, Acutely isolated EGFPhi cells (green box in a) expressed higher levels of cleaved Notch1 (clN1) than EGFPlo/neg cells (blue box in a), of which some expressed low levels (asterisk). Sorting gates were designed to limit overlap between distinct progenitor pools (see Supplementary Fig. 8). c, Primary EGFPhi neocortical cells expressed higher mRNA levels of Notch receptors and targets (Hes1, Hes5) than EGFPlo/neg cells at E14.5, as determined by quantitative RT–PCR. Asterisk, P<0.03. N = 4. df, EGFPhi cells generated more numerous and larger neurospheres (d) than EGFPlo/neg cells (e). f, Quantification of neurosphere frequency and size. Asterisk, P<0.002; n = 3. Similar results were obtained with primary E14.5 neocortical neurospheres (Supplementary Fig. 12). g, Neurospheres derived from EGFPhi cells expressed higher levels of EGFP, Hes1 and Hes5 mRNA than EGFPlo/neg cells, but lower levels of Mash1 and Dlx2 (ref. 9) after one week in vitro, as determined by quantitative RT–PCR. Asterisk, P<0.01; n = 4. Scale bars,10μm (b) and 100μm (d, e).

High resolution image and legend (156K)

We next used adherent cultures to examine the developmental potential of presumptive NSCs and INPs separated by CBF1 activity. CD133+/EGFPhi and CD133+/EGFPlo/neg progenitors isolated from E13.5 neocortex or ganglionic eminences were cultured for two or five days in FGF2-containing medium. BrdU was added either after one day in vitro (DIV1) or four days in vitro (DIV4), and the cells were fixed one day later. Whereas on DIV2 about 60% of both populations were BrdU+, by DIV5 the EGFPhi and EGFPlo/neg cultures had decreased to 47% and 25% BrdU+, respectively (Fig. 4a). Concurrently, whereas TuJ1+ cells (neurons) in EGFPhi cultures remained at about 25% from DIV2 to DIV5, TuJ1+ cells in EGFPlo/neg cultures increased from 33% to 51% (Fig. 4a). After withdrawal of FGF2, EGFPlo/neg cells generated mostly neurons (TuJ1+), whereas EGFPhi cells generated mostly astrocytes (positive for glial fibrillary acidic protein (GFAP+); Fig. 4b, c, and Supplementary Fig. 10). These data support the idea that EGFPlo/neg cells are indeed neurogenic INPs.

Figure 4: Differentiation analysis of EGFPhi and EGFPlo/neg cells.
Figure 4 : Differentiation analysis of EGFPhi and EGFPlo/neg cells. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

a, EGFPlo/neg cells spontaneously differentiate into neurons in mitogenic culture containing FGF2. Asterisk, P<0.003; n = 3. b, c, Differentiation of EGFPhi and EGFPlo/neg cells in vitro revealed that the former generate more astrocytes (GFAP+), and the latter more neurons (TuJ1+). b, Quantification of GFAP and TuJ1 expression status of EGFPhi and EGFPlo/neg cells. c, Representative images showing GFAP and TuJ1 stainings after in vitro differentiation. Similar results were obtained at E12.5 (Supplementary Fig. 10). Asterisk, P<0.001; n = 3. d, The cell transplantation scheme used. e, The location and morphology of donor cells in the host tissue was determined by using PLAP histochemistry. f, After cell transplantation, EGFPlo/neg cells generate more neurons and fewer astrocytes (astros) in vivo than EGFPhi cells. Oligos, oligodendrocytes. Asterisk, P<0.01; n = 15. Scale bars,50μm. All error bars indicate s.d.

High resolution image and legend (253K)

To compare the developmental potential of EGFPhi and EGFPlo/neg cells in a physiological setting we used in vivo cell transplantation. Donor cells were derived from TNR embryos containing a second transgene ubiquitously expressing human placental alkaline phosphatase (PLAP)20 to permit morphological characterization after transplantation. CD133+/EGFPhi and CD133+/EGFPlo/neg cells were isolated by FACS from the ganglionic eminences of E14.5 double-transgenic embryos and were transplanted into the forebrains of E14.5 embryos in utero (Fig. 4d). Postnatal analysis revealed that whereas EGFPhi cells differentiated into neurons, astrocytes and oligodendrocytes with similar efficiencies, EGFPlo/neg cells differentiated predominantly into neurons and oligodendrocytes (Fig. 4e, f). This finding is consistent with previous work showing that Notch signalling promotes astrogliogenesis21, 22, 23 and with our finding that EGFPlo/neg cells generate fewer astrocytes in vitro (Fig. 4b,c). Our data also support the existence of progenitors with primarily neuronal and oligodendroglial potential24, 25 and suggest that INPs are these cells. Notably, many neurons generated from both EGFPhi and EGFPlo/neg cells incorporated BrdU after transplantation (40% and 29%, respectively) and were therefore derived from donor cell divisions in the host. All told, our in vivo and in vitro differentiation data show that telencephalic NSCs and INPs can be prospectively distinguished on the basis of CBF1 signalling.

Finally, whereas CBF1 knockdown in vivo promoted the conversion of NSCs into INPs (see Fig. 2f), the question remained: would forced activation of CBF1 in INPs promote their conversion back into NSCs? To test this possibility, we took advantage of the fact that NSCs generate predominantly astrocytes in vitro, and INPs generate predominantly neurons. CD133+/EGFPhi and CD133+/EGFPlo/neg cells were isolated by FACS, plated into FGF2-containing medium and transduced (with the use of lipofection) on DIV1 with plasmids expressing NICD1, CBF1–VP16 or nucleus-localized β-galactosidase. On DIV5, FGF2 was withdrawn and the cells were allowed to differentiate for one week, at which point lacZ+ cells were scored as either TuJ1+ or GFAP+. The introduction of NICD1 into EGFPlo/neg cells had no significant effect on the percentage of differentiated cells expressing TuJ1 or GFAP (Fig. 2j and Supplementary Fig. 11). Although the expression of CBF1–VP16 did cause a modest shift towards astrocyte fate, as a population EGFPlo/neg cells expressing CBF1–VP16 still differentiated mostly into neurons (Fig. 2j and Supplementary Fig. 11). Thus, once CBF1 signalling has been attenuated in INPs (EGFPlo/neg cells), they become largely refractory to subsequent CBF1 activation.

Here we show that telencephalic progenitors are heterogeneous with respect to Notch signal transduction. Supplementary Fig. 1 summarizes the main result of this paper. In NSCs, Notch inhibits exit from the proliferative zone and activates CBF1 targets, whereas in INPs, Notch also inhibits exit from the proliferative zone but does not activate CBF1 targets. Knockdown of CBF1 promotes the conversion of NSCs into INPs, suggesting that the heritable blockade to CBF1 activation we have observed in INPs has a causal role in this transition. However, forced activation of CBF1 in INPs is ineffective at reverting them to NSC character, suggesting that INPs have limited plasticity. The importance of these findings beyond the nervous system has been supported by a recent haematopoietic stem cell (HSC) study using the TNR line: within a cell fraction highly enriched for HSCs, EGFPhi cells were mostly HSCs, whereas EGFPlo/neg cells were mostly unipotent progenitors4. The parallels between that work and the present study indicate that the differential use of Notch signalling, in particular with respect to CBF1 activation, may be a mechanism used to distinguish between stem and progenitor cell subtypes in many tissues.


Methods Summary


All mice were maintained in accordance with the Institutional Animal Care and Use Committee (IACUC) at Johns Hopkins University School of Medicine. The plug date was designated as E0.5. The TNR mouse line was generated by the Johns Hopkins Transgenic Core facility by pronuclear DNA injection, and is maintained and distributed by The Jackson Laboratory (http://jaxmice.jax.org/strain/005854.html). The construct used contains four CBF1-binding sites derived from Epstein–Barr virus upstream of the basal SV40 promoter (provided by D. Hayward). Gene transfer into CD1 or TNR mouse brains in utero was performed as described previously26, 27 with the Nepagene CUY21EDIT electroporator. Each electroporation result was reproduced in multiple brains derived from at least three litters.

For cell transplantation, donor cells were isolated from E14.5 ganglionic eminences, and trypsin-EDTA was used during dissociation. Each E14.5 host received (1–3)×104 cells in about 1μl. Hosts were analysed at postnatal date (P) 21 or P30, and tissue sections were stained histochemically to detect PLAP+ donor cells. Cellular phenotype was evaluated by morphology, and a total of about 2,000 forebrain cells were scored.

Cell culture

Neurosphere and adherent progenitor cultures were established from the neocortex or ganglionic eminences (lateral and medial) of E12.5–E14.5 embryos. Dissected tissue was dissociated by trituration in the absence of trypsin-EDTA. For adherent cultures, CD133+/EGFPhi or CD133+/GFPlo/neg cells were isolated by FACS and plated in eight-well chamber slides (Nunc). Cells were cultured in basal serum-free medium28 including 10ngml-1 FGF2.

CBF1 shRNA knockdown

The target sequences used to knockdown CBF1 by shRNA interference were 5′-AAGAACTACTGCACAGCCAAA-3′ (CBF1 no. 1 ), and 5′-AAGCAGACGGCATTACTGGAT-3′ (CBF1 no. 2), and were first validated in NIH 3T3 cells. The target sequence of luciferase used for the control shRNA-Luc was 5′-CTTACGCTGAGTACTTCGATT-3′; it has previously been shown to knock down luciferase29.

Full methods accompany this paper.



  1. Temple, S. The development of neural stem cells. Nature 414, 112–117 (2001) | Article | PubMed | ISI | ChemPort |
  2. Gal, J. S. et al. Molecular and morphological heterogeneity of neural precursors in the mouse neocortical proliferative zones. J. Neurosci. 26, 1045–1056 (2006) | Article | PubMed | ISI | ChemPort |
  3. Merkle, F. T. & Alvarez-Buylla, A. Neural stem cells in mammalian development. Curr. Opin. Cell Biol. 18, 704–709 (2006) | Article | PubMed | ISI | ChemPort |
  4. Duncan, A. W. et al. Integration of Notch and Wnt signaling in hematopoietic stem cell maintenance. Nature Immunol. 6, 314–322 (2005) | Article |
  5. Bolos, V., Grego-Bessa, J. & de la Pompa, J. L. Notch signaling in development and cancer. Endocr. Rev. 28, 339–363 (2007) | Article | PubMed | ISI | ChemPort |
  6. Yoon, K. & Gaiano, N. Notch signaling in the mammalian central nervous system: insights from mouse mutants. Nature Neurosci. 8, 709–715 (2005) | Article |
  7. Chiba, S. Notch signaling in stem cell regulation. Stem Cells 24, 2437–2447 (2006) | Article | PubMed | ISI | ChemPort |
  8. Iso, T., Kedes, L. & Hamamori, Y. HES and HERP families: multiple effectors of the Notch signaling pathway. J. Cell. Physiol. 194, 237–255 (2003) | Article | PubMed | ISI | ChemPort |
  9. Yun, K. et al. Modulation of the notch signaling by Mash1 and Dlx1/2 regulates sequential specification and differentiation of progenitor cell types in the subcortical telencephalon. Development 129, 5029–5040 (2002) | PubMed | ISI | ChemPort |
  10. Mason, H. A. et al. Notch signaling coordinates the patterning of striatal compartments. Development 132, 4247–4258 (2005) | Article | PubMed | ISI | ChemPort |
  11. Hsieh, J. J. et al. Truncated mammalian Notch1 activates CBF1/RBPJk-repressed genes by a mechanism resembling that of Epstein–Barr virus EBNA2. Mol. Cell. Biol. 16, 952–959 (1996) | PubMed | ISI | ChemPort |
  12. Estrach, S., Ambler, C. A., Lo Celso, C., Hozumi, K. & Watt, F. M. Jagged 1 is a beta-catenin target gene required for ectopic hair follicle formation in adult epidermis. Development 133, 4427–4438 (2006) | Article | PubMed | ISI | ChemPort |
  13. Hellstrom, M. et al. Dll4 signalling through Notch1 regulates formation of tip cells during angiogenesis. Nature 445, 776–780 (2007) | Article | PubMed | ISI | ChemPort |
  14. Lee, A. et al. Isolation of neural stem cells from the postnatal cerebellum. Nature Neurosci. 8, 723–729 (2005) | Article |
  15. Tokunaga, A. et al. Mapping spatio-temporal activation of Notch signaling during neurogenesis and gliogenesis in the developing mouse brain. J. Neurochem. 90, 142–154 (2004) | Article | PubMed | ISI | ChemPort |
  16. Niwa, H., Yamamura, K. & Miyazaki, J. Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108, 193–199 (1991) | Article | PubMed | ISI | ChemPort |
  17. Waltzer, L., Bourillot, P. Y., Sergeant, A. & Manet, E. RBP-Jkappa repression activity is mediated by a co-repressor and antagonized by the Epstein–Barr virus transcription factor EBNA2. Nucleic Acids Res. 23, 4939–4945 (1995) | Article | PubMed | ISI | ChemPort |
  18. Sawamoto, K. et al. Direct isolation of committed neuronal progenitor cells from transgenic mice coexpressing spectrally distinct fluorescent proteins regulated by stage-specific neural promoters. J. Neurosci. Res. 65, 220–227 (2001) | Article | PubMed | ISI | ChemPort |
  19. Ohtsuka, T. et al. Visualization of embryonic neural stem cells using Hes promoters in transgenic mice. Mol. Cell. Neurosci. 31, 109–122 (2006) | Article | PubMed | ISI | ChemPort |
  20. DePrimo, S. E., Stambrook, P. J. & Stringer, J. R. Human placental alkaline phosphatase as a histochemical marker of gene expression in transgenic mice. Transgenic Res. 5, 459–466 (1996) | Article | PubMed | ISI | ChemPort |
  21. Gaiano, N., Nye, J. S. & Fishell, G. Radial glial identity is promoted by Notch1 signaling in the murine forebrain. Neuron 26, 395–404 (2000) | Article | PubMed | ISI | ChemPort |
  22. Tanigaki, K. et al. Notch1 and Notch3 instructively restrict bFGF-responsive multipotent neural progenitor cells to an astroglial fate. Neuron 29, 45–55 (2001) | Article | PubMed | ISI | ChemPort |
  23. Gaiano, N. & Fishell, G. The role of notch in promoting glial and neural stem cell fates. Annu. Rev. Neurosci. 25, 471–490 (2002) | Article | PubMed | ISI | ChemPort |
  24. He, W., Ingraham, C., Rising, L., Goderie, S. & Temple, S. Multipotent stem cells from the mouse basal forebrain contribute GABAergic neurons and oligodendrocytes to the cerebral cortex during embryogenesis. J. Neurosci. 21, 8854–8862 (2001) | PubMed | ISI | ChemPort |
  25. Battiste, J. et al. Ascl1 defines sequentially generated lineage-restricted neuronal and oligodendrocyte precursor cells in the spinal cord. Development 134, 285–293 (2007) | Article | PubMed | ISI | ChemPort |
  26. Mizutani, K. & Saito, T. Progenitors resume generating neurons after temporary inhibition of neurogenesis by Notch activation in the mammalian cerebral cortex. Development 132, 1295–1304 (2005) | Article | PubMed | ISI | ChemPort |
  27. Saito, T. & Nakatsuji, N. Efficient gene transfer into the embryonic mouse brain using in vivo electroporation. Dev. Biol. 240, 237–246 (2001) | Article | PubMed | ISI | ChemPort |
  28. Nakashima, K. et al. Synergistic signaling in fetal brain by STAT3–Smad1 complex bridged by p300. Science 284, 479–482 (1999) | Article | PubMed | ISI | ChemPort |
  29. Elbashir, S. M. et al. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498 (2001) | Article | PubMed | ISI | ChemPort |

Supplementary Information

Supplementary information accompanies this paper.



We thank D. Hayward, D. Johns, E. Manet, T. Haydar, A. Ayoub and T. Ohtsuka for plasmids; J. Corbin for the PLAP mice; L. Blosser and A. Tam for cell sorting; R.-J. Zhao and J. Kim for technical assistance; and R. Wechsler-Reya, T. Reya, M. Starz-Gaiano, T. Haydar and S. Temple for discussions. K.M. was supported by a fellowship from the Japan Society for the Promotion of Science. This work was supported by grants from the Burroughs Wellcome Fund, the Sidney Kimmel Foundation for Cancer Research, and the National Institute of Neurological Disorders and Stroke (all to N.G.).

Author Contributions K.M. performed in utero electroporations, adherent and neurosphere cell culture experiments, quantitative RT–PCR analysis, γ-secretase inhibition and shRNA experiments. K.Y. generated and validated the TNR line, established flow cytometry staining protocols, performed γ-secretase inhibition and MEF experiments, and did the Nestin tissue staining. L.D. characterized the in vivo expression pattern of EGFP in the telencephalon of TNR embryos, performed in vivo cell transplantations, and did the CD133 tissue staining. A.T. performed clN1, EGFP and CD133 immunostainings. N.G. conceived of the TNR line, oversaw the project, and prepared the manuscript.


Competing interests statement

The authors declare  competing financial interests.


Online Methods


For gene transfer into CD1 or TNR mouse brains, 1–2μl of DNA solution in PBS (0.1–4.0mgml-1, depending on the construct) was injected into the lateral ventricle of E12.5 or E13.5 embryos with the use of a mouth-controlled pulled capillary micropipette. Five square electric pulses (30V for E12.5; 33V for E13.5) were delivered at one pulse per second (50-ms pulse followed by 950-ms gap) to embryos through the uterus with forceps-type electrodes ( CUY650P5, 5-mm diameter platinum round plates; Nepagene), while the uterus was kept wet with PBS.

Cell culture

Neurospheres were grown in serum-free medium containing B27 without vitamin A (Gibco) and 10ngml-1 FGF2. Trypsin-EDTA was used to dissociate primary neurospheres for the generation of secondary neurospheres. Neurosphere frequency and size were scored after one week.

For adherent cultures, CD133+/EGFPhi (the brightest 5% in the CD133+ range) or CD133+/GFPlo/neg (the dimmest 10% in the CD133+ range) cells were isolated by FACS and plated at a concentration of 5×104 cells per chamber in eight-well chamber slides (Nunc) precoated with 15μgml-1 poly-(l-ornithine) (Sigma) and 1μgml-1 fibronectin (Sigma). Cells were cultured in basal serum-free medium28 including 10ngml-1 FGF2.

Cell staining and flow cytometry

FACS was performed at the Johns Hopkins Flow Cytometry Core facility with a Becton Dickinson FACS-Vantage. For CD133 staining, phycoerythrin-conjugated CD133 antibodies were used (from eBioscience at 1:20 dilution, or from Miltenyi at 1:50 dilution). Cells were incubated with antibody for 30min on ice, washed three times with cold PBS and then sorted in cold PBS. For anti-Nestin (Developmental Studies Hybridoma Bank) staining of fixed cells a modified version of the Fix and Perm protocol (Caltag) was used.

Tissue and adherent culture staining

The antibodies used were mouse anti-βIII-tubulin ( TuJ1; Covance), rabbit anti-GFAP (Sigma), rat anti-BrdU (Accurate), goat anti-EGFP (Rockland), rabbit anti-DsRed (Clontech), goat anti-β-galactosidase (Biogenesis), rabbit anti-cleaved Notch1 (Cell Signalling) and mouse anti-CD133 (eBioscience and Miltenyi). Alexa Fluor-conjugated secondary antibodies (Molecular Probes) were used. For double immunostaining with DsRed2 and BrdU, BrdU was administered three times at 5-h intervals before harvesting of embryos. Sections were treated with 1M HCl for 30min at 37°C, and signal amplification was used to detect BrdU. Stainings were revealed with a Zeiss Axioskop with an Axiocam, or a Zeiss LSM 510. Images were processed with Adobe Photoshop.

Quantitative RT–PCR

Total RNA was isolated from samples with the use of RNeasy (Qiagen) and was reverse transcribed into complementary DNA, which was then quantified with an ND1000 spectrophotometer (NanoDrop). Quantitative real-time PCR was performed with an ABI 7900 Sequence Detection System (Applied Biosystems) and the SYBR green labelling system (Applied Biosystems). Primer sequences used are available from the authors on request. GAPDH expression was used to normalize the samples, and each sample was run in triplicate.

CBF1 shRNA knockdown

Short hairpins designed to target sequences of luciferase were subcloned into the pSiren retroviral vector (Clontech), which includes the reporter gene ZsGreen. Versions were also made containing DsRed2 as the reporter. Retroviral infection was used to knock down CBF1 in neurospheres, whereas in utero electroporation of the retroviral plasmids was used in vivo.

Statistical analysis

Entire litters (10–15 embryos) were pooled for each in vitro experiment, and experiments were performed on at least three different days. Thus, whereas in some of the presented histogram data n = 3 for each class, each of those n values represents the mean value derived from a population. In cases where n = 3, P values were determined with a stringent two-tailed heteroscedastic t-test, which, although not ideal for such n values, uniformly yielded highly significant P values. To corroborate these we used a non-parametric two-tailed Mann–Whitney U-test and obtained the best possible P value with this test for n = 3 (P = 0.1). In cases where n4, the two-tailed Mann–Whitney U-test was used to generate the P values shown. To compare gene expression levels, log2 fold changes were used.


These links to content published by NPG are automatically generated.


Numb, neurogenesis and epithelial polarity

Nature Neuroscience News and Views (01 Jul 2007)

Readers' Comments

If you find something abusive or inappropriate or which does not otherwise comply with our Terms and Conditions or Community Guidelines, please select the relevant 'Report this comment' link.

There are currently no comments.

Add your own comment

This is a public forum. Please keep to our Community Guidelines. You can be controversial, but please don't get personal or offensive and do keep it brief. Remember our threads are for feedback and discussion - not for publishing papers, press releases or advertisements.

You need to be registered with Nature and agree to our Community Guidelines to leave a comment. Please log in or register as a new user. You will be re-directed back to this page.