View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Dispatch R349

Neuronal differentiation: Proneural genes inhibit Sean J. Morrison

Proneural basic helix–loop–helix factors inhibit alternative cell fates in addition to promoting such as Neurogenin are activators of neuronal gene lineage-specific differentiation [5]. The demonstration that expression. Recent studies show they reinforce proneural bHLH proteins inhibit gliogenesis suggests neuronal differentiation by also inhibiting the that it may be a general strategy for lineage-determining expression of glial genes, and are required in vivo to transcription factors to inhibit alternative cell fates while prevent premature and excessive gliogenesis. acting as transcriptional activators to promote lineage-spe- cific differentiation. These observations may also provide Address: Howard Hughes Medical Institute, Departments of Internal Medicine and Cell and Developmental Biology, 3215 CCGC, insight into the mechanism behind the recent discovery University of Michigan, Ann Arbor, Michigan, 48109-0934, USA. that Notch activation can promote glial lineage determina- tion by neural stem cells [6–9]. As Notch inhibits the Current Biology 2001, 11:R349–R351 expression of proneural bHLH genes, including Mash1 [6] 0960-9822/01/$ – see front matter and Neurogenin [10], these recent papers suggest that © 2001 Elsevier Science Ltd. All rights reserved. Notch may promote gliogenesis in stem cells by relieving the inhibition imposed by proneural bHLH genes. Do you ever wonder how it is possible to make a commit- ment? For example, how do neural stem cells commit to What happens to cortical progenitors in the absence of a neuronal fate, steadfastly avoiding the temptations of proneural bHLH transcription factors? Nieto et al. [3] ana- gliogenesis? Our answer has been that stem cells express lyzed mice that were deficient for Mash1 and/or Neuro- master regulators of neuronal differentiation, the proneural genin2 and concluded that, in addition to reduced basic helix–loop–helix (bHLH) transcription factors, which , glial precursors are generated earlier and in promote the expression of a cascade of other genes that greater numbers in the double mutant mice. However, create neuronal identity [1]. The proneural bHLH genes only a minority of animals showed premature glial fibril- are necessary for the formation of many different types of lary acidic protein (GFAP) expression, suggesting that in the central and peripheral nervous systems. But loss of Mash1 and Neurogenin2 may accelerate glial is it enough for proneural genes to induce the expression lineage determination but that other factors determine of neuronal genes, or must they also employ other mecha- the timing of overt differentiation. Tomita et al. [2] nisms to keep stem cells from seductive second thoughts observed accelerated gliogenesis in the tectum and hind- of gliogenesis? Recent studies [2–4] indicate that, in addi- , using both early and late glial markers, but not in tion to promoting neuronal differentiation, proneural the of Mash1/Math3-deficient mice. Thus gliogen- bHLH transcription factors also act through independent esis is accelerated to different extents in different regions mechanisms to inhibit gliogenesis. of the , depending on which proneural genes are deleted. Proneural bHLH proteins regulate the onset and extent of gliogenesis in vivo. Tomita and colleagues [2] studied mice In vitro clonal analysis of cortical progenitors from that were deficient for the proneural genes Mash1 and Neurogenin2 and/or Mash1 deficient mice suggests that Math3 and observed increased gliogenesis in the tectum, Mash1 deficiency has a greater effect than Neurogenin2 hindbrain and , in addition to reduced neurogenesis. deficiency on glial lineage determination. Nieto et al. Nieto and colleagues [3] studied mice that were deficient [3] examined the relative frequencies of progenitors that for the proneural genes Mash1 and Neurogenin2 and observed formed neuron-only, glial-only, or mixed — multipotent — evidence of increased cortical gliogenesis in addition to colonies. Neurogenin2-deficient progenitors do not differ reduced neurogenesis. And now Sun and colleagues [4] in the types of colony they form relative to wild-type cells, have provided a mechanism by which proneural bHLH suggesting that Mash-1 compensates for the lack of Neu- factors may regulate gliogenesis: Neurogenin1 inhibits gli- rogenin2 in these mice. In Mash-1 deficient mice, the ogenesis by binding the CBP/Smad1 or p300/Smad1 tran- Neurogenin2-expressing subset of progenitors also form scriptional co-activator and sequestering it away from the colonies similar to wild-type cells, but the subset of cells promoters of glial genes, preventing transcription. that do not express Neurogenin2 exhibit a dramatic increase in clones that make only . This suggests that These studies have two important implications for under- Mash-1 and Neurogenin2 can largely compensate for standing how lineage determination occurs in stem cells. each other, but that in the absence of both proneural Transcription factors that are important regulators of lineage genes, at least one subset of cortical progenitors exhibits determination in the hematopoietic system consistently increased gliogenesis. R350 Current Biology Vol 11 No 9

Figure 1 and very few [3]. More data will be required to understand the significance of these observations. Notch LIF-R BMP-R A complementary paper by Sun et al. [4] describes mecha- nisms by which Neurogenin1 may inhibit glial lineage determination by multipotent progenitors. Sun et al. Neurogenin1 observed that Neurogenin1 strongly inhibits leukemia inhibitory factor (LIF)-induced glial differentiation by cor- STAT1/3 Smad1 tical progenitors in culture, even in cells that fail to undergo overt neuronal differentiation. This suggests that Neuro- CBP/ genin1 may inhibit gliogenesis independently of its ability p300 GFAP to promote neuronal differentiation. Within the GFAP Stat promoter, the STAT — ‘signal transducer and activator binding NeuroD of transcription’ — binding site is nec- site E-box essary for the full Neurogenin1-mediated inhibition of Glial GFAP expression; Neurogenin1 binding sites are not nec- differentiation Neuronal essary. This suggests that inhibition does not require DNA differentiation Current Biology binding by Neurogenin1.

Neurogenin1 inhibits gliogenesis (red arrow pathway) by binding and sequestering the CBP/p300/Smad1 transcriptional co-activator. In the The STAT binding site promotes GFAP transcription by absence of Neurogenin1, the CBP/p300/Smad1 complex is recruited binding a complex of STAT1/3, activated by LIF/ciliary to the promoters of glial genes by binding to activated STAT1/3. This neurotrophic factor (CNTF) signaling, the ubiquitously results in the expression of genes such as GFAP and glial expressed transcriptional coactivators CBP or p300, and differentiation. When present, Neurogenin1 binds the CBP/p300/Smad1 complex, recruiting it to the promoters of neuronal Smad1, activated by bone morphogenetic protein (BMP) genes (green arrow pathway), and sequestering it from the promoters signaling. Sun et al. [4] found that Neurogenin1 inhibits of glial genes. Neurogenin1 also inhibits the activation of STAT1/3. By the assembly of this complex by binding to both CBP and these two mechanisms, Neurogenin1 not only promotes the expression Smad1, interfering with the ability of CBP/Smad1 to bind of neuronal genes but inhibits the expression of glial genes. Consistent with this model, bone morphogenetic proteins (blue arrow) promote STAT3 (see Figure 1). On top of this effect, Neurogenin1 neurogenesis in the presence of Neurogenin1 and gliogenesis in the also blocks LIF signaling by preventing STAT activation. absence of Neurogenin1. Notch activation may promote glial lineage Thus, Neurogenin1 may inhibit glial lineage determina- determination by inhibiting the expression of proneural bHLH genes, tion by sequestering the CBP/p300/Smad1 transcriptional such as Neurogenin, and perhaps through other proneural gene- complex away from the promoters of glial genes, as well as independent mechanisms. This figure was adapted from Sun et al. [4]. LIF-R, leukemia inhibitory factor receptor; BMP-R, bone morphogenetic by suppressing STAT-mediated signal transduction. protein receptor; NeuroD, a neurogenic bHLH transcription factor, downstream of Neurogenin1. In addition to implicating Neurogenin1 as an inhibitor of gliogenesis, the work reported by Sun et al. [4] may also explain why BMPs promote neurogenesis by stem cells in So what happens in double mutant mice? Mice deficient some cases, while promoting gliogenesis in other cases. In for both Mash-1 and Neurogenin2 exhibit an increase in progenitors that express high levels of Neurogenin1, BMP mixed clones and a decrease in neuronal clones but no promotes neurogenesis, whereas in progenitors that express change in clones containing only glia [3]. Given the appear- low levels of Neurogenin1, BMP promotes gliogenesis. ance of an increase in the number of immature glia in This suggests that, in the presence of Neurogenin1, all the double mutants in vivo, and the increased gliogenesis in endogenous CBP/p300/Smad1 is bound by Neurogenin one subset of progenitors from the Mash-1-deficient mice, and diverted to neuronal promoters, preventing it from this failure to observe increased glial commitment in the interacting with STATs at glial promoters. In the absence double mutant is puzzling. One possible explanation is of Neurogenin1, CBP/p300/Smad1 is free to interact with that proneural genes are necessary to avoid premature glial STATs to activate glial promoters. lineage determination, but loss of proneural genes is not sufficient to cause glial commitment. Perhaps the progeni- Of course these interpretations require endogenous tors from double mutant mice are specified to undergo gli- CBP/p300 levels to be limiting. While Sun et al. [4] present ogenesis but have not yet committed to the glial lineage, considerable evidence to support this model, it has not so they form mixed colonies in culture but behave like yet been demonstrated that endogenous CBP/p300 levels immature glia in vivo. Consistent with this hypothesis is are limiting or that endogenous Neurogenin can bind to the observation that, unlike wild-type mixed clones, the all available CBP/p300. Nonetheless, the available data mixed clones from double mutant mice contain mainly glia suggest that Neurogenin1 is a molecular switch in stem Dispatch R351

cells that determines whether BMP signaling is gliogenic References or neurogenic. 1. Lo L, Tiveron M-C, Anderson DJ: MASH1 activates expression of the paired homeodomain transcription factor Phox2a, and couples pan-neuronal and subtype-specific components of autonomic The inhibition of gliogenesis by proneural bHLH genes neuronal identity. Development 1998, 125:609-620. 2. Tomita K, Moriyoshi K, Nakanishi S, Guillemot F, Kageyama R: has implications for understanding how Notch activation Mammalian achaete-scute and atonal homologs regulate promotes gliogenesis in stem cells [6,9]. As Notch inhibits neuronal versus glial fate determination in the central nervous the expression of proneural bHLH genes, Notch activa- system. EMBO J 2000, 19:5460-5472. 3. Nieto M, Schuurmans C, Britz O, Guillemot F: Neural bHLH genes tion may inhibit neurogenesis and relieve the inhibition control the neuronal versus glial fate decision in cortical of gliogenesis. But is this the whole story? Even transient progenitors. Neuron 2001, 29:401-403 Notch activation accelerates glial lineage determination 4. Sun Y, Nadal-Vicens M, Misono S, Lin MZ, Zubiaga A, Hua X, Fan G, Greenberg ME: Neurogenin promotes neurogenesis and inhibits in both neural crest stem cells [6] and central nervous glial differentiation by independent mechanisms. Cell 2001, system stem cells [9]. Could simply inhibiting neurogene- 104:365-376. 5. Orkin SH: Diversification of haematopoietic stem cells to specific sis and relieving the inhibition of gliogenesis account for lineages. Nature Rev Genet 2000, 1:57-64. irreversible and accelerated gliogenesis? Or must Notch 6. Morrison SJ, Perez S, Verdi JM, Hicks C, Weinmaster G, Anderson DJ: promote gliogenesis in additional ways that are indepen- Transient Notch activation initiates an irreversible switch from neurogenesis to gliogenesis by neural crest stem cells. Cell 2000, dent of its inhibition of proneural bHLH genes? 101:499-510. 7. Gaiano N, Nye JS, Fishell G: Radial glial identity is promoted by By analogy to other lineage determination factors, Notch Notch1 signaling in the murine forbrain. Neuron 2000, 26:395-404. 8. Furukawa T, Mukherjee S, Bao Z-Z, Morrow EM, Cepko CL: pathway genes might directly activate the transcription of rax, Hes1, and notch1 promote the formation of muller glia by glial genes in addition to inhibiting alternative fates. Tani- postnatal retinal progenitor cells. Neuron 2000, 26:383-394. 9. Tanigaki K, Nogaki F, Takahashi J, Tashiro K, Kurooka H, Honjo T: gaki et al. [9] found that constitutively active Notch is Notch1 and Notch3 intructively restrict bFGF-responsive capable of slightly promoting GFAP expression in the multipotent neural progenitor cells to an astroglial fate. Neuron absence of CNTF, and that this promotion does not 2001, 29:45-55. 10. Ma Q, Chen ZF, Barrantes IB, de la Pompa JL, Anderson DJ: require STAT3 activation or even the STAT binding Neurogenin 1 is essential for the determination of neuronal site in the GFAP promoter. This suggests that the pro- precursors for proximal cranial sensory ganglia. Neuron 1998, motion of gliogenesis by Notch does not depend on sig- 20:469-482. naling through the STAT pathway and is not entirely based on inhibiting the sequestration by Neurogenin of CBP/p300/Smad1 from the STAT binding site [4]. Thus Notch may promote gliogenesis in additional ways beyond the inhibition of proneural gene expression. As all things Notch related turn out to be unendingly complicated this should not be surprising.

It seems that commitment is engendered by the impetus to traverse one path, and the inability to access other paths. Proneural genes drive neuronal differentiation of stem cells while inhibiting glial differentiation. But any relationship is complex, and many details remain to be added to the picture sketched by the Greenberg, Guille- mot and Kageyama laboratories [2–4]. Can Mash1 sequester the CBP/p300/Smad1 transcriptional or inhibit STAT activation just like Neurogenin1? Do other proneural bHLH proteins have similar activities? Does Notch promote gliogenesis by inhibiting proneural gene expres- sion? Can Neurogenin2 or Mash1 play some positive role regulating glial differentiation or reinforcing commitment after glial lineage determination has occurred? This is a big step forward, but more work will be required before we really know how to make a commitment.

Acknowledgements I am supported as an Assistant Investigator of the Howard Hughes Medical Institute and am grateful for critical comments from David Anderson, Ben Barres, Genevieve Kruger, Theodora Ross, David Turner, Francois Guille- mot, Ryoichiro Kageyama, Yi Sun and Michael Greenberg.