JNK (c-Jun N-terminal kinase) is a member of the MAPK (mitogen-activated protein kinase) family that regulates a range of biological processes implicated in tumorigenesis and neurodegenerative disorders. For example, genetic studies have demonstrated that the removal of specific Jnk genes can reduce neuronal death associated with cerebral ischaemia. As such, targeting JNK signalling constitutes an obvious opportunity for therapeutic intervention. However, MAPK inhibitors can display toxic effects. Consequently, dual-specificity MKKs (MAPK kinases) may represent more attractive targets. In particular, evidence that blocking JNK activation by removing MKK4 offers an effective therapy to treat pathological conditions has started to emerge. MKK4 was the first JNK activator identified. The remaining level of JNK activity in cells lacking MKK4 expression led to the discovery of a second activator of JNK, named MKK7. Distinct phenotypic abnormalities associated with the targeted deletion of Mkk4 and Mkk7 in mice have revealed that MKK4 and MKK7 have non-redundant function in vivo. Further insights into the specific functions of the JNK activators in cancer cells and in neurons will be of critical importance to validate MKK4 and MKK7 as promising drug targets.
Cells have the ability to communicate with each other and modify their behaviour according to changes in their environment. The information received at the cell surface is transmitted to different compartments within the cells via complex networks of signalling pathways. One of them, the JNK (c-Jun N-terminal kinase) cascade, also referred to as the stress-activated signalling pathway, has been the focus of many laboratories. JNK is a MAPK (mitogen-activated protein kinase) mostly implicated in mediating the apoptotic response of cells to pro-inflammatory cytokines, and genotoxic and environmental stresses. However, JNK activation is also associated with the regulation of cell proliferation, survival and differentiation . Differences in signal intensity and duration may discriminate between these seemingly antagonistic functions of JNK signalling, with evidence that transient JNK activation promotes cell survival, whereas prolonged JNK activation induces cellular apoptosis . Furthermore, the formation of specific molecular complexes via scaffold proteins may allow distinct responses by enabling different stimuli to activate a particular JNK isoform. Three Jnk genes, Jnk1, Jnk2 and Jnk3, encoding ten isoforms, have been identified . JNK1 and JNK2 appear to have redundant functions during development that are distinct from that of JNK3. Furthermore, whereas JNK3 is predominantly detected in brain, testis and heart, JNK1 and JNK2 are ubiquitously expressed. The present review describes the progress made in understanding the role of JNK signalling in diseases and how pharmacological inhibition of JNK activation, rather than JNK activity, may be a potential therapeutic strategy that avoids the unwanted side effects associated with direct targeting of JNK isoforms.
Organization of the JNK signalling cascade
As with other members of the MAPK family, JNK is activated upon dual phosphorylation on a Thr-Pro-Tyr motif located within the activation loop in protein kinase subdomain VIII by an MKK (MAPK kinase), which in turn is activated by phosphorylation on serine and threonine residues by an upstream kinase . Two JNK activators, MKK4 and MKK7, have been identified. Whereas MKK7 is a specific activator of JNK, MKK4 can also phosphorylate p38 MAPK, although, in vivo, MKK3 and MKK6 are the main activators of p38 MAPK . MKK4 and MKK7 function in a non-redundant manner, as targeted deletion of either the Mkk4 or Mkk7 genes results in early embryonic death, and have distinct affinities for JNK, with MKK4 and MKK7 preferentially phosphorylating JNK on tyrosine and threonine residues respectively . This has led to the idea that full JNK activation by MKKs occurs in a synergistic manner . Consistently, genetic analysis of MEFs (mouse embryonic fibroblasts) demonstrated that the loss of MKK4 or MKK7 reduced JNK activation in response to UV and anisomycin . However, MKK7 is essential for activating JNK in cells incubated with pro-inflammatory cytokines . The specific activation of MKK4 and MKK7 by extracellular stimuli is consistent with evidence that MKK4 and MKK7 display different affinities for scaffold proteins . For example, MKK7 binds JIP (JNK-interacting protein) 1 together with JNK and members of the MLK (mixed lineage kinase) family, whereas a spliced variant of JIP3, JSAP1, scaffolds a MEKK1 (MAPK/extracellular-signal-regulated kinase kinase 1)–MKK4–JNK module .
JNK signalling in brain development and neurological disorders
The first demonstration that JNK signalling had a pivotal role in mediating neuronal apoptosis was provided in 1995 by Xia et al.  who showed that the removal of neurotrophic factors from cultures of neuron-like PC12 cells caused a strong activation of JNK which correlated with increased apoptosis. The apoptotic effect of JNK is mediated through the direct regulation of Bcl-2 family members and cytochrome c release from the mitochondria, and activation of the AP-1 (activator protein 1) transcription factor family member c-Jun . The involvement of JNK in programmed cell death was subsequently confirmed by the analysis of knockout mouse models. Mice deficient in a single JNK isoform survived, as did Jnk1−/−Jnk3−/− or Jnk2−/−Jnk3−/− mice, whereas mice lacking Jnk1 and Jnk2 genes died during embryogenesis and displayed severe deregulation of apoptosis in the brain . However, the developmental defect of the brain associated with the absence of JNK1 and JNK2 was not observed following the loss of JNK activity caused by the targeted deletion of the Mkk4 or Mkk7 genes in neural cells. MKK4 mutant newborns were indistinguishable from their control littermates, but stopped growing a few days later and died prematurely, displaying severe neurological defects associated with the misalignment of the Purkinje cells in the cerebellum . In contrast, Mkk7brain−/− embryos died at birth due to an inability to breathe . The hypothesis that MKK4 and MKK7 regulate different subcellular pools of JNK (nuclear compared with cytoplasmic) and different JNK isoforms due to their distinct binding affinity with scaffold proteins may explain why the down-regulation of JNK signalling following the removal of MKK4 or MKK7 in the brain causes very distinct developmental defects. This is consistent with the idea that MKK4 and MKK7 have non-redundant functions in vivo.
In the adult mouse brain, JNK3, and not JNK1 or JNK2, was required for the apoptosis of hippocampal neurons induced by glutamate receptor agonist kainic acid  and the loss of neurons following ischaemic injury , clearly establishing that JNK3 is essential for mediating the apoptotic response of neurons to stress. The importance of this finding to the molecular understanding of neurodegenerative disorders was provided by evidence that the brains of patients affected by AD (Alzheimer's disease), PD (Parkinson's disease), HD (Huntington's disease) and ALS (amyotrophic lateral sclerosis) displayed significant increases in regional apoptosis [16,17]. Various transgenic and animal models of neurological conditions mirror this up-regulation of the apoptotic machinery. For example, treatment of mice with the selective nigrostriatal dopaminergic neurotoxin MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) to recreate the pathology of PD increased brain apoptosis . Likewise, the Aβ (amyloid β-peptide) induces apoptotic cell death in transgenic mice and in cultures of cortical or hippocampal neurons [19,20]. Consistent with evidence that JNK3 is required for mediating the neurotoxic effect of Aβ in vitro , we have recently discovered that Aβ-induced neuronal death is significantly reduced in cortical neurons lacking MKK4 and MKK7 expression . Furthermore, we have found that JNK signalling is required for the formation of amyloid plaques in vivo. The phosphorylation of APP (amyloid precursor protein) at Thr668 by JNK downstream of MKK4 and MKK7 promotes the amyloidogenic cleavage of APP, thereby increasing Aβ production.
JNK signalling in cancer
The differential role of JNK isoforms in disease is probably most clearly demonstrated in cancer. For example, JNK1, but not JNK2, deficiency significantly decreased HCC (hepatocelluar carcinoma) in the DEN (diethylnitrosamine)-induced HCC mouse model . Likewise, JNK2 deficiency prevented skin cancer formation induced by DMBA (7,12-dimethylbenz[a]anthracene) and PMA treatment, whereas JNK1 deficiency increased tumour incidence [24,25]. Taken together, these studies suggest that JNK1 acts as a tumour suppressor, whereas JNK2 functions as a tumour promoter. Consistently, gene array profiling of Jnk1−/− and Jnk2−/− MEFs, under both basal and PMA stimulation, demonstrated that the loss of JNK1 enhanced the expression of anti-apoptotic genes, whereas cells lacking JNK2 display increased expression of genes related to tumour suppression and induction of cell differentiation, apoptosis or cell growth . More recently, we showed that increased JNK2 expression in the epidermis treated with DMBA/PMA was prevented by the loss of MKK4 . The discovery that skin-specific MKK4-deficient mice were resistant to tumorigenesis associated with oncogenic activation of the H-ras gene provided the first genetic evidence that signalling downstream of MKK4 is essential for tumour formation in the skin . Further evidence supporting the pro-oncogenic function of MKK4 include the demonstration that the pancreatic PL5 cell line lacking MKK4 produced fewer lung metastases when injected into athymic mice  and that silencing of MKK4 in breast cancer cell lines decreased anchorage-independent growth and suppressed in vivo tumour growth . Furthermore, higher MKK4 and MKK7 expression is associated with higher-stage prostate cancer .
However, like JNK, the role of MKK4 in cancer is complex, with evidence that MKK4 acts as a tumour suppressor. This mostly relies on the analysis of clinical human biopsies. A polymorphism in the promoter of the MKK4 gene (−1304T>G) was shown to increase MKK4 mRNA and protein levels that correlated with a reduced risk of AML (acute myeloid leukaemia), lung and colorectal cancer [31–33]. In addition, a small percentage of human cancers from various origins such as the pancreas, lung, breast, colon, prostate and ovary display a functional loss of MKK4 . MKK4 also functions as a metastatic suppressor by preventing the metastatic colonization of disseminated cancer cells. Subsequently, the introduction of MKK4 into the highly metastatic AT6.1 rat prostate cancer cells or human ovarian cancer cell line SKOV3ip.1 followed by injection into immunocompromised mice prevented the proliferation of metastatic cells at secondary sites. This correlated with increased expression of the cell cycle inhibitor p21CIP1 . Interestingly, high-grade prostate cancers that lack MKK4 expression and are subsequently highly metastatic do so not by gene silencing, but through repressing the translation of MKK4 mRNA . It has been shown that MKK7, but not MKK4, was specifically required for mediating H-rasV12-induced autophagy by up-regulating Atg5 . Furthermore, genetic deletion of the Mkk7 gene in the lung promoted oncogenic K-rasG12D-induced tumour formation, with an overall mean survival time of the Mkk7−/−/K-rasG12D mice reduced from 185 to 102 days . Together, these findings are consistent with the idea that autophagic-mediated cell death is one mechanism by which oncogenic stress protects cells from malignant progression.
Small-molecule inhibitors of components of the JNK pathway
Kinases are attractive targets for drug design, and the importance of the JNK pathway in the pathology of numerous diseases has triggered extensive effort in identifying JNK inhibitors suitable for clinical application. SP600125 is one of the most extensively studied ATP-competitive JNK inhibitors . Although it lacks specificity by inhibiting all three JNK isoforms and, at higher doses, also blocks p38 signalling, a number of in vivo experiments using SP600125 have demonstrated the potential of directly inhibiting JNK for therapeutic intervention. For example, intraperitoneal injection of SP600125 was effective in the MPTP mouse model of PD . However, the ability of SP600125 to inhibit all JNK isoforms could be problematic, particularly in cancer where JNK1 and JNK2 may have opposing functions. Furthermore, the clinical efficacy of SP600125 is limited due to poor water solubility. With a similar mechanism of action to SP600125, AS601245 blocks JNK3 with a greater potency than JNK1 and JNK2. Consistent with the role of JNK3 in neuronal death, AS601245 provided significant protection against the delayed loss of hippocampal CA1 neurons in a gerbil model of transient global ischaemia and in rats after focal cerebral ischaemia . AS601245 also displayed anti-inflammatory activity in an experimental model of rheumatoid arthritis . Despite these promising findings, AS601245 did not show sufficient efficiency to undergo clinical trials for neurodegenerative diseases.
A second strategy in inhibiting JNK activity is the use of peptide inhibitors that target the substrate-binding site or regulatory protein site of the kinase. JIPs function as scaffold proteins enabling the formation of mulitprotein complexes consisting of components of the JNK signalling cascade . Endogenous JIP1 interacts with JNK and facilitates JNK activation; however, overexpression of JIP1 blocks JNK activity in mammalian cells probably by disrupting the stoichiometry of complex formation. Consequently, a cell-penetrating JNK peptide inhibitor (D-JNKI-1) was generated by fusing the amino acid sequence corresponding to the JNK-binding site of JIP1 to the HIV-TAT (transactivator of transcription) transporter sequence. Various inhibitor peptides have been generated with differing specificity for JNK isoforms. Peptides corresponding to residues 143–153 of JIP1 displayed no isoform specificity, whereas a longer peptide corresponding to residues 143–162 of JIP1 displayed some specificity for JNK3 . Further manipulation of the composition of the JIP1 peptide and the position of where the cell-permeable peptide is conjugated enabled the generation of a relatively specific JNK2-inhibitor peptide . This work is significant as JNK2, and not JNK1, has been proposed to have a dominant role in skin tumorigenesis. However, further development to improve in vivo penetration and pharmacokinetic and biodistribution parameters of the therapeutic peptides will be required before their clinical application is feasible .
Although the majority of efforts in the identification of inhibitors for the JNK pathway have focused on the JNK proteins themselves, suppression of basal JNK activity could have unfavourable effects on normal cellular homoeostasis. Therefore specific inhibitors of MKK4 and MKK7 may be a better strategy to block JNK signalling with less toxic effects. Recently the crystal structure for MKK4 bound to a p38α peptide has been solved . Solving the crystal structure for MKK7 would enable direct comparisons of how these kinases interact with substrates and perhaps enable specific MKK4 and MKK7 inhibitors to be designed. Interestingly a number of organic compounds including THIF (7,3′,4′-trihydroxyisoflavone), the major metabolite of the soy isoflavones Daidzein, and the plant pigment Cyanidin have been shown to have some specificity for MKK4 [46,47]. THIF inhibited MKK4 in both in vitro kinase assays and in cell-based experiments without affecting the activity of MKK3, MKK6, p38α or JNK1. A single topical administration of THIF to the skin of an SKH-1 hairless mouse before a 27 week course of UVB exposure, significantly reduced the number of tumours per mouse and the overall tumour volume . These results are highly encouraging and correlate with the in vivo findings that genetic deletion of MKK4 prevents skin cancer formation . Whether this compound is specific for MKK4 over MKK7 has yet to be determined. Nevertheless, it provides synthetic chemists with a structure to modify and adapt.
Non-conventional targets for regulating MKK4 and MKK7
Identifying distinct binding proteins that regulate MKK4/MKK7 kinase activity and disrupting this interaction could be a novel mechanism to target the JNK pathway for inhibition. For example, the RASSF7 protein inhibits MKK7 activity . Treatment of cells with prolonged stresses including UV, oxidative stress and the DNA-alkylating agent methyl methanesulfonate induces the ubiquitin-mediated degradation of RASSF7, increases phophorylation of JNK and promotes cellular apoptosis . Interestingly, RASSF7 is up-regulated in cancer . Therefore these findings suggests that peptides that block RASSF7–MKK7 interactions could be used to sensitize cells to JNK-mediated apoptosis in response to chemotherapeutic agents.
The early embryonic lethality caused by the deletion of components of the JNK cascade has demonstrated the critical importance of this pathway during development. However, the function of JNK signalling in disease has been very difficult to predict, particularly because of the seemingly contradictory role of JNK in promoting cell survival and proliferation on one hand and apoptosis on the other. With the help of mouse models, several JNK inhibitors have been identified that could partly or fully rescue neurons from degeneration. However, before these compounds can be developed into successful therapies, a deeper understanding of the biochemical mechanisms presiding over programmed cell death in pathological situations needs to be reached. In the context of cancer, the inconsistencies in whether MKK4/JNK signalling functions as a tumour promoter or suppressor may reflect differences in biological settings. It is therefore imperative that further studies aimed at testing the effect of conditional genetic deletion of components of the JNK cascade in well-defined mouse models of human cancer are performed. Together, these future studies will be key in deciding whether therapeutic targeting of the JNK pathway will be of clinical benefit, and, more importantly, how best to specifically modulate by small-molecule compounds a particular JNK-mediated cellular event.
Signalling 2011: a Biochemical Society Centenary Celebration: A Biochemical Society Focused Meeting held at the University of Edinburgh, U.K., 8–10 June 2011. Organized and Edited by Nicholas Brindle (Leicester, U.K.), Simon Cook (The Babraham Institute, U.K.), Jeff McIlhinney (Oxford, U.K.), Simon Morley (University of Sussex, U.K.), Sandip Patel (University College London, U.K.), Susan Pyne (University of Strathclyde, U.K.), Colin Taylor (Cambridge, U.K.), Alan Wallace (AstraZeneca, U.K.) and Stephen Yarwood (Glasgow, U.K.).
amyloid precursor protein
c-Jun N-terminal kinase
mitogen-activated protein kinase
mouse embryonic fibroblast
Our work is supported by project grants from Cancer Research UK [grant number C18267/A11727], Association for International Cancer Research [grant number 10-0134] and Alzheimer's Research UK [grant number ART-PG2011-10].