Fig. 13.3 Molecular regulation of MEKK1. MEKK1 e^js^s in the cytoplasm in an inactive form. Its activity is induced by various extra- or intracellular signals that activate the MEKK1 N-terminal domain-associated upstream regulators, including HPK, PKG, NIK., GLK, leading to MEKK1 phosphorylation and activation. Alternatively, the N-terminal domain of MEKK1 can interact with Grb2 upon EGF stimulation, which causes MHKK1 membrane localization, conformational change and autophosphorylation. Simultaneous interaction of MEKK1 with RhoA, JNK, actinin and MEK4 allows efficient signal transduction to regulate actin cytoskeleton, a likely pathway used by TGFI to induce actin stress fiber formation.

Such scaffolding function of MEKK1 organizes a signaling complex for the RhoA signal efficiently transmitting to MEKK1 and to the downstream MEK4-JNK pathway that is linked to the actin cytoskeleton. This pathway is likely to be utilized by TGFp, which activates RhoA, to control c-Jun phosphorylation and the organization of actin cytoskeleton (Gallagher et al., 2004; Atfi et al., 1997; Zhang et al., 2005). Through interaction with axin, MEKK1 may mediate Wnt signal in JNK activation and planar polarity determination (Zhang et al., 1999b). Alternatively, several protein kinases, including HPK, NIK, PKG and GLK, have been shown to directly associate with and phosphorylate MEKK1 in vitro, acting as the possible upstream activators for MEKK1 (Su et al., 1997; Diener et al., 1997; Soh et al., 2001). The complexity of MEKK1 regulation provides a promising angle to explain signal integration and segregation that takes place on MAPK activation. As far as the c-Jun phosphorylation is concerned, each MAPKKK may be responsible for transducing a subset of signals to the activation of the MAPK-c-Jun, whereas, a particular signal-mediated by MAPKKKs can lead to the activation of not only JNK-c-Jun, but also other downstream effectors.

With this idea in mind, it is not surprising to find that each MAPKKK knockout mice have rather distinct phenotypes. Like the Jnkl(-/-)Jnk2(-/-) and c-Jun(-/-) mice that are embryonic lethal, some MAPKKK knockout mice, including Raf-1(-/-), Mekk3(-/-), and Mekk4(-/-), die in embryogenesis, but by different reasons (Kuan CY, 1999; Johnson, R.S, 1993). The Raf-1(-/-) fetuses show vascular defects in the yolksac and placenta as well as increased apoptosis of embryonic tissues (Huser et al., 2001), the Mekk3(-/~) embryos die due to impaired blood vessels development (Yang et al., 2000), while the Mekk4(-/-) mice die from neural tube defects that are associated with massively elevated apoptosis before and during neural tube closure (Chi et al., 2005). Conversely, some MAPKKKs, such as ASK1 Tpl2 and MEKK2, are dispensable for embryonic development, as their knockout mice are all born alive with no overt developmental defects. Only under certain pathological or environmental conditions, these MAPKKK are required for a cell type specific response. The Askl(-f-) embryonic fibroblasts are resistant to apoptosis induced by TNFa and H(2)0(2) (Tobiume et al., 2001), the Tpl2 (-/-) mice produce low levels of TNFa when exposed to lipopolysaccharide (LPS) (Dumitru et al., 2000), and the Mekk2(-/-) T cells show increased proliferation and susceptibility to apoptosis induced by T cell receptor cross-linking (Guo et al., 2002). The MEKK1 -deficient mice also survive

Cell Proliferation

The very first clue for the involvement of c-Jun in cell growth stems from the observation that c-Jun expression takes place in quiescent cells at the GO to G1 transition (Ryseck et al, 1988) and upon exposure to mitogenic signals at the G1 to S transition (Carter et al, 1994; Mayo et al., 1994). A more clear indication derives from studies in c-Jun-null fibroblasts and in erythroleukemia cells expressing c-Jun antisense sequences. Both cells display greatly reduced growth rates and blockage of Gl-to-S-phase cell cycle progression (Johnson et al., 1993; Smith and Prochownik, 1992; Schreiber et al., 1999). In addition to its expression, c-Jun activity also fluctuates in a cell cycle dependent manner, possibly subjected to regulation by cyclin-dependent kinase (cdk) inhibitor, p27(Kipl). In quiescent cells, p27(Kipl) is phosphorylated at serine 10 and the phosphorylated p27(Kip) binds to Jabl, preventing Jab 1-c-Jun interaction. Upon quiescence exit, p27(Kipl) dephosphorylates, releasing Jabl from its suppression state to activate c-Jun (Chopra et al., 2002). Once activated, c-Jun must turn on gene expression programs crucial for cell progression through the Gl-S checkpoint.

One such gene regulated by c-Jun is cyclin Dl. Having an AP-1 binding site in its promoter, cyclin Dl transcripton is directly induced by c-Jun, which recruits p300 for transcription activation, but is suppressed by adenovirus El A, which competes for p300 (Wisdom et al., 1999; Albanese et al., 1999). Cyclin Dl in turn forms complex with cdks for RB phosphorylation. The phosphorylated RB releases from E2F, allowing E2F to activate genes that enable cells to advance into late Gi and S phases (Grana and Reddy, 1995; Zetterberg et al, 1995). Another cell cycle regulator is the tumor suppressor p53, which is known to inhibit cell cycle progression by transcriptional activation of p21, the inhibitor of cyclin/cdks (Levine, 1997; Agarwal et al, 1995). p53 null fibroblasts have accelerated proliferation rate (Harvey et al, 1993), in contrast, c-Jun null cells show growth defects. It turns out that c-Jun is a negative regulator for the transcription of p53 and p21, an effect mediated by direct binding of c-Jun to a variant AP-1 site in the p53 promoter or indirectly through a SP-1 site in the p21 promoter (Wang et al, 2000). In the absence of c-Jun, an elevated expression of p53 and its target genep21 leads to growth arrest (Schreiber et al, 1999).

Despite the crucial role of c-Jun in cell proliferation, c-Jun null embryos survive until mid-gestation without overt abnormalities before death and c-Jun-null embryonic stem cells grow normally in culture and contribute to most tissues in chimeric mice (Hilberg et al, 1993; Hilberg and Wagner, 1992; Johnson et al, 1993). Hence, the function of c-Jun in cell proliferation is likely cell type specific, and it is required only at late developmental stages or under certain environmental conditions, such as following UV irradiation, to trigger cell cycle re-entry (Shaulian and Karin, 2002).

Apoptotic Cell Death

The pro-apoptotic function of c-Jun was first observed in neuronal cells, from which nerve growth factor (NGF) withdrawal caused cell death, correlating with an induction of c-Jun expression and an increase of AP-1 DNA binding activity (Ham et al, 1995). Expression of a dominant negative c-Jun mutant protects these cells from death. Apoptosis is an intrinsic program to activate caspase-mediated proteolysis pathways to eliminate cells that have suffered serious damage, a process required for normal development and homeostasis maintenance. Apoptotic cells have a unique morphological pattern characterized by chromatin condensation, membrane blebbing and DNA fragmentation. The property of c-Jun of being involved in both cell proliferation and apoptosis is in fact common to a number of oncogene products, including Ras, c-Myc and E2F (Sears and Nevins, 2002). These proteins, originally identified as positive regulators of cell growth, were subsequently found to be involved in apoptosis. It actually makes sense for cells to employ an overlapping system to regulate cell growth and death, as it provides the most convenient way to switch cell functions in response to environmental changes. c-Jun is apparently part of the program for life-death decisions of neuronal cells depending on the presence or absence of NGF.

Neurotoxicity is mediated by JNK phosphorylation of c-Jun (Xia et al, 1995). Supporting this view is the finding that mice deficient in the neural specific JNK3 and mice of which the endogenous Jun is replaced by c-Jun 63/73 AA are protected from excitatory amino acid kainite induced neuronal cell apoptosis (Yang et al, 1997; Behrens et al, 1999). We now know that the JNK-c-Jun pathway is widely implicated in apoptosis induced by many stress stimuli, including DNA damaging agents, microtubule inhibitors, protein synthesis inhibitors, cytokines and lipid mediators; its effect is not limited to the neuronal cells, but also applies to other cell types, such as T cells, macrophages and epithelial cells (Chen et al, 1996; Park et al, 1997; Mosieniak et al, 1997; Verheij et al, 1996).

Apoptosis requires the participation of both the transactivation domain and the bZIP domain of c-Jun and is prevented by a c-Jun N-terminal truncation mutant TAM67, c-Jun antisense oligonucleotides, and c-Jun genetic knockout, suggesting that c-Jun transcriptional activities are needed for the expression of apoptosis effectors (Fan et al., 2001; Sawai et al., 1995; Shaulian et al, 2000). Candidate effectors include FasL, CD95L and the death receptor 4(DR4), all having AP-1 binding sites in their gene promoters and being induced by a number of stress stimuli in a JNK dependent manner (Eichhorst et al., 2000; Fans et al., 1998; Guan et al., 2002). Their induction initiates apoptosis by the death-receptor mediated pathways. Another effector is BIM, a member of the proapoptotic BCL-2 family, whose expression is induced by NGF withdrawal in a c-Jun dependent manner (Toh et al., 2004). BIM can in turn be phosphorylated by JNK for a further enhancement of the proapoptotic activity, allowing signal amplification that activates the mitochondria death pathway (Putcha et al., 2003).

Not all death response requires c-Jun and gene transcription. JNK can potentiate apoptosis by transcription independent modulation of the 14-3-3 protein, whose unphosporylated form serves as a cytoplasmic anchor for c-Abl and BAX (Yoshida et al., 2005; Tsuruta et al., 2004). Phosphorylation of 14-3-3 by JNK triggers the release of c-Abl to the nucleus to activate apoptotic genes and of BAX to mitochondria to induce cytochrome c release (Tournier et al., 2000; Koo et al., 2002). The multiple facets of JNK in apoptosis through both c-Jun dependent and independent mechanisms explain its unique role in the apoptotic response during tissue development. Developmental defects associated with JNK inactivation, including dysregulation of brain cell apoptosis in Jnkl(-/-)Jnk2(-/-) mice and reduction of double positive thymocyte death in dominant negative JNK transgenic mice, are not observed in mice of which c-Jun activity is compromised, such as TAM-67 or c-Jun 63/73 AA transgenic mice (Kingei al., 1999; Behrens et al., 1999; Kuan et al., 1999; Rincon et al., 1998).

Interestingly, c-Jun appears also to have anti-apoptotic functions, probably by the induction of the anti-apoptotic factor BCL3 (Rebollo et al., 2000). c-Jun mutant fetuses harbor increased numbers of apoptotic cells in liver (Eferl et al., 1999) and c-Jun-null primary embryonic fibroblasts show enhanced sensitivity to UV-induced apoptosis (Wisdom et al., 1999). Correspondingly, suppression of AP-1 activity by the glucocorticoid receptor induces leukemia cell apoptosis (Helmberg et al., 1995; Chen et al., 1996). The observations that AP-1 is involved in Ras-induced survival of only RB-null fibroblasts (Young and Longmore, 2004) and that JNK potentiates the growth of only p53-deficient tumor cells (Potapova et al., 2000), suggest that the cross talk to the tumor suppressor pathway may have a lot to do with life and death decision made by the JNK-c-Jun pathway. An example of such cross-talk is the rapid and sustained activation of ASK1 and JNK, which leads to c-Jun phosphorylation, when rhabdomyosarcoma cells are subjected to rapamycin or amino acid deprivation. In the absence of p53, this pathway leads to apoptosis, while in the presence of wild type p53 or p21(Cip), cells arrest in G1 and remain viable (Huang et al., 2003). The apparent contradictory roles for c-Jun in p53-dependent survival and apoptosis, together with the previously discussed role for c-Jun in the regulation of p53 expression and cell cycle progression, strongly indicate that c-Jun does not act alone and that its function in cell fate determination must be subjected to the regulation by a complex system in a cell type and stimuli specific fashion (Potapova et al., 2000; Huang et al., 2003; Schreiber et al., 1999; Shaulian and Karin, 2002).

How are the apoptotic signals transduced to JNK? NGF deprivation signals are known to be organized by a protein called POSH, which stands for plenty of SH3s. POSH acts as a scaffold for the formation of a complex consisting of the activated Racl/Cdc42, MLKs, MEK4/7, JNK and c-Jun, allowing the sequential activation of the signaling cascade that leads to neuronal death (Xu et al., 2003). Although it has been suggested that the TNFa signal is mediated by ASK1 and the DNA-damaging signals are transduced by BRCA1 (Ichijo et al., 1997; Harkin et al., 1999), the precise mechanisms involved in JNK activation during these signaling processes are still unknown.

0 0

Post a comment