- Top of page
- Materials and methods
The duration of intracellular signaling is thought to be a critical component in effecting specific biological responses. This paradigm is demonstrated by growth factor activation of the extracellular signal-regulated kinase (ERK) signaling cascade in the rat pheochromocytoma cell line (PC12 cells). In this model, sustained ERK activation induced by nerve growth factor (NGF) results in differentiation, whereas transient ERK activation induced by epidermal growth factor (EGF) results in proliferation in these cells. Recently, the immediate early gene product c-fos has been proposed to be a sensor for ERK signaling duration in fibroblasts. In this study, we ask whether this is true for NGF and EGF stimulation of PC12 cells. We show that NGF, but not EGF, can regulate both c-fos stability and activation in an ERK-dependent manner in PC12 cells. This is achieved through ERK-dependent phosphorylation of c-fos. Interestingly, distinct sites regulate enhanced stability and transactivation of c-fos. Phosphorylation of Thr325 and Thr331 are required for maximal NGF-dependent transactivation of c-fos. In addition, a consensus ERK binding site (DEF domain) is also required for c-fos transactivation. However, stability is controlled by ERK-dependent phosphorylation of Ser374, while phosphorylation of Ser362 can induce conformational changes in protein structure. We also provide evidence that sustained ERK activation is required for proper post-translational regulation of c-fos following NGF treatment of PC12 cells. Because these ERK-dependent phosphorylations are required for proper c-fos function, and occur sequentially, we propose that c-fos is a sensor for ERK signaling duration in the neuronal-like cell line PC12.
Nerve growth factor (NGF) triggers neuronal differentiation in the PC12 cell model through the sustained activation of the mitogen-activated protein (MAP) kinase or extracellular signal-regulated kinase (ERK). Neuronal differentiation by NGF is characterized by the induction of immediate early genes that encode transcription factors that promote the transcriptional activation of a set of NGF-responsive genes. Because of the requirement of sustained ERK activation for differentiation, NGF action on PC12 cells has served as a model for the role of duration of intracellular signals in dictating physiological responses. However, the mechanism by which transcription factors sense sustained ERK activation in PC12 cells is not known.
Other cell types have provided insight into how incremental changes in the duration of ERK activation can have profound effects on cellular responses (Weber et al. 1997; Bottazzi et al. 1999; Roovers et al. 1999; Adachi et al. 2002; Murphy et al. 2002; Koike et al. 2003; Werlen et al. 2003). For example, one the one hand, in Swiss 3T3 fibroblast cells, the sustained activation of ERKs is required for growth factor-induced proliferation by platelet-derived growth factor (PDGF). On the other hand, transient activation of ERKs by epidermal growth factor (EGF) is not mitogenic in these cells. One ERK target, the transcription factor c-fos, has been proposed to mediate this action (Murphy et al. 2002). c-fos is a proto-oncogene that, unlike its oncogenic counterpart, v-fos, requires additional signals to achieve maximal proliferative potential (Chen et al. 1993; Okazaki and Sagata 1995; Chen et al. 1996; Monje et al. 2003). ERK-dependent signals stimulate c-fos at multiple levels (Monje et al. 2005), but perhaps the best studied of these actions is the stimulation of c-fos transcription (Monje et al. 2003; Tanos et al. 2005). This is achieved by phosphorylation of the transcription factor Elk-1 which functions with other serum response factors to turn on the c-fos promoter (Gille et al. 1992; Hipskind et al. 1994).
Direct phosphorylation of c-fos protein by ERKs can also enhance c-fos function at AP-1 promoters (Sutherland et al. 1992; Monje et al. 2003). This occurs via two interdependent mechanisms. First, ERK-dependent phosphorylations on Ser362 and Ser374 within the C-terminus (called ‘priming’ phosphorylations) can stabilize c-fos, possibly by interfering with degradation signals within the c-fos protein (Okazaki and Sagata 1995; Ferrara et al. 2003). The Ser374 site has been shown to be phosphorylated by ERKs in vivo and in vitro (Chen et al. 1996). However, ERK does not phosphorylate c-fos at Ser362 in vitro (Monje et al. 2003), and additional kinases have been proposed for the phosphorylation of the Ser362 site (Tratner et al. 1992; Chen et al. 1993). RSK2, an ERK-dependent kinase, has been implicated in growth factor-induced phosphorylation of this site. Most previous studies examining these sites within c-fos utilized c-fos double mutants that were mutated at both sites (Chen et al. 1993; Okazaki and Sagata 1995; Chen et al. 1996; Murphy et al. 2002, 2004). By examining individual mutants of these sites, we show that phosphorylations at these sites have distinct contributions to c-fos stability and priming of c-fos for further phosphorylations.
Second, these phosphorylations prime additional ERK phosphorylations within the transactivation domain (TAD) of c-fos that Gutkind and colleagues have shown potentiate AP-1-dependent transcription (Monje et al. 2003). These phosphorylations on Thr325 and Thr331 (and possibly Thr232) are thought to contribute to the retarded electrophoretic mobility shift associated with elevated c-fos phosphorylation and function (Monje et al. 2003). TAD phosphorylation by ERKs may be enhanced by directing the binding of ERKs to an ERK targeting domain, also known as a DEF domain, which has been identified near these sites (Murphy et al. 2002).
In this study, we test whether c-fos is a sensor of sustained ERK activation in PC12 cells. We show that phosphorylations of Thr325 and Thr331 are required for maximal NGF-dependent transactivation of c-fos in PC12 cells. Like in fibroblasts, c-fos requires an intact DEF domain for transactivation. c-fos stabilization and conformational changes in protein structure are also regulated by ERK-dependent sites, namely Ser374 and Ser362, respectively. We provide evidence that sustained ERK activation is required for post-translational regulation of c-fos during differentiation of PC12 cells. Thus, our data suggest that both neuronal and non-neuronal cell types utilize c-fos as a sensor for ERK signaling duration despite the fact that this transcription factor is coupled to distinct physiological outcomes, differentiation and proliferation, respectively, in these cell types.
- Top of page
- Materials and methods
We show in PC12 cells that NGF induces the stability and transactivation of c-fos. We show that this phenomenon is dependent on the sustained activation of ERKs and is blocked by pharmacological inhibitors of ERK activation. This establishes that c-fos is a sensor for sustained ERK activation in PC12 cells. This is similar to the model proposed for c-fos activation by PDGF in fibroblasts (Murphy et al. 2002). Importantly, PDGF is a proliferative agent in these cells. Therefore, our study demonstrates that the requirement of sustained ERK activation for c-fos function is independent of the physiological outcome of c-fos activation. Recently, studies in fibroblasts have demonstrated that c-myc, Egr-1, Fra-1, Fra-2, and c-jun are also targets for activation by ERKs (Murphy et al. 2004). Indeed, a recent report demonstrates that sustained ERK1/2 activation is required for an AP-1-dependent mitogenic signal by growth factors in fibroblasts (Yamamoto et al. 2006). It is likely that some of these proteins also respond to sustained activation of ERKs by NGF to promote the induction of NGF-responsive genes in neuronal cells (Pap and Szeberenyi 1998; Cosgaya and Aranda 1999; Riccio et al. 1999; Groot et al. 2000; Boss et al. 2001).
We were able to observe a slight, but significant, increase in c-fos transactivation and protein stability by EGF. This effect of EGF can partially be attributed to increased basal levels of transfected Flag-Fos under control of the CMV promoter, which allows transient ERK1/2 activation to immediately act on pre-made c-fos protein. However, small increases of EGF in c-fos transactivation are unable to reach the threshold for induction of differentiation, demonstrating the requirement for sustained ERK activation to initiate this paradigm.
Several reports have examined residues in c-fos that are important for transactivation in fibroblasts (Murphy et al. 2002; Monje et al. 2003; Tanos et al. 2005). In this study, we show that both Thr325 and Thr331 are required for maximal NGF-dependent activation of c-fos in PC12 cells. Our evidence suggests that both residues contribute to c-fos transactivation, although inhibition of Thr325 phosphorylation reduces transactivation more strongly. However, phosphorylation of these residues has a synergistic effect. This may explain why both residues are required for migration of the slowest mobility (top band). Blenis and colleagues showed a robust phosphorylation of Thr325 upon PDGF treatment in non-neuronal cells. This phosphorylation was also dependent on sustained ERK activation and an intact DEF domain. However, when they mutated these same residues, they saw a very modest effect on basal AP-1 activity. Although this difference in results could be cell-type specific, it is more likely as a result of differences in basal versus stimulated c-fos activity. Our data confirm those of Blenis and co-workers that demonstrated the importance of the DEF domain in promoting ERK-dependent phosphorylations critical for c-fos transactivation. At this time, we cannot rule out the possibility that this domain directs the ERK-dependent phosphorylation of additional target proteins.
It has been assumed that phosphorylation at both sites functions similarly to enhance c-fos stability and transformation, largely because previous studies have examined c-fos mutated simultaneously at both sites. By examining a more complete series of mutants, we demonstrate that phosphorylations at these two serine sites contribute to the function of c-fos in distinct ways. Phosphorylation of Ser374 is required for the stabilizing effects of NGF stimulation, as evidenced by the finding that inhibition of Ser374 phosphorylation by UO126 correlates with decreased stability of c-fos protein. This can be shown by the absence of NGF-induced stabilization in Fos mutants lacking this serine (FosSA, FosDA and FosAA). Phosphorylation at this site appears to be sufficient for stabilization as phosphomimetic mutations at residue Ser374 produce c-fos mutants (FosSD and Fos AD) that are stable even under resting conditions. Phosphorylation of Ser362 appears to contribute significantly to the intermediate mobility shift, even in the absence of phosphorylation of Ser374 (FosDA). This likely reflects the contribution of Ser362 in priming additional phosphorylations. The exact mechanism by which phosphorylation of Ser362 promotes subsequent phosphorylations is not known. It is possible that this is achieved in part by promoting ERK binding to the DEF domain in c-fos that has been proposed to contribute to the hyperphosphorylation and shift. This model is illustrated in Fig. 8.
The requirement for sustained ERK activation was seen not only in the activation of wild-type c-fos proteins but also in c-fos mutants in which priming sites were replaced with aspartates (FosDD). FosDD is basally stable and no longer requires ERK-dependent phosphorylations at the ‘priming’ sites. Even in this mutant, transient ERK activation is not sufficient to induce c-fos activation. The inability of EGF to activate the stable mutant FosDD suggests that sustained activation of ERKs is required to maintain the phosphorylation status of the activating threonines (Thr325 and Thr331). It is likely that these sites are tightly regulated by active phosphatases. This ensures that c-fos requires sustained ERK activation at multiple levels of its regulation.
In summary, the ability of NGF to stimulate c-fos function requires sustained ERK activation. This requirement is as a result of two ERK-dependent phosphorylation sites within the C-terminus (Ser362 and Ser374) that have been proposed to confer stability on nascent c-fos protein and also serve as priming sites for subsequent phosphorylations within the TAD region. We show that phosphorylation of Ser374 is required for stabilization. As previously suggested, phosphorylation of this site may interfere with the well-characterized degradation sequence surrounding this site (Okazaki and Sagata 1995; Ferrara et al. 2003). Phosphorylation of Ser362 is critical in inducing the intermediate mobility shift, and, with some contribution of phosphorylation of Ser374, enables c-fos to be targeted for additional ERK-dependent phosphorylations. These additional phosphorylations, specifically Thr325 and Thr331, are required for c-fos transactivation in this model. Moreover, an intact ERK binding site is also needed for stimulated c-fos activity, presumably for these ‘activating’ phosphorylations. Because c-fos requires multiple ERK-dependent phosphorylations to be stabilized, shifted, and hyperphosphorylated, c-fos is strictly dependent on sustained ERK activation for stability and function. This requirement for sustained ERK activation serves as a paradigm for the distinct actions of NGF versus EGF on c-fos-dependent PC12 gene expression.