SAPK/JNK (stress-activated protein kinase/c-Jun N-terminal kinase) belongs to the MAPK (mitogen-activated protein kinase) family and is important in many biological contexts. JNK activation is regulated by phosphorylation of specific tyrosine and threonine residues sequentially catalysed by MKK4 and MKK7, which are both dual-specificity MAPKKs (MAPK kinases). Previously, we reported that tyrosine-phosphorylation of JNK by MKK4 precedes threonine-phosphorylation by MKK7, and that both are required for synergistic JNK activation. In the present study, we identify the actin-binding protein-280 (Filamin A) as a presumed ‘binder’ protein that can bind to MKK7, as well as to MKK4, connecting them in close proximity. We show that Filamin family members A, B and C interact with MKK4 and MKK7, but not with JNK. Filamin A binds to an N-terminal region (residues 31–60) present in the MKK7γ and MKK7β splice isoforms, but cannot bind to MKK7α which lacks these amino acids. This same N-terminal region is crucial for the intracellular co-localization of MKK7γ with actin stress fibres and Filamin A. Experiments using Filamin-A-deletion mutants revealed that the MKK7-binding region of Filamin A differs from its MKK4-binding region, and that MKK7γ (but not MKK7α) can form a complex with Filamin A and MKK4. Finally, we used Filamin-A-deficient cells to show that Filamin A enhances MKK7 activation and is important for synergistic stress-induced JNK activation in vivo. Thus Filamin A is a novel member of the group of scaffold proteins whose function is to link two MAPKKs together and promote JNK activation.

INTRODUCTION

JNK (c-Jun N-terminal kinase) is a member of the family of MAPKs (mitogen-activated protein kinases), which are ubiquitously expressed and evolutionarily conserved. JNK is activated not only by many types of external stress, including changes in osmolarity, heat shock and UV-irradiation, but also by LPA (lysophosphatidic acid) and inflammatory cytokines. Activated JNK phosphorylates the transcription factors c-Jun, Jun D and ATF-2 (activating transcription factor 2) to regulate gene expression governing stress responses. JNK signalling is also involved in the normal physiological processes of cell proliferation, apoptosis, differentiation and cell migration [1,2].

Activation of JNK requires the phosphorylation of the tyrosine and threonine residues located in a threonine-proline-tyrosine motif in the activation loop between regions VII and VIII of the kinase domain. This phosphorylation is catalysed by two dual-specificity kinases, MKK4 (SEK1) and MKK7. MKK4 and MKK7 are MAPKKs (MAPK kinases) that are activated by various MAPKKKs (MAPKK kinases), including MLKs (mixed lineage protein kinases), MEKK1 {MEK [MAPK/ERK (extracellular-signal-regulated kinase) kinase] kinase}, TAK1 [TGF (transforming growth factor)-β-activated kinase 1] and ASK1 (apoptosis signal-regulating kinase) [1,2]. In-depth studies of JNK activation have shown that MKK4 preferentially phosphorylates the tyrosine residue of the threonine-proline-tyrosine motif, whereas MKK7 preferentially phosphorylates the threonine residue. Phosphorylation of both residues in vitro results in synergistic activation of JNK [35]. We obtained strong in vivo support for this latter activation mechanism from our studies of murine ES (embryonic stem) cells bearing disruptions of the mkk4 and/or mkk7 genes [68]. No JNK activation was observed in mkk4−/−mkk7−/− ES cells [9]. In mkk7−/− ES cells, a severe impairment of JNK activation was observed that was accompanied by loss of phosphorylation of the threonine residue of JNK; however, there was no significant reduction in the phosphorylation of the tyrosine residue of JNK. In mkk4−/− ES cells, reductions in the phosphorylation of both the tyrosine and threonine residues of JNK were noted. These results indicated that tyrosine phosphorylation by MKK4, followed by threonine phosphorylation by MKK7, leads to synergistic JNK activation in stress-stimulated ES cells. However, the molecular mechanism underlying the sequential phosphorylation of JNK by MKK4 and MKK7 remains to be elucidated.

Recent studies have shown that scaffold proteins mediate the structural and functional organization of a three-tier MAPK activation module which involves a MAPKKK, a MAPKK and a MAPK [10]. These MAPK-specific scaffold proteins link these kinases into a multienzyme complex and provide an insulated physical conduit through which signals from a MAPK can be transmitted to the appropriate spatiotemporal cellular loci. The scaffold proteins then modulate the signalling strength of their cognate MAPK module by regulating the amplitude and duration of signalling. Several scaffold proteins involved in mammalian JNK signalling modules have been identified, including JIP (JNK-interacting protein) 1, JIP2, JSAP1 [JNK/SAPK (stress-activated protein kinase)-associated protein 1]/JIP3, JLP (JNK-associated leucine-zipper protein) and POSH [plenty of SH3s (Src homology 3)] and their various splice variants. JIP1, JIP2 and JSAP1 bind to JNK, MKK7 and various MLKs. JSAP1 associates with JNK, MKK4 and MEKK1, whereas JLP links Max with c-Myc, and JNK with p38, MKK4 or MEKK3. In addition, multiple upstream MAPKKKs can act as scaffold proteins, as well as exert their intrinsic kinase activities. For example, MEKK1 binds to and regulates MKK4. Despite this flexibility, theoretical considerations have dictated that a single JIP-based MAPK module containing MKK4 and MKK7 physically cannot catalyse the sequential phosphorylation of JNK by these kinases [11]. We therefore speculated that additional scaffold molecules must exist that can bind to and connect two sets of complexes, one containing MKK4 and the other containing MKK7. To identify these presumed ‘binder’ molecules, we used MKK7 and the yeast two-hybrid method to screen a human leucocyte cDNA library. We isolated Filamin A, which has been previously reported to interact with MKK4 [12], as a predicted ‘binder’ protein that can also interact with MKK7. Our results show that different MKK7 splice isoforms [13] have different scaffold-binding properties, and that Filamin A plays an important role in synergistic JNK activation.

EXPERIMENTAL

Cell culture

HEK (human embryonic kidney)-293T cells, HeLa cells and non-transfected (M2) or stably transfected (A7) human melanoma cell lines were maintained in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% FCS (fetal calf serum), 0.16% NaHCO3 and 0.6 mg·ml−1L-glutamine. To maintain Filamin A expression, A7 cells were cultured in the presence of 0.5 mg/ml G418 (Sigma).

Antibodies and GFP (green fluorescent protein) vector

Abs (antibodies) against human Filamin A (MAB1680) were from Chemicon. Abs against SAPK/JNK1 (C-17 and FL) or phospho-SAPK/JNK (#9251) were from Santa Cruz Biotechnology and Cell Signaling Technology respectively. Anti-FLAG (M2) and anti-c-Myc (9E10) Abs were from Sigma–Aldrich. Anti-HA (haemagglutinin) high-affinity (3F10) Ab was from Roche Diagnostics. The rat anti-MKK7 (KN-004) mAb (monoclonal Ab) used for immunoprecipitation and immunoblotting was prepared in our laboratory as previously described [7]. Abs against MKK4 (sc-837) or phospho-MKK4 (#9151) were from Santa Cruz Biotechnology and Cell Signaling Technology respectively. The pEGFP-C1 vector was from BD Biosciences.

Construction of plasmids

cDNAs encoding FLAG-tagged versions of MKK4, MKK7α1, MKK7β1, MKK7γ1, MKK7γ2, JNK1 and full-length Filamin A were cloned into the mammalian expression vector pCMV5. Plasmids expressing Myc-tagged full-length Filamin A, the Myc-tagged CT (C-terminus) of Filamin A (residues 2282–2647), or Myc–Filamin B, Myc–Filamin C or HA-tagged MKK4 were also cloned into pCMV5. Myc-tagged Filamin A deletion mutants A, B and C were generated using PCR and subcloned into pCMV5. FLAG-tagged MKK7 deletion mutants A, B and C were constructed using a One Day Mutagenesis kit (QuikChange® kit from Stratagene) and cloned into pCMV5.

Transfection

For gene expression analysis, HEK-293T cells were plated at 6×106 cells in a 100-mm dish, or at 2×106 cells in a 60-mm dish. Cells were transfected 1 day later with 5 μg (100-mm dish) or 1.8 μg (60-mm dish) of plasmid DNA using 5 μl of Lipofectamine™ 2000 (Invitrogen). M2 cells were plated at 1.25×106 cells in a 60-mm dish and transfected 1 day later with 2.5 μg of plasmid DNA as above. After 48 h in culture, cell extracts were prepared and subjected to immunoprecipitation as described below.

Immunoprecipitation and immunoblotting

Transfected HEK-293T cells were resuspended in lysis buffer A [20 mM Hepes/KOH (pH 7.4), 40 mM NaCl, 1% Triton X-100, 1 mM dithiothreitol, 1 mM EDTA, 2 mM EGTA and 0.1 mM PMSF] at 4 °C. Cell lysates were incubated with anti-FLAG M2 agarose (Sigma) at 4 °C for 2 h. The immunocomplexes were washed several times with lysis buffer A and eluted with 200 μg/ml FLAG peptides. The eluted samples were fractionated by SDS/PAGE and proteins were electrophoretically transferred on to a PVDF membrane (Bio-Rad). Membranes were immunoblotted with anti-FLAG, anti-Myc, anti-HA or anti-(Filamin A) Abs. Bands were visualized using SuperSignal West Pico Chemiluminescent Substrate for the development of immunoblots and a HRP (horseradish peroxidase)-conjugated secondary Ab, according to the manufacturer's instructions (Pierce).

Assay of JNK and MKK7 activities

Confluent melanoma cells were treated at 37 °C for 20 min with sorbitol (50, 100, 150, 200, 300 or 500 mM). Treated cells were washed and resuspended in lysis buffer B [100 mM NaCl, 40 mM Tris/HCl (pH 8.0), 1% Nonidet P40, 0.05% 2-mercaptoethanol, 1 mM EDTA, 1 mM EGTA and 4 μg/ml aprotinin]. Extracts were incubated for 2 h at 4 °C with anti-JNK pAb (polyclonal antibody; C-17, Santa Cruz Biotechnology) to immunoprecipitate JNK proteins. Endogenous MKK7 proteins were immunoprecipitated in a similar manner with anti-MKK7 (KN-004) mAb. Immunocomplexes were washed three times with lysis buffer B and three times with kinase reaction buffer [10 mM MgCl2, 50 mM Tris/HCl (pH 7.5) and 1 mM EGTA]. JNK activity in immunoprecipitates was measured using a previously described in vitro kinase assay [6,7] employing 60 μM [γ-32P]ATP and GST (glutathione transferase)–c-Jun as the substrate. MKK7 activity on beads was measured using an in vitro MAPK kinase assay [6,7] employing 100 μM unlabelled ATP with GST–JNK1 as the substrate; the product was detected using an anti-phospho-JNK Ab. For both in vitro kinase assays, the reactions were terminated after 30 min at 30 °C by the addition of 10 μl of 4×SDS gel sample buffer.

To assay JNK activity in transfected M2 cells, cells were resuspended in lysis buffer C [150 mM NaCl, 40 mM Hepes (pH 7.4), 1% Nonidet P40, 0.05% 2-mercaptoethanol, 1 mM EDTA, 1 mM EGTA, 10 mM MgCl2 and 4 μg/ml aprotinin]. The resulting cell lysates were analysed by SDS/PAGE and immunoblotting in which JNK activation was detected using an anti-phospho-JNK Ab.

Yeast two-hybrid assay

A human brain cDNA library (Clontech) and the MatchMaker™ GAL4 Two-Hybrid System 3 (Clontech) were used, as described previously [14], for a yeast two-hybrid assay in which MKK7 served as the bait.

Confocal microscopy

HeLa cells transiently expressing GFP–MKK7γ1 were cultured on a polylysine-coated glass coverslip (15-mm diameter) and washed three times with PBS before fixation with 4% (w/v) paraformaldehyde in PBS for 15 min at 4 °C. After treatment with 0.1 mM glycine in PBS for 15 min, the cells were permeabilized with 0.1% Triton X-100 in blocking solution (3% BSA in PBS) before incubation for 1 h at room temperature (25 °C) with rhodamine-conjugated phalloidin diluted in blocking solution. After three PBS washes, the coverslip was mounted on to a glass slide in Permafluor-mounting medium (Immunon) and viewed using a confocal microscope (Carl Zeiss) with LSM510 software. Excitation wavelengths of 488 nm or 546 nm were used. The images were merged using Photoshop (Adobe Systems).

For overexpression of GFP-tagged Filamin A, GFP–MKK7α1, GFP–MKK7γ1 or GFP–MKK7 deletion B, and for co-expression of GFP-tagged MKK7γ1 and a fusion protein composed of red fluorescent protein (DsRed) plus Filamin A, transfected HeLa cells were cultured for 48 h in a glass-based dish (35-mm diameter; Iwaki) and examined by confocal microscopy as described previously [14].

RESULTS

Interaction of Filamin A with MKK4 and MKK7

To identify proteins that interacted with both MKK4 and MKK7, a human leucocyte cDNA library was screened with MKK7 in a yeast two-hybrid system. Among various MKK7-binding cDNAs isolated were several encoding the CT of ABP-280 (actin-binding protein-280, Filamin A), which was previously reported to bind to MKK4 [12]. To determine whether endogenous full-length Filamin A could interact with both MKK4 and MKK7 in vivo, we transiently overexpressed FLAG–MKK4, FLAG–MKK7γ2 and FLAG–JNK1 in HEK-293T cells and carried out co-immunoprecipitation experiments. As shown in Figure 1(A), a protein of ~280 kDa co-immunoprecipitated with FLAG–MKK4 and FLAG–MKK7γ2, but not with FLAG–JNK1. We confirmed that this 280 kDa protein was indeed Filamin A using Western blotting employing anti-(Filamin A) Ab (Figure 1B). We next used FLAG and Myc tagging to examine whether two other members of the Filamin protein family, Filamin B and Filamin C, could bind to MKK4 and MKK7 in vivo. When overexpressed in HEK-293T cells, both Filamin B and Filamin C co-immunoprecipitated with both MKK4 and MKK7γ2 (Figure 1C). These results show that Filamin family proteins can interact with at least two MAPKKs, MKK4 and MKK7γ2, and demonstrate a novel connection between an MKK7 enzyme and a Filamin protein known to interact with MKK4.

Interaction of Filamin A with MKK4 and MKK7

Figure 1
Interaction of Filamin A with MKK4 and MKK7

(A and B) HEK-293T cells were transfected with 5 μg of pCMV5/FLAG (lanes 1), FLAG–MKK4 (lanes 2), FLAG–MKK7γ2 (lanes 3) or FLAG–JNK1 (lanes 4). Cell lysates were prepared (B, panels c and d) and immunoprecipitated (IP) with anti-FLAG M2 agarose (A and B, panels a and b). (A) Co-immunoprecipitated proteins (*) were visualized using Coomassie Blue. The molecular mass in kDa is indicated on the left-hand side. (B) Co-immunoprecipitated Filamin A and immunoprecipitated FLAG–MKK4, FLAG–MKK7γ2 and FLAG–JNK1 were identified using anti-(Filamin A) (panel a) and anti-FLAG M2 (panel b) Abs respectively. Expression of Filamin A and FLAG–MKK4, FLAG–MKK7γ2 and FLAG–JNK1 were determined by immunoblotting (IB) using anti-(Filamin A) (panel c) and anti-FLAG M2 (panel d) Abs respectively. (C) HEK-293T cells were transfected with 0.9 μg of pCMV5/Myc-Filamin A (CT) (lanes 1, 4 and 7), Myc–Filamin B (lanes 2, 5 and 8) or Myc–Filamin C (lanes 3, 6 and 9), together with 0.9 μg of pCMV5/FLAG (lanes 1–3), FLAG–MKK4 (lanes 4–6) or FLAG–MKK7γ2 (lanes 7–9). Cell lysates were prepared (panels c and d) and immunoprecipitated (IP) with anti-FLAG M2 agarose (panels a and b). Co-immunoprecipitated Myc–Filamin proteins and immunoprecipitated FLAG–MKK4 and FLAG–MKK7γ2 were determined using anti-c-Myc (panel a) and anti-FLAG M2 (panel b) Abs respectively. Expression of Myc–Filamin proteins and FLAG–MKK4 and FLAG–MKK7γ2 were determined using anti-c-Myc (panel c) and anti-FLAG M2 (panel d) Abs respectively. IB, immunoblot.

Figure 1
Interaction of Filamin A with MKK4 and MKK7

(A and B) HEK-293T cells were transfected with 5 μg of pCMV5/FLAG (lanes 1), FLAG–MKK4 (lanes 2), FLAG–MKK7γ2 (lanes 3) or FLAG–JNK1 (lanes 4). Cell lysates were prepared (B, panels c and d) and immunoprecipitated (IP) with anti-FLAG M2 agarose (A and B, panels a and b). (A) Co-immunoprecipitated proteins (*) were visualized using Coomassie Blue. The molecular mass in kDa is indicated on the left-hand side. (B) Co-immunoprecipitated Filamin A and immunoprecipitated FLAG–MKK4, FLAG–MKK7γ2 and FLAG–JNK1 were identified using anti-(Filamin A) (panel a) and anti-FLAG M2 (panel b) Abs respectively. Expression of Filamin A and FLAG–MKK4, FLAG–MKK7γ2 and FLAG–JNK1 were determined by immunoblotting (IB) using anti-(Filamin A) (panel c) and anti-FLAG M2 (panel d) Abs respectively. (C) HEK-293T cells were transfected with 0.9 μg of pCMV5/Myc-Filamin A (CT) (lanes 1, 4 and 7), Myc–Filamin B (lanes 2, 5 and 8) or Myc–Filamin C (lanes 3, 6 and 9), together with 0.9 μg of pCMV5/FLAG (lanes 1–3), FLAG–MKK4 (lanes 4–6) or FLAG–MKK7γ2 (lanes 7–9). Cell lysates were prepared (panels c and d) and immunoprecipitated (IP) with anti-FLAG M2 agarose (panels a and b). Co-immunoprecipitated Myc–Filamin proteins and immunoprecipitated FLAG–MKK4 and FLAG–MKK7γ2 were determined using anti-c-Myc (panel a) and anti-FLAG M2 (panel b) Abs respectively. Expression of Myc–Filamin proteins and FLAG–MKK4 and FLAG–MKK7γ2 were determined using anti-c-Myc (panel c) and anti-FLAG M2 (panel d) Abs respectively. IB, immunoblot.

Isoform-specific interaction of MKK7 with Filamin A

To date, six isoforms of MKK7 have been identified [12]. To test whether four of these MKK7 isoforms (Figure 2A) interacted differentially with Filamin A, HEK-293T cells were transiently co-transfected with individual FLAG–MKK7 isoforms plus Myc-tagged Filamin A (CT). As shown in Figure 2(B), Myc–Filamin A (CT) co-immunoprecipitated with FLAG–MKK7γ2, FLAG–MKK7γ1 and FLAG–MKK7β1, but not with FLAG–MKK7α1. We next constructed three deletion mutants of MKK7γ1 (MKK7γ1 deletion A, B and C; Figure 2A) that affected the presumed Filamin-A-binding site in MKK7 and repeated the co-immunoprecipitation experiments. Myc–Filamin A (CT) co-immunoprecipitated with FLAG–MKK7γ1 deletion A and deletion C, but not with FLAG–MKK7γ1 deletion B or FLAG–MKK7α1 (Figure 2C). Thus a short N-terminal region (encompassing amino acids 31–60) that is present in MKK7γ and MKK7β, but absent in MKK7α1, is required for MKK7 binding to Filamin A.

Isoform-specific interaction of MKK7 with Filamin A

Figure 2
Isoform-specific interaction of MKK7 with Filamin A

(A) Schematic diagram of MKK7 isoforms and deletion mutants (A, B and C). The ability of each protein to interact with Filamin A is indicated. a.a., amino acid. (B) HEK-293T cells were co-transfected with 0.9 μg of pCMV5/FLAG (lane 1), or FLAG–MKK7 isoforms γ2 (lane 2), γ1 (lane 3), β1 (lane 4) or α1 (lane 5), together with 0.9 μg of pCMV5/Myc-Filamin A (CT). (C) HEK-293T cells were transfected with 0.9 μg of pCMV5/FLAG (lane 1), or FLAG–MKK7 isoforms α1 (lane 2), γ1 (lane 3), γ1 deletion A (lane 4), γ1 deletion B (lane 5) or γ1 deletion C (lane 6), together with 0.9 μg of pCMV5/Myc-Filamin A (CT). Co-immunoprecipitated Myc–Filamin A (CT) and immunoprecipitated FLAG–MKK7s were identified using anti-c-Myc (panel a) and anti-FLAG M2 (panel b) Abs respectively. Expression of Myc–Filamin A (CT) was determined using anti-c-Myc (panel c). IB, immunoblot; IP, immunoprecipitation.

Figure 2
Isoform-specific interaction of MKK7 with Filamin A

(A) Schematic diagram of MKK7 isoforms and deletion mutants (A, B and C). The ability of each protein to interact with Filamin A is indicated. a.a., amino acid. (B) HEK-293T cells were co-transfected with 0.9 μg of pCMV5/FLAG (lane 1), or FLAG–MKK7 isoforms γ2 (lane 2), γ1 (lane 3), β1 (lane 4) or α1 (lane 5), together with 0.9 μg of pCMV5/Myc-Filamin A (CT). (C) HEK-293T cells were transfected with 0.9 μg of pCMV5/FLAG (lane 1), or FLAG–MKK7 isoforms α1 (lane 2), γ1 (lane 3), γ1 deletion A (lane 4), γ1 deletion B (lane 5) or γ1 deletion C (lane 6), together with 0.9 μg of pCMV5/Myc-Filamin A (CT). Co-immunoprecipitated Myc–Filamin A (CT) and immunoprecipitated FLAG–MKK7s were identified using anti-c-Myc (panel a) and anti-FLAG M2 (panel b) Abs respectively. Expression of Myc–Filamin A (CT) was determined using anti-c-Myc (panel c). IB, immunoblot; IP, immunoprecipitation.

Co-localization of MKK7 with Filamin A

To investigate whether MKK7 co-localizes with Filamin A in cells, we first examined the subcellular localization of MKK7γ1 and found that this isoform associated with fibre-like structures. To determine whether these fibres were actin stress fibres, HeLa cells were transiently transfected with GFP-tagged MKK7γ1, fixed, stained with phalloidin to visualize actin distribution, and examined by confocal microscopy. We observed that MKK7γ1 did indeed co-localize with actin stress fibres (Figure 3A). Because Filamin A is known to cross-link actin filaments, these results suggested that MKK7γ1 co-localized with Filamin A. We tested this possibility by transfecting HeLa cells with GFP–MKK7γ1 and DsRed–Filamin A. As expected, MKK7γ1 co-localized with Filamin A in selected areas (Figure 3B). To investigate whether the apparent association of MKK7γ1 with actin stress fibres was due to the interaction of MKK7γ1 with Filamin A, HeLa cells were co-transfected with GFP–Filamin A and GFP–MKK7γ1, GFP–MKK7γ1 deletion B or GFP–phospho-MKK7α1. Confocal microscopy showed that GFP–MKK7γ1 and GFP–Filamin A co-localized on fibre-like structures, whereas GFP–MKK7α1 and GFP–MKK7γ1 deletion B (which do not interact with Filamin A) were diffusely distributed throughout the cytoplasm (Figure 3C). Thus MKK7γ1 and Filamin A co-localize in cells, and MKK7γ1 associates with actin stress fibres due to its interaction with Filamin A.

Co-localization of MKK7γ isoforms with Filamin A

Figure 3
Co-localization of MKK7γ isoforms with Filamin A

(A) HeLa cells were transiently transfected with 1.2 μg of GFP–MKK7γ1 and cultured for 48 h. Actin was identified by rhodamine–phalloidin staining, and individual and merged images (yellow) were visualized using confocal scanning microscopy. (B) HeLa cells were transiently co-transfected with 0.6 μg of GFP–MKK7γ1 plus 0.6 μg of DsRed–Filamin A. Images were visualized as for (A). (C) HeLa cells were transiently transfected with 1.2 μg of GFP–MKK7γ1 (panel a), GFP–Filamin A (panel b), GFP–MKK7α1 (panel c) or GFP–MKK7γ1 deletion B (panel d), and cultured for 48 h prior to examination by confocal microscopy.

Figure 3
Co-localization of MKK7γ isoforms with Filamin A

(A) HeLa cells were transiently transfected with 1.2 μg of GFP–MKK7γ1 and cultured for 48 h. Actin was identified by rhodamine–phalloidin staining, and individual and merged images (yellow) were visualized using confocal scanning microscopy. (B) HeLa cells were transiently co-transfected with 0.6 μg of GFP–MKK7γ1 plus 0.6 μg of DsRed–Filamin A. Images were visualized as for (A). (C) HeLa cells were transiently transfected with 1.2 μg of GFP–MKK7γ1 (panel a), GFP–Filamin A (panel b), GFP–MKK7α1 (panel c) or GFP–MKK7γ1 deletion B (panel d), and cultured for 48 h prior to examination by confocal microscopy.

Filamin A mediates a connection between MKK4 and MKK7γ

The above experiments showed that Filamin A interacts not only with MKK4, but also with MKK7γ> and MKKβ isoforms. To identify the regions of Filamin A that can interact with MKK4 or MKK7γ, HEK-293T cells were transiently co-transfected with either Myc–Filamin A (CT), including its MKK4-binding region (amino acids 2282–2454), or a series of Filamin A deletion mutant proteins lacking different portions of the MKK4-binding region (Figure 4A), together with FLAG–MKK4 or FLAG–MKK7γ2. Co-immunoprecipitation assays showed that amino acids 2297–2311 of Filamin A were required for its interaction with MKK7γ2 (Figure 4B). Thus the region of Filamin A required for binding to MKK7γ2 is distinct from that needed for the interaction with MKK4. We then investigated whether Filamin A could interact simultaneously with MKK4 and MKK7γ2 by transiently co-transfecting HEK-293T cells with HA-tagged MKK4 plus FLAG–MKK7γ2 or FLAG–MKK7α1. HA–MKK4 and endogenous Filamin A co-immunoprecipitated with FLAG–MKK7γ2, but not with FLAG–MKK7α1 (Figure 4C). Because this experiment did not exclude the possibility that MKK4 might interact directly with MKK7γ2 (rather than with Filamin A), we repeated these co-immunoprecipitation experiments using a human melanoma cell line M2 that has spontaneously lost expression of Filamin A [15]. M2 cells were transiently co-transfected with HA–MKK4 and FLAG–MKK7γ2, together with increasing amounts of Myc–Filamin A (CT). Interaction of MKK4 with MKK7 was enhanced when higher amounts of Filamin A (CT) were present (Figure 4D), suggesting that Filamin A, MKK4 and MKK7γ2 form a complex in which Filamin A connects MKK4 and MKK7γ2.

Filamin-A-mediated connection of MKK4 and MKK7γ

Figure 4
Filamin-A-mediated connection of MKK4 and MKK7γ

(A and B) Deletion analysis identifying the MKK7-binding region of Filamin A. (A) Schematic diagram of Filamin A proteins, including full-length, CT and deletion mutants A, B and C. Relevant domains are indicated. (B) HEK-293T cells were co-transfected with 0.9 μg of pCMV5/Myc-Filamin A (CT) (lanes 1, 2, and 6), or Myc–Filamin A deletion A (lanes 3 and 7), Myc–Filamin A deletion B (lanes 4 and 8) or Myc–Filamin A deletion C (lanes 5 and 9), together with 0.9 μg of pCMV5/FLAG (lane 1), FLAG–MKK7γ2 (lanes 2–5) or FLAG–MKK4 (lanes 6–9). Co-immunoprecipitated Myc–Filamin A proteins and immunoprecipitated FLAG–MKK7γ2 or FLAG–MKK4 were identified using anti-c-Myc (panel a) and anti-FLAG M2 (panel b) Abs respectively. Expression of Myc–Filamin A proteins and FLAG–MKK7γ2 or FLAG–MKK4 were determined using anti-c-Myc (panel c) and anti-FLAG M2 (panel d) Abs respectively. (C) HEK-293T cells were transfected with 2 μg of pCMV5/HA-tagged MKK4, together with 3 μg of pCMV5/FLAG (lane 1), pCMV5/FLAG-MKK7γ2 (lane 2) or pCMV5/FLAG-MKK7α1 (lane 3). Co-immunoprecipitated HA–MKK4, endogenous Filamin A and immunoprecipitated FLAG–MKK7 were determined using anti-HA (panel a), anti-Filamin A (panel b) and anti-FLAG M2 (panel c) Abs respectively. Expression levels of HA–MKK4, endogenous Filamin A and FLAG–MKK7 proteins were determined using anti-HA (panel d), anti-Filamin A (panel e) and anti-FLAG M2 (panel f) Abs respectively. (D) M2 human melanoma cells were co-transfected with various amounts of pCMV5/Myc-Filamin A (CT) (0, 0, 1, 2 and 2 μg in lanes 1–5 respectively), plus 3 μg of pCMV5/FLAG (lane 1), pCMV5/FLAG-MKK7γ2 (lanes 2–4) or pCMV5/FLAG-MKK7α1 (lane 5), together with 2 μg of pCMV5/HA-tagged MKK4. Co-immunoprecipitated HA–MKK4, Myc–Filamin A (CT) and immunoprecipitated FLAG–MKK7 proteins were determined using anti-HA (panel a), anti-c-Myc (panel b) and anti-FLAG M2 (panel c) Abs respectively. Expression of HA–MKK4, Myc–Filamin A and FLAG–MKK7 proteins was determined using anti-HA (panel d), anti-c-Myc (panel e) and anti-FLAG M2 (panel f) Abs respectively. IB, immunoblot; IP, immunoprecipitation.

Figure 4
Filamin-A-mediated connection of MKK4 and MKK7γ

(A and B) Deletion analysis identifying the MKK7-binding region of Filamin A. (A) Schematic diagram of Filamin A proteins, including full-length, CT and deletion mutants A, B and C. Relevant domains are indicated. (B) HEK-293T cells were co-transfected with 0.9 μg of pCMV5/Myc-Filamin A (CT) (lanes 1, 2, and 6), or Myc–Filamin A deletion A (lanes 3 and 7), Myc–Filamin A deletion B (lanes 4 and 8) or Myc–Filamin A deletion C (lanes 5 and 9), together with 0.9 μg of pCMV5/FLAG (lane 1), FLAG–MKK7γ2 (lanes 2–5) or FLAG–MKK4 (lanes 6–9). Co-immunoprecipitated Myc–Filamin A proteins and immunoprecipitated FLAG–MKK7γ2 or FLAG–MKK4 were identified using anti-c-Myc (panel a) and anti-FLAG M2 (panel b) Abs respectively. Expression of Myc–Filamin A proteins and FLAG–MKK7γ2 or FLAG–MKK4 were determined using anti-c-Myc (panel c) and anti-FLAG M2 (panel d) Abs respectively. (C) HEK-293T cells were transfected with 2 μg of pCMV5/HA-tagged MKK4, together with 3 μg of pCMV5/FLAG (lane 1), pCMV5/FLAG-MKK7γ2 (lane 2) or pCMV5/FLAG-MKK7α1 (lane 3). Co-immunoprecipitated HA–MKK4, endogenous Filamin A and immunoprecipitated FLAG–MKK7 were determined using anti-HA (panel a), anti-Filamin A (panel b) and anti-FLAG M2 (panel c) Abs respectively. Expression levels of HA–MKK4, endogenous Filamin A and FLAG–MKK7 proteins were determined using anti-HA (panel d), anti-Filamin A (panel e) and anti-FLAG M2 (panel f) Abs respectively. (D) M2 human melanoma cells were co-transfected with various amounts of pCMV5/Myc-Filamin A (CT) (0, 0, 1, 2 and 2 μg in lanes 1–5 respectively), plus 3 μg of pCMV5/FLAG (lane 1), pCMV5/FLAG-MKK7γ2 (lanes 2–4) or pCMV5/FLAG-MKK7α1 (lane 5), together with 2 μg of pCMV5/HA-tagged MKK4. Co-immunoprecipitated HA–MKK4, Myc–Filamin A (CT) and immunoprecipitated FLAG–MKK7 proteins were determined using anti-HA (panel a), anti-c-Myc (panel b) and anti-FLAG M2 (panel c) Abs respectively. Expression of HA–MKK4, Myc–Filamin A and FLAG–MKK7 proteins was determined using anti-HA (panel d), anti-c-Myc (panel e) and anti-FLAG M2 (panel f) Abs respectively. IB, immunoblot; IP, immunoprecipitation.

Filamin A enhances the activation of MKK7 and JNK

The above results suggested that Filamin A might be one of the ‘binder’ proteins predicted to closely connect two MAPKKs within a MAPK module in living cells [11]. To examine the effect of Filamin A on JNK activation in vivo, we compared JNK activation in M2 cells (no Filamin A expression) with that in A7 cells (M2 cells stably transfected with a vector expressing full-length human Filamin A; [16]). In response to increasing concentrations of sorbitol, JNK was activated much more strongly in A7 cells than in M2 cells (Figure 5A). When we analysed the activation state of MKK7 in these sorbitol-treated M2 and A7 cells, we found that 500 mM sorbitol induced a 3.3-fold activation of MKK7 in A7 cells, but only a 1.4-fold activation of MKK7 in M2 cells (Figure 5B). The activation of MKK4 is also higher in A7 cells than in M2 cells (Figure 5C). Thus, in living cells, Filamin A promotes optimal stress-induced JNK activation by enhancing MKK7 and MKK4 activation.

Impaired synergistic JNK activation in Filamin-A-deficient cells

Figure 5
Impaired synergistic JNK activation in Filamin-A-deficient cells

(A) Inset: confluent M2 and A7 human melanoma cells were treated with sorbitol (0.05, 0.1, 0.15, 0.2 or 0.3 M in lanes 1–6 respectively) at 37 °C for 20 min. Lysates were immunoprecipitated with anti-JNK Ab, and JNK activity in the immunoprecipitates was measured using an in vitro kinase assay with GST–c-Jun as the substrate. 32P-labelled phosphorylated GST–c-Jun (panels a and c) and total immunoprecipitated JNK (panels b and d) are shown. Graph: levels of JNK activity in treated M2 (Δ) and A7 (●) cells were expressed as the fold-stimulation compared with levels in untreated M2 or A7 cells. (B) Confluent M2 and A7 cells were treated with 0.5 M sorbitol at 37 °C for 20 min, and lysates were immunoprecipitated with an anti-MKK7 mAb. MKK7 activity in the immunoprecipitates was measured using an in vitro kinase assay with GST–JNK1 as the substrate. Levels of phosphorylated JNK were determined using an anti-phospho-JNK Ab (upper panel). Total immunoprecipitated MKK7 was determined using an anti-MKK7 Ab (lower panel). Histogram: MKK7 activity in lysates of M2 cells and A7 cells that were not treated (NT) or treated with 0.5 M sorbitol (Sor) were expressed relative to activities of NT in M2 cells. Bars indicate the means±S.E.M. for three independent experiments. (C) MKK4 activity was measured using an anti-phospho-MKK4 (upper panel) Ab. Total MKK4 was determined using an anti-MKK4 Ab (lower panel). Histogram: MKK4 activity in lysates of M2 cells and A7 cells that were not treated (NT) or treated with 0.5 M sorbitol (Sor) were expressed relative to activities of NT in M2 cells as described in (B). (D) M2 cells were co-transfected with various amounts of pCMV5/FLAG-Filamin A (0, 1, 0, 0.5 and 1 μg in lanes 1–5 respectively) and 0.5 μg of pCMV5/HA-JNK1, together with (lanes 3–5) or without (lanes 1 and 2) 0.5 μg of pCMV5/FLAG-MKK4 and FLAG-MKK7γ2. M2 cells were co-transfected with various amounts of pCMV5/FLAG-Filamin A (0, 0.5 and 1 μg in lanes 8–10 and 11–13 respectively) and 0.5 μg of pCMV5/HA-JNK1, with 0.5 μg of pCMV5/FLAG-MKK4 (lanes 8–10) or 0.5 μg of pCMV5/FLAG-MKK7γ2 (lanes 11–13). After 48 h culture, cell lysates were analysed for JNK phosphorylation (activation) by immunoblotting with anti-phospho-JNK Ab. Total HA–JNK1, FLAG–Filamin A, FLAG–MKK4 and FLAG–MKK7 expressions were determined using anti-HA and anti-FLAG Abs. Histogram: densitometry of the bands in the upper panel, expressed as the fold increase over phosphorylation in the absence of MKK4, MKK7 proteins and Filamin A. After 48 h culture, cell lysates were analysed for JNK phosphorylation (activation) by immunoblotting with anti-phospho-JNK Ab. Total HA–JNK1 expression was determined using anti-HA Abs. Histogram: densitometry of the bands in the upper panel, expressed as the fold increase over phosphorylation in the absence of MKK4, MKK7 proteins and Filamin A. Results were obtained from at least three independent experiments. IB, immunoblot.

Figure 5
Impaired synergistic JNK activation in Filamin-A-deficient cells

(A) Inset: confluent M2 and A7 human melanoma cells were treated with sorbitol (0.05, 0.1, 0.15, 0.2 or 0.3 M in lanes 1–6 respectively) at 37 °C for 20 min. Lysates were immunoprecipitated with anti-JNK Ab, and JNK activity in the immunoprecipitates was measured using an in vitro kinase assay with GST–c-Jun as the substrate. 32P-labelled phosphorylated GST–c-Jun (panels a and c) and total immunoprecipitated JNK (panels b and d) are shown. Graph: levels of JNK activity in treated M2 (Δ) and A7 (●) cells were expressed as the fold-stimulation compared with levels in untreated M2 or A7 cells. (B) Confluent M2 and A7 cells were treated with 0.5 M sorbitol at 37 °C for 20 min, and lysates were immunoprecipitated with an anti-MKK7 mAb. MKK7 activity in the immunoprecipitates was measured using an in vitro kinase assay with GST–JNK1 as the substrate. Levels of phosphorylated JNK were determined using an anti-phospho-JNK Ab (upper panel). Total immunoprecipitated MKK7 was determined using an anti-MKK7 Ab (lower panel). Histogram: MKK7 activity in lysates of M2 cells and A7 cells that were not treated (NT) or treated with 0.5 M sorbitol (Sor) were expressed relative to activities of NT in M2 cells. Bars indicate the means±S.E.M. for three independent experiments. (C) MKK4 activity was measured using an anti-phospho-MKK4 (upper panel) Ab. Total MKK4 was determined using an anti-MKK4 Ab (lower panel). Histogram: MKK4 activity in lysates of M2 cells and A7 cells that were not treated (NT) or treated with 0.5 M sorbitol (Sor) were expressed relative to activities of NT in M2 cells as described in (B). (D) M2 cells were co-transfected with various amounts of pCMV5/FLAG-Filamin A (0, 1, 0, 0.5 and 1 μg in lanes 1–5 respectively) and 0.5 μg of pCMV5/HA-JNK1, together with (lanes 3–5) or without (lanes 1 and 2) 0.5 μg of pCMV5/FLAG-MKK4 and FLAG-MKK7γ2. M2 cells were co-transfected with various amounts of pCMV5/FLAG-Filamin A (0, 0.5 and 1 μg in lanes 8–10 and 11–13 respectively) and 0.5 μg of pCMV5/HA-JNK1, with 0.5 μg of pCMV5/FLAG-MKK4 (lanes 8–10) or 0.5 μg of pCMV5/FLAG-MKK7γ2 (lanes 11–13). After 48 h culture, cell lysates were analysed for JNK phosphorylation (activation) by immunoblotting with anti-phospho-JNK Ab. Total HA–JNK1, FLAG–Filamin A, FLAG–MKK4 and FLAG–MKK7 expressions were determined using anti-HA and anti-FLAG Abs. Histogram: densitometry of the bands in the upper panel, expressed as the fold increase over phosphorylation in the absence of MKK4, MKK7 proteins and Filamin A. After 48 h culture, cell lysates were analysed for JNK phosphorylation (activation) by immunoblotting with anti-phospho-JNK Ab. Total HA–JNK1 expression was determined using anti-HA Abs. Histogram: densitometry of the bands in the upper panel, expressed as the fold increase over phosphorylation in the absence of MKK4, MKK7 proteins and Filamin A. Results were obtained from at least three independent experiments. IB, immunoblot.

Finally, we analysed the effect of Filamin A on JNK activation induced by overexpression of MKK4 and/or MKK7γ2. M2 cells were transiently transfected with FLAG–MKK4 and FLAG–MKK7γ2, together with various amounts of FLAG–Filamin A. Overexpression of MKK7γ2 and MKK4 induced JNK activation which was greatly enhanced when the amount of transfected Filamin A was increased (Figure 5D, upper panel). However, MKK4- or MKK7γ2-induced JNK activation was not enhanced when the amount of transfected Filamin A was increased (Figure 5D, lower panel). These results support our contention that, by physically connecting MKK4 and MKK7γ2, Filamin A facilitates synergistic JNK activation dependent on these MAPKKs.

DISCUSSION

In vitro experiments have shown that synergistic activation of JNK requires the phosphorylation of both the threonine and tyrosine residues within the threonine-proline-tyrosine motif of JNK, and that this phosphorylation is mediated by two different enzymes, MKK4 (SEK1) and MKK7 [35]. Although both of these MAPKKs can catalyse the phosphorylation of both tyrosine and threonine, MKK4 preferentially phosphorylates the tyrosine residue, whereas MKK7 preferentially phosphorylates the threonine residue [3]. In a previous study of stress-stimulated murine ES cells, we presented in vivo data confirming that these key tyrosine and threonine residues of JNK are sequentially phosphorylated by MKK4 and MKK7 respectively [7,11] (Figure 6A). The present study provides evidence that Filamin A is one of the ‘binder’ molecules presumed to directly and closely connect MKK4 and MKK7 so that they can mediate this tyrosine/threonine phosphorylation. We showed that Filamin A (as well as Filamin B and C) associate with MKK7 and MKK4, but not with JNK1 itself (Figure 1). Furthermore, this association is isoform-specific, since Filamin A interacted with MKK7γ and MKK7β, but not with MKK7α (Figure 2). In addition, MKK7γ (but not MKK7α) co-localized with actin filaments in a manner dependent on Filamin A binding (Figure 3). We also showed that the MKK7-binding site of Filamin A was different from that for MKK4, and that the formation of the MKK7γ–MKK4 complex was mediated by Filamin A (Figure 4). Lastly, we demonstrated that Filamin A was essential for strong JNK, MKK7 and MKK4 activation in response to sorbitol, as well as for induction of JNK activation by MKK4 and MKK7 (Figure 5). Thus we present a novel model in which MKK4 and MKK7γ use Filamin A as a scaffold to support their sequential tyrosine/threonine phosphorylation and thus synergistic activation of JNK (Figure 6B).

Schematic model of the role of Filamin A in the sequential phosphorylation of JNK1 by MKK4 and MKK7γ

Figure 6
Schematic model of the role of Filamin A in the sequential phosphorylation of JNK1 by MKK4 and MKK7γ

(A) Synergistic activation of JNK through sequential tyrosine/threonine phosphorylation by MKK4 and MKK7 in murine cells (as reported in our previous study; [7]). In a murine ES cell subjected to stress, activated (phosphorylated) MKK4 mediates the phosphorylation of the tyrosine residue of the threonine-proline-tyrosine motif (TPY) of JNK (step 1), followed by threonine phosphorylation of the same JNK molecule by activated (phosphorylated) MKK7 (step 2). JNK activity is synergistically enhanced. (B) MKK4 and MKK7γ are connected by their mutual interaction with Filamin A. As demonstrated in the present study, Filamin A associates with actin filaments comprising the cytoskeleton. Filamin A also contains distinct binding sites for MKK4 and MKK7, and can interact simultaneously with these MAPKKs. The interaction of all three proteins with JNK induces synergistic levels of JNK phosphorylation and thus activation.

Figure 6
Schematic model of the role of Filamin A in the sequential phosphorylation of JNK1 by MKK4 and MKK7γ

(A) Synergistic activation of JNK through sequential tyrosine/threonine phosphorylation by MKK4 and MKK7 in murine cells (as reported in our previous study; [7]). In a murine ES cell subjected to stress, activated (phosphorylated) MKK4 mediates the phosphorylation of the tyrosine residue of the threonine-proline-tyrosine motif (TPY) of JNK (step 1), followed by threonine phosphorylation of the same JNK molecule by activated (phosphorylated) MKK7 (step 2). JNK activity is synergistically enhanced. (B) MKK4 and MKK7γ are connected by their mutual interaction with Filamin A. As demonstrated in the present study, Filamin A associates with actin filaments comprising the cytoskeleton. Filamin A also contains distinct binding sites for MKK4 and MKK7, and can interact simultaneously with these MAPKKs. The interaction of all three proteins with JNK induces synergistic levels of JNK phosphorylation and thus activation.

Our work clarifies that there are at least three types of JNK activation mechanisms: (i) MKK4-mediated JNK phosphorylation, (ii) MKK7-mediated JNK phosphorylation, and (iii) JNK phosphorylation mediated by both MKK4 and MKK7. Molecular mechanisms (i) and (ii) depend on the JNK scaffold proteins JIP1, JIP2, JSAP1 and JLP, which bind to JNK and MKK4 or MKK7, but not to all three proteins [10]. In contrast, mechanism (iii) depends on a Filamin protein (A, B or C), rather than on a JIP-type protein. We speculate that, in living cells, there may be a wide variety of JNK activation and signalling modules that involve either JIP-type or Filamin-type molecules, or a combination of these scaffold proteins depending on different cell types. The ratios of the JNK signalling modules seem to vary in different cell types.

Filamin A is an actin cross-linking protein and possesses an actin-binding domain and a homodimerization domain that allow it to determine the submembranous cytoskeletal architecture of cells. Consistent with these properties, Filamin A is required for cell adhesion and migration [15]. Previous studies have shown that Filamin A interacts with MKK4 and TRAF2 [TNF (tumour necrosis factor)-receptor-associated factor]. TRAF2 is an intracellular adaptor protein that is involved in signal transduction from the TNF receptor and related receptors, and is required for TNF-induced JNK activation [12,17]. Other work has indicated that Filamin A binds to small GTPases such as Rac, Rho, Cdc42 and RalA [16]. Thus many signalling molecules, including small GTPases, TRAF2 and MKK4, accumulate on Filamin A. These observations suggest that, as well as regulating cytoskeletal architecture, Filamin A may function as a signalsome that connects diverse signalling pathways, including MAPK modules. Conversely, in addition to its role in stress signalling, JNK has been implicated in cell migration and cytoskeletal reorganization. Huang et al. [18] reported that JNK1 phosphorylates paxillin, a focal adhesion adaptor, both in vitro and in intact fish and rat cells. Similarly, Otto et al. [19] identified p150-Spir, a Drosophila JNK-interacting protein, as belonging to the Wiscott–Aldrich syndrome protein homology domain-2 family of proteins involved in actin reorganization. p150-Spir is phosphorylated by JNK both in vitro and in vivo, indicating that p150-Spir is a downstream target of JNK function and providing a direct link between JNK activity and actin reorganization. Thus the downstream targets of JNK include not only transcription factors, but also cytoskeletal regulators. Another intriguing observation is that the receptor tyrosine kinase Ror2 mediates Wnt5a-induced polarized cell migration by activating JNK via Filamin A [20]. Lastly, Filamin A is not the only Filamin family member involved in JNK signalling, as Filamin B functions as a scaffold linking activated Rac1, MEKK1 and MKK4 to JNK during type I interferon signalling [21,22]. Taken together, these data point towards multiple connections between signalling modules and Filamin proteins.

The present study has revealed interesting differences among MKK7 isoforms. Tournier et al. [13] showed that MKK7 isoforms are present both in the cytoplasm and the nucleus, and are selectively regulated by specific extracellular stimuli. We have previously reported that treatment of ES cells with sorbitol (which affects the plasma membrane) induces stronger activation of MKK7γ1 than MKK7α1 [6]. In the present study, we discovered that MKK7γ1 co-localizes with actin on stress fibres and with Filamin A at the cell surface (Figures 3A and 3B). On the other hand, we found that MKK7α1 and MKK7γ1 deletion B, neither of which can interact with Filamin A, were diffusely distributed throughout the cytoplasm (Figure 3C). These data suggest the possibility that MKK7γ1, which co-localizes with the actin cytoskeleton right up to the plasma membrane, is involved in signal transduction from this membrane, whereas MKK7α1, which is distributed throughout the cytoplasm, is involved in a cytoplasmic signalling pathway. Thus differences in the intracellular distribution of MKK7 isoforms may correspond to different MAPK signalling modules mediating distinct cellular responses.

In conclusion, the present study has demonstrated that Filamin A interacts with specific MKK7 isoforms and is a candidate for a presumed ‘binder’ protein needed to form the JNK–MKK4–MKK7 MAPK module. Filamin A may support or induce the sequential phosphorylation of JNK by MKK4 and MKK7, and thus function as a signalsome involved in cytoskeletal events. Future work will establish whether Filamin A is a key factor supporting additional JNK signalling cascades.

Abbreviations

     
  • Ab

    antibody

  •  
  • CT

    C-terminus

  •  
  • ES

    embryonic stem

  •  
  • GFP

    green fluorescent protein

  •  
  • GST

    glutathione transferase

  •  
  • HA

    haemagglutinin

  •  
  • HEK

    human embryonic kidney

  •  
  • JNK

    c-Jun N-terminal kinase

  •  
  • JIP

    JNK-interacting protein

  •  
  • JLP

    JNK-associated leucine-zipper protein

  •  
  • mAb

    monoclonal antibody

  •  
  • MAPK

    mitogen-activated protein kinase

  •  
  • MAPKK (MKK)

    MAPK kinase

  •  
  • MAPKKK

    MAPKK kinase

  •  
  • MEKK1

    MEK [MAPK/ERK (extracellular-signal-regulated kinase) kinase] kinase

  •  
  • MLK

    mixed lineage protein kinase

  •  
  • SAPK

    stress-activated protein kinase

  •  
  • JSAP1

    JNK/SAPK-associated protein 1

  •  
  • TNF

    tumour necrosis factor

  •  
  • TRAF2

    TNF-receptor-associated factor

AUTHOR CONTRIBUTION

Kentaro Nakagawa, Misato Sugahara, Tokiwa Yamasaki and Hiroaki Kajiho performed the experiments. Shinya Takahashi and Jun Hirayama analysed the data. Yasuhiro Minami, Yasutaka Ohta and Toshio Watanabe provided essential reagents. Yutaka Hata, Toshiaki Katada and Hiroshi Nishina designed the experiments, provided funding and wrote the paper.

We are grateful to numerous members of the Nishina and Katada laboratories for helpful discussions, expert technical assistance and critical reading of this manuscript.

FUNDING

This work was supported, in part, by a Grant-in-Aid for Scientific Research on a Priority Area from the Ministry of Education, Culture, Sport, Science and Technology of Japan and the Ministry of Health, Labour and Welfare of Japan [grant numbers 17081005 and 20390022].

References

References
1
Davis
R. J.
Signal transduction by the JNK group of MAP kinases
Cell
2000
, vol. 
103
 (pg. 
239
-
252
)
2
Chang
L.
Karin
M.
Mammalian MAP kinase signalling cascades
Nature
2001
, vol. 
410
 (pg. 
37
-
40
)
3
Lawler
S.
Fleming
Y.
Goedert
M.
Cohen
P.
Synergistic activation of SAPK1/JNK1 by two MAP kinase kinases in vitro
Curr. Biol.
1998
, vol. 
8
 (pg. 
1387
-
1390
)
4
Fleming
Y.
Armstrong
C. G.
Morrice
N.
Paterson
A.
Goedert
M.
Cohen
P.
Synergistic activation of stress-activated protein kinase 1/c-Jun N-terminal kinase (SAPK1/JNK) isoforms by mitogen-activated protein kinase kinase 4 (MKK4) and MKK7
Biochem. J.
2000
, vol. 
352
 (pg. 
145
-
154
)
5
Lisnock
J.
Griffin
P.
Calaycay
J.
Franz
B.
Parsons
J.
O'Keefe
S. J.
LoGrasso
P.
Activation of JNK3 α1 requires both MKK4 and MKK7: kinetic characterization of in vitro phosphorylated JNK3 α1
Biochemistry
2000
, vol. 
39
 (pg. 
3141
-
3148
)
6
Wada
T.
Nakagawa
K.
Watanabe
T.
Nishitai
G.
Seo
J.
Kishimoto
H.
Kitagawa
D.
Sasaki
T.
Penninger
J. M.
Nishina
H.
Katada
T.
Impaired synergistic activation of stress-activated protein kinase SAPK/JNK in mouse embryonic stem cells lacking SEK1/MKK4: different contribution of SEK2/MKK7 isoforms to the synergistic activation
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
30892
-
30897
)
7
Kishimoto
H.
Nakagawa
K.
Watanabe
T.
Kitagawa
D.
Momose
H.
Seo
J.
Nishitai
G.
Shimizu
N.
Ohata
S.
Tanemura
S.
, et al. 
Different properties of SEK1 and MKK7 in dual phosphorylation of stress-induced activated protein kinase SAPK/JNK in embryonic stem cells
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
16595
-
16601
)
8
Nishina
H.
Wada
T.
Katada
T.
Physiological roles of SAPK/JNK signaling pathway
J. Biochem.
2004
, vol. 
136
 (pg. 
123
-
126
)
9
Nishitai
G.
Shimizu
N.
Negishi
T.
Kishimoto
H.
Nakagawa
K.
Kitagawa
D.
Watanabe
T.
Momose
H.
Ohata
S.
Tanemura
S.
, et al. 
Stress induces mitochondria-mediated apoptosis independent of SAPK/JNK activation in embryonic stem cells
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
1621
-
1626
)
10
Dhanasekaran
D. N.
Kashef
K.
Lee
C. M.
Xu
H.
Reddy
E. P.
Scaffold proteins of MAP-kinase modules
Oncogene
2007
, vol. 
26
 (pg. 
3185
-
3202
)
11
Nishina
H.
Nakagawa
K.
Azuma
N.
Katada
T.
Activation mechanism and physiological roles of stress-activated protein kinase/c-Jun NH2-terminal kinase in mammalian cells
J. Biol. Regul. Homeost. Agents
2003
, vol. 
17
 (pg. 
295
-
302
)
12
Marti
A.
Luo
Z.
Cunningham
C.
Ohta
Y.
Hartwig
J.
Stossel
T. P.
Kyriakis
J. M.
Avruch
J.
Actin-binding protein-280 binds the stress-activated protein kinase (SAPK) activator SEK-1 and is required for tumor necrosis factor-α activation of SAPK in melanoma cells
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
2620
-
2628
)
13
Tournier
C.
Whitmarsh
A. J.
Cavanagh
J.
Barrett
T.
Davis
R. J.
The MKK7 gene encodes a group of c-Jun NH2-terminal kinase kinases
Mol. Cell. Biol.
1999
, vol. 
19
 (pg. 
1569
-
1581
)
14
Kitagawa
D.
Kajiho
H.
Negishi
T.
Ura
S.
Watanabe
T.
Wada
T.
Ichijo
H.
Katada
T.
Nishina
H.
Release of RASSF1C from the nucleus by Daxx degradation links DNA damage and SAPK/JNK activation
EMBO J.
2006
, vol. 
25
 (pg. 
3286
-
3297
)
15
Cunningham
C. C.
Gorlin
J. B.
Kwiatkowski
D. J.
Hartwig
J. H.
Janmey
P. A.
Byers
H. R.
Stossel
T. P.
Actin-binding protein requirement for cortical stability and efficient locomotion
Science
1992
, vol. 
255
 (pg. 
325
-
327
)
16
Ohta
Y.
Suzuki
N.
Nakamura
S.
Hartwig
J. H.
Stossel
T. P.
The small GTPase RalA targets filamin to induce filopodia
Proc. Natl. Acad. Sci. U.S.A.
1999
, vol. 
96
 (pg. 
2122
-
2128
)
17
Leonardi
A.
Ellinger-Ziegelbauer
H.
Franzoso
G.
Brown
K.
Siebenlist
U.
Physical and functional interaction of filamin (actin-binding protein-280) and tumor necrosis factor receptor-associated factor 2
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
271
-
278
)
18
Huang
C.
Rajfur
Z.
Borchers
C.
Schaller
M. D.
Jacobson
K.
JNK phosphorylates paxillin and regulates cell migration
Nature
2003
, vol. 
424
 (pg. 
219
-
223
)
19
Otto
I. M.
Raabe
T.
Rennefahrt
U. E.
Bork
P.
Rapp
U. R.
Kerkhoff
E.
The p150-Spir protein provides a link between c-Jun N-terminal kinase function and actin reorganization
Curr. Biol.
2000
, vol. 
10
 (pg. 
345
-
348
)
20
Nomachi
A.
Nishita
M.
Inaba
D.
Enomoto
M.
Hamasaki
M.
Minami
Y.
Receptor tyrosine kinase Ror2 mediates Wnt5a-induced polarized cell migration by activating c-Jun N-terminal kinase via actin-binding protein filamin A
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
27973
-
27981
)
21
Jeon
Y. J.
Choi
J. S.
Lee
J. Y.
Yu
K. R.
Ka
S. H.
Cho
Y.
Choi
E-J.
Baek
S. H.
Seol
J. H.
Park
D.
, et al. 
Filamin B serves as a molecular scaffold for type I interferon-induced c-Jun NH2-terminal kinase signaling pathway
Mol. Biol. Cell
2008
, vol. 
19
 (pg. 
5116
-
5130
)
22
Jeon
Y. J.
Choi
J. S.
Lee
J. Y.
Yu
K. R.
Kim
S. M.
Ka
S. H.
Oh
K. H.
Kim
K. I.
Zhang
D-E.
Bang
O.S.
Chung
C.H.
ISG15 modification of filamin B negatively regulates the type I interferon-induced JNK signalling pathway
EMBO Rep.
2009
, vol. 
10
 (pg. 
374
-
380
)

Author notes

1

These authors contributed equally to the present study.