Mitogen-activated protein kinases (MAPKs) are evolutionarily conserved proteins that regulate multiple intracellular processes in all eukaryotic species, from protozoa to plants and vertebrates. Activation of the MAPK family is achieved by three distinct tiers of upstream kinases: The first tier is MAP kinase kinase kinases (MAPKKKs or MAP3Ks), which can respond to various extracellular signals such as mechanical stresses, oxidative stresses, and growth factors, and transduces these extracellular signals into intracellular signaling cascades that regulate various cellular responses. MAPKKKs include MEKK1/2/3/4 and A-/B-/C-Raf (Boulton et al. 1991; Widmann et al. 1999). These kinases activate their downstream target kinases, the second tier, MAP kinase kinases (MAPKKs or MAP2Ks) or MEKs by phosphorylating specific serine (Ser; S) and threonine (Thr; T) residues of MEK. Activated MEKs in turn phosphorylate a T or tyrosine (Tyr; Y) residue in the conserved T-x-Y motif (T-E-Y motif) of the last tier of kinases (MAP kinases, MAPKs) (Widmann et al. 1999; Zhang and Dong 2007). As the consequence, the MAPK phosphorylates specific downstream effector molecules such as transcription factors, structural proteins, phospholipids, and cytoplasmic enzymes to generate specific intracellular responses such as cell proliferation, differentiation, migrations, survival, and apoptosis (Chang and Karin 2001; Lewis et al. 1998; Qi and Elion 2005; Yang et al. 2003; Yoon and Seger 2006).
There are four MAPK cascades in mammalian cells: the extracellular signal-regulated kinase 1/2 (ERK1/2), c-Jun NH2-terminal kinases 1/2/3 (JNK1/2/3), p38MAPKα/β/γ/δ, and ERK5 or BMK1 (Abe et al. 1996; Chang and Karin 2001; Nishida and Gotoh 1993; Robinson and Cobb 1997). ERK5 is the newest member, and as such it is least studied. In this chapter, we will focus on the structure, activation, and function of MEK5 and ERK5. We will also summarize the upstream and downstream effectors as well as the significance of the MEK5-ERK5 signaling pathway in the context of the EC function and pathophysiology. Last but not least, we will discuss the interconnection between the MEK5-ERK5 and MEK1/2-ERK1/2 pathways.
MEK5α and MEK5β: Distribution and Function
The amino acid sequence of MEK5 was found to be more closely related to MEK1 and MEK2 than to MEK3 (MKK3) and MEK4 (MKK4). MEK5 is activated by MAP kinase kinases, MEKK2 and MEKK3, through direct binding and phosphorylation (Chao et al. 1999; Nakamura and Johnson 2003; Sun et al. 2001), and MEK5 directly activates ERK5 (Mody et al. 2003). Alternative splicing of mek5 mRNA at the 5′-end (N-terminus) results in two isoforms: the shorter beta isoform (MEK5β, 40 kDa) is ubiquitously expressed and is primarily cytosolic, and the longer alpha isoform (MEK5α, 50 kDa) is richly expressed in liver and brain and is associated with the insoluble fraction of the cell (English et al. 1995). Overexpression of a constitutively active form of MEK5α (CA-MEK5α), but not MEK5β (CA-MEK5β), causes ERK5 activation and nuclear translocation (Cameron et al. 2004). When MEK5 phosphorylation sites, S311 and T315, were mutated to aspartic acid (Asp, D), the overexpression of this MEK5α (S311D/T315D) mutant also caused the activation and nuclear translocation of ERK5. However, the corresponding dual mutations in MEK5β failed to affect ERK5 in the same manner (Cameron et al. 2004). In addition, it was found that MEK5β competed with MEK5α for the same binding site on ERK5 and exerted a dominant negative effect on ERK5. Indeed, MEK5β inhibits ERK5 activation by epidermal growth factor (EGF) as well as by overexpression of the constitutively active form of MEK5α (S311D/T315D), suggesting that MEK5β inhibits activated MEK5α-mediated ERK5 activation (Cameron et al. 2004). In summary, MEK5β is an endogenous dominant negative form of MEK5α and that MEK5α but not MEK5β is involved in ERK5 activation and ERK5 nuclear translocation.
Compared to the shorter MEK5β isoform, the full-length isoform of MEK5α has extra 89 amino acids at its N-terminus, and this portion (the head domain) contains the Phox and Bem1P (PB1) domain (amino acids 12–85) (Cameron et al. 2004). The PB1 domain was first defined in p67phox and Bem1p proteins and shown to mediate protein-protein heterodimerization (Ponting et al. 2002). The PB1 domain is conserved in plants, amoebas, fungi, and all animals. Cameron et al. (2004) found that ERK activation was not achieved when the amino acids 1–50 were deleted in the constitutively active MEK5α (S311D/T315D) dual phosphorylation site mutant. The PB1 domain is also present in atypical PKC (aPKC), cdc24, Par6B, and Par6C (Diaz-Meco and Moscat 2001). In addition, the interaction between MEK5α and aPKC has been found to be critical for MEK5α kinase activity and subsequent ERK5 activation (Cameron et al. 2004; Diaz-Meco and Moscat 2001; Nakamura and Johnson 2003, 2007; Nakamura et al. 2006; Pearson et al. 2001). Taken together, these observations suggest that the PB1 domain of MEK5α is required for ERK5 activation (Cameron et al. 2004; English et al. 1995).
In MCF7 breast cancer cells, MEK5β expression is detected only in growth-arrested cells and not in actively proliferating cells. Consistent with this result, overexpression of MEK5β inhibits breast carcinoma cell proliferation (Cameron et al. 2004). Overall, these studies indicate that MEK5α is the positive regulator for ERK5 activation, ERK5 nuclear translocation, and subsequent cell proliferation while MEK5β is a dominant-negative MEK5α.
MEK5 Upstream Regulators
MEKK2 and MEKK3
As we described earlier, MEKK2 and MEKK3 (MAPK kinase kinases), MEK5 (MAPK kinase), and ERK5 (MAPK) are members of the three-kinase cascade for ERK5 activation (Seger and Krebs 1995), in which MEKK2 and MEKK3 are the upstream regulators that activate MEK5 (Chao et al. 1999; Sun et al. 2001). MEKK2 and MEKK3 have similar kinase domains but different regulatory domains located in their N-termini (Sun et al. 2001). MEK5 can be activated by MEKK2 and MEKK3, but not by MEKK1, an upstream activator of the JNK pathway (Chang and Karin 2001; Chao et al. 1999; Chayama et al. 2001; Sun et al. 2001). Activated MEK5 then phosphorylates and activates ERK5 (Nakamura and Johnson 2003). This last process requires formation of the MEK5-ERK5 complex (Seyfried et al. 2005). Sun et al. (2001) have demonstrated that the N-termini of MEKK2 and MEKK3 can bind to MEK5, but compared to MEKK3, MEKK2 has a higher binding affinity for MEK5 (Chao et al. 1999; Sun et al. 2001), which is suggested to be a reason why MEKK2 is much more effective than MEKK3 in activating ERK5 (Sun et al. 2003).
MEK5, MEKK2, and MEKK3 contain a PB1 domain in their N-termini (Lamark et al. 2003; Wilson et al. 2003). Nakamura and Johnson (2003) reported that the PB1 domain of MEKK2 and MEKK3 specifically participated in their binding with MEK5 to form a heterodimer. Subsequently, it was shown that the MEK5 PB1 domain not only heterodimerized with the PB1 domain of MEKK2 (or MEKK3), but also associated with ERK5 thereby forming a MEKK2 (or MEKK3)-MEK5-ERK5 trimer complex, which was required for ERK5 activation (Nakamura and Johnson 2003; Nakamura et al. 2006; Sumimoto et al. 2007). MEKK2 and MEK5 coimmunoprecipitated, and the MEKK2-MEK5 complex formation was abolished when the MEKK2 PB1 domain was deleted or mutated, indicating that the PB1 domain of MEKK2 is required for the MEKK2-MEK5 interaction (Nakamura and Johnson 2007). In addition, overexpression of the PB1 domain or the MEKK2 PB1 domain inhibited ERK5 activation, whereas expression of a PB1 domain-deleted mutant MEKK2 is unable to bind MEK5, indicating that the MEKK2-MEK5 association is mediated by the PB1 domain of both kinases and that the MEKK2 and MEK5 PB1 domains are critical for the regulation of the ERK5 pathway (Nakamura and Johnson 2007).
MEKK2 and MEKK3 have been the only known MAPKKK that functions as direct upstream regulators of MEK5 (Chayama et al. 2001; Nakamura and Johnson 2003; Sun et al. 2001). Song et al. (2004), however, identified signal transducer and activator of transcription 3 (Stat3), which modulated the expression level of MEK5. They showed that Stat3 upregulated MEK5 by directly binding to the MEK5 gene promoter and upregulating its transcription. The increased expression of MEK5 enhanced the activity of ERK5 and certain oncogenic proteins (i.e., MMP-1, MMP-9, Bcl-XL, cyclin D1).
Atypical Protein Kinase Cs (aPKCs)
The phospholipid-dependent serine/threonine protein kinase C (PKC) family is classified into three PKC isoform subfamilies: conventional or classical (cPKCs: α, βI, βII and γ), novel or non-classical (nPKCs: δ, ε, η and θ), and atypical (aPKCs: ζ, ι and λ) based on their structure and activation (Steinberg 2008). The aPKC isoforms (PKCζ, ι and λ) participate in disparate signaling cascades required for cell proliferation and apoptosis (Akimoto et al. 1996, 1998; Berra et al. 1993; Diaz-Meco and Moscat 2001) and can interact with different protein modulators (Colledge and Scott 1999; Mochly-Rosen 1995; Mochly-Rosen and Gordon 1998). For example, only the aPKCs but not the classical or the novel PKC isoforms interact with the proapoptotic protein Par-4 (Diaz-Meco et al. 1996, 1999; Wang et al. 1999) and a scaffold protein p62 (Sanchez et al. 1998; Sanz et al. 1999). The aPKCs interact with p62 through a small acidic stretch of 44 amino acid residues termed the aPKC interaction domain (AID) (English et al. 1995; Zhou et al. 1995). Because MEK5 possesses a stretch of sequence highly homologous to AID, it is suggested that MEK5 might be an aPKC-binding partner. Using the MEK5 wild type (WT) and a MEK5 AID deleted mutant (MEK5 DC) in a coimmunoprecipitation experiment, Diaz-Meco and Moscat (2001)) found that the aPKC isoforms indeed interacted with the MEK5 AID domain. MEK5–ERK5 signaling plays a critical role in the EGF-induced mitogenic activation (Kato et al. 1998). Further studies by Diaz-Meco and Moscat (2001) revealed that the aPKCs-MEK5 interaction was direct and occurred only in cells treated with EGF, which triggered a robust aPKC-MEK5 interaction, which is essential for MEK5 activation. However, how aPKCs activate MEK5 without enzymatic activity is not yet completely understood (Diaz-Meco and Moscat 2001).
In 1995, ERK5 was identified by two independent research groups. By screening a human placenta cDNA library using degenerate PCR, Lee et al. (1995) discovered a novel MAPK gene, which they named big MAPK (BMK)-1 because of its large size of the C-terminus relative to ERK1/2. Zhou et al. (1995), on the other hand, first identified MEK5 and then used it in a yeast two-hybrid screen to identify ERK5 as a substrate of MEK5. ERK5 and BMK1 were later confirmed to be identical. ERK5 is expressed in many tissue types, especially abundant in brain, heart, skeletal muscle, placenta, lungs, and kidneys, and is localized both in the cytoplasm and the nucleus (Buschbeck and Ullrich 2005; Lee et al. 1995; Raviv et al. 2004; Zhou et al. 1995).
In human, the erk5 (also called mapk7 and bmk1) gene has a total of 5824 bases with an open reading frame of 2451 base pairs encoding a protein with 816 amino acids (Fig. 1a). The ERK5 N-terminus contains binding sites for ATP (amino acid 84), small-ubiquitin like modifier (SUMO) (amino acids K6 and K22), MEK5 (amino acids 78–140), and other oligomers (amino acids 140–406) and the MAPK-conserved T-E-Y phosphorylation motif in the activation loop (T218EY220) (Mody et al. 2003; Yan et al. 2001). The kinase domain of ERK5 shares approximately 2/3 sequence identify with the kinase domain of ERK1/2 (amino acids 78–406) (Zhou et al. 1995). ERK5 has a molecular weight of approximately 102 kDa, which is more than twice the size of other MAPKs due to its long C-terminus tail containing 410 amino acids. Within the ERK5 C-terminus, there is a nuclear localization signal (NLS; amino acids 505–539) needed for its nuclear translocation. The C-terminus also contains two proline-rich (PR) domains, PR1 (amino acids 434–465) and PR2 (amino acids 578–701), which are thought to be the binding sites for Src-homolog 3 (SH3)-domain-containing proteins, and the myocyte enhancer factor 2 (MEF2)-interacting region (Yan et al. 2001). The ERK5 C-terminus with the transcriptional activity domain contains a binding site for activated 90 kDa ribosomal S6 kinase (p90RSK) (amino acids 571–807), which inhibits ERK5 transcriptional activity by phosphorylating ERK5 at S496 (Heo et al. 2015, 2016; Le et al. 2013) (Fig. 1b). The ERK5 C-terminus also exhibits autoinhibitory effect over its N-terminus kinase domain, as the deletion of the last 100 amino acids in the ERK5 tail increases its kinase activity (Buschbeck and Ullrich 2005).
ERK5 Upstream Regulator: MEK5
ERK5 is activated by several growth factors including vascular endothelial growth factor (VEGF), basic fibroblast growth factor (FGF), platelet-derived growth factor (PDGF), and epidermal growth factor (EGF) as well as by osmotic stress and UV light. ERK5 is also activated by a variety of trophic factors trafficking in neurons such as brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) and specific inflammatory cytokines such as interleukin 6 (IL-6) (Abe et al. 1996; Cavanaugh et al. 2001; Hayashi and Lee 2004; Kato et al. 1998; Kesavan et al. 2004).
To date, MEK5 appears to be the only MAPKK family member known as an immediate upstream that specifically activates ERK5. The MEK5-mediated ERK5 activation requires the MEK5-ERK5 interaction achieved by the MEK5 PB1 domain (Nakamura et al. 2006; Seyfried et al. 2005). Studies have identified MEK5 as an ERK5 binding partner and have shown that ERK5 does not interact with either MEK1 or MEK2, suggesting that the MEK5-ERK5 interaction is a unique event in the canonical ERK5 signaling pathway (Lee et al. 1995; Nishimoto and Nishida 2006; Zhou et al. 1995). Once activated, MEK5 initiates signaling by phosphorylating ERK5 at its T-E-Y dual phosphorylation motif (T218 and Y220) (English et al. 1995; Zhou et al. 1995; Mody et al. 2003). However, it has also been noted that during mitosis, ERK5 is phosphorylated at multiple sites on its C-terminus by cyclin dependent kinase 1 (CDK1) (Inesta-Vaquera et al. 2010), which will be fully discussed in another section.
ERK5 Nuclear Translocation
ERK5 Intramolecular Interaction and N-Terminus Activation
ERK5 nuclear translocation is crucial for transmitting growth factor-initiated signaling to the nucleus. In quiescent cells, overexpressed ERK5 is localized in the cytoplasm, but it translocates to the nucleus when ERK5 is coexpressed with a constitutively active form of MEK5 (Kato et al. 1997; Yan et al. 2001). Investigate the mechanism that modulates ERK5 nuclear translocation upon MEK5 activation, Kondoh et al. (2006) found that both endogenous and overexpressed ERK5 translocated into the nucleus by a stimulus-dependent mechanism and that the subcellular localization of ERK5 was regulated by both active nuclear import and export mechanisms. In their study, they found that the N-terminus of ERK5 was directly bound to its C-terminus and that this binding was required for ERK5 cytoplasmic localization. By using cells coexpressing both N- and C-terminus domains of ERK5, they demonstrated that the complex of the N- and C- termini were exported out of the nucleus while neither the N-terminus nor the C-terminus of ERK5 alone was excluded from the nucleus. Moreover, ERK5 lacking the intramolecularly interacting region fails to localize in the cytoplasm regardless of stimulation, indicating that the intramolecular interaction between the N- and C-termini causes ERK5 nuclear export. When MEK5 is activated, it phosphorylates the ERK5 T-E-Y motif at its N-terminus and disrupts the interaction of the two termini, causing its nuclear translocation and accumulation in the nucleus. Although the NLS is constitutively active, the nuclear export activity is thought to overwhelm the action of NLS, thus the balance is tipped in favor of cytoplasmic localization of ERK5. This localization of ERK changes when MEK5 is activated and phosphorylates ERK5, which disrupts the intramolecular binding, causing ERK5 to translocate into the nucleus (Kondoh et al. 2006; Plotnikov et al. 2011; Yan et al. 2001). What remains to be determined are how the folded ERK5 is exported from the nucleus and which part of the molecule is responsible for this.
The Super-Chaperone Hsp90-Cdc37 and ERK5 Nuclear Translocation (C-Terminus Activation)
The proper folding and maturation of proteins depend on the action of chaperones. Heat shock protein 90 (Hsp90) is an essential chaperone in the cytoplasm. Hsp90 contains an ATP-binding site, which regulates the folding, maturation, and stability of substrate proteins (McClellan et al. 2007; Pearl and Prodromou 2006). When the Hsp90 ATP-binding site is blocked by Hsp90 inhibitors, substrate proteins are released and subjected to degradation by proteasome (Prodromou et al. 1997). For this reason, Hsp90 inhibitors show anticancer effects because Hsp90 is known to stabilize many of its oncogene substrates such as Akt, Raf, ERBB2, and epidermal growth factor receptor (EGFR). In addition to Hsp90, the kinase substrates of Hsp90 require an additional chaperone to maintain their stability. These chaperones are called cochaperones; one of which is the cochaperone cell division-cycle 37 (Cdc37). Cdc37 forms a ternary complex with Hsp90 and the substrate kinase (kinase-Cdc37-Hsp90). The N-terminus domain of Cdc37 interacts with the kinase substrate while the C-terminus domain recruits Hsp90 (Roe et al. 2004). This ternary complex either keeps the substrate kinase active or in a conformation apt for activation (Karnitz and Felts 2007). Interestingly, Cdc37 was found to be overexpressed in many cancer types and to promote the stabilization of various oncogenes (Gray et al. 2007, 2008).
By tandem affinity purification, Erazo et al. (2013) identified the Hsp90-Cdc37 chaperone as an ERK5 interacting partner in the cytoplasm of resting human cells (HEK293, HeLa, PC-3). Geldanamycin and radicicol are Hsp90 inhibitors that bind the ATP binding pocket of Hsp90 (Piper 2001; Piper et al. 2003; Schulte et al. 1998, 1999). Interestingly, when these investigators treated SH-SY5Y and HeLa cells with one of these inhibitors, ERK5 expression was downregulated in a dose-dependent manner. They also found that higher concentration of the drugs completely inhibited ERK5 expression while the expression of non-Hsp90 client MAP kinases, ERK1 and ERK2, was not affected. Similarly, silencing Cdc37 by siRNA resulted in the reduced expression of ERK5 and Akt without affecting the level of ERK1 and ERK2 (Erazo et al. 2013). In addition, they also found that the loss of ERK5 protein induced by Hsp90 inhibition or Cdc37 silencing was due to increased ERK5 ubiquitination and proteasome-mediated degradation. These results indicate the importance of Hsp90 and Cdc37 in maintaining ERK5 stability in the cytoplasm of resting cells. Erazo et al. (2013) suggested that ERK5 activation induced Hsp90 dissociation from the ERK5-Cdc37 complex. This dissociation leads to ERK5 nuclear translocation and consequently increases its transcriptional activity, which was independent of TEY motif phosphorylation, but was dependent on auto-phosphorylation at its C-terminal tail.
Several studies have shown that ERK5 resides in the cytoplasm and migrates to the nucleus in response to EGF (Esparis-Ogando et al. 2002; Kondoh et al. 2006) or active MEK5 overexpression (Kato et al. 1997) while other studies indicate that ERK5 localization depends on the cell type and can even be nuclear in unstimulated cells (Borges et al. 2007; Raviv et al. 2004). As we have discussed earlier, Kondoh et al. (2006) suggest that ERK5 subcellular localization depends on the relative rates of nuclear export vs import, which are determined by the extent of interaction between the N- and C-termini. As Erazo et al. (2013) have suggested the N- and C- termini are bound in unstimulated cells and that this ERK5 conformation exposes its NES and stabilizes in the cytoplasm by forming the trimer with the chaperon proteins. When cells are stimulated, ERK5 autophosphorylates its C-terminus, an event that disrupts the intramolecular interaction between the C- and N-termini, and moves into the nucleus. Because the dissociation of the intramolecular association causes ERK5 to lose its NES activity, it will remain in the nucleus.
ERK5 Phosphorylation in Mitosis
In addition to its known functions to regulate cell proliferation, differentiation, migration, survival, apoptosis, and stress responses, ERK5 regulates cell cycle progression by controlling the G2–M transition. When constitutively active MEK5 and ERK5 were cooverexpressed, the mitotic index increased (Cude et al. 2007). Girio et al. (2007) reported that the level of ERK5 phosphorylation was higher during mitosis when compared with G1 and S phases of the cell cycle. Inhibition of ERK5 by pharmacological means or by RNAi diminishes cell survival in mitosis (Girio et al. 2007). Bim, a BH3-only protein of the Bcl-2 family, is phosphorylated in mitosis in a MEK5-dependent manner (Graos et al. 2005). Girio et al. (2007) found that Bim in mitotic cells was able to coimmunoprecipitate with ERK5. They also found that ERK5 phosphorylated GST-Bim in an in vitro kinase reaction and that inhibition of ERK5 during mitosis caused Bim dephosphorylation, suggesting that phosphorylation of ERK5 and Bim is a decisive step for the survival of proliferating cells.
Using mass spectrometry, different groups of investigators found five phosphorylation sites in ERK5 in mitotic cells: S567, S720, S731, T733, and S803, all of which are located at the ERK5 C-terminus (Mody et al. 2003; Morimoto et al. 2007) (Dephoure et al. 2008; Inesta-Vaquera et al. 2010). Although ERK5 autophosphorylation at the C-terminus increased its kinase and transcriptional activities (Buschbeck and Ullrich 2005; Mody et al. 2003; Morimoto et al. 2007; Terasawa et al. 2003), Inesta-Vaquera et al. (2010) found that the same ERK5 autophosphorylation inhibited both ERK5 kinase and transcriptional activities, suggesting that the regulation of ERK5 activity by ERK5 C-terminus phosphorylation is not that simple. Further studies by Inesta-Vaquera and colleagues showed that ERK5 kinase activity, but not transcriptional activity, was downregulated when all of the five autophosphorylatable residues were mutated to unphosphorylatable alanine (ERK5 (5A)), but both ERK5 kinase and transcriptional activities decreased when the same five residues were mutated to the phosphorylatiom mimic glutamine (ERK5(5E)). These results demonstrate that although ERK5 is phosphorylated at the C-terminus, it may not be active during mitosis. However, much more study is necessary to determine the role of ERK5 C-terminus phosphorylation.
Inesta-Vaquera et al. (2010) reported that the C-terminus phosphorylation of ERK5 was independent of MEK5 but was dependent on cyclin-dependent kinase 1 (CDK1). Indeed, when cells were treated with a CDK1 inhibitor, RO3306, ERK5 phosphorylation during mitosis was aborted (Diaz-Rodriguez and Pandiella 2010). These results suggest that there is crosstalk between CDK1 and ERK5 pathways during mitosis, which could be crucial for the correct cell cycle progression. In nonmitotic cells, the wild type ERK5 and ERK5(5A) were excluded from the nulceus, but ERK5(5E) was mainly localized to the nucleus (Inesta-Vaquera et al. 2010).
A study by Diaz-Rodriguez and Pandiella (2010) identified S706, T732, S753, and S773 as the mitotic ERK5 phosphorylation sites, which are different from the sites identified by others. Borges et al. (2007) reported that the truncation of the C-terminus region containing these four residues (ERK5Δ699) prevented ERK5 nuclear exclusion. However, the Erk5Δ699 mutant, although it was in the nucleus, did not have transcriptional activity (Diaz-Rodriguez and Pandiella 2010). On the other hand, when these phosphorylation sites were substituted by glutamate (phosphorylation mimic residue), such mutant ERK5 was found in the nucleus with increased transacriptional activity in interphase cells (Diaz-Rodriguez and Pandiella 2010). It is interesting to note that phosphorylation of the mitotic phosphorylation sites regulates ERK5 transcriptional activity in nonmitotic cells. It is possible that other posttranslational modifications (PTMs) of ERK5 during mitosis may modify the role of ERK5 C-terminus phosphorylation. The interplay among various PTMs during mitosis is an area not yet studied and awaits investigation.
SUMOylation contributes to the complexity of eukaryotic regulation of proteomes, causes transient changes in signal transduction, and regulates a multitude of cellular processes including proliferation, differentiation, and motility (Wilkinson and Henley 2010). At the molecular level, SUMOylation is known to alter the binding property, enzymatic activity, localization, and stability of proteins. SUMO modification is seen not only in cytosolic proteins but also in those associated with cell organelles including the nucleus, the plasma and intracellular membrane system, and mitochondria. Emerging evidence suggests critical roles of SUMOylation in human diseases such as malignant, cardiovascular, and neurological diseases (Geiss-Friedlander and Melchior 2007; Heo et al. 2015; Hilgarth et al. 2004; Le et al. 2012a).
The ERK5 C-terminus contains a MEF2-interacting domain and also a potent transcriptional activation domain (Kasler et al. 2000), which is negatively regulated by its N-terminus (Akaike et al. 2004). Woo et al. (2008) found that there were two SUMOylation sites (K6 and K22) in the ERK5 N-terminus (Fig. 1b). These SUMOylation sites are near the ATP-binding site and are within the cytoplasmic targeting domain. Reflecting the importance of these SUMOylation sites, they are conserved in the human, mouse, and rat ERK5. Since SUMOylation is known to negatively regulate the activity of certain transcriptional factors (Verger et al. 2003), ERK5 SUMOylation may affect its transactivation. Indeed, ERK5 transcriptional activity is inhibited by increased SUMOylation by overexpressing Ubc9 or PIAS1, a SUMO E3 ligase (Woo et al. 2008). Although Ubc9 inhibits ERK5 transcriptional activity, it does not inhibit the phosphorylation and kinase activity of ERK5, suggesting that ERK5 SUMOylation does not regulate ERK5 phosphorylation and kinase activity (Woo et al. 2008). These authors also reported that ERK5 SUMOylation inhibited MEF2-KLF2 signaling, which is known to play a role in the development and function of the vascular system including inflammation (Nuez et al. 1995; SenBanerjee et al. 2004) (Fig. 1b). Because EC inflammation is among the major initiators of atherosclerosis (Tousoulis et al. 2006), inhibition of ERK5 SUMOylation may be a new therapeutic strategy to treat atherosclerosis.
MEK5/ERK5 Signaling Pathway
MEK5-ERK5 Signaling Pathway in Disease
Endothelial MEK5-ERK5 Signaling in Vascular Inflammation, Dysfunction, and Atherosclerotic Plaque Formation
Shear stress is imposed directly on the luminal surface of blood vessels covered by a thin monolayer of ECs. There are two types of flow, which can differentially modulate endothelial structure and function via regulating local mechanotransduction mechanisms. It has been well established that these local mechanotransduction mechanisms can ultimately activate the shear stress response promoter elements and transcription factors, and determine the phenotypes of endothelial cells via modulating diferent types of endothelial gene expression (Davis et al. 2003; Huddleson et al. 2004; Nagel et al. 1999; Urbich et al. 2003). For example, atherosclerotic plaques are rare in areas exposed to steady laminar flow (s-flow). S-flow (10–20 dyn/cm2) induces anti-inflammatory signaling and antiatherogenic gene expression (Topper et al. 1997; Traub and Berk 1998). S-flow suppresses the expression of inflammatory chemokines and adhesion molecules (VCAM1, ICAM1, E selectin) while maintaining the production of atheroprotective factors (NO, eNOS) (Akaike et al. 2004; Parmar et al. 2006). It also protects vessels from inflammation by inhibiting TNFα-induced activation of the apoptosis signal-regulation kinase1 (ASK1)-JNK signaling (Lerner-Marmarosh et al. 2003; Liu et al. 2001; Surapisitchat et al. 2001; Yamawaki et al. 2003, 2005). In fact, it has been reported that the MEK5-ERK5 signaling, but not ERK1/2 signaling, plays a role in the s-flow-induced anti-inflammatory effects via inhibiting JNK activation in ECs (Li et al. 2008). Furthermore, it has been suggested that Kruppel-like factor 2 (KLF2) plays a crucial role in the s-flow-induced anti-inflammatory effect via ERK5 activation (Akaike et al. 2004; Parmar et al. 2006) (Fig. 1b).
Atherosclerotic plaques localize to areas of disturbed flow (d-flow) found at regions where vessels curve acutely, bifurcate or branch. D-flow has been shown to be proatherogenic; induces inflammation, apoptosis, and proliferation of ECs; and reduces vascular reactivity. Underlying regulatory mechanisms for the d-flow-induced EC inflammation and subsequent atherosclerotic plaque formation have been described in a number of studies (Cecchi et al. 2011). Recently, our group has reported that d-flow induces phosphorylation of serine/threonine kinase p90RSK, which then allows p90RSK to bind the ERK5 C-terminus region (amino acids 571–807). This binding enables p90RSK to phosphorylate ERK5 at S496 (Le et al. 2013), which then inhibits the transcriptional activity of ERK5, thereby ablating KLF2 promoter activity and eNOS activity. This series of molecular events lead to EC inflammation and dysfunction. Moreover, being phosphorylated by d-flow, p90RSK phosphorylates sentrin-specific protease 2 (SENP2), a de-SUMOylation enzyme, at T368. This phosphorylation promotes SENP2 nuclear export, leading to increased ERK5 SUMOylation in the nucleus (Heo et al. 2015, 2016). Because SUMOylation inhibits ERK5 transcriptional activity (Woo et al. 2008), the d-flow-induced ERK5 SUMOylation through p90RSK-induced SENP2 T368 phosphorylation and subsequent SENP2 nuclear export should lead to inhibition of ERK5 transcriptional activity and promote EC inflammation (Heo et al. 2015, 2016).
To investigate the functional role of ERK5 in EC inflammation in vivo, we generated tamoxifen-inducible EC-specific ERK5 knockout mice (ERK5-EKO) and examined leukocyte rolling and acetylcholine-induced vasodilation in the mesenteric arteries. We observed increased leukocyte rolling and impaired vessel reactivity in the ERK5-EKO mice compared to the control mice. In addition, the extent of atherosclerosis a in the aortic root of the inducible ERK5-EKO-LDLR-KO mice was significantly increased when compared to that of the control mice (Le et al. 2013). Overall, these studies suggest that endothelial ERK5 plays a crucial role in determining EC dysfunction and subsequent atherosclerotic plaque formation.
MEK5-ERK5 Signaling in Epithelial Cancers
In addition to its function in the normal cell growth, survival, and differentiation, the MEK5-ERK5 pathway has been reported to activate oncogenes (Drew et al. 2012). Drug resistance is one of the major obstacles in breast cancer therapy. Although some mechanistic aspects underlying drug resistance of tumors have been characterized, reasons for highly variable responses of breast cancers to chemotherapy are not well-understood. Epithelial-mesenchymal transition (EMT) is known to play a role in the drug resistance of cancers. A study carried out by Zhou et al. (2008) demonstrated that MEK5 overexpression (hence ERK5 activation) increased EMT markers in breast cancer cells. In other studies, MEK5 upregulation was also shown to increase the expression of EMT marker genes such as ZEB-1 and SNAIL2 (Weldon et al. 2002; Zhou et al. 2008). Using gene expression microarray, Weldon et al. (2002) found that out of 1186 genes, mek5 was expressed at a significantly high level (22-fold increase) in apoptosis resistant (APO-) MCF-7 cells compared to apoptosis sensitive (APO+)-MCF7 cells. Similarly, when examining 127 cases of prostate cancer and 20 cases of benign prostatic hypertrophy, Mehta et al. (2003) noted a higher level of MEK5 in the prostate cancer tissue. Furthermore, MEK5 upregulation was associated with metastases to the bone and worsened patient outcome. An in vitro study showed that overexpression of MEK5 in prostate cancer cells increased the proliferation and invasion potential of these cells. MEK5 expression correlates with an increase in MMP-9, a matrix metallopeptidase that is involved in the degradation of extracellular matrix, and AP-1, an activator protein required for transcriptional activation of MMP-9 (Mehta et al. 2003). Taken together, current data indicate the potential role of MEK5-ERK5 in oncogenesis.
MEK5-ERK5 Downstream Targets
Recently, much interest has been focused on the mechanism of ERK5 transcriptional activity in which ERK5 autophosphorylation on the C-terminus plays a critical role (Morimoto et al. 2007). As an activator of transcription, ERK5 regulates a wide array of transcriptional factors in different cell types.
Bone is a dynamic tissue that is constantly being remodeled. There are three process of bone remodeling: (1) remove discrete packets of old bone, (2) replace these packets by newly synthesized proteinaceous matrix, and (3) mineralize the matrix to form new bone (Clarke 2008). These processes prevent the accumulation of microdamages within the bone and maintain its strength and mineral homeostasis. The cells responsible for the bone renewal are osteoclasts and osteoblasts, which are involved in bone resorption and bone formation, respectively. Bone formation is increased when mechanical loading is increased, while bone resorption is increased when mechanical loading diminishes (Hillam and Skerry 1995). Mechanical loading is thought to generate fluid flow in the lacunar-canalicular system of the lamellar bone, which exerts shear stress to thin extensions of the osteocyte in canaliculi. Osteocytes are thought to sense this fluid shear stress (estimated to be 12 dyne/cm(2)) and generate biochemical signals that influence the metabolism of osteoblasts and osteocytes (Bacabac et al. 2004; Fritton and Weinbaum 2009; Hillsley and Frangos 1994).
Bone Formation: Cyclooxygenase 2 (COX-2) converts arachidonic acid to prostaglandin E2 (PGE2), which is involved in bone remodeling. COX-2 induction appears to be critical for the anabolic effect of mechanical loading, and COX-2 inhibitors terminate mechanically induced bone remodeling in vivo (Forwood 1996; Klein et al. 2007; Kujubu et al. 1991; McAllister et al. 2000; Ogasawara et al. 2001). Several studies have suggested that ERK5 is involved in flow-induced responses in osteoblasts such as proliferation, differentiation, and inhibition of apoptosis, indicating a potential link between ERK5 and COX-2 induction (Amano et al. 2015; Zhao et al. 2014). To examine this possibility, Jiang et al. (2015) exposed MC3T3-E1 osteoblastic cells to flow and found that stimulation by laminar flow increased COX-2 activity as well as ERK5, CREB, and NF-κB phosphorylation and that these flow effects were suppressed when cells were treated with ERK5 siRNA.
Osteoblast Differentiation: To explore the role of the MEK5-ERK5 signaling pathway in the regulation of Runt-related transcription factor-2 (Runx-2) expression and osteoblast differentiation, both of which are stimulated by intermittent fluid shear stress, Zhao et al. (2014) applied laminar flow to MC3T3-E1 and obsevered a marked increase in ERK5 phosphorylation, Runx-2 expression and phenotypic markers of osteoblast differentiation including alkaline phosphatase (ALP) activity and expression of osteopontin (OPN) and osteocalcin (OCN). When the cells were treated with BIX02189 (MEK5 inhibitor), the expression of Runx-2, OPN, and OCN was reduced (Zhao et al. 2014), indicating that the MEK5-ERK5 pathway plays an important role in mechanically stimulated osteoblast differentiation.
Osteoblast Apoptosis: Bin et al. (2015) found that TNF-α-induced apoptosis of MC3T3-E1 osteoblastic cells was averted when they were exposed to flow 1 h and that this effect was reversed by XMD8–92, an ERK5 inhibitor. Under the same flow condition, Bad phosphorylation was increased while Caspase-3 activity was decreased, and XMD8–92 blocked these effects (Bin et al. 2015). These results suggest that the antiapoptotic effect of flow in osteoblasts is mediated by the MEK5-ERK5 signaling pathway, in which Bad is a crucial downstream target.
ERK5 in Endothelial Cell Survival
Global ERK5-knockout in mice is embryonic lethal at around day 10 due to defective blood vessel formation and impaired cardiac development (Regan et al. 2002; Yan et al. 2003). Relatively similar to global ERK5-knockout mice, EC-specific ERK5-knockout mice also exhibit cardiovascular defects and die at around day 10 (Hayashi et al. 2004). However, ERK5-knockout specific in cardiomyocytes, hepatocyte, and neuronal cells show no severe defects in development (Hayashi et al. 2004; Hayashi and Lee 2004; Kimura et al. 2010; Zou et al. 2012, 2013), suggesting that endothelial ERK5 expression is critical for the development of the cardiovascular system.
ECs play a crucial role in regulating vascular reactivity and permeability, leukocyte adhesion, thromboresistance, vascular cells inflammation, and proliferation (Deanfield et al. 2007; Gimbrone 1999a, b; Nagel et al. 1999; Topper and Gimbrone 1999). Stimuli such as proinflammatory cytokines and blood flow modulate EC phenotype, thereby affecting the vascular physiology. KLF2 is known to play an important role in regulating EC biology such as antithrombotic function, proliferation, migration, angiogenesis, and leukocyte adhesion (Atkins and Jain 2007; Lin et al. 2005). In cultured ECs, KLF2 is inhibited by the inflammatory cytokine interleukin-1β (IL-1β) and is activated by s-flow. HUVECs overexpressing KLF2 robustly induces eNOS expression and total enzymatic activity, while potently inhibiting the induction of VCAM1 and E-selectin in response to various proinflammatory cytokines (Dekker et al. 2002, 2005; SenBanerjee et al. 2004). S-flow and d-flow differently regulate KLF2: s-flow causes sustained KLF2 induction but d-flow causes prolonged KLF2 suppression after a transient induction (Wang et al. 2006a). The s-flow-induced upregulation of KLF2 expression is through activating the MEK5-ERK5-MEF2C signaling cascade (Parmar et al. 2006).
ERK5 in Endothelial Migration and Morphology
New blood vessel formation requires highly coordinated changes in ECs structure/morphology and extracellular matrix (ECM), in which directed EC migration is a crucial step and is dependent on cellular signaling. Spiering et al. (2009) examined the role of MEK5-ERK5 signaling in EC migration and found that the activation of the MEK5-ERK5 pathway induced drastic EC morphological changes including increased focal contacts and changes in actin filament organization. As the consequence, cells increased stiffness and cell motility was significantly reduced. The observed phenotype was likely due to the decreased focal contact turnover caused by the decreased expression of p130Cas, a key player in directed cell migration (Spiering et al. 2009). This study demonstrates that the MEK5-ERK5 signaling pathway plays roles in not only cell survival and stress response but also in EC migration and morphology.
The regulation of cell migration and cytoskeletal restructuring by MEK5-ERK5 is cell-type specific. Overexpressing CA-MEK5 in ECs leads to the inhibition of cell migration and causes cell morphological changes as we have described above (Spiering et al. 2009). In keratinocytes, however, MEK5-ERK5 activation promotes cell migration (Arnoux et al. 2008). During cutaneous wound healing, re-epithelialization requires various signaling that results in basal keratinocyte activation, spreading, and migration, all of which require loosening of cell–cell adhesions. The heightened cell motility is linked to the transcriptional factor Slug activity (Savagner et al. 2005). When coexpressing a reporter gene driven by the Slug promoter with the depletion of ERK5, Arnoux et al. (2008) noted the activation of this reporter gene. In addition, when HaCaT human keratinocyte cells were treated with EGF, ERK5 phosphorylation was induced together with increased Slug mRNA expression and accelerated wound healing. In contrast, the depletion of ERK5 by shRNA in keratinocytes led to reduced Slug expression, and cell migration was totally blocked (Arnoux et al. 2008).
Cardiomyocyte apoptosis is one of the key events in the development and progression of heart failure. Tomita et al. (2003) reported that the overexpression of inducible cAMP early repressor (ICER), an endogenous inhibitor of CRE-mediated transcription, induced cardiomyocyte apoptosis. Antisense inhibition of ICER significantly inhibited cardiac myocyte apoptosis, suggesting that ICER functions as a proapoptosis regulator in cardiomyocytes. ICER is ubiquitinated and degraded by C-terminus of Hsc70-interacting protein (CHIP), an E3 ubiquitin ligase enzyme with cardio-protective function (Zhang et al. 2005). In the heart, activation of ERK5 induces its association with CHIP, upregulates CHIP ubiquitin ligase activity, thereby causing ICER degradation and inhibiting cardiomyocyte apoptosis (Woo et al. 2010). We have shown that p90RSK and CHIP share a common binding site on the ERK5 C-terminus (amino acids 571–807) and that activation of p90RSK inhibits the ERK5-CHIP association and reduces CHIP ubiquitin ligase activity (Le et al. 2012b). This enhances ICER expression and promotes cardiomyocyte apoptosis. Similar to p90RSK activation in the endothelium, the activation of p90RSK in the heart also results in increased ERK5 S496 phosphorylation, leading to the increased ICER expression and cardiomyocyte apoptosis. In this study, we have also shown that ERK5 has both proapoptotic and antiapoptotic functions depending on the site of phosphorylation. While phosphorylation of ERK5 at the T-E-Y motif inhibits cell death, phosphorylation at S496 promotes cell death (Le et al. 2012b).
Because cardiomyocytes do not divide beyond birth in mammals, hypertrophic growth of cardiomyocytes is seen durings both normal and stress-induced remodeling (MacLellan and Schneider 2000). Roles of MAPKs in cardiomyocyte hypertrophic growth have been suggested (Aoki et al. 2000; Zhang et al. 2003). Bueno et al. (2000) reported that specific activation of ERK1/2, but not JNK1/2 or p38, was associated with physiologic hypertrophy of the heart, whose cardiac function and resistance to apoptotic stimuli were increased. To investigate the role of the MEK5–ERK5 pathway in cardiac hypertrophy, Nicol et al. (2001) overexpressed CA-MEK5 in cardiomyocytes in vitro and found that MEK5 activation induced elongated cell morphology, a form of hypertrophic phenotype in vitro. Leukemia inhibitory factor (LIF)-mediated MEK5 activation evoked a similar morphological response. Overexpression of dominant negative MEK5, on the other hand, inhibited the LIF-induced cardiomyocyte elongation and expression of fetal cardiac genes without inhibiting other aspects of LIF-induced hypertrophy. In vivo, transgenic mice–expressing cardiac-specific MEK5 exhibited eccentric cardiac hypertrophy that progressed to dilated cardiomyopathy and sudden death (Nicol et al. 2001). These studies appear to link ERK5 activation in cardiomyocytes to cardiomyopathy and inhibition of cardiomyocyte apoptosis.
Immune System (Macrophages)
Macrophages are the central cells for the innate immunity and inflammation. Colony-stimulating factor 1 (CSF-1, or macrophage CSF or M-CSF) is one of the important agonists to regulate the growth, survival, and differentiation of macrophages (Stanley et al. 1997). CSF-1 specifically binds its high-affinity receptor CSF-1R (encoded by the c-fms protooncogene). CSF-1R is tyrosine kinase receptor, and the binding to CSF-1 ligand increases CSF-1R tyrosine phosphorylation, and consequently activates STAT1, 3, and 5 transcriptional activities via upregulating ERK1/2 and PI3-K activation (Hamilton 1997). To examine the underlying mechanism of CSF-1-mediated sustained proliferation and survival of macrophages, Rovida et al. (2008b) treated peripheral blood-derived human macrophages and BAC1.2F5 cells (a cell line derived from mouse macrophages) with CSF-1 and found a rapid and sustained increase in ERK5 T-E-Y motif phosphorylation in a dose-dependent manner. Treatment of BAC1.2F5 cells with GM-CSF, another mitogen, also resulted in ERK5 T-E-Y motif phosphorylation, but CSF-1 appears to be the most potent ERK5 activator. Potent macrophage activators such as LPS or IFN-γ, and other inflammatory cytokines such as IL1 or IL6, however, fail to induce ERK5 phosphorylation, suggesting that proliferative but not activating signals stimulate ERK5 phosphorylation in macrophages. To examine the role of ERK5 in CSF-1-induced macrophage proliferation, macrophage ERK5 was depleted in BAC1.2F5 cells by ERK5-specific siRNA, and [3H] thymidine incorporation was measured. They found that the CSF-1-stimulated increase in [3H] thymidine incorporation by the ERK5 siRNA treated cells was inhibited although their basal [3H] thymidine incorporation was not affected by the ERK5 null condition. Moreover, in the ERK5 siRNA treated cells, CSF-1-induced c-Jun expression and phosphorylation was decreased while the expression of p27 was increased. As p27 inhibits CSF-1-mediated proliferation (Kato et al. 1994), the effects of ERK5 depletion on c-Jun and p27 expression levels may explain why ERK5 depletion impairs CSF-1-induced mitogenesis. As we have described earlier, Lee and States (2006) reported that the CSF-1-dependent myeloid cell proliferation depended on ERK1/2 and PI3K. The depletion of ERK5 by siRNA does not alter either PI3K activation, at least on the basis of Akt phosphorylation, or ERK1/2 phosphorylation (Rovida et al. 2008a, b). It appears, therefore, that ERK5 does not appear to be a directly upstream regulator of PI3K or ERK1/2 in CSF-1- or PDGF-stimulated cells (Rovida et al. 2008b). These results show that macrophage ERK5 plays an indispensable role in CSF-1-mediated macrophage proliferation and survival.
Throughout the progression of atherosclerotic lesion development, macrophages undergo apoptosis and they are cleared by live macrophages (efferocytosis). In time, these macrophages also die and need to be cleared. In advanced atherosclerotic lesions, it is thought that efferocytotic clearance cannot keep pace with increasing numbers of dying macrophages and that apoptotic macrophages accumulate and become a necrotic core. Our understanding of the molecular mechanisms for both normal and defective efferocytosis is severely limited. Recently, our group has uncovered the crucial role of macrophage ERK5 activation in controlling efferocytosis (Heo et al. 2014). When macrophages were fed apoptotic cells in vitro, efferocytotic capacity is upregulated in an ERK5 activity dependent manner. As the consequence, efferocytosis-related signaling is also upregulated. Macrophages isolated from LysM-Cre-specific ERK5-knockout mice (myeloid cells specific deletion) display reduced efferocytosis together with a reduced expression level of both mRNA and protein of efferocytosis-related genes. When LysM-Cre-specific ERK5-knockout mice in the LDLR-KO background were fed a high cholesterol diet, the formation of atherosclerotic plaques was increased with more advance and vulnerable plaque features.
We have shown that pitavastatin activates ERK5 (Le et al. 2014). When cultured macrophages were fed apoptotic cells in the presence of pitavastatin, efferocytosis was upregulated. In addition, pitavastatin treatment reduced the extent of atherosclerotic plaque formation in mice (Heo et al. 2014). Our studies suggest that macrophage ERK5 is atheroprotective and that activation of macrophage ERK5 and its signaling cascade may be a good target for treating cardiovascular diseases.
Although ERK5 is ubiquitously expressed, the level of expression is not the same in different organs, and the highest level is found in the brain (Yan et al. 2003) with the most robust activation during embryonic development (Liu et al. 2003). In response to various neurotrophic prosurvival stimuli, ERK5 appears to be a crucial mediator for neuronal growth and survival (Liu et al. 2003; Wang et al. 2006b), which require target-derived neurotrophins. Neurotrophins are growth factors that play important roles in neuronal cell development, differentiation, and survival (Korsching 1993). Dysregulation of neurotrophins has been implicated in various neurodegenerative disorders. High-affinity receptors for neurotrophins are the Trk (tropomyosin receptor kinase) receptor protein tyrosine kinases, which consists of three members: TrkA, TrkB, and TrkC (Trk A/B/C) (Green et al. 2012). It has been demonstrated that ligand binding activates Trk receptors locally within the distal axon and initiates signaling. The ligand-receptor complex is endocytosed and retrogradely carried inside the long axon to the cell body (Ginty and Segal 2002). In mass cultures of DRG (dorsal root ganglia) neurons, neurotrophin upregulates both ERK5 kinase activity and ERK5 T-E-Y motif phosphorylation, which can be inhibited by PD98059, a MEK5 inhibitor (Watson et al. 2001). In addition, neurotrophin treatment induces ERK5 nuclear translocation. When a Trk kinase inhibitor K252a was applied to either the distal axon or the cell body, both treatments blocked ERK5 activation in the cell body following neurotrophin stimulation of the distal axon (Watson et al. 2001). These results suggest that the ERK5 signaling pathway is critically involved in the transmission of the retrograde survival signal initiated by neurotrophins at the axon tip.
Using rat cortical neurons, Wang et al. (2006b) have shown that BDNF (brain-derived neurotrophic factor) mediates sustained activation of ERK5 as well as that of a small GTPase Rap1 and MEKK2 (Wang et al. 2006b). Activation of either Rap1 or MEKK2 was sufficient to induce ERK5 activation, whereas inhibition of either Rap1 or MEKK2 prevented sustained ERK5 activation. Moreover, the BDNF-mediated MEKK2 activation was Rap1 dependent, indicating that Rap1 and MEKK2 are critical upstream signaling molecules for ERK5 activation by BDNF in CNS (Wang et al. 2006b). To study the role of ERK5 during the development of the peripheral nervous system (PNS), Finegan et al. (2009) deleted erk5 gene in cultured sympathetic neurons. They found that ERK5 deletion increased cell death which was associated with the elevated expression of Bad and Bim, the BH3-only family members of the Bcl-2 family. These results show that ERK5 suppresses the transcription of the bad and bim genes.
Dysregulation of BDNF, a member of the neurotrophin family, has been implicated in the long-term potentiation and cognition deficiency and may be involved in the development of Alzheimer’s disease. Su et al. (2011) found that the BDNF mRNA level was increased when C6 cells and primary astrocytes were treated with U0126, an inhibitor of both ERK5 and ERK1/2. Interestingly, treatment of cells with PD184352, a MEK1/2-specific inhibitor, did not increase BDNF mRNA expression levels, suggesting that ERK5 negatively regulates BDNF expression. Consistently, RNAi-mediated ERK5 depletion resulted in increased BDNF expression while constitutively acitve-MEK5 transfection led to reduced BDNF expression (Su et al. 2011; Yu et al. 2012).
MEK5-ERK5 and MEK1/2-ERK1/2 Pathways Are Interconnected
ERK5 and ERK1/2, although they have different upstream kinases (MEK5 vs MEK1/2, respectively), share some features in common including the ability to regulate cell cycle progression, the amino acid sequence (approximately 2/3 identical), and being activated by stress stimuli (Nishimoto and Nishida 2006). However, they do function differently. For example, ERK1/2 plays an important role in regulating intestinal epithelial proliferation and differentiation, but ERK5’s role is not known (Osaki and Gama 2013). When investigating a potential crosstalk between ERK1/2 and other MAPK signaling pathways in intestinal epithelial cells and colorectal cancer, de Jong et al. (2016) made an unexpected discovery: the loss of ERK1/2 in intestinal epithelial cells negatively affected cell migration and differentiation but not cell proliferation. This observation suggests the existence of some compensatory mechanisms for cell proliferation. Genetic deletion of erk1/2 or pharmacological inhibition of MEK1/2 resulted in supraphysiological activity of ERK5 in intestinal epithelial cells. Indeed, cell proliferation was suppressed in mouse intestinal organoids and human colorectal cancer cell lines when both ERK5 and ERK1/2 were inhibited. These results reveal the presence of interplay between ERK5 and ERK1/2 signaling pathways.
The discovery of the MAPK family has provided important information regarding major signaling pathways involved in both normal and disease development. A huge effort has been made to understand the biology of ERK1/2, JNK, and p38, but it is more recent that the ERK5 system has become a target of focused investigation. As such, we have made progress toward understanding the MEK5-ERK5 regulatory mechanism, the biological targets of the pathway, as well as strategies for the development of effective pharmacological and molecular interventions specifically targeted to ERK5. In this review, we have highlighted the structure, activation, and function of the two key players in the ERK5 signaling pathway. The MEK5-ERK5 signaling cascade plays a critical role in making cell fate decisions such as proliferation, differentiation, migration, survival, and death of cells (Drew et al. 2012; Hayashi and Lee 2004; Kato et al. 1998; Roberts et al. 2009; Wang and Tournier 2006).
Due to its unique structure with the transcriptional activity domain in the long C-terminus, ERK5 can act as both a positive and negative regulator of cell apoptosis. This function can be achieved by regulating its multiphosphorylation sites. In the heart, activation of the ERK5 T-E-Y motif induces ERK5-CHIP association, upregulating CHIP ubiquitin ligase activity. This causes ICER ubiquitination and degradation and inhibits cardiac apoptosis. The activation of p90RSK leads to ERK5 S496 phosphorylation, which leads to increased ICER protein expression and myocyte apoptosis. In the endothelium, s-flow activates the MEK5-ERK5 signaling pathway by phosphorylating ERK5 at the T-E-Y phosphorylation motif, and the phosphorylation of this site promotes anti-inflammatory and antiatherogenic effects of ERK5. On the other hand, d-flow induces p90RSK-dependent ERK5 S496 phosphorylation and promotes EC inflammation and apoptosis.
ERK5 exhibits a dual residency: in the cytoplasm and the nucleus, responding accordingly to different extracellular signals. Although available information regarding ERK5 nuclear translocation and functionality is summarized, our understanding on the subject remains incomplete. Further studies are necessary to determine the regulatory mechanisms of ERK5 cellular localization, which is critical in regulating kinase and transcriptional activity. Although emerging evidence has highlighted the existence of a noncanonical, CDK1-dependent, ERK5 phosphorylation pathway during mitosis, further studies are required to gain more insights into the role of ERK5 phosphorylation during mitosis. In addition, as there is a link between ERK5 expression and patient prognosis in cancer, from the clinical point of view, it would be important to examine the correlation between the state of ERK5 phosphorylation at various sites at its C-terminus and oncogenesis.
In this review, we discussed various cellular functions of ERK5. For example, we discussed the role of MEK5-ERK5 signaling in maintaining the integrity of the existing vasculature, angiogenesis, EMT, and shear stress-mediated antiapoptotic and inflammatory effects in osteoblasts and ECs. However, there are many controversies regarding the functional role of ERK5, and we believe that the controversy is due to ERK5’s ability to be posttranslationally modified at multiple sites in various manners including phosphorylation, SUMOylation, and ubiquitination. The functional interplay among these posttranslational modifications needs further investigation to clarify the existing controversies.
- Akaike M, Che W, Marmarosh NL, Ohta S, Osawa M, Ding B, Berk BC, Yan C, Abe J. The hinge-helix 1 region of peroxisome proliferator-activated receptor gamma1 (PPARgamma1) mediates interaction with extracellular signal-regulated kinase 5 and PPARgamma1 transcriptional activation: involvement in flow-induced PPARgamma activation in endothelial cells. Mol Cell Biol. 2004;24:8691–704.PubMedPubMedCentralCrossRefGoogle Scholar
- Boulton TG, Nye SH, Robbins DJ, Ip NY, Radziejewska E, Morgenbesser SD, DePinho RA, Panayotatos N, Cobb MH, Yancopoulos GD. ERKs: a family of protein-serine/threonine kinases that are activated and tyrosine phosphorylated in response to insulin and NGF. Cell. 1991;65:663–75.PubMedCrossRefGoogle Scholar
- Chayama K, Papst PJ, Garrington TP, Pratt JC, Ishizuka T, Webb S, Ganiatsas S, Zon LI, Sun W, Johnson GL, Gelfand EW. Role of MEKK2-MEK5 in the regulation of TNF-alpha gene expression and MEKK2-MKK7 in the activation of c-Jun N-terminal kinase in mast cells. Proc Natl Acad Sci U S A. 2001;98:4599–604.PubMedPubMedCentralCrossRefGoogle Scholar
- Dekker RJ, van Thienen JV, Rohlena J, de Jager SC, Elderkamp YW, Seppen J, de Vries CJ, Biessen EA, van Berkel TJ, Pannekoek H, Horrevoets AJ. Endothelial KLF2 links local arterial shear stress levels to the expression of vascular tone-regulating genes. Am J Pathol. 2005;167:609–18.PubMedPubMedCentralCrossRefGoogle Scholar
- Erazo T, Moreno A, Ruiz-Babot G, Rodriguez-Asiain A, Morrice NA, Espadamala J, Bayascas JR, Gomez N, Lizcano JM. Canonical and kinase activity-independent mechanisms for extracellular signal-regulated kinase 5 (ERK5) nuclear translocation require dissociation of Hsp90 from the ERK5-Cdc37 complex. Mol Cell Biol. 2013;33:1671–86.PubMedPubMedCentralCrossRefGoogle Scholar
- Gao, C., W. Huang, K. Kanasaki, and Y. Xu. 2014. The role of ubiquitination and sumoylation in diabetic nephropathy. BioMed Res Int. 2014;160692.Google Scholar
- Karnitz LM, Felts SJ. Cdc37 regulation of the kinome: when to hold ’em and when to fold ’em. Sci STKE. 2007;pe22.Google Scholar
- Kimura TE, Jin J, Zi M, Prehar S, Liu W, Oceandy D, Abe J, Neyses L, Weston AH, Cartwright EJ, Wang X. Targeted deletion of the extracellular signal-regulated protein kinase 5 attenuates hypertrophic response and promotes pressure overload-induced apoptosis in the heart. Circ Res. 2010;106:961–70.PubMedPubMedCentralCrossRefGoogle Scholar
- Lee AW. Atypical protein kinase Cs promote CSF-1-dependent Erk activation and proliferation in myeloid cells. Blood. 2006;108:4227.Google Scholar
- Le NT, Corsetti JP, Dehoff-Sparks JL, Sparks CE, Fujiwara K, Abe J. Reactive oxygen species, SUMOylation, and endothelial inflammation. Int J Inflamm. 2012a;678190.Google Scholar
- Le NT, Takei Y, Shishido T, Woo CH, Chang E, Heo KS, Lee H, Lu Y, Morrell C, Oikawa M, McClain C, Wang X, Tournier C, Molina CA, Taunton J, Yan C, Fujiwara K, Patterson C, Yang J, Abe J. p90RSK targets the ERK5-CHIP ubiquitin E3 ligase activity in diabetic hearts and promotes cardiac apoptosis and dysfunction. Circ Res. 2012b;110:536–50.PubMedPubMedCentralCrossRefGoogle Scholar
- Le NT, Heo KS, Takei Y, Lee H, Woo CH, Chang E, McClain C, Hurley C, Wang X, Li F, Xu H, Morrell C, Sullivan MA, Cohen MS, Serafimova IM, Taunton J, Fujiwara K, Abe J. A crucial role for p90RSK-mediated reduction of ERK5 transcriptional activity in endothelial dysfunction and atherosclerosis. Circulation. 2013;127:486–99.PubMedCrossRefGoogle Scholar
- Le NT, Takei Y, Izawa-Ishizawa Y, Heo KS, Lee H, Smrcka AV, Miller BL, Ko KA, Ture S, Morrell C, Fujiwara K, Akaike M, Abe J. Identification of activators of ERK5 transcriptional activity by high-throughput screening and the role of endothelial ERK5 in vasoprotective effects induced by statins and antimalarial agents. J Immunol. 2014;193:3803–15.PubMedPubMedCentralCrossRefGoogle Scholar
- Lee JD, Ulevitch RJ, Han J. Primary structure of BMK1: a new mammalian map kinase. Biochem Biophys. 1995;213:715–24.Google Scholar
- Lerner-Marmarosh N, Yoshizumi M, Che W, Surapisitchat J, Kawakatsu H, Akaike M, Ding B, Huang Q, Yan C, Berk BC, Abe J. Inhibition of tumor necrosis factor-[alpha]-induced SHP-2 phosphatase activity by shear stress: a mechanism to reduce endothelial inflammation. Arterioscler Thromb Vasc Biol. 2003;23:1775–81.PubMedCrossRefGoogle Scholar
- Moscat J, Diaz-Meco MT. The atypical protein kinase Cs. Functional specificity mediated by specific protein adapters. 2000. Review.Google Scholar
- Ogasawara A, Arakawa T, Kaneda T, Takuma T, Sato T, Kaneko H, Kumegawa M, Hakeda Y. Fluid shear stress-induced cyclooxygenase-2 expression is mediated by C/EBP beta, cAMP-response element-binding protein, and AP-1 in osteoblastic MC3T3-E1 cells. J Biol Chem. 2001;276:7048–54.PubMedCrossRefGoogle Scholar
- Piper PW, Millson SH, Mollapour M, Panaretou B, Siligardi G, Pearl LH, Prodromou C. Sensitivity to Hsp90-targeting drugs can arise with mutation to the Hsp90 chaperone, cochaperones and plasma membrane ATP binding cassette transporters of yeast. Eur J Biochem. 2003;270:4689–95.PubMedCrossRefGoogle Scholar
- SenBanerjee S, Lin Z, Atkins GB, Greif DM, Rao RM, Kumar A, Feinberg MW, Chen Z, Simon DI, Luscinskas FW, Michel TM, Gimbrone Jr MA, Garcia-Cardena G, Jain MK. KLF2 is a novel transcriptional regulator of endothelial proinflammatory activation. J Exp Med. 2004;199:1305–15.PubMedPubMedCentralCrossRefGoogle Scholar
- Sumimoto H, Kamakura S, Ito T. Structure and function of the PB1 domain, a protein interaction module conserved in animals, fungi, amoebas, and plants. Sci STKE. 2007;re6.Google Scholar
- Sun W, Wei X, Kesavan K, Garrington TP, Fan R, Mei J, Anderson SM, Gelfand EW, Johnson GL. MEK kinase 2 and the adaptor protein Lad regulate extracellular signal-regulated kinase 5 activation by epidermal growth factor via Src. Mol Cell Biol. 2003;23:2298–308.PubMedPubMedCentralCrossRefGoogle Scholar
- Surapisitchat J, Hoefen RJ, Pi X, Yoshizumi M, Yan C, Berk BC. Fluid shear stress inhibits TNF-alpha activation of JNK but not ERK1/2 or p38 in human umbilical vein endothelial cells: Inhibitory crosstalk among MAPK family members. Proc Natl Acad Sci USA. 2001;98:6476–81.PubMedPubMedCentralCrossRefGoogle Scholar
- Tomita H, Nazmy M, Kajimoto K, Yehia G, Molina CA, Sadoshima J. Inducible cAMP early repressor (ICER) is a negative-feedback regulator of cardiac hypertrophy and an important mediator of cardiac myocyte apoptosis in response to beta-adrenergic receptor stimulation. Circ Res. 2003;93:12–22.PubMedCrossRefGoogle Scholar
- Topper JN, Cai J, Qiu Y, Anderson KR, Xu YY, Deeds JD, Feeley R, Gimeno CJ, Woolf EA, Tayber O, Mays GG, Sampson BA, Schoen FJ, Gimbrone Jr MA, Falb D. Vascular MADs: two novel MAD-related genes selectively inducible by flow in human vascular endothelium. Proc Natl Acad Sci U S A. 1997;94:9314–9.PubMedPubMedCentralCrossRefGoogle Scholar
- Weldon CB, Scandurro AB, Rolfe KW, Clayton JL, Elliott S, Butler NN, Melnik LI, Alam J, McLachlan JA, Jaffe BM, Beckman BS, Burow ME. Identification of mitogen-activated protein kinase kinase as a chemoresistant pathway in MCF-7 cells by using gene expression microarray. Surgery. 2002;132:293–301.PubMedCrossRefGoogle Scholar
- Woo CH, Shishido T, McClain C, Lim JH, Li JD, Yang J, Yan C, Abe J. Extracellular signal-regulated kinase 5 SUMOylation antagonizes shear stress-induced antiinflammatory response and endothelial nitric oxide synthase expression in endothelial cells. Circ Res. 2008;102:538–45.PubMedCrossRefGoogle Scholar
- Woo CH, Le NT, Shishido T, Chang E, Lee H, Heo KS, Mickelsen DM, Lu Y, McClain C, Spangenberg T, Yan C, Molina CA, Yang J, Patterson C, Abe J. Novel role of C terminus of Hsc70-interacting protein (CHIP) ubiquitin ligase on inhibiting cardiac apoptosis and dysfunction via regulating ERK5-mediated degradation of inducible cAMP early repressor. FASEB J. 2010;24:4917–28.PubMedPubMedCentralCrossRefGoogle Scholar
- Yu SJ, Grider JR, Gulick MA, Xia CM, Shen S, Qiao LY. Up-regulation of brain-derived neurotrophic factor is regulated by extracellular signal-regulated protein kinase 5 and by nerve growth factor retrograde signaling in colonic afferent neurons in colitis. Exp Neurol. 2012;238:209–17.PubMedPubMedCentralCrossRefGoogle Scholar
- Zhou C, Nitschke AM, Xiong W, Zhang Q, Tang Y, Bloch M, Elliott S, Zhu Y, Bazzone L, Yu D, Weldon CB, Schiff R, McLachlan JA, Beckman BS, Wiese TE, Nephew KP, Shan B, Burow ME, Wang G. Proteomic analysis of tumor necrosis factor-alpha resistant human breast cancer cells reveals a MEK5/Erk5-mediated epithelial-mesenchymal transition phenotype. Breast Cancer Res. 2008;10:R105.PubMedPubMedCentralCrossRefGoogle Scholar
- Zou J, Pan YW, Wang Z, Chang SY, Wang W, Wang X, Tournier C, Storm DR, Xia Z. Targeted deletion of ERK5 MAP kinase in the developing nervous system impairs development of GABAergic interneurons in the main olfactory bulb and behavioral discrimination between structurally similar odorants. J Neurosci. 2012;32:4118–32.PubMedPubMedCentralCrossRefGoogle Scholar