Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2010 Jan 1.
Published in final edited form as: J Neurochem. 2008 Nov 15;108(1):11–22. doi: 10.1111/j.1471-4159.2008.05749.x

The Transcription Factor ATF5: Role In Neurodevelopment And Neural Tumors

Lloyd A Greene 1,*, Hae Young Lee 2,*, James M Angelastro 3,*
PMCID: PMC2680418  NIHMSID: NIHMS106318  PMID: 19046351

Abstract

We review recent findings regarding the properties of ATF5 and the major roles that this transcription factor plays in development of the nervous system and in survival of neural tumors. ATF5 is a widely expressed basic leucine zipper protein that has been subject to limited characterization. It is highly expressed in zones of neuroprogenitor cell proliferation. In vitro and in vivo studies indicate that it functions there to promote neuroprogenitor cell expansion and to suppress their differentiation into neurons or glia. ATF5 expression is down regulated by trophic factors and this is required for their capacity to promote neuroprogenitor cell cycle exit and differentiation into either neurons, oligodendroglia or astrocytes. ATF5 is also highly expressed in a number of tumor types, including neural tumors such as neuroblastomas, medulloblastomas and glioblastomas. Examination of the role of ATF5 in glioblastoma cells indicates that interference with its expression or activity causes them to undergo apoptotic death. In contrast, normal astrocytes and neurons do not appear to require ATF5 for survival, indicating that it may be a selective target for treatment of glioblastomas and other neural neoplasias. Further studies are needed to identify the transcriptional targets of ATF5 and the mechanisms by which its expression is regulated in neuroprogenitors and tumors.

Keywords: ATF5, transcription factor, glioma, neural progenitor, neurotrophic factors, development

Introduction

An exceptionally critical event in formation of the nervous system is the transition between proliferating progenitors and differentiated neurons and glia. Inappropriate regulation of this step will result in either too few or too many neural cells. Over the past several decades, it has become manifestly clear that the proliferation and differentiation of neuroprogenitor cells is regulated by specific sets of transcription factors and that these are themselves subject to regulation by extracellular signals (reviewed by Doe 2008). In addition, it has also become evident that such transcription factors and their regulatory pathways often play major roles in the generation, growth and survival of neural tumors (Clark et al., 2007). While a number of transcription factors that govern neuroprogenitor proliferation and differentiation have been identified, still others remain to be found and/or characterized in depth.

In this context, the transcription factor ATF5 has recently emerged as an important regulator of neuroprogenitor cell proliferation and neural tumor cell survival. The overall aim of this review is therefore to highlight what is currently known about ATF5 and its roles in neural tissues. In addition, because studies of ATF5 are relatively limited, this article will also aim to point out the many specific gaps in our knowledge about this transcription factor and its actions. It is the desire of the authors that these gaps will be seen as opportunities for those interested in better understanding nervous system development and/or treatment of neural tumors.

General properties of ATF5

This section considers the general properties of ATF5 and what is known about its regulation and actions. Although the bulk of such findings have been obtained in non-neuronal tissues, they nevertheless are likely to be relevant to ATF5's actions in neuroprogenitor cells and neural tumors.

Properties of ATF5 protein and transcripts and ATF5 tissue expression

First described by Nishizawa and Nigata in 1992, ATF5 (also initially referred to in some publications as ATFx) has been classified as a member of the activating transcription factor (ATF)/cAMP response-element binding protein (CREB) family (Hai and Hartman, 2001; Vinson et al. 2002). As other members of this family, ATF5 protein contains a C-terminal valine/leucine zipper (with 5 heptad repeats) that directs homophilic dimerization (Peters et al. 2001) as well as interactions with certain other leucine zipper proteins (Vinson et al. 2002; Al Sarraj et al. 2005). Also characteristic of the family, ATF5 possesses a basic DNA-binding domain that is N-terminal to the leucine zipper. Thus, ATF5 belongs to the group of basic-region leucine zipper (bZIP) proteins. ATF5 additionally contains a central proline-rich domain suggested to function in DNA transactivation (Hansen et al. 2002) and a non-conserved N-terminal sequence. Sequence comparison reveals that ATF5 is most similar to the bZIP protein ATF4, primarily within the leucine zipper domains, which are approximately 55% identical. On this basis, it has been proposed that ATF5 should be classified in the ATF4 subfamily of bZIP proteins (Vinson et al. 2002)

Two mRNAs (ATF5α and ATF5β) encoding the same ATF5 protein have been described that differ in their 5' UTRs (Hansen et al. 2002). The most 5' 182 nucleotides in the alpha transcript are replaced by a different 177-nucleotide sequence in the beta transcript. The alpha transcript appears to be predominant during development and in the adult. The significance of the two transcripts is not presently known, but as described below, they differ in their translational regulation.

Examination of the nucleotide sequence of ATF5 mRNA reveals two potential in-frame Kozak start sites encoding proteins of approximately 20-22 and 28-30 kDa (Angelastro et al. 2003). An endogenous protein of the former apparent mol wt is detected in PC12 cells, indicating preferential use of the second start site in this line (Angelastro et al., 2003). However, an endogenous ATF5 protein of apparent mol wt approximately 30 kDa was reported in other cell types including differentiated murine embryonic stem cells (Sampath et al. 2008) and HeLa cells (Wei et al. 2008). The differences in cellular distribution and function of the two translation products remain to be investigated.

ATF5 transcripts and protein are expressed in a wide variety of tissues during development and in the adult, with particularly high expression of transcripts in adult liver (Hansen et al. 2002; Pascual et al. 2008; Peters et al. 2001). As discussed below, within the brain, ATF5 is expressed by populations of progenitor cells. It is also present in a variety of tumor cell types and this topic will also be further addressed below.

Regulation of ATF5 expression

The means by which ATF5 expression is regulated at the transcriptional level are only partially understood at present. As reviewed below, both ATF5 mRNA and protein are down-regulated in PC12 and neuroprogenitor cells by differentiating growth factors (Angelastro et al. 2000; 2003). In contrast, serum-stimulated fibroblasts and interleukin-treated lymphocytic and bone marrow cells are reported to express ATF5 mRNA only during the G1/S phase of the cell cycle, while withdrawal of serum or interleukins results in a profound loss of ATF5 transcripts (Persengiev et al. 2002; Persengiev and Green 2003).

Within the adrenal glands and liver, ATF5 transcripts undergo a circadian change in expression (Lemos et al. 2007). This appears to be directly regulated by binding of the clock gene heterodimer CLOCK/BMAL1 to an E-box motif in the ATF5 promoter, leading to repression of ATF5 transcription (Lemos et al. 2007). Arsenite also induces ATF5 transcripts and interestingly, both the basal and stimulated levels of ATF5 message were much lower in cells null for ATF4, implicating the latter as a regulator of ATF5 expression (Zhou et al. 2008). Outside of such examples, the molecular bases for regulation of ATF5 transcripts are unknown and it will be of interest to further examine the mechanism by which growth factors regulate ATF5 message levels.

In addition to transcriptional control, ATF5 is subject to translational regulation. Recent findings reveal that as in the case of ATF4, ATF5 is markedly induced by a variety of stresses including heat shock (Wang et al. 2007), ER stress (Zhou et al. 2008), arsenite exposure (Zhou et al. 2008), and amino acid starvation (Watatani et al. 2007; 2008). In at least several of these instances regulation occurs by a novel mechanism involving translational control (Zhou et al. 2008; Watatani et al. 2007; 2008). Similarly to ATF4, the 5' UTR of the ATF5 alpha transcript has two untranslated open reading frames. The first promotes ATF5 translation while the second represses translation by reducing association of ATF5 transcripts with ribosomes. During stress conditions (and in response to phosphorylation of the eukaryotic initiation factor eIF2), the second site is bypassed, leading to efficient translation of ATF5 messages (Zhou et al. 2008; Watatani et al. 2007; 2008). The translational efficiency and ribosomal association of ATF5 transcripts are also greatly enhanced when embryonic stem cells are stimulated to differentiate into embryoid bodies by removal of leukemia inhibitory factor (Sampath et al. 2008). Watatani et al. (2008) examined whether the ATF5β transcript, which lacks the same open reading frames present in the α transcript, is subject to a similar mechanism of translational regulation and found that this is not the case. The forms of ATF5 transcripts present in neuroprogenitor cells and in tumors have not been reported at present and it will be of interest to determine whether they are differentially responsive to growth factor manipulation, to stresses and/or to translational control and what the consequences of this might be regarding differentiation and survival.

In addition to regulation at the level of translation, ATF5 expression is regulated by degradation. Pati et al. (1999) reported that ATF5 is targeted by the E3 ligases Cdc34 and Rad6 for proteasomal turnover. It was also recently shown that Cdc34-dependent degradation of ATF5 is inhibited by cisplatin (Wei et al. 2008).

Post-translational regulation of ATF5 activity

The capacity for transcriptional activation by certain members of the ATF/CREB family such as CREB and ATF2 requires their phosphorylation at Ser/Thr residues in the transactivation domain (reviewed by Lonze and Ginty 2002). In contrast, as ATF4, ATF5 contains a constitutively active transcriptional activation domain that does not appear to require phosphorylation (Al Sarraj et al. 2005). On the other hand, the observations (Peters et al. 2001) that ATF5 can be tyrosine phosphorylated in vitro by src and that it binds and can be dephosphorylated (at the src site(s)) by PRL-1, a tyrosine phosphatase, raise the possibility that ATF5 may be affected by phosphorylation at Tyr residues. Other potential post-translational mechanisms of ATF5 regulation have not been reported.

Transcriptional regulation by ATF5

Several studies have begun to examine potential transcriptional targets of ATF5. A number of ATF family members bind to conserved CREs (cAMP Response Elements) (Hai and Hartman, 2001). Pati et al. (1999) reported that ATF5 blocks activation of a CRE reporter by CREB, indicating that ATF5 represses CRE-dependent transcription, and Peters et al. (2001) found that ATF5 bound to a CRE sequence as a homodimer. In line with a repressive action on CRE sites, ATF5 over-expression repressed baseline activation of a CRE reporter in NGF-treated PC12 cells while the capacity of ATF5 to suppress neurite outgrowth in such cells was over-ridden by a constitutively active CREB construct, VP16-CREB (Angelastro et al. 2003). On the other hand, in another study, ATF5 failed to activate a CRE reporter (Al Sarraj et al. 2005). These findings are thus consistent with the possibility that ATF5 does not activate, but can repress genes with CRE sites or blocks their activation by CREB and other CRE binding transcription factors. In contrast to such a conclusion, however, it was reported that ATF5 over-expression elevates transcripts for Hsp27 in cardiac myocyte H9c2 cells and activates an Hsp27 promoter-reporter, in part via a CRE site (Wang et al. 2007). Evidence was also provided that ATF5 binds the Hsp27 promoter when over-expressed and that silencing of ATF5 blocks induction of Hsp27 in response to heat shock (Wang et al., 2007). These findings identify Hsp27 as a possible direct target of ATF5 and indicate that Hsp27 regulation is at least in part mediated by CRE sites. The role of CRE sites in ATF5 actions thus needs to be further defined and could be heavily dependent on cellular context, an issue of particular importance in the nervous system.

Over-expression studies have found that ATF5, as ATF4, stimulates expression of a promoter-reporter for arginine synthetase and by binding to a nutrient sensing response element (NRSE) that is distinct from CRE sites (Al Sarraj et al. 2005). Such findings have not been confirmed under more physiologic conditions. In another set of studies, ATF5 over-expression in cultured hepatocytes co-operated with the constitutive androstane receptor to elevate transcription of CYP2B6, a member of the P450 family (Pascual et al. 2008). However, it was not established whether CYP2B6 is a direct or indirect transcriptional target of ATF5 and if such regulation occurs with endogenous ATF5. Yet another investigation found that over-expressed ATF5 induced Cyclin D3 expression as well as a Cyclin D3 promoter-reporter (Wei et al., 2006). In this case, activation of the reporter was abolished by mutation of an E2F1 binding site, leading the authors to suggest that ATF5 cooperated with E2F1 to induce Cyclin D3. Here again, however, it was unclear whether the effect was direct or indirect and whether it also occurs in response to the endogenous protein.

As a leucine zipper transcription factor family member, it can be anticipated that ATF5 regulates transcription as a homodimer and/or by forming heterodimers with other members of the leucine zipper protein family (Vinson et al. 2002). This may be especially important since the binding partners for ATF5 may define its transcriptional activities in various contexts. In the initial discovery of ATF5, it was found as a binding partner (in blot studies) of the G-CSF gene promoter element 1-binding protein (GPE1-BP), a leucine zipper motif protein (Nishizawa and Nigata 1992). Studies with over-expression of d/n forms indicate that ATF5 can heterodimerize with ATF4 and C/EBP (another group of bZIP transcription factors), but not with CREB, Fos, Jun or ATF2 (Al Sarraj et al. 2005). Of the C/EBPs, a coiled-coil protein array analysis indicated strongest interaction with C/EBPγ (Newman and Keating 2003). Studies of endogenous ATF5 partners, however, have not been reported. Thus, in summary, the DNA domains to which ATF5 binds, its direct gene targets and its transcriptional binding partners remain to be more definitively characterized both within and outside of the nervous system.

Non-transcriptional binding partners for ATF5

Aside from associating with potential transcription factor partners, there are also several reports that ATF5 interacts with additional types of proteins. Two-hybrid screens revealed that ATF5 binds to GABAb receptors (White et al. 2000), to the tyrosine phosphatase PRL-1 (Peters et al. 2001) and to Disrupted in Schizophrenia 1 (Disc1) (Morris et al. 2003), a product of a gene that has been implicated in schizophrenia and bipolar disorder. In each case, binding appeared to depend on the ATF5 leucine zipper domain and it is not surprising therefore that ATF4 was also found to interact with GABAb receptors and Disc1. The PRL-1 interaction could reflect ATF5 as a substrate for this phosphatase, particularly in view of the demonstration that the latter can dephosphorylate ATF5 in vitro (Peters et al. 2001). The association of ATF5 with GABAb receptors and Disc1 is intriguing, but it has not been established whether such interactions of the endogenous proteins actually occur in neurons.

Functional roles for ATF5

Apart from its actions in neuroprogenitor cells that will be described below, little is known about the functional roles of ATF5. Persengiev et al. (2002) noted that ATF5 levels dramatically fall in lymphocytes and HeLa cells after withdrawal of trophic support and that this correlates with the onset of apoptotic death. In support of a role for ATF5 in promoting survival of such cells, constitutive ATF5 expression was protective while transfection with a d/n ATF5 induced death. On this basis, it was proposed that ATF5 plays an essential role in cell survival (Persengiev et al. 2002; Persengiev and Green 2003). A role for ATF5 in cell survival is also suggested by its capacity to protect cardiomyocytes from death induced by heat shock (Wang et al. 2007). However, while a pro-survival action of ATF5 appears to pertain to certain cell types and for at least several types of tumor cells (Angelastro et al. 2006; Monaco et al. 2007; see below), this does not appear to be universally the case and there are a number of examples in which cells survive in absence of apparent ATF5 function or expression (Angelastro et al. 2003; 2005; 2006; Mason et al. 2005; Monaco et al. 2007). Another potential role for ATF5 is in cell proliferation. While manipulation of ATF5 levels and activity in lymphocytes and HeLa cells did not show any effect on cell cycle (Persengiev et al. 2002), as discussed below, ATF5 does regulate proliferation of neuroprogenitor cells and of astrocytes (Angelastro et al. 2003; 2005; 2006; Mason et al. 2005). A third cellular activity that ATF5 may affect is state of differentiation. As reviewed in a following section, ATF5 appears to inhibit differentiation of neural progenitors into neurons and glia. In contrast, ATF5 levels fall as hepatocytes de-differentiate in culture and it has therefore been suggested to play a positive role in their differentiation (Pascual et al. 2008). ATF5 expression also increases during the early phase of chondrogenesis (Shinomura et al. 2006) and during embryonic stem cell differentiation to embryoid bodies (Sampath et al. 2008) and in each case, it has been speculated that this induction contributes to the differentiation process. The contrasts in the functional roles of ATF5 in different cell types again underscores the potential importance of cellular context in its actions.

The structural similarity between ATF4 and ATF5 raises the intriguing question as to whether they may perform similar or redundant functions. However, there have been no systematic studies to address this issue. For instance, while ATF4 plays a major role in cellular/neuronal responses to stresses (Ameri and Harris 2008; Lange et al. 2008), there is currently no evidence that ATF5, which can also be induced by stress, has similar actions. Conversely, there are presently no findings suggesting that, as described below for ATF5, that ATF4 participates in neuronal development or survival of glioblastoma cells.

Expression and functional roles of ATF5 in the nervous system

Precisely regulated spatiotemporal differentiation of neural progenitors is essential for proper nervous system development. This portion of the review will focus on emerging functional studies indicating that ATF5 plays a key role in regulating neural progenitor cell differentiation and proliferation.

ATF5 transcript expression in sensory odorant neurons of the olfactory epithelium

Expression of ATF5 in the developing mouse nervous system was initially detected in the olfactory epithelium (OE) by in situ hybridization (Hansen et al. 2002). ATF5 transcripts were observed at E11.5 in cells lining the olfactory placode, which constitutes the primordium of the OE. Expression remains robust in the OE and in the vomeronasal organ at least through day E18.5. At postnatal day 5, strong ATF5 expression remains in the OE. This is confined to layers of sensory neurons in the middle of the OE and is excluded from the non-neuronal supporting cells in the apical zone and in the progenitor cells in the basal layer. In contrast, ATF5 signal is low to undetectable in the vomeronasal organ at P5. ATF5 RNA expression was not examined in the adult OE.

On the basis of its spatiotemporal expression, Hansen et al. (2002) suggested that ATF5 may play a role in odorant sensory neuron differentiation and function. To date there have been no follow-up studies on this intriguing possibility. The rodent OE is capable of prolonged neurogenesis, beginning at E10 and continuing throughout adulthood (Murdoch and Roskams 2007). Given the apparent importance of ATF5 in regulating neuronal progenitor proliferation and differentiation in the telencephalon as reviewed below, it would be interesting to determine if ATF5 also plays a role in development as well as regeneration of adult odorant sensory neurons in the OE.

Hansen et al. (2002) also noted that they did not detect ATF5 transcripts in tissues other than the OE and vomeronasal organ in mid to late gestation mouse embryos. This contrasts with reports of the presence of ATF5 transcripts and protein in the developing CNS (Angelastro et al., 2003; 2005; Mason et al., 2005) and of ATF5 transcripts in the developing limb bud (Shinomura et al. 2006). This discrepancy may be due to a very high level of AFT5 transcripts in the OE. During initial in situ characterization of ATF5 expression in brain (Angelastro et al., 2003), a very marked hybridization was observed in the OE that was much stronger than in the telencephalic ventricular zone (JMA, C Mendelsohn, and LAG, unpublished data). This suggests that, for reasons that are currently not understood, ATF5 expression may be unusually high in the developing OE compared with other tissues.

ATF5 in PC12 cells and its implication in NGF-promoted neuronal differentiation

The potential involvement of ATF5 in neuronal differentiation was first suggested by a serial analysis of gene expression (SAGE) study that compared PC12 pheochromocytoma cells before and after a 9-day exposure to NGF. In absence of NGF (Angelastro et al. 2000), PC12 cells proliferate and resemble the immature progenitors of adrenal chromaffin cells and sympathetic neurons. In presence of NGF, they exit the cell cycle and attain many properties associated with post-mitotic sympathetic neurons (Greene and Tischler 1976). Among the identified regulated genes, ATF5 transcripts, which were among the most highly expressed in the cells before NGF exposure, fell by 25-fold following NGF treatment.

A follow-up study (Angelastro et al. 2003) revealed that the temporal decrease in ATF5 transcripts in PC12 cells that occurs with NGF treatment is paralleled by a fall in expression of ATF5 protein as well as by the outgrowth of neurites. To evaluate the functional role of ATF5 down-regulation in NGF-induced neuronal differentiation, the protein was constitutively expressed in PC12 cells. This blocked cell cycle exit and neuron differentiation.

Loss-of-function experiments with a d/n-ATF5 were also carried out. The d/n protein lacks N-terminal acidic activation and DNA binding domains and contains an enhanced b-Zip domain. Though incapable of directly binding DNA, this protein should strongly interact with endogenous ATF5 and its heterologous binding partners, thereby blocking ATF5 function (Angelastro et al. 2003; Vinson et al. 2002). Consistent with an inhibitory role for ATF5 in neuronal differentiation, the d/n-ATF5 accelerated NGF-promoted neurite outgrowth (Angelastro et al. 2003). Because the d/n construct should bind the same partners as wt ATF5, this result also rules out the possibility that the responses to ATF5 over-expression are merely due to a non-specific “squelching” effect due to sequestration of other transcription factors. Significantly, ATF5 loss-of-function was not sufficient on its own to promote neuronal differentiation without the presence of NGF. Thus, down-regulation of ATF5 appears to be necessary, but not sufficient for the differentiation process. Taken together, the studies of ATF5 in PC12 cells indicate that ATF5 is a negative regulator of neuronal differentiation and that down-regulation of ATF5 by an extrinsic factor (in this case, NGF) is necessary for neuronal differentiation.

An initial study of the mechanism by which ATF5 inhibits neuronal differentiation indicated a role for repression of CRE-mediated gene transactivation (Angelastro et al. 2003). In line with the previous reports that ATF5 binds CRE sites (Peters et al. 2001) and blocks cAMP-mediated activation of a CRE reporter (Pati et al. 1999), over-expressed ATF5 (but not of d/n-ATF5) repressed the basal activity of a CRE reporter in PC12 cells. Moreover, a strongly driven constitutively active CREB construct (VP16-CREB) over-rode not only this effect, but also the capacity of over-expressed ATF5 to suppress neurite outgrowth.

ATF5 and Neural progenitors

The potential similarity of PC12 cells to neural progenitors led to an examination of expression and function of ATF5 in the developing telencephalon. Angelastro et al., (2003) detected ATF5 mRNA expression in E12-E15 rat embryonic brain utilizing in situ hybridization. ATF5 transcript expression was highest in the ventricular zone (VZ) of the neural epithelium, and was decreased in overlying structures containing migrating and postmitotic neurons. Examination of ATF5 protein expression by immunohistochemistry showed strong nuclear expression in the VZ of the E12 (Figure 1), E14, and E17 rat telencephalon. In contrast, expression was undetectable towards the surface of the developing cortex, where cells express the postmitotic neuronal marker, Tubulin βIII (TUJ1). ATF5 also appears to be undetectable in neurons in the post-natal brain (Angelastro et al. 2003; 2005; Mason et al. 2005). During the early postnatal period, ATF5 expression persists in the remaining VZ and is present in a subpopulation of cells in the subventricular zone (SVZ) (Angelastro et al. 2005). Strong expression of ATF5 protein has also been observed in the post-natal cerebellar external granule cell layer, indicating that ATF5 expression in neural progenitors is not confined to the telencephalon (HYL, JMA, Carol Mason and LAG, unpublished). As in the cortex, ATF5 does not appear to be expressed in post-mitotic neurons in the cerebellum.

Figure 1.

Figure 1

Open in a new tab

Expression of ATF5 in E12 rat telencephalon as shown by immunostaining. Note the high expression of ATF5 (red) in the ventricular zone and its exclusion form the cortical area of high of tubulin βIII expression (green), a marker for differentiated neurons. Coronal section.

ATF5 as a marker for neural stem cells was further supported by its presence in AC133 and nestin positive progenitor cells at the center of neurospheres (Angelastro et al. 2003). In such cultures, ATF5 expression was nuclear and was excluded from cells expressing a postmitotic neuronal marker.

Gain- and loss-of-function experiments were carried out to assess the function of ATF5 in neural progenitors (Angelastro et al., 2003). Similar to its effect on PC12 cells, expression of exogenous ATF5 in cultured embryonic (E14) telencephalon progenitors suppressed neurogenesis. Furthermore, constitutive expression of ATF5 inhibited the appearance of the postmitotic neuronal marker tubulin βIII (TUJ1), but increased the proportion of cells with the progenitor marker nestin. In contrast, loss-of-function promoted either by d/n-ATF5 or with ATF5-siRNA significantly enhanced the proportion of neurite-bearing TUJ1-postive cells in the cultures.

To investigate the role of neurotrophic factors in regulating ATF5 expression in cortical neural progenitor cells, cultures were exposed to NT3, a factor that promotes their differentiation into neurons (Ghosh and Greenberg 1995). NT3 led to down-regulation of endogenous ATF5 expression and to a concomitant 3-fold enhancement of neurogenesis (Angelastro et al. 2003). Expression of exogenous ATF5 suppressed NT3-induced neurogenesis while d/n-ATF5 substantially accelerated neuronal differentiation in the absence of NT3, with no significant effect on neurogenesis in the presence of NT3. Thus, in parallel with PC12 cells, in telencephalic progenitor cells, ATF5 blocks neuronal differentiation and drives proliferation while a neurotrophic factor down-regulates ATF5, thereby permitting neuronal differentiation.

ATF5 regulates differentiation of astrocyte progenitors

In addition to neurons, a portion of VZ neural stem and neuroprogenitor cells differentiate into astrocytes. The universal presence of ATF5 in VZ cells thus led to a consideration of its possible role in astrocyte genesis (Angelastro et al. 2005). Examination of postnatal rat forebrain (P9) revealed that ATF5 continues to display nuclear expression in the VZ as well as within a subpopulation of cells in the SVZ, a structure that produces a second wave of astrocytic and oligodendrocyte precursors as well as neurons (Marshall et al., 2003). The Identity of the ATF5 positive cells in the SVZ was not examined and it would be of particular interest to know whether these constitute a distinctive sub-population. ATF5 expression was not seen in mature cortical astrocytes.

To assess the potential role of ATF5 in the astrocyte lineage, its expression was studied in plated neurospheres cultured under conditions that permit neuronal and astrocytic differentiation (Angelastro et al. 2003). Whereas undifferentiated neural stem cells showed high ATF5 expression, mature GFAP positive astrocytes showed none. However, a population was also observed that had low levels both of nuclear ATF5 and of cytoplasmic GFAP. This was interpreted to indicate that such cells were in transition between undifferentiated neural progenitors and differentiated astrocytes and that ATF5 was therefore present in progenitor cells destined to become astrocytes. Subsequent in vitro experiments with cultured E14/E15 telencephalic neural progenitors revealed that constitutive expression of ATF5 suppressed their differentiation into GFAP+ astrocytes and maintained them in a proliferating, nestin positive state. Conversely, loss of ATF5 function accelerated the differentiation of such cells into astrocytes (Angelastro et al. 2003).

As in the cases of NGF and NT3 in neuron differentiation, CNTF a factor that promotes differentiation of neural progenitors into astrocytes, down-regulated expression of ATF5 in cultured neural progenitors. Constitutive expression of ATF5 also blocked the capacity of CNTF to drive astrocytic differentiation. The effects of CNTF and of ATF5 loss-of-function on astrocyte differentiation were similar in that both promoted astrocytic differentiation of neural progenitors and were not additive when applied together. These observations are consistent with the conclusion that ATF5 blocks neuroprogenitor differentiation into astrocytes (as well into neurons) and that CNTF drives astrocyte differentiation at least in part by down-regulating ATF5 expression.

Manipulation of ATF5 was also carried out with PSA-NCAM positive neural progenitor cells cultured from the P1 rat forebrain SVZ. As with VZ cells, d/n-ATF5 promoted differentiation of SVZ cells into non-proliferative neurons and astrocytes while constitutive expression of ATF5 blocked such differentiation and maintained the cells in a proliferative state. Surprisingly, the SVZ cells with constitutive ATF5 expression appeared to revert to a VZ-cell-like phenotype. Thus, these cells were all positive for β-catenin, a VZ cell marker largely absent from SVZ cells, and many expressed nestin and GFAP, which are also much more commonly expressed in the VZ than the SVZ.

Cell culture experiments were extended to the intact developing CNS by infecting proliferating P1 rat forebrain SVZ cells with retroviruses expressing ATF5 and d/n-ATF5 (Angelastro et al. 2005). After 7 days, cells infected with a control virus were distributed among the SVZ, corpus callosum, cortex, and olfactory bulb, with none detected in the VZ. With ATF5-expressing retrovirus, in contrast, nearly 70% of cells resided in the VZ, and the remaining population was within the SVZ. Characterization by morphology and with markers further revealed that, as in vitro, constitutive expression of ATF5 in SVZ cells in vivo blocks their differentiation and causes them to revert to a VZ-like neural progenitor phenotype. Because both VZ and SVZ progenitor cells normally express ATF5, it is presently unclear why constitutive expression of this protein in SVZ cells appears to be sufficient to revert them to VZ cells. One possibility is that this is due to over-expression and that there are graded effects of the level of ATF5 expression on progenitor cell properties. Another is that the effect is due to timing and that prolonged ATF5 expression favors a VZ phenotype. It will be illuminating to determine the underlying mechanism of this effect, especially with respect to understanding the relationship between VZ and SVZ cells and the transition between these two populations. In addition, this phenomenon may be relevant to the role of ATF5 in neural tumors.

The effect of sustained in vivo expression of ATF5 was also examined in SVZ cells (Angelastro et al. 2005) After 3.5 months, cells infected with control virus were relatively few in number and most had moved to the white matter or cortex with the morphology of mature glial cells. In contrast, the majority of ATF5-infected cells remained next to the ventricle and proliferated to form a large multilayered, hyperplasic mass. The cells in these masses stained positively for β-catenin, nestin and GFAP, supporting the idea that they represent a sustained hyperplasic growth of VZ-like neural progenitor cells.

Retroviral expression of d/n-ATF5 also altered the phenotype and distribution of infected SVZ cells in vivo so that they showed accelerated glial differentiation and the nonmigratory behavior of mature astrocytes (Angelastro et al. 2005). When examined one week after infection, most had the shape of process-bearing oligodendrocytes and astrocytes. Fewer (compared with controls) expressed β-catenin and more were GFAP positive. Moreover, unlike cells infected with control virus that migrated into white matter, striatum and cortex, cells infected with d/n-ATF5 were largely restricted to the border between the SVZ and the white matter.

ATF5 and Oligodendrocyte progenitors

Like neurons and astrocytes, oligodendroglia are derived from neuroprogenitor cells in the VZ and SVZ (Nicolay et al. 2007; Bongarzone 2008). Among the earliest proliferating and migratory precursors committed to the oligodendrocyte lineage are those bearing the markers O4 and NG2. These differentiate into O1+ oligodendrocytes, which then mature into non-proliferating myelin-forming oligodendroglia characterized by markers such as CC1.

Examination of ATF5 expression in oligodendrocytes in neonatal rat brain revealed nuclear expression in O4+ precursors, but no detectable expression in CC1+ mature oligodendroglia (Mason et al. 2005). O4+ cells also showed nuclear ATF5 expression in culture. When these differentiated into O1+ cells in vitro, they retained ATF5 expression, but this was almost exclusively cytoplasmic. It was unclear whether such expression by O1+ cells might be an artifact of culture, but in any case ATF5 would be unable to act as a transcription factor with such a localization.

As in the case of progenitors for neurons and astrocytes, ATF5 appeared to maintain oligodendrocyte precursors in a proliferative state while loss of ATF5 function accelerated their cell cycle exit and differentiation. Constitutive expression of ATF5 in cultured O4+ cells prevented their differentiation into O1+ cells and significantly increased the proportion that remained in the cell cycle (Mason et al. 2005). However, unlike the effect of constitutive expression of ATF5 in SVZ progenitors, the O4+ cells expressing exogenous ATF5 did not revert to a VZ-like phenotype nor did they express nestin or β-catenin. In contrast to the effects of over-expression, loss-of-ATF5-function caused cultured O4+ progenitors to leave the cell cycle and enhanced their differentiation into O1+ oligodendrocytes.

Stereotactic injection of retroviruses expressing ATF5 or d/n-ATF5 into the SVZ of P2 rat forebrain was used to investigate the role of ATF5 in oligodendrocyte development in vivo (Mason et al. 2005). As in the study by Angelastro et al. (2005), constitutive ATF5 expression blocked oligodendrocyte differentiation and the cells appeared to revert to a VZ neural progenitor phenotype. For cells infected with d/n-ATF5 virus for one week, none (in contrast to 30% of controls) expressed the oligo precursor marker NG2 while there was a doubling in proportion of those expressing CC1, a marker of mature oligodendroglia (Mason et al. 2005).

Taken together, in vitro and in vivo observations thus indicate that the role of ATF5 in NG2+/O4+ oligodendrocyte precursor cells is to maintain them in a proliferative state and to block their further differentiation to O1+ oligodendrocytes and into mature CC1+ oligodendroglia. Interestingly, once the O4+ state of differentiation is reached, constitutive ATF5 expression does not revert the cells back to a VZ-like phenotype as it does in the case of SVZ neural progenitors. ATF5 loss-of-function on the other hand, appears to cause pre-mature differentiation of O4+ progenitors and in vivo, presumably due to timing, this interferes with their proper migration.

General aspects of the role of ATF5 in neural development

During development of at least the forebrain, present findings indicate that ATF5 plays an important role in promoting proliferation and suppressing differentiation of progenitors for all three neural lineages. Figure 2 shows a scheme of ATF5 expression in various neural progenitors and mature neural cells. Significantly, ATF5 is down-regulated by growth factors that promote neuronal and glial differentiation. Nevertheless, ATF5 remains expressed in proliferating precursor cells committed to the oligodendrocyte lineage. Thus, there appears to be a degree of variation with respect to the effects of ATF5 expression on progenitor cell differentiation. The bases for these differences remain to be explored as well as the reasons why neural progenitor cells are able to enter the oligodendrocyte lineage despite expressing ATF5.

Figure 2.

Figure 2

Open in a new tab

Scheme depicting expression of ATF5 by various cell types during development of the cortex. Red nuclei depicts those with positive ATF5 expression; black nuclei are those without detectable ATF5 expression. Adapted from Mason et al. (2005).

An additional unsettled issue regards the relationship of ATF5 to other transcription factors that have been implicated in blocking cell cycle exit and differentiation of neural progenitor cells. This includes inhibitory members of the basic helix-loop-helix (bHLH) family such as Hes and Id proteins (reviewed by Ross et al. 2003) and the B1 group of Sox genes (Bylund et al, 2003). It will be important to establish whether ATF5 is up- or downstream of such factors or whether it works independently of them. Hes, Id and SoxB1 proteins all appear to act, at least in part, by interfering with the expression and actions of proneural bHLH genes such as MASH1 and neurogenins (Ross et al. 2003; Bylund et al. 2003) and it will therefore be of further importance to determine whether ATF5 has a similar activity.

ATF5 in neural tumors

Reasons for suspecting the presence of ATF5 in neural tumors

Findings that showed abundant ATF5 expression in the proliferative neural stem/progenitor cell population, but little or none in postmitotic neurons and nonproliferative astrocytes and oligodendrocytes, led to the hypothesis that ATF5 may also be highly and constitutively expressed in neural tumors. This was supported by the observation that constitutive expression of ATF5 in SVZ cells caused them to continue to proliferate to form a hyperplastic mass (Angelastro et al. 2005). Neural tumors are currently thought to arise from cancer stem cells (Vescovi et al. 2006) that can self-renew, but that can also give rise to tumor cell progeny that display arrested neuronal or glial differentiation. Taken together, the similarity between neural cancer stem cells and dividing, ATF5-positive neural progenitor cells suggested that ATF5 might be an appropriate candidate for exploration in the context of neural tumors.

ATF5 is expressed by gliomas and proliferating astrocytes

Initial immunofluorescence staining studies revealed nuclear ATF5 expression in all eight of rat and human glioma cell lines examined (Angelastro et al. 2006). This was confirmed by Western immunoblotting that detected the 22 kDa form of ATF5. In some of the glioma lines examined, virtually all cells were positive for ATF5 expression. ATF5 was also detected in nuclei of glioma cells undergoing mitosis. This suggests that in contrast to several other cell lines studied (Persengiev et al. 2002), ATF5 expression in glioma cells is not necessarily limited to the G1/S phase of the cell cycle.

The findings with glioma lines were extended to sections of human glioblastomas in which all 29 of the tumors examined were positive for ATF5 immunostaining (Angelastro et al. 2006). An example such expression in a human glioblastoma is shown in Figure 3. In contrast, there was no immunostaining of neurons in the same sections. Dong et al. (2005) also reported the presence of ATF5 in glioblastomas based on cDNA microarray screening. ATF5 was highly expressed in perinecrotic palisades which represent the most aggressive forms of malignant gliomas, and in a series of 28 tumors without perinecrotic palisades, the level of ATF5 expression, along with that of 4 other genes, negatively correlated with time of patient survival.

Figure 3.

Figure 3

Open in a new tab

Expression of ATF5 in a human glioblastoma. Sections from an excised, paraffin-imbedded human glioblastoma were immunostained for expression of ATF5 (brown).

The CD133 antigen is a stem cell marker that has been suggested to be a potential marker for glioma stem cells (Singh et al. 2004). Schrot et al. (2007) examined CD133 and ATF5 expression in a formalin-fixed post-mortem brain harboring a glioblastoma. The two antigens were co-expressed within the ipsilateral hemispheric ependymal and subependymal zones medial to the tumor, but not in the non-tumorous contralateral hemisphere. Such findings are consistent with the hypothesis that at least some gliomas may arise from enhanced generation of CD133+/ATF5+ stem cells in the lateral ventricles.

Immunostaining of human glioblastomas revealed some ATF5 positive cells at the periphery of the tumors with activated astrocyte-like morphology (Angelastro et al. 2006). To examine this further, cultured primary rat astrocytes were examined for ATF5 expression by immunostaining and Western immunoblotting (Angelastro et al., 2006). Astrocytes that had been passaged at least five times and that were highly proliferative, expressed ATF5. In contrast, astrocytes that had been passaged two times or less and that were not proliferative, did not express ATF5. This is consistent with the apparent absence of ATF5 expression by mature astrocytes (Angelastro et al. 2005) and with the idea that ATF5 expression correlates with astrocyte activation and proliferation and with glioblastoma cell proliferation.

Interference with ATF5 function or expression causes apoptotic death of glioblastoma cells in vitro and in vivo

Transfection with d/n-ATF5 was used to assess the role of ATF5 in cultured human and rat glioma cell lines and revealed that all seven lines assessed underwent what appeared to be massive apoptotic death within 5 days (Angelastro et al. 2006). Similar findings were achieved with siRNA-ATF5 and shRNA-ATF5, thus ruling out non-specific actions of the d/n-ATF5 construct. In contrast, d/n-ATF5 or si/shRNA-ATF5 did not cause death of several other non-gliomal cell lines, further indicating that the observed death of glioma cells was not due to non-specific actions of the ATF5 reagents. The pan-caspase inhibitor BAF suppressed such death, which further supported an apoptotic mechanism. The apoptotic effect of ATF5 was p53-independent because many of the lines assessed had mutated, non-functional p53 genes. Moreover, in the one cell line tested, expression of d/n-p53 did not protect from d/n-ATF5.

Interference with ATF5 function has also been carried out in an in vivo glioblastoma model in which striatal tumors formed by introduced C6 glioma cells were injected with control or d/n-ATF5-expressing retroviruses and then examined 3 days later for cell death via TUNEL staining (Angelastro et al. 2006). This revealed that 96% of cells in tumors infected with d/n-ATF5 virus were undergoing death as compared to less than 1% of cells infected with control retrovirus. In contrast, outside the periphery of the tumors, only 2% of cells (presumably activated astrocytes) infected with d/n-ATF5 were TUNEL positive. These findings thus indicate that interfering with ATF5 function efficiently kills glioma cells in an in vivo setting, but does not affect survival of proliferating cells outside the tumors. Additional experiments are ongoing to test whether an inducible d/n-ATF5 can suppress or eradicate induced endogenous gliomas in mice (JMA, unpublished).

Interference with ATF5 function or expression does not kill proliferating astrocytes, but causes them to exit the cell cycle

Proliferating astrocytes that had undergone multiple passages in culture and that expressed ATF5 were also subjected to transfection with d/n-ATF5 or to ATF5-siRNA. Unlike glioma cells, the proliferating astrocytes were spared from death caused by these manipulations (Angelastro et al. 2006). Such findings are consistent with the in vivo findings reviewed above that the survival of proliferating cells (most likely activated astrocytes based on morphology and distribution) outside the periphery of gliomas is also unaffected by interfering with ATF5 function. These observations, along with the absence of detectable ATF5 in brain neurons, indicate that inhibition of ATF5 function or expression in brain may lead to selective death of glioma cells, but not of non-tumor cells.

Multi-passaged astrocytes regain ATF5 expression, raising the question of whether ATF5 is required for their proliferation (Angelastro et al. 2006). Transfection with ATF5 promoted DNA synthesis by astrocytes whereas this was strongly inhibited by d/n-ATF5. In contrast, d/n-ATF5 did not diminish DNA synthesis by C6 glioma cells. These observations raise the possibility that one reason for the differential effect of ATF5 loss-of-function on survival of astrocytes and glioma cells is that the latter attempt to proliferate in the absence of ATF5 and that this triggers apoptotic events.

ATF5 and neural tumors in addition to gliomas

Preliminary studies of patient tissue have also revealed expression of ATF5 by neuroblastomas (JMA and LAG, unpublished observations) and medulloblastomas (HYL, JMA and LAG, unpublished observations). Intriguingly, ATF5 expression was absent from differentiated neuron-like cells within neuroblastomas (JMA and LAG, unpublished observations). The effect of interfering with ATF5 function on the survival of cells from such tumors remains to be determined although in one preliminary experiment, d/n-ATF5 promoted death of cultured SH-SY5Y human neuroblastoma cells (JMA and LAG, unpublished observations).

ATF5 expression in non-neural tumors and effects on survival

Immunohistochemical staining has also been used to examine ATF5 expression in breast cancer paraffin sections and a tumor paraffin section microarray (TMA) (Monaco et al. 2007). The TMA was composed of different types of tumors as well as of corresponding non-neoplastic tissue from different organs. Approximately 63% of all tumors tested were positive for nuclear ATF5 staining (defined as greater than 25% of tumor cells staining) as compared to only 32% of non-neoplastic tissues. Other studies have also reported ATF5 expression in follicular carcinoma of the thyroid (Barden et al. 2003), and in B-cell chronic lymphocytic leukemia (Mittal et al. 2007).

Detailed analysis of breast cancers revealed that ductal carcinomas (invasive and in situ) as well as lobular carcinomas (invasive and in situ) contain approximately 44% more ATF5+ nuclei than normal non-pathological breast tissue (Monaco et al. 2007). In normal breast tissue, ATF5 was found in the cytokeratin 5-luminal progenitor cells and in the more differentiated cytokeratin18 luminal ductal epithelium cells, but not in the stroma cells that surround the lumen (Monaco et al. 2007).

Determination of whether loss of ATF5-function can trigger apoptosis of breast cancer cells indicated that d/n-ATF5 triggered death of 5 neoplastic breast cell lines (Monaco et al. 2007). In contrast, there was little death in the two non-neoplastic lines that were assessed. Thus, as in the case of gliomas, neoplastic breast cancer cells appear to require ATF5 function for survival, whereas their non-transformed counterparts do not. It will be interesting in future to determine whether this distinction will hold for neoplastic and non-neoplastic cells of additional origins.

Is ATF5 a general regulator of cell survival and proliferation?

Given the roles of ATF5 in neuroprogenitor cells and tumors, it is useful to briefly return to the issues of whether ATF5 is a general requirement for cell survival and proliferation. The work of Persengiev et al. (2002) showed that ATF5 is essential for survival of FL5.12 lymphoid and HeLa cells and subsequent studies reviewed above revealed that it is also required for survival of a number of glioma and breast cancer cell lines (Angelastro et al. 2006; Monaco et al. 2007). However, there are now a number of counter examples in which ATF5 loss-of-function does not promote cell death including neuroprogenitor cells, PC12 cells, 293 cells, CAD cells, astrocytes, oligodendrocyte precursors, mouse ES cells and breast epithelial cells (Angelastro et al., 2003; 2005; 2006; Mason et al. 2005; Monaco et al., 2007). While in some cases there appear to be distinctions in the survival requirement for ATF5 based on the state of transformation or non-transformation, this is not a general rule. For instance, PC12 cells are neoplastic and do not appear to require ATF5 expression for survival while FL5.12 cells are non-transformed, but dependent on ATF5 for survival. Thus, it appears that the ATF5 survival requirement may well be cell type dependent, and further work to define the extent of this requirement seems highly warranted, especially in light of the potential of ATF5 as a therapeutic target.

With respect to the ATF5 requirement for cell proliferation, consideration of present findings support the conclusion that this too is dependent on cell type. ATF5 loss-of-function suppresses proliferation of neuroprogenitor cells, oligodendrocyte precursors and mature astrocytes while forced ATF5 expression enhances proliferation of the same cell types (Angelastro et al. 2003, 2005, 2006; Mason et al. 2006). The capacity of ATF5 to regulate the cell cycle gene Cyclin D3 also supports a role in proliferation (Wei et al., 2006). On the other hand, a d/n-ATF5 had no apparent effect on proliferation of HeLa and FL5.12 cells (Persengiev et al. 2002) or of C6 glioma cells (Angelastro et al. 2006). The mechanisms underlying the difference from cell type to cell type for the requirement of ATF5 for proliferation and survival remain to be defined. One possible basis is a cell-type-dependent difference in available binding partners, which may in turn influence the cellular responses that ATF5 engenders.

Summary and final remarks

Over the past several years, ATF5 has emerged as a key player in development of the nervous system and in survival of gliomas and other tumors. Nevertheless, our understanding of its biological functions and mechanisms of action are only fragmentary and there are many outstanding questions in this regard. For example, what role does ATF5 have in development of nervous system structures in addition to the cortex? When is ATF5 first present in the developing nervous system and what promotes its expression? What genes does ATF5 regulate that are responsible for maintaining neural progenitors in a proliferative state and in blocking their differentiation? What are the binding partners for ATF5 in regulation of transcription? Does this vary from cell type to cell type? What are the regulatory DNA elements with which ATF5 interacts? How do growth and differentiation factors down-regulate ATF5 in neural progenitors during development? Does ATF5 play a role in neurogenesis in the adult nervous system? Is it expressed by oligodendrocyte precursors that are present in the adult nervous system (Polito and Reynolds 2005) and does it plays a role in their maintenance and differentiation? If ATF5 is present in adult progenitors for neurons and oligodendroglia, might it be a potential target for manipulating their expansion and/or differentiation to drive neurogenesis or re-myelination for therapeutic purposes? What promotes the expression of ATF5 in tumors? By what mechanisms does ATF5 maintain the survival of tumor cells and by what mechanisms do they die in its absence? The answers to these and related questions are likely to provide us with deeper insight as to how neural progenitor cell proliferation and differentiation are regulated and may open the door to targeting ATF5 as a means manipulate neural stem cell expansion and to treat neural and other types of tumors.

Acknowledgements

Portions of the studies described here were supported by grants from the NIH-NINDS (LAG and JMA). HYL is supported by NIH Training Grant “Hormones: Biochemistry and Molecular Biology”, DK 07328.

Abbreviations

ATF

activating transcription factor

BAF

Asp(O-methyl)CH2F

CNTF

ciliary neurotrophic factor

CRE

cyclic AMP response element

CREB

cyclic-AMP response-element binding protein

ER

endoplasmic reticulum

GFAP

glial fibrillary acidic protein

OE

olfactory epithelium

SVZ

subventricular zone

VZ

ventricular zone

References

  1. Al Sarraj J, Vinson C, Thiel G. Regulation of asparagine synthetase gene transcription by the basic region leucine zipper transcription factors ATF5 and CHOP. Biol Chem. 2005;386:873–879. doi: 10.1515/BC.2005.102. [DOI] [PubMed] [Google Scholar]
  2. Ameri K, Harris AL. Activating transcription factor 4. Int J Biochem Cell Biol. 2008;40:14–21. doi: 10.1016/j.biocel.2007.01.020. [DOI] [PubMed] [Google Scholar]
  3. Angelastro JM, Klimaschewski L, Tang S, Vitolo OV, Weissman TA, Donlin LT, Shelanski ML, Greene LA. Identification of diverse nerve growth factorregulated genes by serial analysis of gene expression (SAGE) profiling. Proc Natl Acad Sci U S A. 2000;97:10424–10429. doi: 10.1073/pnas.97.19.10424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Angelastro JM, Ignatova TN, Kukekov VG, Steindler DA, Stengren GB, Mendelsohn C, Greene LA. Regulated expression of ATF5 is required for the progression of neural progenitor cells to neurons. J Neurosci. 2003;23:4590–4600. doi: 10.1523/JNEUROSCI.23-11-04590.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Angelastro JM, Mason JL, Ignatova TN, Kukekov VG, Stengren GB, Goldman JE, Greene LA. Downregulation of activating transcription factor 5 is required for differentiation of neural progenitor cells into astrocytes. J Neurosci. 2005;25:3889–3899. doi: 10.1523/JNEUROSCI.3447-04.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Angelastro JM, Canoll PD, Kuo J, Weicker M, Costa A, Bruce JN, Greene LA. Selective destruction of glioblastoma cells by interference with the activity or expression of ATF5. Oncogene. 2006;25:907–916. doi: 10.1038/sj.onc.1209116. [DOI] [PubMed] [Google Scholar]
  7. Barden CB, Shister KW, Zhu B, Guiter G, Greenblatt DY, Zeiger MA, Fahey TJ., 3rd Classification of follicular thyroid tumors by molecular signature: results of gene profiling. Clin Cancer Res. 2003;9:1792–1800. [PubMed] [Google Scholar]
  8. Bongarzone ER. Induction of oligodendrocyte fate during the formation of the vertebrate neural tube. Neurochem Res. 2002;27:1361–1369. doi: 10.1023/a:1021675716848. [DOI] [PubMed] [Google Scholar]
  9. Bylund M, Andersson E, Novitch BG, Muhr J. Vertebrate neurogenesis is counteracted by Sox1-3 activity. Nat Neurosci. 2003;6:1162–1168. doi: 10.1038/nn1131. [DOI] [PubMed] [Google Scholar]
  10. Clark PA, Treisman DM, Ebben J, Kuo JS. Developmental signaling pathways in brain-tumor-derived stem-like cells. Dev Dyn. 2007;236:3297–3308. doi: 10.1002/dvdy.21381. [DOI] [PubMed] [Google Scholar]
  11. Doe CQ. Neural stem cells: balancing self-renewal with differentiation. Development. 2008;135:1575–1587. doi: 10.1242/dev.014977. [DOI] [PubMed] [Google Scholar]
  12. Dong S, Nutt CL, Betensky RA, Stemmer-Rachamimov AO, Denko NC, Ligon KL, Rowitch DH, Louis DN. Histology-based expression profiling yields novel prognostic markers in human glioblastoma. J Neuropathol Exp Neurol. 2005;64:948–955. doi: 10.1097/01.jnen.0000186940.14779.90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Ghosh A, Greenberg ME. Distinct roles for bFGF and NT-3 in the regulation of cortical neurogenesis. Neuron. 1995;15:89–103. doi: 10.1016/0896-6273(95)90067-5. [DOI] [PubMed] [Google Scholar]
  14. Greene LA, Tischler AS. Establishment of a noradrenergic clonal line of rat adrenal pheochromocytoma cells which respond to nerve growth factor. Proc Natl Acad Sci U S A. 1976;73:2424–2428. doi: 10.1073/pnas.73.7.2424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hai T, Hartman MG. The molecular biology and nomenclature of the activating transcription factor/cAMP responsive element binding family of transcription factors: activating transcription factor proteins and homeostasis. Gene. 2001;273:1–11. doi: 10.1016/s0378-1119(01)00551-0. [DOI] [PubMed] [Google Scholar]
  16. Hansen MB, Mitchelmore C, Kjaerulff KM, Rasmussen TE, Pedersen KM, Jensen NA. Mouse Atf5: molecular cloning of two novel mRNAs, genomic organization, and odorant sensory neuron localization. Genomics. 2002;80:344–350. doi: 10.1006/geno.2002.6838. [DOI] [PubMed] [Google Scholar]
  17. Lange PS, Chavez JC, Pinto JT, Coppola G, Sun CW, Townes TM, Geschwind DH, Ratan RR. ATF4 is an oxidative stress-inducible, prodeath transcription factor in neurons in vitro and in vivo. J Exp Med. 2008;205:1227–1242. doi: 10.1084/jem.20071460. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Lemos DR, Goodspeed L, Tonelli L, Antoch MP, Ojeda SR, Urbanski HF. Evidence for circadian regulation of activating transcription factor 5 but not tyrosine hydroxylase by the chromaffin cell clock. Endocrinology. 2007;148:5811–5821. doi: 10.1210/en.2007-0610. [DOI] [PubMed] [Google Scholar]
  19. Lonze BE, Ginty DD. Function and regulation of CREB family transcription factors in the nervous system. Neuron. 2002;35:605–623. doi: 10.1016/s0896-6273(02)00828-0. [DOI] [PubMed] [Google Scholar]
  20. Marshall CA, Suzuki SO, Goldman JE. Gliogenic and neurogenic progenitors of the subventricular zone: who are they, where did they come from, and where are they going? Glia. 2003;43:52–61. doi: 10.1002/glia.10213. [DOI] [PubMed] [Google Scholar]
  21. Mason JL, Angelastro JM, Ignatova TN, Kukekov VG, Lin G, Greene LA, Goldman JE. ATF5 regulates the proliferation and differentiation of oligodendrocytes. Mol Cell Neurosci. 2005;29:372–380. doi: 10.1016/j.mcn.2005.03.004. [DOI] [PubMed] [Google Scholar]
  22. Mittal AK, Hegde GV, Aoun P, Bociek RG, Dave BJ, Joshi AD, Sanger WG, Weisenburger DD, Joshi SS. Molecular basis of aggressive disease in chronic lymphocytic leukemia patients with 11q deletion and trisomy 12 chromosomal abnormalities. Int J Mol Med. 2007;20:461–469. [PubMed] [Google Scholar]
  23. Monaco SE, Angelastro JM, Szabolcs M, Greene LA. The transcription factor ATF5 is widely expressed in carcinomas, and interference with its function selectively kills neoplastic, but not nontransformed, breast cell lines. Int J Cancer. 2007;120:1883–1890. doi: 10.1002/ijc.22469. [DOI] [PubMed] [Google Scholar]
  24. Morris JA, Kandpal G, Ma L, Austin CP. DISC1 (Disrupted-In-Schizophrenia 1) is a centrosome-associated protein that interacts with MAP1A, MIPT3, ATF4/5 and NUDEL: regulation and loss of interaction with mutation. Hum Mol Genet. 2003;12:1591–1608. doi: 10.1093/hmg/ddg162. [DOI] [PubMed] [Google Scholar]
  25. Murdoch B, Roskams AJ. Olfactory epithelium progenitors: insights from transgenic mice and in vitro biology. J Mol Histol. 2007;38:581–599. doi: 10.1007/s10735-007-9141-2. [DOI] [PubMed] [Google Scholar]
  26. Newman JR, Keating AE. Comprehensive identification of human bZIP interactions with coiled-coil arrays. Science. 2003;300:2097–2101. doi: 10.1126/science.1084648. [DOI] [PubMed] [Google Scholar]
  27. Nicolay DJ, Doucette JR, Nazarali AJ. Transcriptional control of oligodendrogenesis. Glia. 2007;55:1287–1299. doi: 10.1002/glia.20540. [DOI] [PubMed] [Google Scholar]
  28. Nishizawa M, Nagata S. cDNA clones encoding leucine-zipper proteins which interact with G-CSF gene promoter element 1-binding protein. FEBS Lett. 1992;299:36–38. doi: 10.1016/0014-5793(92)80094-w. [DOI] [PubMed] [Google Scholar]
  29. Pascual M, Gomez-Lechon MJ, Castell JV, Jover R. ATF5 is a highly abundant liver-enriched transcription factor that cooperates with constitutive androstane receptor in the transactivation of CYP2B6: implications in hepatic stress responses. Drug Metab Dispos. 2008;36:1063–1072. doi: 10.1124/dmd.107.019380. [DOI] [PubMed] [Google Scholar]
  30. Pati D, Meistrich ML, Plon SE. Human Cdc34 and Rad6B ubiquitinconjugating enzymes target repressors of cyclic AMP-induced transcription for proteolysis. Mol Cell Biol. 1999;19:5001–5013. doi: 10.1128/mcb.19.7.5001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Persengiev SP, Green MR. The role of ATF/CREB family members in cell growth, survival and apoptosis. Apoptosis. 2003;8:225–228. doi: 10.1023/a:1023633704132. [DOI] [PubMed] [Google Scholar]
  32. Persengiev SP, Devireddy LR, Green MR. Inhibition of apoptosis by ATFx: a novel role for a member of the ATF/CREB family of mammalian bZIP transcription factors. Genes Dev. 2002;16:1806–1814. doi: 10.1101/gad.992202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Peters CS, Liang X, Li S, Kannan S, Peng Y, Taub R, Diamond RH. ATF-7, a novel bZIP protein, interacts with the PRL-1 protein-tyrosine phosphatase. J Biol Chem. 2001;276:13718–13726. doi: 10.1074/jbc.M011562200. [DOI] [PubMed] [Google Scholar]
  34. Polito A, Reynolds R. NG2-expressing cells as oligodendrocyte progenitors in the normal and demyelinated adult central nervous system. J Anat. 2005;207:707–716. doi: 10.1111/j.1469-7580.2005.00454.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ross SE, Greenberg ME, Stiles CD. Basic helix-loop-helix factors in cortical development. Neuron. 2003;39:13–25. doi: 10.1016/s0896-6273(03)00365-9. [DOI] [PubMed] [Google Scholar]
  36. Sampath P, Pritchard DK, Pabon L, Reinecke H, Schwartz SM, Morris DR, Murry CE. A hierarchical network controls protein translation during murine embryonic stem cell self-renewal and differentiation. Cell Stem Cell. 2008;2:448–460. doi: 10.1016/j.stem.2008.03.013. [DOI] [PubMed] [Google Scholar]
  37. Schrot RJ, Ma JH, Greco CM, Arias AD, Angelastro JM. Organotypic distribution of stem cell markers in formalin-fixed brain harboring glioblastoma multiforme. J Neurooncol. 2007;85:149–157. doi: 10.1007/s11060-007-9401-8. [DOI] [PubMed] [Google Scholar]
  38. Shinomura T, Ito K, Kimura JH, Hook M. Screening for genes preferentially expressed in the early phase of chondrogenesis. Biochem Biophys Res Commun. 2006;341:167–174. doi: 10.1016/j.bbrc.2005.12.169. [DOI] [PubMed] [Google Scholar]
  39. Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, Hide T, Henkelman RM, Cusimano MD, Dirks PB. Identification of human brain tumour initiating cells. Nature. 2004;432:396–401. doi: 10.1038/nature03128. [DOI] [PubMed] [Google Scholar]
  40. Vescovi AL, Galli R, Reynolds BA. Brain tumour stem cells. Nat Rev Cancer. 2006;6:425–436. doi: 10.1038/nrc1889. [DOI] [PubMed] [Google Scholar]
  41. Vinson C, Myakishev M, Acharya A, Mir AA, Moll JR, Bonovich M. Classification of human B-ZIP proteins based on dimerization properties. Mol Cell Biol. 2002;22:6321–6335. doi: 10.1128/MCB.22.18.6321-6335.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Wang H, Lin G, Zhang Z. ATF5 promotes cell survival through transcriptional activation of Hsp27 in H9c2 cells. Cell Biol Int. 2007;31:1309–1315. doi: 10.1016/j.cellbi.2007.05.002. [DOI] [PubMed] [Google Scholar]
  43. Watatani Y, Kimura N, Shimizu YI, Akiyama I, Tonaki D, Hirose H, Takahashi S, Takahashi Y. Amino acid limitation induces expression of ATF5 mRNA at the post-transcriptional level. Life Sci. 2007;80:879–885. doi: 10.1016/j.lfs.2006.11.013. [DOI] [PubMed] [Google Scholar]
  44. Watatani Y, Ichikawa K, Nakanishi N, Fujimoto M, Takeda H, Kimura N, Hirose H, Takahashi S, Takahashi Y. Stress-induced translation of ATF5 mRNA is regulated by the 5'-untranslated region. J Biol Chem. 2008;283:2543–2553. doi: 10.1074/jbc.M707781200. [DOI] [PubMed] [Google Scholar]
  45. Wei Y, Jiang J, Sun M, Chen X, Wang H, Gu J. ATF5 increases cisplatin-induced apoptosis through up-regulation of cyclin D3 transcription in HeLa cells. Biochem Biophys Res Commun. 2006;339:591–596. doi: 10.1016/j.bbrc.2005.11.054. [DOI] [PubMed] [Google Scholar]
  46. Wei Y, Jiang J, Liu D, Zhou J, Chen X, Zhang S, Zong H, Yun X, Gu J. Cdc34-mediated degradation of ATF5 is blocked by cisplatin. J Biol Chem. 2008 doi: 10.1074/jbc.M707879200. [DOI] [PubMed] [Google Scholar]
  47. White JH, McIllhinney RA, Wise A, Ciruela F, Chan WY, Emson PC, Billinton A, Marshall FH. The GABAB receptor interacts directly with the related transcription factors CREB2 and ATFx. Proc Natl Acad Sci U S A. 2000;97:13967–13972. doi: 10.1073/pnas.240452197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Zhou D, Palam LR, Jiang L, Narasimhan J, Staschke KA, Wek RC. Phosphorylation of eIF2 directs ATF5 translational control in response to diverse stress conditions. J Biol Chem. 2008;283:7064–7073. doi: 10.1074/jbc.M708530200. [DOI] [PubMed] [Google Scholar]

RESOURCES