Notch as a molecular switch in neural stem cells


  • Zhi-Dong Zhou,

    1. National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore 308433
    Search for more papers by this author
    • These authors contribute equally to this review.

  • Udhaya Kumari,

    1. National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore 308433
    Search for more papers by this author
    • These authors contribute equally to this review.

  • Zhi-Cheng Xiao,

    1. Institute of Molecular and Clinical Medicine, Kunming Medical College, China
    2. R&D China, GlaxoSmithKline, Shanghai, China
    Search for more papers by this author
  • Eng-King Tan

    Corresponding author
    1. National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore 308433
    2. Department of Neurology, Singapore General Hospital, Outram Road, Singapore 169608
    3. Duke-NUS Graduate Medical School, 8 College Road, Singapore 169857
    • Department of Neurology, Singapore General Hospital, Outram Road, Singapore 169608. Fax: 65 62203321
    Search for more papers by this author
    • Fax: 65 62203321

  • Recent unpublished data proves that inhibition of Notch activity by Caspr 4/LNX 2 signaling promote neurogenesis.


Notch signaling pathway, originally discovered in Drosophila, impinges on a wide array of cellular processes including cell proliferation, cell differentiation, and cell apoptosis. Recent accumulating evidences implicated the important roles of Notch signal pathway in different aspects of stem cell biology of neural stem cell (NSC). In vivo and in vitro studies illustrated that Notch signal pathway could promote gliogenesis, inhibit premature neurogenesis, and be involved in self-renew of NSC. This short review summarizes the roles of the Notch signaling pathway on gliogenesis, neurogenesis, and self-renew of NSC and their underlying molecular mechanisms. © 2010 IUBMB IUBMB Life, 62(8): 618–623, 2010.


In 1917, the Notch gene was discovered by Thomas Hunt Morgan and his colleagues who observed that partial loss of function of this gene results in Notches at the wing margin of Drosophila melanogaster (1). The Notch gene, which was cloned in the mid 1980s by groups of Artavanis-Tsakonas, encodes a type I single-pass transmembrane receptor with molecular weight about 300 kDa (2). There are four Notch receptors (Notch 1–4) and five transmembrane ligands (Jagged1, Jagged2, Delta-like1, Delta-like3, and Delta-like4) in mammals. Notch-1 (human chromosome 9q34) and Notch-2 (human chromosome 1p13-p11) are quite similar in structure, whereas Notch-3 (human chromosome 19p13.2-p13.1) lacks some of the domains found in other family members and encodes a considerably shorter intracellular domain (3, 4). In Drosophila, there is only one Notch-encoding gene, one Delta, and one Jagged homologue. In Caenorhabditis elegans, there are two genes encoding for Notch (Lin-12 and Glp-1) and three genes for DSL (Delta, Serrate, and Lag-2) ligand has APX-1, LAG-2, and DSL-1 (4, 5).

Notch molecule was composed with extracellular and intracellular regions. The extracellular region of the Notch receptor contains 10–36 epidermal growth factor (EGF)-like repeats essential for ligand binding and three copies of juxtamembrane repeats motif known as Lin-12-Notch Repeats (LNR) which modulate interactions between the extracellular and intracellular domains of Notch (6). The intracellular domain of Notch is involved in intracellular Notch activity. The intracellular region of Notch includes 6–7 ankyrin repeats flanked by nuclear localization signals, a proline, glutamine, serine, threonine-rich (PEST) domain, and a transactivation domain (TAD) (Fig. 1). Notch receptors are synthesized as single precursor proteins that will be cleaved in the Golgi by a Furin-like convertase during their transport to the cell surface. The cleavage of Notch protein by Furin-like convertase will yield an extracellular domain and a membrane-tethered intracellular domain. Then the two fragments are held together noncovalently and form heterodimers (7). On the continuous transportation through the trans-Golgi network, the bipartite form of Notch receptor will be further modified by glycosylation. After addition of first fucose by the enzyme O-fucosyl transferase, Fringe glycosyltransferases could further add N-acetylglucosamine to EGF-like repeats (8, 9). Thus the matured Notch receptor appears as a heterodimeric molecule at the cell surface with glycosylation imprints (10). The detailed structure of Notch receptors and their legends was illustrated in Fig. 1.

Figure 1.

Notch receptors and their ligands. Notch Receptors: There are four Notch receptors (Notch 1 – 4) in mammals. Notch receptors are expressed on the cell surface as heterodimeric proteins linked noncovalently. Their extracellular portion contains 36 EGF-like repeats that are associated with ligand binding, followed by three cysteine-rich LNR that prevent ligand-independent signaling and a heterodimerization domain. The intracellular portion of the receptor harbours two protein interaction domains, the RAM domain and seven ankyrin repeats (ANK), two nuclear localizations signals (NLS) and a TAD and a PEST sequence. Notch Ligands: Two transmembrane-bound ligands for Notch have been identified in Drosophila, named Delta and Serrate. Mammals have five DSL ligands (Jagged 1-2 homologous of Serrate and Delta-like 1, 3, and 4 homologous to Delta). Notch ligands, Delta and Jagged/Serrate, are composed of a DSL region responsible for the interaction with the Notch receptor and several EGF repeats. Jagged/Serrate also contains an extracellular cystein-rich region.


Notch signaling is activated on direct cell-to-cell contact as a result of interactions between Notch receptors and their ligands (Delta or Jagged) (11). At the molecular level, triggering of Notch receptor by ligand binding promotes two proteolytic cleavage events at the Notch receptor (Fig. 2) (11). The first proteolytic step after binding of Notch receptors to their ligands is carried out by metalloprotease tumor necrosis factor α-converting enzyme, also known as ADAM17, on extracellular part of the receptor. The cleaved extracellular subunit of the receptor is “trans-endocytosed” by the neighbouring ligand-expressing cells (12). This process seems to be controlled by Neutralized or Mindbomb E3 ubiquitin ligases (13). Binding of extracellular ligand to Notch also induces the second proteolytic cleavage event at the transmembrane region by a γ-secretase that is dependent on presenilin-1 (14, 15). This cleavage could release a membrane tethered form of the Notch intracellular domain (NICD) (16).

Figure 2.

Molecular events in Notch signaling pathway. Notch ligands and Notch receptors represent proteins with a single transmembrane spanning domain. During Notch protein maturation, the precursor of the Notch receptor is cleaved (S1) by a furin-like convertase in the trans-Golgi network and the resulting fragments are transported to the cell surface as a noncovalently linked heterodimeric receptor molecule. On binding of Notch receptors to ligand, an ADAM-type metalloprotease catalyzes a specific cleavage step (S2) at the extracellular part of Notch receptor. Subsequently, a further cleavage step (S3) catalyzed by a γ-secretase containing complex, take place at the transmembrane region. This will release the NICD which migrates to nucleus. In nucleus NICD forms a complex with the DNA-binding protein RBP-J. The latter act as a transcriptional repressor as it recruits histone deacetylase before NICD binding. But after complexed with NICD, the histone deacetylase is displaced and histone acetylase molecules are recruited which act to promote an alteration of chromatin conformation favourable to transcription of Notch targeting genes.

The liberated NICD is released from the plasma membrane and then translocates into the nucleus where NICD binds to recombination signal sequence-binding protein J (RBP-J, also called CSL (CBF-1, Su(H), and LAG-1)) via RAM domain and ankyrin repeats of NICD to activate the transcription of Notch targeting genes (17). In the absence of NICD, the DNA-binding protein RBP-J recruits corepressor complexes to represses transcription of Notch targeting genes. RBP-J interacts with the central corepressor protein SHARP (SMRT/Histone deacetylase Associated Repressor Protein-1(HDAC-1)). The SHARP serves as a protein interaction platform recruiting CtIP (CtBP-interacting protein)/CtBP (C-terminus binding protein) and additional corepressors and histone modifying enzymes (18). Thus, the expression of Notch targeting genes will be repressed in the absence of activated Notch signal. However, in the presence of NICD, the NICD interacts with RBP-J and recruits a coactivator complex composed of mastermind-like proteins (MAML-1) and other chromatin modifying transcription factors resulting in the transcriptional activation of Notch targeting genes (19). The Notch targeting genes included the Hairy-Enhancer of Split (HES) and HES-related proteins (HERP) genes, (20, 21) which are basic helix-loop-helix (bHLH) transcriptional regulators. The HES and HERP proteins could function to antagonize the activity of proneural genes like Mash1, Math, NeuroD, and Neurogenins (Ngn), whereas phosphorylation of NICD-facilitated by the mediator components Cyclin C/Cdk 8 kinase or ubiquitinylation of NICD via the E3 ubiquitin ligase Fbw 7/Sel 10 results in rapid NICD degradation and turnover of the NICD coactivator complex (22). Subsequently, RBP-J recruited corepressor complexes will form and shut-down the transcription of Notch targeting genes. The detailed molecular events for Notch signaling pathway was illustrated in Fig. 2.


Neural stem cells (NSCs) are defined as self-renewing, multipotent cells that can generate all major cell types of the adult central nervous system (CNS) (23). In 1992, NSC were first isolated and characterized in vitro by Reynolds and Weiss (24). There are two main types of NSC, CNS stem cells and neural crest stem cells (NCSCs). CNS stem cells are defined by their ability to give rise to neurons, astrocytes, and oligodendrocytes. NCSC are defined by their ability to give rise to neurons and glia of the peripheral nervous system (25). NSC is commonly cultured as neurospheres composed of a heterogenous population of undifferentiated NSC and cells at various stages of differentiation. NSCs can undergo differentiation via two major directions: Neurogenesis and Gliogenesis. In the developing mammalian nervous system, neurogenesis precedes gliogenesis. Consistent with this sequence, the glial differentiation pathways seems to be actively suppressed during the neurogenic period (26). Notch signal pathway is an important determinant of NSC cell fate during development and it is known to have multiple critical roles in the regulation of NSC differentiation throughout the neurogenic to gliogenic “switch” (27, 28). Accumulating evidence proves that Notch signal pathway promotes glial differentiation while it prohibits premature neuronal differentiation (29, 30).

Notch signal pathway could promote gliogenesis and inhibit neurogenesis of NSC directly or via cross-talk with STAT-signal pathway activated by diffusible signaling factors such as LIF (leukemia inhibitory factor), CNTF (ciliary neurotrophic factor), and BMPs (bone morphogenetic proteins) (31–35). The activation of Notch signal will lead to the release of NICD which then interacts with the DNA-binding protein, CSL, to activate the expression of HES genes and GFAP (Glial fibrillary acidic protein) (36–39). The HES genes inhibit neurogenesis and promote gliogenesis by hampering the functions of proneural genes: Mash1, Ngn1, and Ngn2 (35, 40, 41).

Gliogenesis can also be achieved with the aid of diffusible factors such as CNTF, LIF, and BMPs. When CNTF and LIF bind to their respective receptors, they trigger the phosphorylation of STAT1/3. The binding of BMP to its receptor induces the phosphorylation of Smad 1/5/8. Both of these phosphorylated transcriptional factors complexes together forming a common transcriptional complex (STAT-CBP (CREB binding protein)/p300-Smad1). This complex binds to the STAT responsive element in GFAP promoter which then leads to the expression of GFAP (42). On the contrary, the Ngn1 could function to inhibit JAK-STAT signaling pathway (35, 43). The detailed molecular mechanisms for Notch signaling pathway, JAK-STAT signaling pathway and gliogenesis of NSC are illustrated in Fig. 3.

Figure 3.

Molecular mechanisms of gliogenesis of NSC, regulated by Notch signaling pathway. NSC differentiate into glia in response to a variety of extracellular signals such as CNTF, BMPs, and Notch signaling. The activating of Notch signal will lead to the release of NICD, which could translocate to the nucleus and induce expression of HES and GFAP genes. The induced expression of HES could inhibit neurogenesis and promote gliogenesis via inhibition of proneuron transcription factor Ngn and cross-talk with JAK-STAT-signal pathway. The JAK-STAT-signal pathway can be activated by binding of LIF, CNTF, BMPs to their receptors, respectively. Binding of CNTF and LIF to their receptors could trigger the phosphorylation of STAT 1/3 while binding of BMPs to their receptors could induce phosphorylation of Smad 1/5/8. Finally in both cases, a common transcriptional complex (STAT-CBP (CREB binding protein)/p300-Smad1) will be formed which activate the expression of GFAP by binding to the STAT responsive element in GFAP promoter.

Apart from this, Notch pathway might directly activate the transcription of glial genes in addition to inhibiting alternative fates. In 2001, Tanigaki et al. found that constitutively active Notch is capable of slightly promoting GFAP expression in the absence of CNTF, and that this promotion does not require STAT3 activation or even the STAT binding site in the GFAP promoter (44). This suggests that the promotion of gliogenesis by Notch does not solely depend on the STAT signaling pathway and that is not entirely based on inhibiting the segregation by Ngn of CBP/p300/Smad1 from the STAT binding site (35). Thus, Notch may promote gliogenesis in additional ways beyond the inhibition of proneural gene expression.


Although the Notch signal pathway was implicated to promote gliogenesis of NSC, studies of the Notch signal pathway in in vivo mouse models also led to a prevailing view that Notch could function to maintain the self-renewable state of NSC (45). The Notch signaling alone was found to be enough to support the clonal self-renewing growth of NSC independently of exogenous and endogenous growth factors (GFs) (46). Furthermore, the Notch pathway is found to be required for both embryonic and adult forebrain NSC. At embryonic day (E) 14.5, Notch signaling is critical for all NSC to expand their populations by undergoing symmetric self-renewing divisions; this function is independent of the spatial context. Using presenilin-1(PS1) mutants as a model for decreased Notch signaling, the decreased Notch signal activity lead to reduced capacity of ganglionic eminence derived NSC to undergo symmetrical self-renewing divisions (47). It was found that over expression of Notch1 and HES1 could promote cell proliferation and self-renewal of neural precursors (37, 48). It was also found that HES1 and HES5 could promote self-renew of NSC, increase NSC proliferation in the embryonic telencephalon but inhibit their differentiation (49, 50). The HES genes can inhibit the activity of proneural genes such as Mash1, Math, and Ngn1 and allow NSC to undergo asymmetric division, simultaneously giving rise to neurons and more NSC (51). Therefore, regulation of the expression level of HES family of repressive bHLH transcription factors by activating Notch signaling pathway was proposed to be one molecular mechanism for regulation of self-renew of NSC by Notch. Second, the Notch signaling could also mediate the self-renewal of adult NSCs via modifying the length of the cell cycle. The cell cycle could serve as the gatekeeper to self-renew of NSC (47). The cell cycle itself not only controls the rate of NSC proliferation but also serves as a biological switch regulating self-renew and cell fate determination (47). Loss of Notch signaling in NSC could lead to increased cell cycle exit and decreased progenitor pool, whereas over activation of Notch signaling could decrease cell cycle exit and increase the progenitor pool. Third, in actively proliferating NSC, the Notch signaling pathway could contribute to the maintenance of the undifferentiated state and active self-renewing growth of NSC via collaboration with GFs or cytokines (52–55). It was demonstrated that FGF1 and FGF2 could upregulate the expression of Notch and decrease expression of Delta1 in cells (55). The CNTF and LIF, which activate the JAK-STAT signaling pathway in cells, were reported to promote the maintenance of NSCs in adult brain (52, 54). Furthermore, activation of JAK-STAT signaling pathway by CNTF in NSCs could rapidly increased Notch1 expression and activation of Notch1 signals (54). The regulation of the phosphorylation status of two critical residues of STAT3, S727 and Y705, was proposed to be one of the mechanisms underlying the cross talk between Notch, GF, and CNTF signals (54). The Notch signaling pathways to maintain self-renew, to regulate neurogenesis and gliogenesis of NSC was briefly summarized in Fig. 4.

Figure 4.

Notch signaling pathway regulates self-renewal, neurogenesis, and gliogenesis of NSC. Notch promotes NSC self-renewal as well as regulates gliogenesis and neurogenesis. Notch could contribute to maintenance of undifferentiated state and active self-renewing growth of NSC via collaboration with growth factors (FGF, Wnt, Shh, and TGF-β). On the other hand, activating Notch signaling pathway can promote gliogenesis and inhibit neurogenesis of NSC. Activation of Notch signaling pathway could directly promote gliogenesis of NSC or cross-talk with CNTF, LIF, and BMPs signaling pathway to facilitate gliogenesis. However, decreased Notch signaling activity will promote neurogenesis of NSC.


Up to till now, Notch mediated signaling pathways are found to be crucial for mammalian CNS development. Notch signaling pathway could not only function to inhibit differentiation and maintain NSC at progenitor state but also function to promote glial fate differentiation and inhibit neuronal commitment (56). These distinguishing effects of Notch activity on NSC can be explained as follows: First, modifications to cell cycle may alter the mode of division, which alleviates the inconsistency; second, the function of Notch signaling in cells could be dependent on the niche/environment in which it is acting. In fact, others have reported that the influence of Notch signaling on gliogenesis and self-renewal of NSC differs depending on the environmental context (45, 57). Third, the effects of Notch signaling in the embryo and adult also appear different. In embryos, Notch signaling is required for all NSC to undergo expansionary symmetric divisions, regardless of the cellular environment. However, in adult Notch signaling modulates the cell cycle time to prevent brain NSC exhaustion (47).

Recent findings suggest that Notch signaling pathway could be implicated to pathogenesis of neurodegenerative disorders (58). Therefore, further understanding of molecular mechanisms of Notch signaling in NSC biology should be significant to therapeutic strategies for neurodegenerative diseases in the future. Recent studies of in vivo animal models had implicated that agents targeting Notch signaling may work as therapeutic interventions to treat some neurodegenerative disorders (59). Therefore, further study of detailed roles of Notch signaling pathway in NSC biology may not only disclose the underlying molecular mechanisms of Notch signals in the process of NSC differentiation and self-renew but also provide new therapeutic agents to treat neurodegenerative disorders in the future.


Eng-King Tan, Zhi-Dong Zhou, Udahaya Kumari are supported by National Medical Research Council, Singapore Millennium Foundation, Singapore General Hospital and Duke-NUS Graduate Medical School.