Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Review
. 2016 Jul 27;80(3):793-835.
doi: 10.1128/MMBR.00043-14. Print 2016 Sep.

JNK Signaling: Regulation and Functions Based on Complex Protein-Protein Partnerships

András Zeke  1 , Mariya Misheva  2 , Attila Reményi  3 , Marie A Bogoyevitch  4
Affiliations
Review

JNK Signaling: Regulation and Functions Based on Complex Protein-Protein Partnerships

András Zeke et al. Microbiol Mol Biol Rev. .

Abstract

The c-Jun N-terminal kinases (JNKs), as members of the mitogen-activated protein kinase (MAPK) family, mediate eukaryotic cell responses to a wide range of abiotic and biotic stress insults. JNKs also regulate important physiological processes, including neuronal functions, immunological actions, and embryonic development, via their impact on gene expression, cytoskeletal protein dynamics, and cell death/survival pathways. Although the JNK pathway has been under study for >20 years, its complexity is still perplexing, with multiple protein partners of JNKs underlying the diversity of actions. Here we review the current knowledge of JNK structure and isoforms as well as the partnerships of JNKs with a range of intracellular proteins. Many of these proteins are direct substrates of the JNKs. We analyzed almost 100 of these target proteins in detail within a framework of their classification based on their regulation by JNKs. Examples of these JNK substrates include a diverse assortment of nuclear transcription factors (Jun, ATF2, Myc, Elk1), cytoplasmic proteins involved in cytoskeleton regulation (DCX, Tau, WDR62) or vesicular transport (JIP1, JIP3), cell membrane receptors (BMPR2), and mitochondrial proteins (Mcl1, Bim). In addition, because upstream signaling components impact JNK activity, we critically assessed the involvement of signaling scaffolds and the roles of feedback mechanisms in the JNK pathway. Despite a clarification of many regulatory events in JNK-dependent signaling during the past decade, many other structural and mechanistic insights are just beginning to be revealed. These advances open new opportunities to understand the role of JNK signaling in diverse physiological and pathophysiological states.

PubMed Disclaimer

Figures

FIG 1
FIG 1
Overall organization of JNK signaling pathways. JNK pathways are activated by a variety of extracellular stimuli (e.g., cytokines, pathogens, morphogenic factors, hormones) as well as intracellular stimuli (e.g., oxidative stress, DNA damage), converging on the three JNKs. These phosphorylate a variety of cytoplasmic as well as nuclear substrates and engage in direct (e.g., phosphorylation of MAP3Ks) as well as indirect (e.g., expression of the dual-specificity phosphatases MKP1 and MKP5) feedback circuits. The protein kinase members of the core MAPK pathway are displayed in red, while critical proteins directly controlling MAP3K activation are shown in green. Proteins further upstream of the pathway are colored turquoise, MKPs are blue, and substrates are yellow. Note that, for the purposes of clarity, not all the known proteins or possible pathways are shown here. Continuous arrows imply direct binding or direct enzymatic reactions, while dotted arrows show either indirect, multistep reactions or connections where the exact mechanism(s) is uncertain. Abbreviations (other than the protein names defined in the main text): GPCR, G-protein-coupled receptor; Ubi, ubiquitin (usually nondegradative, with Lys63 linkage).
FIG 2
FIG 2
Structure and function of JIP1 acting within the JNK pathway. (A) The domain architecture of the JIP1 protein. The N-terminal regulatory “tail” is largely disordered, while the C-terminal half of JIP1 contains three folded domains as well as a kinesin light chain (KLC)-binding linear motif. The precise function of the intrinsically disordered N terminus (with its conserved acidic motifs) is unknown, yet it is highly phosphorylated by JNK in a D-motif-dependent manner. Currently, only two target sites (T103 and S421) are known to have a role in JNK-dependent physiological regulation of JIP1. This model was built by combining domain signature searches (PFAM), folding tendency predictors (IUPRED), and conservation analyses (multiple alignments among vertebrate proteins) as well as curated data from the literature. The lower line shows the results of conservation analyses (red, highly conserved sequence; blue, nonconserved sequence), when sequences of vertebrate JIP1 and the closely related JIP2 proteins are aligned with each other. Structural domains and key motifs are preserved in both proteins (including the JNK-binding D-motif), while most regulatory phosphorylation sites differ between JIP1 and JIP2. (B) A model of JIP1 actions on the microtubule-dependent transport processes in neurons. The JIP1/2 dimers (turquoise) are capable of transporting a diverse set of membrane-associated proteins (e.g., β-APP, APoE2-R) as well as certain MAP2Ks (MKK7 [red]) and inactive MAP3Ks (MLK3, DLK [magenta]). These complexes are moved along the microtubule filaments with the help of kinesin 1-kinesin light chain 1 motors (blue). At the end of their journey, the transport complexes are uncoupled by a JNK-dependent phosphorylation of JIP1. Since this step also results in the release of upstream components and activators belonging to the JNK pathway, it leads to a positive feedback loop and helps to maintain subcellular compartments with high local JNK activity. The JIP1/2 proteins uncoupled from their cargo are also transported in a reverse direction (likely through a dynein-driven process), although the structural details of the latter complex are poorly known.
FIG 3
FIG 3
Splice isoforms of the JNK1, JNK2, and JNK3 proteins. (A) The structure of JNKs. The generic structure of the JNK proteins is displayed in beige (represented by the crystal structure of JNK1β1; PDB ID 2XRW), and the variable regions (alternative splice isoforms) are highlighted in green, red, and magenta. The catalytic site of the kinase domain, ATP, is indicated in yellow. Regions that are unstructured or flexible are drawn with dotted lines. (B) All human JNK genes encode multiple splice isoforms. Apart from transcripts lacking a complete kinase domain (and therefore likely not yielding a functional protein), there are two variable regions for JNK1/2 and three for JNK3. All these alternative splicing products (as well as those resulting from alternative initiation with JNK3) combine freely and yield four isoforms for JNK1 and JNK2. For JNK3, there are 8 possible isoforms (including the longer [L] and shorter [S] N-terminal extensions), but only 3 isoforms have been characterized to date. However, mRNA sequences from databases (such as ENSEMBL) suggest that, like JNK1/2, JNK3 also contains the same alternative exons in its kinase domain. This hints at the existence of many more uncharacterized JNK3 isoforms (in blue). In the figure, the alternative segments structurally and evolutionarily corresponding to each other are labeled with the same colors: red, within the kinase domain; magenta, C-terminal flexible extension; green, N-terminal flexible extension. (C) Mechanisms of splice isoform generation in vertebrate JNK genes. The JNK3 gene has an upstream ATG codon, resulting in N-terminally extended proteins (green). However, this upstream initiation site has no Kozak consensus sequence, and so this is expected to result in “leaky scanning” by ribosomes, allowing the translational start to stochastically shift downstream to the site shared with all other JNK proteins. All vertebrate JNK genes have a duplicated exon (exon 6a [beige] and exon 6b [red]), where nonregulated splicing 6b is the preferred (major) exon. Their inclusion in the final transcript is mutually exclusive with each other because of the incompatibility of their U2- and U12-recognized splicing sites. Inclusion of the 6a exon depends on the suppression of exon 6b splicing, which can happen when the Nova2 protein binds to its polypyrimidine tract (Py) in JNK2. The ultimate splicing site is also variable, allowing for a 5-nucleotide shift. This results in a frameshift and an early stop codon in the short (p46) isoforms, while allowing the translation of the last exon in full in the case of the long (p54) isoforms. The sequences of the p46 (blue) and p54 (magenta) isoforms in the figure refer to JNK1. Note that the generic intron-exon pattern (colored to match the alternative protein sequences) shown at the top is not proportional to actual intron-exon sizes. The untranslated regions are displayed in light blue.
FIG 4
FIG 4
Structural features and substrate recognition by JNKs. (A) JNK proteins are comprised of a single protein kinase domain (structure on the left). The docking site, consisting of the negatively charged CD region (blue) and the hydrophobic docking groove (beige), plays an important role in partner binding and recruitment of substrates (red). The phosphotransfer reaction from ATP (yellow) takes place at the opposite side of the kinase, where the catalytic residues (pink) are located. Apart from the docking site, the CMGC insert (orange) is also unique to the MAPKs and a few related protein kinases. This also harbors a docking surface called the FxFP site. Although known to be functional in other MAPKs, no FxFP site-dependent substrates have been identified for JNK. The figure is based on the complex of human JNK1 with a docking motif from NFAT4 (PDB ID 2XRW). The peptide chain modeled at the catalytic site is based on the DYRK1A-substrate complex (PDB ID 2WO6; DYRKs are closely related to MAPKs in structure as well as in substrate preference). The rest of the substrate, which is not associated with JNK, is indicated with a dotted red line. Together, the CD region and the docking groove form the major docking site (D-site) of JNK proteins and play a key role in substrate recruitment (shown on the right). The best-characterized substrate proteins either contain a linear motif capable of interacting with the D-site directly (direct substrates [top]) or interact with a third protein having such a motif through heterologous interactions (indirect substrates [bottom]). (B) JNKs bind most of their known partners by engaging a dedicated recruitment site (D-site) that is distinct from their catalytic site. The same docking site is used to interact with activator kinases (MAP2Ks) responsible for the phosphorylation of the JNK activation loop, with phosphatases that dephosphorylate the same residues, as well as with other proteins involved in the regulation of pathway through intracellular compartmentalization and multiprotein complex formation (i.e., scaffolds). Many substrates also utilize the same docking site to provide access to the kinase. Therefore, most partners of JNKs directly compete with each other for binding and access to the catalytic site. Abbreviations: D, docking motif; K, kinesin-binding motif of JIP1.
FIG 5
FIG 5
The two main classes of D-motifs that interact with JNKs. Most of the known JNK-interacting D-motifs (located in diverse partners) belong to one of two distinct structural types, corresponding either to the JIP1 or to the NFAT4 consensus motifs (top). These two structural classes can be described with related, though different, consensus motifs (middle). Despite the differences, all these motifs bind to the same docking site. A large number of known JNK interactors, together with their evolutionarily closely related paralogs, harbor docking motifs showing sequence similarity to either the docking motif of JIP1 or to the docking motif of NFAT4 (bottom panels). Many of these docking motifs were characterized in in vitro experiments only, a few motifs do not satisfy the complete consensus, and some proteins (e.g., BMPR2, ATF2, ATF7, MKK7) contain more than one motif of the same or different type. (Structural panels were made by using crystal structures of JNK-peptide complexes: PDB IDs 4H39, 4H3B, and 2XRW for JNK3-pepJip1, JNK3-pepSab, and JNK1-pepNFAT4, respectively.)
FIG 6
FIG 6
Cooperation between JNK and other kinases in substrate phosphorylation. (A) Phosphorylation by JNK can serve as a priming site for GSK3 enzymes in several substrates. The latter targets the site 4 amino acids upstream of the priming site, with a preference for proteins where the upstream Ser/Thr is also followed by a small amino acid, such as Pro. The double-phosphorylated region can often act as a phosphodegron, as in the case of c-Jun, where the motif is subsequently recognized by the cullin/F-box ubiquitin ligase FBW7. (B) Sites phosphorylated by JNK can also be recognized by casein kinase 1 (CK1). These enzymes phosphorylate Ser/Thr residues 3 amino acids downstream of the original phosphorylation site, with few sequence constraints. Like most other kinases reliant on substrate priming, CK1 can also recognize sites phosphorylated by itself. In the case of NFAT4, this leads to a chain of phosphorylation events initiated by JNK. Multisite phosphorylation of this so-called SRR1 (serine-rich region 1) motif then leads to cytoplasmic anchoring of NFAT4, although its precise binding partners are unknown.
FIG 7
FIG 7
JNK-dependent phosphorylation of c-Jun elicits diverse effects. (A) The three Jun proteins (c-Jun, JunB, and JunD) were the first JNK substrates to be described and are still perhaps the best-known targets of the JNK pathways. The c-Jun protein can be phosphorylated at several sites by JNK; the most important regulatory sites are located in three different linear motifs. Phosphorylation of the two transactivator motifs located directly next to the docking motif can elicit transcriptional activation, probably due to phosphorylation-dependent recruitment of unknown effector proteins. This results in transcriptional activation of genes containing a Jun (AP-1)-binding element on their promoters, including c-Jun itself. On the other hand, phosphorylation of a much more distant phosphodegron motif in c-Jun, either by JNK alone or by cooperating with GSK3, represents an opposing regulatory mechanism. The latter provides a way for negative regulation of c-Jun levels by JNK, the mechanism for which is interestingly absent in the oncogenic v-Jun. (B) Under low or transient JNK activity, the c-Jun phospho-sites directly adjacent to the D-motif are the first to be phosphorylated, due to their favorable stereochemistry and strong coupling with the docking site. This results in a sharp rise of c-Jun mRNA and protein levels, due to autoinduction of the c-Jun gene. (C) If JNK activity is high and/or sustained over several hours, efficient phosphorylation of the distant C-terminal sites may also occur. The result will be the well-known attenuation of c-Jun levels due to ubiquitinylation and proteasome-dependent degradation of phospho–AP-1 complexes.
FIG 8
FIG 8
Negative phospho-switches and autoinhibitory switches driven by JNK. (A) A simple negative switch, illustrated by Bim. Phosphorylation of a dynein light chain (DLC)-binding motif in the BH3-only apoptosis regulator protein Bim impedes its binding to DLC-DIC (dynein intermediate chain) complexes. This shifts the equilibrium toward free Bim molecules, eventually completely releasing them into the cytoplasm. (B) A complex negative switch occurring on PPARγ. The disordered N terminus of the peroxisome proliferator-activated receptor γ contains a WW domain ligand and an overlapping PDSM (phosphorylation-dependent SUMOylation motif). Phosphorylation by JNK can therefore elicit SUMOylation of PPARγ, making its association with WW domain-containing coactivators (such as YAP1) sterically impossible. Ubc9 is a SUMO-conjugating enzyme. (C) The autoinhibitory switch of the ITCH ubiquitin ligase. As typical for the NEDD4 family of E3 ubiquitin ligases, ITCH is subject to autoinhibition. According to studies, the autoinhibitory interface of ITCH is complex, and the C2 domain, linkers, and its WW domains all contribute to maintenance of the “closed” conformation. JNK-dependent phosphorylation of certain residues located in the linker region interfere with the autoinhibitory interactions. This “opens up” the catalytic HECT domain of the ITCH enzyme, allowing the E3 ligase to recruit and ubiquitinylate its substrates.
FIG 9
FIG 9
Positive phospho-switches and allosteric switches controlled by JNK. (A) A simple positive switch acting on Elk1. Phosphorylation of the ETS transcription factor Elk1 by JNK1 (or ERK1/2) on multiple sites at its transactivation region creates a new linear motif. This protein-protein interaction motif binds to the histone acetyltransferase CBP/P300, likely through its second TAZ zinc finger domain. Thus, Elk1 can recruit chromatin-modifying enzymes that enhance transcription of its target genes. (B) A complex positive switch exemplified by phosphodegron systems. Several substrates of JNK, such as c-Myc, contain phosphodegron motifs. Multistep phosphorylation of such a motif (using GSK3) results in the recruitment of an E3 ubiquitin ligase complex containing the recognition subunit FBW7. Subsequent ubiquitinylation will then generate another new protein-protein interaction, this time with the lid of the proteasome. The result is usually the degradation of the ubiquitinylated protein, which in this sense is an inhibitory outcome (despite all protein-protein interactions being positive). (C) An allosteric switch on RXRα elicited by JNK-dependent phosphorylation of a regulatory loop. The retinoid X receptor α is an allosterically sensitive protein which only recruits the coactivator NCoA2 in its ligand (11-cis-retinoic acid [11-CRA])-bound state. The JNK-phosphorylated loop is directly adjacent to the ligand-binding pocket, and all available evidence suggests that it will bind back to the pocket when modified. The consequential distortion of the ligand-binding site would not only elicit release of 11-CRA, but also its coactivators, thus shutting down RXRα-dependent transcription.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Bogoyevitch MA, Kobe B. 2006. Uses for JNK: the many and varied substrates of the c-Jun N-terminal kinases. Microbiol Mol Biol Rev 70:1061–1095. doi:10.1128/MMBR.00025-06. - DOI - PMC - PubMed
    1. Coffey ET. 2014. Nuclear and cytosolic JNK signalling in neurons. Nat Rev Neurosci 15:285–299. doi:10.1038/nrn3729. - DOI - PubMed
    1. Bubici C, Papa S. 2014. JNK signalling in cancer: in need of new, smarter therapeutic targets. Br J Pharmacol 171:24–37. doi:10.1111/bph.12432. - DOI - PMC - PubMed
    1. Bogoyevitch MA, Arthur PG. 2008. Inhibitors of c-Jun N-terminal kinases: JuNK no more? Biochim Biophys Acta 1784:76–93. doi:10.1016/j.bbapap.2007.09.013. - DOI - PMC - PubMed
    1. Sabapathy K. 2012. Role of the JNK pathway in human diseases. Prog Mol Biol Transl Sci 106:145–169. doi:10.1016/B978-0-12-396456-4.00013-4. - DOI - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources