Small molecule ERK5 kinase inhibitors paradoxically activate ERK5 signalling: be careful what you wish for .

ERK5 is a protein kinase that also contains a nuclear localisation signal and a transcriptional transactivation domain. Inhibition of ERK5 has therapeutic potential in cancer and inflammation and this has prompted the development of ERK5 kinase inhibitors (ERK5i). However, few ERK5i programmes have taken account of the ERK5 transactivation domain. We have recently shown that the binding of small molecule ERK5i to the ERK5 kinase domain stimulates nuclear localisation and paradoxical activation of its transactivation domain. Other kinase inhibitors paradoxically activate their intended kinase target, in some cases leading to severe physiological consequences highlighting the importance of mitigating these effects. Here, we review the assays used to monitor ERK5 activities (kinase and transcriptional) in cells, the challenges faced in development of small molecule inhibitors to the ERK5 pathway, and classify the molecular mechanisms of paradoxical activation of protein kinases by kinase inhibitors.

Identification of DYRK1B as a substrate of ERK1/2 and characterisation of the kinase activity of DYRK1B mutants from cancer and metabolic syndrome.

The dual-specificity tyrosine-phosphorylation-regulated kinase, DYRK1B, is expressed de novo during myogenesis, amplified or mutated in certain cancers and mutated in familial cases of metabolic syndrome. DYRK1B is activated by cis auto-phosphorylation on tyrosine-273 (Y273) within the activation loop during translation but few other DYRK1B phosphorylation sites have been characterised to date. Here, we demonstrate that DYRK1B also undergoes trans-autophosphorylation on serine-421 (S421) in vitro and in cells and that this site contributes to DYRK1B kinase activity. Whilst a DYRK1B(S421A) mutant was completely defective for p-S421 in cells, DYRK1B inhibitors caused only a partial loss of p-S421 suggesting the existence of an additional kinase that could also phosphorylate DYRK1B S421. Indeed, a catalytically inactive DYRK1B(D239A) mutant exhibited very low levels of p-S421 in cells but this was increased by KRAS(G12V). In addition, selective activation of the RAF-MEK1/2-ERK1/2 signalling pathway rapidly increased p-S421 in cells whereas activation of the stress kinases JNK or p38 could not. S421 resides within a Ser-Pro phosphoacceptor motif that is typical for ERK1/2 and recombinant ERK2 phosphorylated DYRK1B at S421 in vitro. Our results show that DYRK1B is a novel ERK2 substrate, uncovering new links between two kinases involved in cell fate decisions. Finally, we show that DYRK1B mutants that have recently been described in cancer and metabolic syndrome exhibit normal or reduced intrinsic kinase activity.

A novel DYRK1B inhibitor AZ191 demonstrates that DYRK1B acts independently of GSK3ß to phosphorylate cyclin D1 at Thr(286), not Thr(288).

DYRK1B (dual-specificity tyrosine phosphorylation-regulated kinase 1B) is amplified in certain cancers and may be an oncogene; however, our knowledge of DYRK1B has been limited by the lack of selective inhibitors. In the present study we describe AZ191, a potent small molecule inhibitor that selectively inhibits DYRK1B in vitro and in cells. CCND1 (cyclin D1), a key regulator of the mammalian G1-S-phase transition, is phosphorylated on Thr(286) by GSK3ß (glycogen synthase kinase 3ß) to promote its degradation. DYRK1B has also been proposed to promote CCND1 turnover, but was reported to phosphorylate Thr(288) rather than Thr(286). Using in vitro kinase assays, phospho-specific immunoblot analysis and MS in conjunction with AZ191 we now show that DYRK1B phosphorylates CCND1 at Thr(286), not Thr(288), in vitro and in cells. In HEK (human embryonic kidney)-293 and PANC-1 cells (which exhibit DYRK1B amplification) DYRK1B drives Thr(286) phosphorylation and proteasome-dependent turnover of CCND1 and this is abolished by AZ191 or DYRK1B RNAi, but not by GSK3ß inhibitors or GSK3ß RNAi. DYRK1B expression causes a G1-phase cell-cycle arrest, but overexpression of CCND1 (wild-type or T286A) fails to overcome this; indeed, DYRK1B also promotes the expression of p21CIP1 (21 kDa CDK-interacting protein 1) and p27KIP1 (CDK-inhibitory protein 1). The results of the present study demonstrate for the first time that DYRK1B is a novel Thr(286)-CCND1 kinase that acts independently of GSK3ß to promote CCND1 degradation. Furthermore, we anticipate that AZ191 may prove useful in defining further substrates and biological functions of DYRK1B.

ERK5 and its role in tumour development.

The MEK5 [MAPK (mitogen-activated protein kinase)/ERK (extracellular-signal-regulated kinase) kinase 5]/ERK5 pathway is the least well studied MAPK signalling module. It has been proposed to play a role in the pathology of cancer. In the present paper, we review the role of the MEK5/ERK5 pathway using the 'hallmarks of cancer' as a framework and consider how this pathway is deregulated. As well as playing a key role in endothelial cell survival and tubular morphogenesis during tumour neovascularization, ERK5 is also emerging as a regulator of tumour cell invasion and migration. Several oncogenes can stimulate ERK5 activity, and protein levels are increased by a novel amplification at chromosome locus 17p11 and by down-regulation of the microRNAs miR-143 and miR-145. Together, these finding underscore the case for further investigation into understanding the role of ERK5 in cancer.

ERK5 Signalling and Resistance to ERK1/2 Pathway Therapeutics: The Path Less Travelled?

The RAS-regulated RAF-MEK1/2-ERK1/2 signalling pathway is frequently de-regulated in human cancer. Melanoma in particular exhibits a high incidence of activating BRAFV600E/K and NRASQ61L/K mutations and such cells are addicted to the activity of these mutant oncoproteins. As a result three different BRAF inhibitors (BRAFi) have now been approved for BRAFV600E/K- mutant melanoma and have transformed the treatment of this disease. Despite this, clinical responses are typically transient as tumour cells develop resistance. These resistance mechanisms frequently involve reinstatement of ERK1/2 signalling and BRAFi are now deployed in combination with one of three approved MEK1/2 inhibitors (MEKi) to provide more durable, but still transient, clinical responses. Furthermore, inhibitors to ERK1/2 (ERK1/2i) have also been developed to counteract ERK1/2 signalling. However, recent studies have suggested that BRAFi/MEKi and ERK1/2i resistance can arise through activation of a parallel signalling pathway leading to activation of ERK5, an unusual protein kinase that contains both a kinase domain and a transcriptional transactivation domain. Here we review the evidence supporting ERK5 as a mediator of BRAFi/MEKi and ERK1/2i resistance. We also review the challenges in targeting ERK5 signalling with small molecules, including paradoxical activation of the transcriptional transactivation domain, and discuss new therapeutic modalities that could be employed to target ERK5.

Paradoxical activation of the protein kinase-transcription factor ERK5 by ERK5 kinase inhibitors.

The dual protein kinase-transcription factor, ERK5, is an emerging drug target in cancer and inflammation, and small-molecule ERK5 kinase inhibitors have been developed. However, selective ERK5 kinase inhibitors fail to recapitulate ERK5 genetic ablation phenotypes, suggesting kinase-independent functions for ERK5. Here we show that ERK5 kinase inhibitors cause paradoxical activation of ERK5 transcriptional activity mediated through its unique C-terminal transcriptional activation domain (TAD). Using the ERK5 kinase inhibitor, Compound 26 (ERK5-IN-1), as a paradigm, we have developed kinase-active, drug-resistant mutants of ERK5. With these mutants, we show that induction of ERK5 transcriptional activity requires direct binding of the inhibitor to the kinase domain. This in turn promotes conformational changes in the kinase domain that result in nuclear translocation of ERK5 and stimulation of gene transcription. This shows that both the ERK5 kinase and TAD must be considered when assessing the role of ERK5 and the effectiveness of anti-ERK5 therapeutics.

Identification of a novel orally bioavailable ERK5 inhibitor with selectivity over p38α and BRD4.

Extracellular regulated kinase 5 (ERK5) signalling has been implicated in driving a number of cellular phenotypes including endothelial cell angiogenesis and tumour cell motility. Novel ERK5 inhibitors were identified using high throughput screening, with a series of pyrrole-2-carboxamides substituted at the 4-position with an aroyl group being found to exhibit IC50 values in the micromolar range, but having no selectivity against p38α MAP kinase. Truncation of the N-substituent marginally enhanced potency (∼3-fold) against ERK5, but importantly attenuated inhibition of p38α. Systematic variation of the substituents on the aroyl group led to the selective inhibitor 4-(2-bromo-6-fluorobenzoyl)-N-(pyridin-3-yl)-1H-pyrrole-2-carboxamide (IC50 0.82 µM for ERK5; IC50 > 120 µM for p38α). The crystal structure (PDB 5O7I) of this compound in complex with ERK5 has been solved. This compound was orally bioavailable and inhibited bFGF-driven Matrigel plug angiogenesis and tumour xenograft growth. The selective ERK5 inhibitor described herein provides a lead for further development into a tool compound for more extensive studies seeking to examine the role of ERK5 signalling in cancer and other diseases.

dDYRK2 and Minibrain interact with the chromatin remodelling factors SNR1 and TRX.

The DYRKs (dual specificity tyrosine phosphorylation-regulated kinases) are a conserved family of protein kinases that autophosphorylate a tyrosine residue in their activation loop by an intra-molecular mechanism and phosphorylate exogenous substrates on serine/threonine residues. Little is known about the identity of true substrates for DYRK family members and their binding partners. To address this question, we used full-length dDYRK2 (Drosophila DYRK2) as bait in a yeast two-hybrid screen of a Drosophila embryo cDNA library. Of 14 independent dDYRK2 interacting clones identified, three were derived from the chromatin remodelling factor, SNR1 (Snf5-related 1), and three from the essential chromatin component, TRX (trithorax). The association of dDYRK2 with SNR1 and TRX was confirmed by co-immunoprecipitation studies. Deletion analysis showed that the C-terminus of dDYRK2 modulated the interaction with SNR1 and TRX. DYRK family member MNB (Minibrain) was also found to co-precipitate with SNR1 and TRX, associations that did not require the C-terminus of the molecule. dDYRK2 and MNB were also found to phosphorylate SNR1 at Thr102 in vitro and in vivo. This phosphorylation required the highly conserved DH-box (DYRK homology box) of dDYRK2, whereas the DH-box was not essential for phosphorylation by MNB. This is the first instance of phosphorylation of SNR1 or any of its homologues and implicates the DYRK family of kinases with a role in chromatin remodelling.

Tumor cells with KRAS or BRAF mutations or ERK5/MAPK7 amplification are not addicted to ERK5 activity for cell proliferation.

ERK5, encoded by MAPK7, has been proposed to play a role in cell proliferation, thus attracting interest as a cancer therapeutic target. While oncogenic RAS or BRAF cause sustained activation of the MEK1/2-ERK1/2 pathway, ERK5 is directly activated by MEK5. It has been proposed that RAS and RAF proteins can also promote ERK5 activation. Here we investigated the interplay between RAS-RAF-MEK-ERK and ERK5 signaling and studied the role of ERK5 in tumor cell proliferation in 2 disease-relevant cell models. We demonstrate that although an inducible form of CRAF (CRAF:ER*) can activate ERK5 in fibroblasts, the response is delayed and reflects feed-forward signaling. Additionally, oncogenic KRAS and BRAF do not activate ERK5 in epithelial cells. Although KRAS and BRAF do not couple directly to MEK5-ERK5, ERK5 signaling might still be permissive for proliferation. However, neither the selective MEK5 inhibitor BIX02189 or ERK5 siRNA inhibited proliferation of colorectal cancer cells harbouring KRAS(G12C/G13D) or BRAF(V600E). Furthermore, there was no additive or synergistic effect observed when BIX02189 was combined with the MEK1/2 inhibitor Selumetinib (AZD6244), suggesting that ERK5 was neither required for proliferation nor a driver of innate resistance to MEK1/2 inhibitors. Finally, even cancer cells with MAPK7 amplification were resistant to BIX02189 and ERK5 siRNA, showing that ERK5 amplification does not confer addiction to ERK5 for cell proliferation. Thus ERK5 signaling is unlikely to play a role in tumor cell proliferation downstream of KRAS or BRAF or in tumor cells with ERK5 amplification. These results have important implications for the role of ERK5 as an anti-cancer drug target.

dDYRK2: a novel dual-specificity tyrosine-phosphorylation-regulated kinase in Drosophila.

Dual-specificity tyrosine-phosphorylation-regulated kinases (DYRKs) are an emerging family of protein kinases that have been identified in all eukaryotic organisms examined to date. DYRK family members are involved in regulating key developmental and cellular processes such as neurogenesis, cell proliferation, cytokinesis and cellular differentiation. Two distinct subgroups exist, nuclear and cytosolic. In Drosophila, the founding family member minibrain, whose human orthologue maps to the Down syndrome critical region, belongs to the nuclear subclass and affects post-embryonic neurogenesis. In the present paper, we report the isolation of dDYRK2, a cytosolic DYRK and the putative product of the smell-impaired smi35A gene. This is the second such kinase described in Drosophila, but the first to be characterized at the molecular and biochemical level. dDYRK2 is an 81 kDa dual-specificity kinase that autophosphorylates on tyrosine and serine/threonine residues, but appears to phosphorylate exogenous substrates only on serine/threonine residues. It contains a YXY motif in the activation loop of the kinase domain in the same location as the TXY motif in mitogen-activated protein kinases. dDYRK2 is tyrosine-phosphorylated in vivo, and mutational analysis reveals that the activation loop tyrosines are phosphorylated and are essential for kinase activity. Finally, dDYRK2 is active at all stages of fly development, with elevated levels observed during embryogenesis and pupation.

DYRK1A-mediated Cyclin D1 Degradation in Neural Stem Cells Contributes to the Neurogenic Cortical Defects in Down Syndrome.

Alterations in cerebral cortex connectivity lead to intellectual disability and in Down syndrome, this is associated with a deficit in cortical neurons that arises during prenatal development. However, the pathogenic mechanisms that cause this deficit have not yet been defined. Here we show that the human DYRK1A kinase on chromosome 21 tightly regulates the nuclear levels of Cyclin D1 in embryonic cortical stem (radial glia) cells, and that a modest increase in DYRK1A protein in transgenic embryos lengthens the G1 phase in these progenitors. These alterations promote asymmetric proliferative divisions at the expense of neurogenic divisions, producing a deficit in cortical projection neurons that persists in postnatal stages. Moreover, radial glial progenitors in the Ts65Dn mouse model of Down syndrome have less Cyclin D1, and Dyrk1a is the triplicated gene that causes both early cortical neurogenic defects and decreased nuclear Cyclin D1 levels in this model. These data provide insights into the mechanisms that couple cell cycle regulation and neuron production in cortical neural stem cells, emphasizing that the deleterious effect of DYRK1A triplication in the formation of the cerebral cortex begins at the onset of neurogenesis, which is relevant to the search for early therapeutic interventions in Down syndrome.

CDK1, not ERK1/2 or ERK5, is required for mitotic phosphorylation of BIMEL.

The pro-apoptotic BH3 only protein BIM(EL) is phosphorylated by ERK1/2 and this targets it for proteasome-dependent degradation. A recent study has shown that ERK5, an ERK1/2-related MAPK, is activated during mitosis and phosphorylates BIM(EL) to promote cell survival. Here we show that treatment of cells with nocodazole or paclitaxel does cause phosphorylation of BIM(EL), which is independent of ERK1/2. However, this was not due to ERK5-catalysed phosphorylation, since it was not reversed by the MEK5 inhibitor BIX02189 and proceeded normally in ERK5-/- fibroblasts. Indeed, although ERK5 is phosphorylated at multiple sites in the C-terminal transactivation region during mitosis, these do not include the activation-loop and ERK5 kinase activity does not increase. Mitotic phosphorylation of BIM(EL) occurred at proline-directed phospho-acceptor sites and was abolished by selective inhibition of CDK1. Furthermore, cyclin B1 was able to interact with BIM and cyclin B1/CDK1 complexes could phosphorylate BIM in vitro. Finally, we show that CDK1-dependent phosphorylation of BIM(EL) drives its polyubiquitylation and proteasome-dependent degradation to protect cells during mitotic arrest. These results provide new insights into the regulation of BIM(EL) and may be relevant to the therapeutic use of agents such as paclitaxel.

Characterization of a domain that transiently converts class 2 DYRKs into intramolecular tyrosine kinases.

Dual-specificity tyrosine phosphorylation-regulated kinases (DYRKs) autophosphorylate an essential tyrosine residue in their activation loop and phosphorylate their substrates on serine and threonine residues. Phosphorylation of the activation loop tyrosine occurs intramolecularly, is mediated by a short-lived transitional intermediate during protein maturation, and is required for functional serine-threonine kinase activity of DYRKs. The DYRK family is separated into two subclasses. Through bioinformatics and mutational analyses, we identified a conserved domain in the noncatalytic N terminus of a class 2 DYRK that was required for autophosphorylation of the activation loop tyrosine but not for the phosphorylation of serine or threonine residues in substrates. We propose that this domain, which we term the NAPA domain, provides a chaperone-like function that transiently converts class 2 DYRKs into intramolecular kinases capable of autophosphorylating the activation loop tyrosine. The conservation of the NAPA domain from trypanosomes to humans indicates that this form of intramolecular phosphorylation of the activation loop is ancient and may represent a primordial mechanism for the activation of protein kinases.

Protein kinase activation loop autophosphorylation in cis: overcoming a Catch-22 situation.

Autophosphorylation of the activation loop in cis is an underappreciated and poorly understood mode of activation of protein kinases. Here, I describe four examples of protein kinases that are activated in this way, concentrating on their biochemical properties and how their autophosphorylation in cis is achieved.

A chaperone-dependent GSK3beta transitional intermediate mediates activation-loop autophosphorylation.

Glycogen synthase kinase 3 (GSK3), a key component of the insulin and wnt signaling pathways, is unusual, as it is constitutively active and is inhibited in response to upstream signals. Kinase activity is thought to be increased by intramolecular phosphorylation of a tyrosine in the activation loop (Y216 in GSK3beta), whose timing and mechanism is undefined. We show that GSK3beta autophosphorylates Y216 as a chaperone-dependent transitional intermediate possessing intramolecular tyrosine kinase activity and displaying different sensitivity to small-molecule inhibitors compared to mature GSK3beta. After autophosphorylation, mature GSK3beta is then an intermolecular serine/threonine kinase no longer requiring a chaperone. This shows that autoactivating kinases have adopted different molecular mechanisms for autophosphorylation; and for kinases such as GSK3, inhibitors that affect only the transitional intermediate would be missed in conventional drug screens.

Activation-loop autophosphorylation is mediated by a novel transitional intermediate form of DYRKs.

Autophosphorylation of a critical residue in the activation loop of several protein kinases is an essential maturation event required for full enzyme activity. However, the molecular mechanism by which this happens is unknown. We addressed this question for two dual-specificity tyrosine-phosphorylation-regulated protein kinases (DYRKs), as they autophosphorylate their activation loop on an essential tyrosine but phosphorylate their substrates on serine and threonine. Here we demonstrate that autophosphorylation of the critical activation-loop tyrosine is intramolecular and mediated by the nascent kinase passing through a transitory intermediate form. This DYRK intermediate differs in residue and substrate specificity, as well as sensitivity to small-molecule inhibitors, compared with its mature counterpart. The intermediate's characteristics are lost upon completion of translation, making the critical tyrosine autophosphorylation a "one-off" inceptive event. This mechanism is likely to be shared with other kinases.

Insulin regulation of insulin-like growth factor-binding protein-1 gene expression is dependent on the mammalian target of rapamycin, but independent of ribosomal S6 kinase activity.

Insulin inhibits the expression of the hepatic insulin-like growth factor-binding protein-1 (IGFBP-1) and glucose-6-phosphatase (G6Pase) genes. The signaling pathway that mediates these events requires the activation of phosphatidylinositol 3-kinase, whereas transfection studies have suggested an involvement of Akt (protein kinase B) and FKHR, a transcription factor regulated by Akt. We now demonstrate that insulin repression of endogenous IGFBP-1 gene transcription was blocked by rapamycin or by amino acid starvation. Rapamycin inhibited the mammalian target of rapamycin (mTOR) and the subsequent activation of p70/p85 S6 protein kinase-1 (S6K1) by insulin, whereas amino acid depletion prevented insulin induction of these signaling molecules. Importantly, we demonstrate that insulin regulation of the thymine-rich insulin response element of the IGFBP-1 promoter was also inhibited by rapamycin. However, sustained activation of S6K1 did not repress this promoter. In addition, rapamycin did not affect insulin regulation of G6Pase expression or Akt activation. We propose that these observations indicate that an mTOR-dependent, but S6K-independent mechanism regulates the suppression of IGFBP-1 (but not G6Pase) gene expression by insulin. Therefore, although the insulin-responsive sequence of the G6Pase gene promoter is related to that of the IGFBP-1 promoter, the signaling pathways that mediate suppression of these genes are distinct.