- Email: [email protected]
Dok5 is substrate of TrkB and TrkC receptors and involved in neurotrophin induced MAPK activation
Cellular Signalling 18 (2006) 1995 – 2003 www.elsevier.com/locate/cellsig
Dok5 is substrate of TrkB and TrkC receptors and involved in neurotrophin induced MAPK activation Lei Shi a,1 , Jiping Yue a,1 , Yuangang You a , Bin Yin a , Yanhua Gong a , Caimin Xu a , Boqin Qiang a , Jiangang Yuan a , Yongjian Liu b,⁎, Xiaozhong Peng a,⁎ a
The National Laboratory of Medical Molecular Biology, Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences and Peking Union Medical College, Chinese National Human Genome Center, Beijing 100005, China b Department of Neurology, University of Pittsburgh School of Medicine, PA 15213, USA Received 4 February 2006; received in revised form 3 March 2006; accepted 10 March 2006 Available online 2 May 2006
Abstract Tropomyosin-related kinase (Trk) family receptors are a group of high affinity receptors for neurotrophin growth factors, which have pivotal functions in many physiological processes of nervous system. Trk receptors can dimerize and autophosphorylate upon neurotrophin stimulation, then recruit multiple adaptor proteins to transduct signal. In this report, we identified Dok5, a member of Dok family, as a new substrate of TrkB/C receptors. In yeast two-hybrid assay, Dok5 can interact with intracellular domain of TrkB and TrkC receptor through its PTB domain, but not with that of TrkA receptor. The interaction was then confirmed by GST pull-down assay and Co-IP experiment. Dok5 co-localized with TrkB and TrkC in differentiated PC12 cells, providing another evidence for their interaction. By using mutational analysis, we characterized that Dok5 PTB domain bound to Trk receptor NPQY motif in a kinase-activity-dependent manner. Furthermore, competition experiment indicated that Dok5 competed with N-shc for binding to the receptors at the same site. Finally, we showed that Dok5 was involved in the activation of MAPK pathway induced by neurotrophin stimulation. Taken together, these results suggest that Dok5 acts as substrate of TrkB/C receptors and is involved in neurotrophin induced MAPK signal pathway activation. © 2006 Elsevier Inc. All rights reserved. Keywords: Dok5; PTB domain; Trk receptor; NPQY motif; MAPK
1. Introduction The neurotrophin family growth factors, including nerve growth factor (NGF), brain-derived neurotrophic factors (BDNF), neurotrophin-3 (NT-3) and neurotrophin-4/5 (NT-4/5), have important functions in nervous system. They play multiple roles in vertebrate neural development, differentiation, survival, synaptic plasticity, myelination and many other processes [1,2]. Their high affinity receptors, tropomyosin-related kinase (Trk) family receptors, are cell membrane receptor tyrosine kinase (RTK). This receptor family has three structural conserved members: TrkA, TrkB and TrkC. Each neurotrophin binds to different receptors: NGF binds to TrkA, BDNF binds to TrkB, and NT-3 binds to ⁎ Corresponding authors. Tel.: +86 10 65296411; fax: +86 10 65240529. E-mail address: [email protected] (X. Peng). 1 These two authors contributed equally to this work. 0898-6568/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.cellsig.2006.03.007
TrkC. Also NT-3 can bind to TrkA and NT-4 can bind to TrkB [1,2]. Stimulation of neurotrophins can result in dimerization and activation of Trk receptors. Then the activated receptors phosphorylate several tyrosines of its intracellular domain (ICD). The phosphorylated tyrosines may serve as the docking site for signal molecules and mediate the signal transduction. Many signal molecules such as Src homology 2 domain containing transforming protein (Shc), FGF receptor substrate 2 (FRS2), growth factor receptor bound protein 2 (Grb2), phospholipase Cγ(PLC-γ) have been proved to be substrates of Trk receptors [3–5]. They can be phosphorylated by Trk receptors and cascade the signal to downstream. Downstream of tyrosine kinase/Docking proteins (Dok) are adaptor proteins that can act as substrate of multiple tyrosine kinase including both receptor tyrosine kinases and non-receptor tyrosine kinase [6–10]. The first member of Dok family, Dok1/ p62Dok, was identified by Carpino et al. [11] and Yamanashi and
1996
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
Baldimore [12]. Subsequently other members were identified, and now there are six members in this family: Dok1/p62Dok, Dok2/DokR/p56Dok, Dok3/DokL, Dok4, Dok5, and Dok6/ Dok5 like [13–17]. Dok proteins have classic feature of adaptor proteins: N-terminal pleckstrin homology (PH) domain, central phosphotyrosine binding (PTB) domain and C-terminal region. According to different expression patterns, Dok family proteins are divided into two subgroups. The first subgroup, Dok1– 3 are predominantly expressed in hematopoietic tissues and act as negative regulator in many signal pathways. Dok1/2 inhibit MAPK activation through their association with RasGTPaseactivating protein (RasGAP) [18,19]. Van Slyke et al. showed Dok2 attenuated EGFR signaling by recruiting c-Src and Csk, providing an alternative regulatory way [20]. Dok3, as the negative regulator of the immunoreceptor signaling in B cells and macrophages, exerts its function by the formation of Dok3SHIP1 complex [15,21]. Recently, studies from double or single knockout mice of Dok1 and/or Dok2 provided solid evidences in vivo for their negative roles in signal transduction and suppressant functions in leukemogenesis [22–24]. Another subgroup, Dok4–6 are expressed in non-hematopoietic tissues. Dok4 is widely expressed, especially in tissues of epithelial origin, and strongly inhibit Ret or Fyn signaling [25]. Itoh et al. found Dok4 localized in mitochondrial and
played a role in tumor necrosis factor-alpha (TNF-alpha) mediated reactive oxygen species (ROS) production [26]. Dok6, the newest member, is mainly expressed in nervous system and promotes Ret mediated neurite outgrowth in N2A-α1 cells [17]. And for Dok5, Northern blot and in situ hybrid results showed that Dok5 was mainly expressed in nervous system, especially in neural tube, dorsal root ganglia (DRG) and cranial ganglia. Studies also showed Dok5 was involved in insulin and GDNF signal pathway and it mediated GDNF induced neurite outgrowth in PC12 cells [16,27]. These results suggest Dok5 has very important function in neural development. But as recently identified member, its function remains mostly unclear. In this study, we identified Dok5 as novel substrate of TrkB/ C receptors. We proved Dok5 could selectively bind to NPQY motif of TrkB and TrkC receptors, not that of TrkA receptors in a kinase-activity-dependent manner. Co-localization of Dok5 and TrkB or TrkC receptors in differentiated PC12 cells provide another evidence for their interaction. In competition assay, we showed that Dok5 competed with N-shc for Trk receptor binding. Furthermore, we demonstrated that Dok5 was involved in the activation of MAPK signal pathway upon neurotrophin stimulation. Taken together, we proved that Dok5 could act as substrate of Trk receptors and was involved in neurotrophin induced MAPK activation. These results indicate that Dok5
Fig. 1. Dok5 interacts with TrkB and TrkC ICD in yeast two-hybrid system. (A) Schematic diagram shows the structures of Dok5 and Trk receptors. (B) Dok5 interacts with TrkB and TrkC ICD, not with TrkA ICD in yeast two-hybrid system. Yeast (SFY526) were transformed with Trk ICD in pAS2-1 vector and Dok5 in pACT2 vector, then cultured on SD/-Trp/-Leu plates for 4 days at 30 °C. Colony liftβ-galactosidase assays were performed according to the user manual. (C) Dok5 PTB domain, not PH domain or C-terminal region, binds to TrkB and TrkC ICD in yeast two-hybrid assay.
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
plays an important role in neurotrophin signal pathway, and also provide new clues for Dok5 function in neuronal development and differentiation. 2. Materials and methods
1997
esis Kit (Stratagene) [3,4,28,29]. For prokaryotic expression, the sequence coding for Dok5 PTB domain was subcloned into pGEX6P-1 vector. For eukaryotic expression, wild type and mutant Trk receptors were subcloned into expression vector pcDNA3.1V5. Dok5 full length cDNA and Dok5ΔC-terminal region(Dok5ΔC) were subcloned into expression vector pcDNA3.1Flag vector. For image analysis, TrkB and TrkC full length cDNA was subcloned into pmRFP-N1 vector and Dok5 cDNA were subcloned into pEGFP-N1 vector.
2.1. Reagents and cell culture 2.3. Yeast two-hybrid assay Human recombinant NGF, BDNF and NT-3 were purchased from Peprotech Inc. The anti-V5 polyclonal antibody was purchased from Novus. The anti-Flag monoclonal antibody was from Sigma. The anti-Myc monoclonal antibody (9E10) was purchased from Convance. The anti-Erk and anti-phosphop44/42Erk MAP kinase antibody were purchased from Cell signaling. The anti-pTyr (PY99) monoclonal antibody was product of Santa Cruz Biotechnology. Human Embryonic Kidney (HEK)293 cells were cultured in DMEM medium with 10% fetal bovine serum (Hyclone). PC12 cells were maintained in DMEM medium with 10% horse serum and 5% fetal bovine serum. For transfection, cells were split into 6-well plate and cultured to 90% confluency. Then the transfection was done with Lipofectamine2000 (Invitrogen) according to the user manual.
2.2. Plasmid construction The human TrkA full length cDNA was a kindly gift from Dr. Lloyd Greene. The human TrkB and TrkC full length cDNA were amplified from human fetal brain cDNA library. For yeast two-hybrid assay, the Trk receptors intracellular domains (ICD) were subcloned into pAS2-1 vector. Dok5 full length cDNA and the sequence coding for PH domain, PTB domain and C-terminal region were subcloned into pACT2 vectors, respectively. All the mutant TrkB and TrkC receptors, including TrkBM1(K572A), TrkBM2(Y516F), TrkBM3(Y702A), TrkBM4(Y706D), TrkBM5(Y707E), TrkBM6(Y817F), TrkCM1(K572A), TrkCM2(Y516F), TrkCM3(Y705A), TrkCM4(Y709D), TrkCM5(Y710E) and TrkCM6(Y820F) were generated by using QuikChange Site-Directed Mutagen-
For yeast two-hybrid assay, Trk receptor intracellular domains were fused to the GAL4 DNA binding domain in pAS2-1 vector. Wild type Dok5, or Dok5 PH domain, PTB domain and C-terminal region were fused to GAL4 activation domain in pACT2 vectors, respectively. The co-transformation was performed in yeast strain SFY526 by using TE/LiAc method according to the user manual of MATCHMAKER yeast two-hybrid system 2 (Clontech). Transformed yeast were selected on SD dropout medium/-Leu/-Trp plates at 30 °C for 4 days. Then colony liftβ-galactosidase assays were performed.
2.4. GST pull-down assay pGEX6P-Dok5-PTB plasmid was transformed into BL21 Escherichia coli strain. Then the overnight cultures were diluted 1:500 in 200 ml LB medium and cultured at 37 °C until OD600 to 0.6. The induction was performed by shaking at 30 °C for 4 h with 0.8 mM isopfopyl-β-D-thiogalactopyranoside. GST-Dok5PTB fusion proteins were immobilized onto Sepharose 4B beads according to the protocol (Amersham Biosciences). The HEK293 cells were transient transfected with pcDNA3.1V5-Trk plasmids. 24 h later, the cell were treated with or without neurotrophin for 10 min and lysed. The cell lysates were centrifuged at 12,000×g for 15 min and the supernatants were incubated with the same amount of GSTDok5-PTB Sepharose 4B beads at 4 °C for 2 h. The beads were washed with 1× PBS for 3 times and then resuspended in 1× SDS loading buffer and boiled for 5 min. Protein samples were analyzed by Western blotting or Coomassie Brilliant Blue staining.
Fig. 2. Dok5 PTB domain binds to TrkB and TrkC receptors in vitro. (A) Dok5 PTB domain binds to TrkB and TrkC receptors in GST pull-down assay. HEK293 cells were transfected with pcDNA3.1V5-Trk and treated with or without neurotrophin (100 ng/ml) for 10 min. The lysates were incubated with Dok5-PTB-GST Sepharose 4B beads at 4 °C for 2 h. The bound proteins were then analyzed by Western blotting with anti-V5 antibody (upper panel). Cell lysates were examined by anti-V5 or anti-pTyr antibody to detect the expression and phosphorylation level of Trk receptors (second and third panel). Dok5-PTB-GST proteins were examined by Coomassie Brilliant Blue staining (bottom panel). (B) Alignment of amino acids sequence around NPQY motif of Trk family receptors. Homology level: red, 100%; blue, ≥50%. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 3. Dok5 binds to NPQY motif of TrkB and TrkC receptors by a kinase-activity-dependent manner. (A) Schematic diagram shows the important amino acids of TrkB and TrkC. (B) Mutation analysis of the key amino acids of Trk receptor for Dok5 binding. Yeast two-hybrid assay was performed with pACT2-Dok5-PTB and TrkB or TrkC ICD mutants in pAS2-1 vector, including BM1(K572A), BM2(Y516F), BM3(Y702A), BM4 (Y706D), BM5(Y707E), BM6(Y817F), or CM1(K572A), CM2(Y516F), CM3(Y705A), CM4(Y709D), CM5(Y710E), CM6(Y820F). (C) Dok5 binds to NPQY motif of TrkB and TrkC receptors by a kinase-activitydependent manner in vivo. Co-IP experiments were done in HEK293 cells co-transfected with Dok5 and wild type or mutant receptors. Before harvest, the cells were treated with 100 ng/ml BDNF or NT-3 for 10 min. Protein samples were then detected by Western blotting with anti-V5, anti-Flag or anti-pTyr antibody. (D) Dok5 is phosphorylated by TrkB or TrkC receptors after stimulation. WT, wild type.
1998 L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
2.5. Co-immunoprecipitation HEK293 cells were plated into 6-well plate and cultured to 90% confluency. The cells were then transient co-transfected with pcDNA3.1V5-Trk and pcDNA3.1Flag-Dok5, pcDNA3.1V5-TrkM1 and pcDNA3.1Flag-Dok5, pcDNA3.1V5-TrkM2 and pcDNA3.1Flag-Dok5, or pcDNA3.1V5-TrkM6 and pcDNA3.1Flag-Dok5. 24 h after transfection, stimulate the cells with 100 ng/ml neurotrophin (PeproTech) for 10 min. Then cell were lysed in 750 μl lysis buffer (50 mM Tris–HCl pH 7.5, 1 mM EDTA, 120 mM NaCl, 0.5% NP-40, 10% glycerol, 1 mM Na3VO4, 25 mM NaF) containing protease inhibitors. After incubation on ice for 30 min, detergent-insoluble materials were removed by centrifugation at 4 °C at 12,000×g for 15 min. The 2-μg anti-Flag monoclonal antibody (Sigma) was added to the lysates and incubated at 4 °C rotating for 3 h followed by addition of 30 μl protein G-agarose (Roche). The incubation was continued for 3 h. The mixture was centrifuged at 12,000×g for 1 min, and the pellet was washed 3 times with 1 ml washing buffer (20 mM Tris–HCl pH 7.4, 1 mM EDTA, pH 8.0, 200 mM NaCl, 0.5% NP-40, 100 μM Na3VO4). The samples were analyzed by Western blotting with anti-V5 antibody (Invitrogen), anti-Flag antibody (Sigma) or anti-pTyr antibody (PY99, Santa Cruz).
2.6. Fluorescence microscopy pmRFPN1-TrkB or pmRFPN1-TrkC and pEGFPN1-Dok5 were co-transfected into PC12 cells. Twenty-four hours later, PC12 cells were split into 24-well plate with polylysine coated cover slips and cultured in low serum medium (1% equine serum) with 50 ng/ml BDNF or NT-3 for 48 h. Cells then were fixed with 4% PFA for 30 min at RT and washed with PBS for 3 times. The cover slips were mounted and co-localization was analyzed under a confocal imaging system.
2.7. MAPK signal pathway detection HEK293 cells were co-transfected with pcDNA3.1V5-Trk and pcDNA3. 1Flag, pcDNA3.1V5-Trk and pcDNA3.1Flag-Dok5, or pcDNA3.1V5-Trk and
1999
pcDNA3.1Flag-Dok5ΔC. The cells were treated with 100 ng/ml BDNF or NT-3 and harvested at 0 min, 5 min, 15 min, 30 min and 60 min. The lysates were examined by Western blotting with anti-Erk1 (k-23, Santa Cruz), anti-pErk (E-4, Santa Cruz), anti-Flag antibody (Sigma) or anti-V5 antibody (Invitrogen).
3. Results 3.1. Dok5 PTB domain binds to TrkB and TrkC ICD in yeast two-hybrid system Bioinformatics analysis and our previous structural studies showed that Dok5 has a conserved PTB domain [30], which make it possible for Dok5 to be the potential substrate of multiple tyrosine kinases. In consideration of Dok5 neural specific expression, we try to identify new tyrosine kinase expressed in nerve system that coupled with Dok5. Trk receptors, which are very important RTKs during neural development and differentiation, could be good candidates. To verify this possibility, we use yeast two-hybrid system to examine whether there are interactions between Dok5 and Trk receptors. There was evidence showing that the dimerization of the Gal4 BD domain could activate Trk ICD [5], which enable yeast two-hybrid assay to be used to detect the interaction between substrate and receptors. So we performed yeast two-hybrid with Trk ICD in pAS2-1 vector and Dok5 in pACT2 vector. Interestingly, the result showed that there were interactions between Dok5 and TrkB or TrkC ICD, while there was no obvious interaction between Dok5 and TrkA ICD (Fig. 1B). This implies that Dok5 has binding selectivity among
Fig. 4. Dok5 co-localizes with TrkB or TrkC in differentiated PC12 cells. PC12 cells were transfected with the constructs as indicated and induce by 50 ng/ml BDNF or NT-3 for 2 days.
2000
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
three members of Trk family receptors. In order to determine the interaction domain of Dok5, we further subcloned Dok5 PH domain, PTB domain and C-terminal region into pACT2 vector, respectively, and performed yeast two-hybrid assay again. As shown in Fig. 1C, Dok5 PTB domain is key domain responsible for the interaction with TrkB and TrkC ICD, while neither PH domain nor C-terminal region can interact with the receptors.
fusing protein could not bind to TrkA receptors no matter whether TrkA was activated or not (Fig. 2A). The consistent results from yeast two-hybrid and pull-down assay indicated that Dok5 selectively bind to TrkB and TrkC, but not TrkA.
3.2. Dok5 binds to TrkB and TrkC receptor in GST pull-down
We have proved that Dok5 interacted with the Trk receptors through its PTB domain. As known, PTB domain can recognize the NPQY motif by two manners: dependent or independent on phosphorylation [31,32]. To determine by which tyrosine and by which manner that Dok5 binds to Trk receptors, we constructed serial mutant Trk receptors for yeast two-hybrid assay and Co-IP experiment. As showed in Fig. 3A, Trk receptors have 5 tyrosine residues in their ICD that can be phosphorylated upon activation. For TrkB, they are Tyr516 , Tyr702, Tyr706, Tyr707 and Tyr817 . Each tyrosine mutation would disrupt the interaction between the receptor and the substrate binding at this site. The juxtamembrane Tyr516 lies in the conserved NPQY motif, which is typical binding motif for PTB domain. Many adaptor proteins containing PTB domain, such as Shc and Frs2, bind to Trk receptors at this site. And we supposed that this tyrosine is also the binding site for Dok5. Another key amino residue, Lys572, is essential for tyrosine kinase activity of the receptor. The mutation of this lysine would produce kinase inactive receptor. As expected, the yeast two-hybrid results proved that NPQY motif is the binding site for Dok5, because the mutant receptor Y516F lost the interaction with Dok5. Other tyrosine mutations have no obvious influence on the interaction. And the kinase inactive TrkB (K572A) had no interaction with Dok5 either, which implied that the kinase activity is essential for the interaction (Fig. 3B). There was similar result between Dok5 and TrkC. By Co-IP experiment, we further proved that Dok5 bind to NPQY motif of TrkB depending on phosphorylation. The result showed that wild type TrkB receptors could be immunoprecipitated by Dok5 upon BDNF stimulation; while the mutant receptor, TrkBM1 (K572A) or TrkBM2 (Y516F), could not be co-immunoprecipitated by Dok5 with the stimulation. The unrelated tyrosine mutation, TrkBM6 (Y817F), did not affect the interaction. Similar results were obtained between Dok5 and TrkC with stimulation of NT-3 (Fig. 3C). Moreover, as substrates of TrkB and TrkC, Dok5 phosphorylation level was greatly enhanced upon neurotrophin stimulation (Fig. 3D). These results strongly suggested that Dok5 binds to NPQY motif of TrkB/C receptors in a kinase-activity-dependent manner.
To confirm the results of yeast two-hybrid assay, we used GST pull-down assay to examine whether there are interaction between Dok5 and Trk receptors in vitro. The result proved that Dok5 PTB domain can bind to the activated TrkB and TrkC receptors (Fig. 2A). Upon neurotrophin stimulation, Trk receptors became highly phosphorylated comparing with those without stimulation. And only activated TrkB or TrkC receptors can be pulled down by GST-Dok5-PTB fusion protein. However, GST-Dok5-PTB
3.3. Dok5 binds to NPQY motif of TrkB and TrkC in a kinaseactivity-dependent manner
3.4. Dok5 co-localizes with TrkB and TrkC receptors in differentiated PC12 cells
Fig. 5. Dok5 and N-shc compete for TrkB (A) or TrkC (B) receptors binding. HEK293 cells were transfected with Trk, Dok5 and N-shc and then treated with 100 ng/ml BDNF or NT-3 for 10 min. Lysates were immunoprecipitated with anti-Flag antibody and the precipitated proteins were detected with anti-V5 and anti-Flag antibody. Cell lysates were detected with anti-V5 and anti-Myc antibody to examine the expression of Trk and N-shc.
PC12 cell line has been widely used as a good cell model for researches on neurotrophin and Trk receptors. With induction of NGF, PC12 cells would manifest characteristics of neuronal cells. So we choose PC12 cells for Dok5 and Trk receptors localization study. We constructed EGFPN1-Dok5 and mRFPN1-TrkB/C clones and co-transfected them into PC12 cells. The cells were
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
2001
Fig. 6. Dok5 mediate the activation of MAPK pathway upon neurotrophin stimulation in HEK293 cells. HEK293 cells were transfected with the constructs as indicated. The cells were treated with BDNF (A) or NT-3 (B) and harvested at 0 min, 5 min, 15 min, 30 min and 60 min. Cell lysates were examined by Western blotting with antibody indicated.
induced with 50 ng/ml BDNF or NT-3 in low concentration serum medium (1% equine serum) for 2 days and then analyzed by confocal microscopy. Obviously, Dok5 co-localize with TrkB and TrkC receptors in differentiated PC12 cells (Fig. 4). The colocalization of Dok5 and Trk receptors provided another evidence for their interaction. 3.5. Dok5 competes with N-shc for interaction with the same site of Trk receptor As well known, Shc can bind to the NPQY motif of Trk receptors through its PTB domain. In this study, we showed that Dok5 PTB domain also bind to the same site of Trk receptor. So it is possible that Dok5 competes with Shc for Trk receptor binding. Shc adaptor protein family comprises three members; ShcA, B and C. Among them, ShcC/N-shc is exclusively expressed in brain and also interact with Trk receptors [33–35]. We subcloned N-shc into pcDNA4Myc vector and performed a competition experiment. In competition assay, pcDNA3.1V5Trk, pcDNA3.1Flag-Dok5 and pcDNA4Myc-N-shc were cotransfected into HEK293 cells. The plasmid amounts of Trk and Dok5 were kept equal, with the plasmid amount of N-shc increasing. As indicated in Fig. 5, with the increasing expresion of N-shc, the amount of Trk receptors that interacted with Dok5 reduced gradually (Fig. 5). When there was a huge amount of Nshc, almost no Trk receptors were co-immunoprecipitated by
Dok5. This result provided a good competitive model for Dok5 and N-shc interacting with Trk receptor. 3.6. Dok5 is involved in the activation of MAPK signal pathway induced by BDNF and NT-3 To make clear the function of Dok5 in neurotrophin signaling, we detected its effects on MAPK signal pathway. We cotransfected Dok5 and Trk into HEK293 cells, another group the empty vector and Trk were used as control. Twenty-four hours after transfection, cells were treated with BDNF or NT-3 and then harvest cells at indicated time and detected MAPK pathway. Obviously, the existence of Dok5 enhanced the phosphorylation level of Erk1/2 compared with the control group. Meanwhile Dok5 also prolonged the activation time of MAPK pathway (Fig. 6). Previous studies showed that the C-terminal region of Dok protein was essential for signal transduction. Dok1/2 C-terminal region have multiple tyrosine residues and PXXP motifs that can serve as docking site for SH2/SH3 containing signal proteins [13,36]. The deletion of C-terminal region would disrupt their functions. So we were motivated to construct Dok5ΔC mutant and detect MAPK pathway again. As shown in Fig. 6, Dok5ΔC could inhibit MAPK pathway by reducing the strength and shorten the activation time. Thus, the result demonstrated that Dok5 participated in MAPK pathway activation induced by BDNF and NT-3.
2002
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
4. Discussion In this study, by using yeast two-hybrid and pull-down assay, we identified Dok5 as substrate of TrkB and TrkC receptors. There was no obvious interaction between Dok5 and TrkA. As members of same family, TrkA, B and C share high homology in structure and they all have juxtamembrane NPQY motif for PTB domain binding. By aligning the amino acid sequence around NPQY motif, we found there were several amino acid residue deletions in TrkA compared with TrkB and TrkC receptors (Fig. 2B). These amino residues at the N-terminal or C-terminal of NPQY motif might provide the recognition specificity for the substrates. Interestingly, study from Shi et al. illustrated the structural basis of recognition between Dok1 PTB domain and c-Ret, and they also showed that Dok1 PTB domain did not recognize phosphopeptide including NPQY motif from TrkA [37]. The structural information from Dok1 may provide the clue for Dok5 specific recognition among Trk receptors. Furthermore, Dok5 selectively binding to TrkB and TrkC may suggest its distinct function in BDNF and NT-3 signal pathway. Though evidences showed that there might be functional redundancy among Trk signal pathways, studies also proved TrkA/B/C had nonoverlapping functions in neural development. The significant loss of neurons in BDNF/TrkB and NT-3/TrkC knockout mouse demonstrated their indispensable functions during sensory and motor neurons development [38]. The overlapped expression patterns of Dok5 and TrkB/C receptors and their interaction imply that Dok5 may have important functions in sensory and motor neurons development and survival. Moreover, we proved that Dok5 binds to TrkB and TrkC receptors NPQY motif in a kinase-activity-dependent manner and Dok5 is phosphorylated by Trk receptors upon neurotrophin stimulation. Co-localization of Dok5 and TrkB or TrkC receptors in differentiated PC12 cells provides another evidence for their physical interaction. The identification of Dok5 as substrate of TrkB/C receptors will enrich the signal network of neurotrophin pathway. Besides Dok5, many other adaptor proteins containing PTB domain have been reported to bind to the NPQY motif of Trk receptors, for example Shc, IRS and FRS2 [4,39]. It is hard to imagine that so many kinds of adaptor proteins bind to the receptor at the same site and at the same time. The most possibility is that they compete with each other for receptor binding. And our result established a competitive model between Dok5 and N-shc for Trk binding. This might be a good explanation for several adaptor proteins binding to the receptors at the same site to mediate different downstream signal pathway. But actual situation in vivo may be much more complicated. The different expression tissues and cell types of these adaptor proteins, and their different binding affinity with the receptor and many other factors should be considered. Unlike Dok1 and Dok2 as negative regulator in MAPK pathways [22,40,41], Dok5 mediated GDNF-induced MAPK activation [16]. Here we also showed that overexpression of Dok5 could augment MAPK activation induced by BDNF or NT-3; while the deletion of Dok5 C-terminal region disrupted this effect. This indicates that Dok5 C-terminal region is res-
ponsible for cascading the signal to downstream. However, till now downstream signal molecules that associate with Dok5 Cterminal region have not been identified yet. Compared with Dok1/2, Dok5 has a much shorter C-terminal region. Many SH2/ SH3 containing proteins that associated with Dok1/2, including RasGAP and NCK, showed no interaction with Dok5 [16,27]. Crowder et al. tried to find downstream signal proteins coupled with Dok4/5/6 subgroup by bioinformatics database search, but no functional known proteins were obtained [17]. So the studies of identifying the downstream partners of Dok5 become very important for explaining its function in signal transduction. In summary, we identified Dok5 as the new substrate of TrkB and TrkC receptors and it was involved in MAPK activation induced by BDNF and NT-3. Dok5 bind to the NPQY motif of Trk receptors in a kinase-activity-dependent manner through its PTB domain. These results enrich the neurotrophin signaling network and supply important information about Dok5 function in nervous system. Our studies as well as previous researches have showed that Dok5 was involved in TrkB/C, c-Ret, and insulin signal transduction, however the detailed regulatory mechanism and comprehensive functions of Dok5 in these pathways are yet unknown, more studies need to be done. Acknowledgements This work was supported by grants from the National Program for the Key Basic Research Project (‘973’ Nos. 2001CB510206, 2004CB518604 and 2005CB522507), and the National Sciences Foundation of China (Nos. 30421003 and 30430200). References [1] E.J. Huang, L.F. Reichardt, Annu. Rev. Neurosci. 24 (2001) 677. [2] E.J. Huang, L.F. Reichardt, Annu. Rev. Biochem. Allied Res. India 72 (2003) 609. [3] A. Obermeier, R. Lammers, K.H. Wiesmuller, G. Jung, J. Schlessinger, A. Ullrich, J. Biol. Chem. 268 (1993) 22963. [4] S.O. Meakin, J.I. MacDonald, E.A. Gryz, C.J. Kubu, J.M. Verdi, J. Biol. Chem. 274 (1999) 9861. [5] J.I. MacDonald, E.A. Gryz, C.J. Kubu, J.M. Verdi, S.O. Meakin, J. Biol. Chem. 275 (2000) 18225. [6] N. Jones, S.H. Chen, C. Sturk, Z. Master, J. Tran, R.S. Kerbel, D.J. Dumont, Mol. Cell Biol. 23 (2003) 2658. [7] T.B. van Dijk, E. van Den Akker, M.P. Amelsvoort, H. Mano, B. Lowenberg, M. von Lindern, Blood 96 (2000) 3406. [8] X. Liang, D. Wisniewski, A. Strife, Shivakrupa, B. Clarkson, M.D. Resh, J. Biol. Chem. 277 (2002) 13732. [9] M.J. Wick, L.Q. Dong, D. Hu, P. Langlais, F. Liu, J. Biol. Chem. 276 (2001) 42843. [10] N. Jones, D.J. Dumont, Curr. Biol. 9 (1999) 1057. [11] N. Carpino, D. Wisniewski, A. Strife, D. Marshak, R. Kobayashi, B. Stillman, B. Clarkson, Cell 88 (1997) 197. [12] Y. Yamanashi, D. Baltimore, Cell 88 (1997) 205. [13] A. Di Cristofano, N. Carpino, N. Dunant, G. Friedland, R. Kobayashi, A. Strife, D. Wisniewski, B. Clarkson, P.P. Pandolfi, M.D. Resh, J. Biol. Chem. 273 (1998) 4827. [14] N. Jones, D.J. Dumont, Oncogene 17 (1998) 1097. [15] S. Lemay, D. Davidson, S. Latour, A. Veillette, Mol. Cell Biol. 20 (2000) 2743. [16] J. Grimm, M. Sachs, S. Britsch, S. Di Cesare, T. Schwarz-Romond, K. Alitalo, W. Birchmeier, J. Cell Biol. 154 (2001) 345. [17] R.J. Crowder, H. Enomoto, M. Yang, E.M. Johnson Jr., J. Milbrandt, J. Biol. Chem. 279 (2004) 42072.
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003 [18] R. Gugasyan, C. Quilici, S.T. I., D. Grail, A.M. Verhagen, A. Roberts, T. Kitamura, A.R. Dunn, P. Lock, J. Cell Biol. 158 (2002) 115. [19] I. Tamir, J.C. Stolpa, C.D. Helgason, K. Nakamura, P. Bruhns, M. Daeron, J.C. Cambier, Immunity 12 (2000) 347. [20] P. Van Slyke, M.L. Coll, Z. Master, H. Kim, J. Filmus, D.J. Dumont, Mol. Cell Biol. 25 (2005) 3831. [21] J.D. Robson, D. Davidson, A. Veillette, Mol. Cell Biol. 24 (2004) 2332. [22] T. Yasuda, M. Shirakata, A. Iwama, A. Ishii, Y. Ebihara, M. Osawa, K. Honda, H. Shinohara, K. Sudo, K. Tsuji, H. Nakauchi, Y. Iwakura, H. Hirai, H. Oda, T. Yamamoto, Y. Yamanashi, J. Exp. Med. 200 (2004) 1681. [23] M. Niki, A. Di Cristofano, M. Zhao, H. Honda, H. Hirai, L. Van Aelst, C. Cordon-Cardo, P.P. Pandolfi, J. Exp. Med. 200 (2004) 1689. [24] A. Di Cristofano, M. Niki, M. Zhao, F.G. Karnell, B. Clarkson, W.S. Pear, L. Van Aelst, P.P. Pandolfi, J. Exp. Med. 194 (2001) 275. [25] A. Bedirian, C. Baldwin, J. Abe, T. Takano, S. Lemay, J. Biol. Chem. 279 (2004) 19335. [26] S. Itoh, S. Lemay, M. Osawa, W. Che, Y. Duan, A. Tompkins, P.S. Brookes, S.S. Sheu, J.I. Abe, J. Biol. Chem. 280 (2005) 26383. [27] D. Cai, S. Dhe-Paganon, P.A. Melendez, J. Lee, S.E. Shoelson, J. Biol. Chem. 278 (2003) 25323. [28] H. Yamashita, S. Avraham, S. Jiang, I. Dikic, H. Avraham, J. Biol. Chem. 274 (1999) 15059. [29] R.M. Stephens, D.M. Loeb, T.D. Copeland, T. Pawson, L.A. Greene, D.R. Kaplan, Neuron. 12 (1994) 691.
2003
[30] N. Shi, W. Zhou, K. Tang, Y. Gao, J. Jin, F. Gao, X. Peng, M. Bartlam, B. Qiang, J. Yuan, Z. Rao, Acta Crystallogr., D Biol. Crystallogr. 58 (Pt 12) (2002) 2170. [31] B. Margolis, J. Lab. Clin. Med. 128 (1996) 235. [32] M.T. Uhlik, B. Temple, S. Bencharit, A.J. Kimple, D.P. Siderovski, G.L. Johnson, J. Mol. Biol. 345 (2005) 1. [33] T. Nakamura, R. Sanokawa, Y. Sasaki, D. Ayusawa, M. Oishi, N. Mori, Oncogene 13 (1996) 1111. [34] T. Nakamura, S. Muraoka, R. Sanokawa, N. Mori, J. Biol. Chem. 273 (1998) 6960. [35] H.Y. Liu, S.O. Meakin, J. Biol. Chem. 277 (2002) 26046. [36] F. Cong, B. Yuan, S.P. Goff, Mol. Cell Biol. 19 (1999) 8314. [37] N. Shi, S. Ye, M. Bartlam, M. Yang, J. Wu, Y. Liu, F. Sun, X. Han, X. Peng, B. Qiang, J. Yuan, Z. Rao, J. Biol. Chem. 279 (2004) 4962. [38] R. Klein, FASEB J. 8 (1994) 738. [39] C. Miranda, A. Greco, C. Miele, M.A. Pierotti, E. Van Obberghen, J. Cell Physiol. 186 (2001) 35. [40] H. Shinohara, T. Yasuda, Y. Yamanashi, Genes Cells 9 (2004) 601. [41] H. Shinohara, A. Inoue, N. Toyama-Sorimachi, Y. Nagai, T. Yasuda, H. Suzuki, R. Horai, Y. Iwakura, T. Yamamoto, H. Karasuyama, K. Miyake, Y. Yamanashi, J. Exp. Med. 201 (2005) 333.
Dok5 is substrate of TrkB and TrkC receptors and involved in neurotrophin induced MAPK activation Lei Shi a,1 , Jiping Yue a,1 , Yuangang You a , Bin Yin a , Yanhua Gong a , Caimin Xu a , Boqin Qiang a , Jiangang Yuan a , Yongjian Liu b,⁎, Xiaozhong Peng a,⁎ a
The National Laboratory of Medical Molecular Biology, Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences and Peking Union Medical College, Chinese National Human Genome Center, Beijing 100005, China b Department of Neurology, University of Pittsburgh School of Medicine, PA 15213, USA Received 4 February 2006; received in revised form 3 March 2006; accepted 10 March 2006 Available online 2 May 2006
Abstract Tropomyosin-related kinase (Trk) family receptors are a group of high affinity receptors for neurotrophin growth factors, which have pivotal functions in many physiological processes of nervous system. Trk receptors can dimerize and autophosphorylate upon neurotrophin stimulation, then recruit multiple adaptor proteins to transduct signal. In this report, we identified Dok5, a member of Dok family, as a new substrate of TrkB/C receptors. In yeast two-hybrid assay, Dok5 can interact with intracellular domain of TrkB and TrkC receptor through its PTB domain, but not with that of TrkA receptor. The interaction was then confirmed by GST pull-down assay and Co-IP experiment. Dok5 co-localized with TrkB and TrkC in differentiated PC12 cells, providing another evidence for their interaction. By using mutational analysis, we characterized that Dok5 PTB domain bound to Trk receptor NPQY motif in a kinase-activity-dependent manner. Furthermore, competition experiment indicated that Dok5 competed with N-shc for binding to the receptors at the same site. Finally, we showed that Dok5 was involved in the activation of MAPK pathway induced by neurotrophin stimulation. Taken together, these results suggest that Dok5 acts as substrate of TrkB/C receptors and is involved in neurotrophin induced MAPK signal pathway activation. © 2006 Elsevier Inc. All rights reserved. Keywords: Dok5; PTB domain; Trk receptor; NPQY motif; MAPK
1. Introduction The neurotrophin family growth factors, including nerve growth factor (NGF), brain-derived neurotrophic factors (BDNF), neurotrophin-3 (NT-3) and neurotrophin-4/5 (NT-4/5), have important functions in nervous system. They play multiple roles in vertebrate neural development, differentiation, survival, synaptic plasticity, myelination and many other processes [1,2]. Their high affinity receptors, tropomyosin-related kinase (Trk) family receptors, are cell membrane receptor tyrosine kinase (RTK). This receptor family has three structural conserved members: TrkA, TrkB and TrkC. Each neurotrophin binds to different receptors: NGF binds to TrkA, BDNF binds to TrkB, and NT-3 binds to ⁎ Corresponding authors. Tel.: +86 10 65296411; fax: +86 10 65240529. E-mail address: [email protected] (X. Peng). 1 These two authors contributed equally to this work. 0898-6568/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.cellsig.2006.03.007
TrkC. Also NT-3 can bind to TrkA and NT-4 can bind to TrkB [1,2]. Stimulation of neurotrophins can result in dimerization and activation of Trk receptors. Then the activated receptors phosphorylate several tyrosines of its intracellular domain (ICD). The phosphorylated tyrosines may serve as the docking site for signal molecules and mediate the signal transduction. Many signal molecules such as Src homology 2 domain containing transforming protein (Shc), FGF receptor substrate 2 (FRS2), growth factor receptor bound protein 2 (Grb2), phospholipase Cγ(PLC-γ) have been proved to be substrates of Trk receptors [3–5]. They can be phosphorylated by Trk receptors and cascade the signal to downstream. Downstream of tyrosine kinase/Docking proteins (Dok) are adaptor proteins that can act as substrate of multiple tyrosine kinase including both receptor tyrosine kinases and non-receptor tyrosine kinase [6–10]. The first member of Dok family, Dok1/ p62Dok, was identified by Carpino et al. [11] and Yamanashi and
1996
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
Baldimore [12]. Subsequently other members were identified, and now there are six members in this family: Dok1/p62Dok, Dok2/DokR/p56Dok, Dok3/DokL, Dok4, Dok5, and Dok6/ Dok5 like [13–17]. Dok proteins have classic feature of adaptor proteins: N-terminal pleckstrin homology (PH) domain, central phosphotyrosine binding (PTB) domain and C-terminal region. According to different expression patterns, Dok family proteins are divided into two subgroups. The first subgroup, Dok1– 3 are predominantly expressed in hematopoietic tissues and act as negative regulator in many signal pathways. Dok1/2 inhibit MAPK activation through their association with RasGTPaseactivating protein (RasGAP) [18,19]. Van Slyke et al. showed Dok2 attenuated EGFR signaling by recruiting c-Src and Csk, providing an alternative regulatory way [20]. Dok3, as the negative regulator of the immunoreceptor signaling in B cells and macrophages, exerts its function by the formation of Dok3SHIP1 complex [15,21]. Recently, studies from double or single knockout mice of Dok1 and/or Dok2 provided solid evidences in vivo for their negative roles in signal transduction and suppressant functions in leukemogenesis [22–24]. Another subgroup, Dok4–6 are expressed in non-hematopoietic tissues. Dok4 is widely expressed, especially in tissues of epithelial origin, and strongly inhibit Ret or Fyn signaling [25]. Itoh et al. found Dok4 localized in mitochondrial and
played a role in tumor necrosis factor-alpha (TNF-alpha) mediated reactive oxygen species (ROS) production [26]. Dok6, the newest member, is mainly expressed in nervous system and promotes Ret mediated neurite outgrowth in N2A-α1 cells [17]. And for Dok5, Northern blot and in situ hybrid results showed that Dok5 was mainly expressed in nervous system, especially in neural tube, dorsal root ganglia (DRG) and cranial ganglia. Studies also showed Dok5 was involved in insulin and GDNF signal pathway and it mediated GDNF induced neurite outgrowth in PC12 cells [16,27]. These results suggest Dok5 has very important function in neural development. But as recently identified member, its function remains mostly unclear. In this study, we identified Dok5 as novel substrate of TrkB/ C receptors. We proved Dok5 could selectively bind to NPQY motif of TrkB and TrkC receptors, not that of TrkA receptors in a kinase-activity-dependent manner. Co-localization of Dok5 and TrkB or TrkC receptors in differentiated PC12 cells provide another evidence for their interaction. In competition assay, we showed that Dok5 competed with N-shc for Trk receptor binding. Furthermore, we demonstrated that Dok5 was involved in the activation of MAPK signal pathway upon neurotrophin stimulation. Taken together, we proved that Dok5 could act as substrate of Trk receptors and was involved in neurotrophin induced MAPK activation. These results indicate that Dok5
Fig. 1. Dok5 interacts with TrkB and TrkC ICD in yeast two-hybrid system. (A) Schematic diagram shows the structures of Dok5 and Trk receptors. (B) Dok5 interacts with TrkB and TrkC ICD, not with TrkA ICD in yeast two-hybrid system. Yeast (SFY526) were transformed with Trk ICD in pAS2-1 vector and Dok5 in pACT2 vector, then cultured on SD/-Trp/-Leu plates for 4 days at 30 °C. Colony liftβ-galactosidase assays were performed according to the user manual. (C) Dok5 PTB domain, not PH domain or C-terminal region, binds to TrkB and TrkC ICD in yeast two-hybrid assay.
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
plays an important role in neurotrophin signal pathway, and also provide new clues for Dok5 function in neuronal development and differentiation. 2. Materials and methods
1997
esis Kit (Stratagene) [3,4,28,29]. For prokaryotic expression, the sequence coding for Dok5 PTB domain was subcloned into pGEX6P-1 vector. For eukaryotic expression, wild type and mutant Trk receptors were subcloned into expression vector pcDNA3.1V5. Dok5 full length cDNA and Dok5ΔC-terminal region(Dok5ΔC) were subcloned into expression vector pcDNA3.1Flag vector. For image analysis, TrkB and TrkC full length cDNA was subcloned into pmRFP-N1 vector and Dok5 cDNA were subcloned into pEGFP-N1 vector.
2.1. Reagents and cell culture 2.3. Yeast two-hybrid assay Human recombinant NGF, BDNF and NT-3 were purchased from Peprotech Inc. The anti-V5 polyclonal antibody was purchased from Novus. The anti-Flag monoclonal antibody was from Sigma. The anti-Myc monoclonal antibody (9E10) was purchased from Convance. The anti-Erk and anti-phosphop44/42Erk MAP kinase antibody were purchased from Cell signaling. The anti-pTyr (PY99) monoclonal antibody was product of Santa Cruz Biotechnology. Human Embryonic Kidney (HEK)293 cells were cultured in DMEM medium with 10% fetal bovine serum (Hyclone). PC12 cells were maintained in DMEM medium with 10% horse serum and 5% fetal bovine serum. For transfection, cells were split into 6-well plate and cultured to 90% confluency. Then the transfection was done with Lipofectamine2000 (Invitrogen) according to the user manual.
2.2. Plasmid construction The human TrkA full length cDNA was a kindly gift from Dr. Lloyd Greene. The human TrkB and TrkC full length cDNA were amplified from human fetal brain cDNA library. For yeast two-hybrid assay, the Trk receptors intracellular domains (ICD) were subcloned into pAS2-1 vector. Dok5 full length cDNA and the sequence coding for PH domain, PTB domain and C-terminal region were subcloned into pACT2 vectors, respectively. All the mutant TrkB and TrkC receptors, including TrkBM1(K572A), TrkBM2(Y516F), TrkBM3(Y702A), TrkBM4(Y706D), TrkBM5(Y707E), TrkBM6(Y817F), TrkCM1(K572A), TrkCM2(Y516F), TrkCM3(Y705A), TrkCM4(Y709D), TrkCM5(Y710E) and TrkCM6(Y820F) were generated by using QuikChange Site-Directed Mutagen-
For yeast two-hybrid assay, Trk receptor intracellular domains were fused to the GAL4 DNA binding domain in pAS2-1 vector. Wild type Dok5, or Dok5 PH domain, PTB domain and C-terminal region were fused to GAL4 activation domain in pACT2 vectors, respectively. The co-transformation was performed in yeast strain SFY526 by using TE/LiAc method according to the user manual of MATCHMAKER yeast two-hybrid system 2 (Clontech). Transformed yeast were selected on SD dropout medium/-Leu/-Trp plates at 30 °C for 4 days. Then colony liftβ-galactosidase assays were performed.
2.4. GST pull-down assay pGEX6P-Dok5-PTB plasmid was transformed into BL21 Escherichia coli strain. Then the overnight cultures were diluted 1:500 in 200 ml LB medium and cultured at 37 °C until OD600 to 0.6. The induction was performed by shaking at 30 °C for 4 h with 0.8 mM isopfopyl-β-D-thiogalactopyranoside. GST-Dok5PTB fusion proteins were immobilized onto Sepharose 4B beads according to the protocol (Amersham Biosciences). The HEK293 cells were transient transfected with pcDNA3.1V5-Trk plasmids. 24 h later, the cell were treated with or without neurotrophin for 10 min and lysed. The cell lysates were centrifuged at 12,000×g for 15 min and the supernatants were incubated with the same amount of GSTDok5-PTB Sepharose 4B beads at 4 °C for 2 h. The beads were washed with 1× PBS for 3 times and then resuspended in 1× SDS loading buffer and boiled for 5 min. Protein samples were analyzed by Western blotting or Coomassie Brilliant Blue staining.
Fig. 2. Dok5 PTB domain binds to TrkB and TrkC receptors in vitro. (A) Dok5 PTB domain binds to TrkB and TrkC receptors in GST pull-down assay. HEK293 cells were transfected with pcDNA3.1V5-Trk and treated with or without neurotrophin (100 ng/ml) for 10 min. The lysates were incubated with Dok5-PTB-GST Sepharose 4B beads at 4 °C for 2 h. The bound proteins were then analyzed by Western blotting with anti-V5 antibody (upper panel). Cell lysates were examined by anti-V5 or anti-pTyr antibody to detect the expression and phosphorylation level of Trk receptors (second and third panel). Dok5-PTB-GST proteins were examined by Coomassie Brilliant Blue staining (bottom panel). (B) Alignment of amino acids sequence around NPQY motif of Trk family receptors. Homology level: red, 100%; blue, ≥50%. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 3. Dok5 binds to NPQY motif of TrkB and TrkC receptors by a kinase-activity-dependent manner. (A) Schematic diagram shows the important amino acids of TrkB and TrkC. (B) Mutation analysis of the key amino acids of Trk receptor for Dok5 binding. Yeast two-hybrid assay was performed with pACT2-Dok5-PTB and TrkB or TrkC ICD mutants in pAS2-1 vector, including BM1(K572A), BM2(Y516F), BM3(Y702A), BM4 (Y706D), BM5(Y707E), BM6(Y817F), or CM1(K572A), CM2(Y516F), CM3(Y705A), CM4(Y709D), CM5(Y710E), CM6(Y820F). (C) Dok5 binds to NPQY motif of TrkB and TrkC receptors by a kinase-activitydependent manner in vivo. Co-IP experiments were done in HEK293 cells co-transfected with Dok5 and wild type or mutant receptors. Before harvest, the cells were treated with 100 ng/ml BDNF or NT-3 for 10 min. Protein samples were then detected by Western blotting with anti-V5, anti-Flag or anti-pTyr antibody. (D) Dok5 is phosphorylated by TrkB or TrkC receptors after stimulation. WT, wild type.
1998 L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
2.5. Co-immunoprecipitation HEK293 cells were plated into 6-well plate and cultured to 90% confluency. The cells were then transient co-transfected with pcDNA3.1V5-Trk and pcDNA3.1Flag-Dok5, pcDNA3.1V5-TrkM1 and pcDNA3.1Flag-Dok5, pcDNA3.1V5-TrkM2 and pcDNA3.1Flag-Dok5, or pcDNA3.1V5-TrkM6 and pcDNA3.1Flag-Dok5. 24 h after transfection, stimulate the cells with 100 ng/ml neurotrophin (PeproTech) for 10 min. Then cell were lysed in 750 μl lysis buffer (50 mM Tris–HCl pH 7.5, 1 mM EDTA, 120 mM NaCl, 0.5% NP-40, 10% glycerol, 1 mM Na3VO4, 25 mM NaF) containing protease inhibitors. After incubation on ice for 30 min, detergent-insoluble materials were removed by centrifugation at 4 °C at 12,000×g for 15 min. The 2-μg anti-Flag monoclonal antibody (Sigma) was added to the lysates and incubated at 4 °C rotating for 3 h followed by addition of 30 μl protein G-agarose (Roche). The incubation was continued for 3 h. The mixture was centrifuged at 12,000×g for 1 min, and the pellet was washed 3 times with 1 ml washing buffer (20 mM Tris–HCl pH 7.4, 1 mM EDTA, pH 8.0, 200 mM NaCl, 0.5% NP-40, 100 μM Na3VO4). The samples were analyzed by Western blotting with anti-V5 antibody (Invitrogen), anti-Flag antibody (Sigma) or anti-pTyr antibody (PY99, Santa Cruz).
2.6. Fluorescence microscopy pmRFPN1-TrkB or pmRFPN1-TrkC and pEGFPN1-Dok5 were co-transfected into PC12 cells. Twenty-four hours later, PC12 cells were split into 24-well plate with polylysine coated cover slips and cultured in low serum medium (1% equine serum) with 50 ng/ml BDNF or NT-3 for 48 h. Cells then were fixed with 4% PFA for 30 min at RT and washed with PBS for 3 times. The cover slips were mounted and co-localization was analyzed under a confocal imaging system.
2.7. MAPK signal pathway detection HEK293 cells were co-transfected with pcDNA3.1V5-Trk and pcDNA3. 1Flag, pcDNA3.1V5-Trk and pcDNA3.1Flag-Dok5, or pcDNA3.1V5-Trk and
1999
pcDNA3.1Flag-Dok5ΔC. The cells were treated with 100 ng/ml BDNF or NT-3 and harvested at 0 min, 5 min, 15 min, 30 min and 60 min. The lysates were examined by Western blotting with anti-Erk1 (k-23, Santa Cruz), anti-pErk (E-4, Santa Cruz), anti-Flag antibody (Sigma) or anti-V5 antibody (Invitrogen).
3. Results 3.1. Dok5 PTB domain binds to TrkB and TrkC ICD in yeast two-hybrid system Bioinformatics analysis and our previous structural studies showed that Dok5 has a conserved PTB domain [30], which make it possible for Dok5 to be the potential substrate of multiple tyrosine kinases. In consideration of Dok5 neural specific expression, we try to identify new tyrosine kinase expressed in nerve system that coupled with Dok5. Trk receptors, which are very important RTKs during neural development and differentiation, could be good candidates. To verify this possibility, we use yeast two-hybrid system to examine whether there are interactions between Dok5 and Trk receptors. There was evidence showing that the dimerization of the Gal4 BD domain could activate Trk ICD [5], which enable yeast two-hybrid assay to be used to detect the interaction between substrate and receptors. So we performed yeast two-hybrid with Trk ICD in pAS2-1 vector and Dok5 in pACT2 vector. Interestingly, the result showed that there were interactions between Dok5 and TrkB or TrkC ICD, while there was no obvious interaction between Dok5 and TrkA ICD (Fig. 1B). This implies that Dok5 has binding selectivity among
Fig. 4. Dok5 co-localizes with TrkB or TrkC in differentiated PC12 cells. PC12 cells were transfected with the constructs as indicated and induce by 50 ng/ml BDNF or NT-3 for 2 days.
2000
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
three members of Trk family receptors. In order to determine the interaction domain of Dok5, we further subcloned Dok5 PH domain, PTB domain and C-terminal region into pACT2 vector, respectively, and performed yeast two-hybrid assay again. As shown in Fig. 1C, Dok5 PTB domain is key domain responsible for the interaction with TrkB and TrkC ICD, while neither PH domain nor C-terminal region can interact with the receptors.
fusing protein could not bind to TrkA receptors no matter whether TrkA was activated or not (Fig. 2A). The consistent results from yeast two-hybrid and pull-down assay indicated that Dok5 selectively bind to TrkB and TrkC, but not TrkA.
3.2. Dok5 binds to TrkB and TrkC receptor in GST pull-down
We have proved that Dok5 interacted with the Trk receptors through its PTB domain. As known, PTB domain can recognize the NPQY motif by two manners: dependent or independent on phosphorylation [31,32]. To determine by which tyrosine and by which manner that Dok5 binds to Trk receptors, we constructed serial mutant Trk receptors for yeast two-hybrid assay and Co-IP experiment. As showed in Fig. 3A, Trk receptors have 5 tyrosine residues in their ICD that can be phosphorylated upon activation. For TrkB, they are Tyr516 , Tyr702, Tyr706, Tyr707 and Tyr817 . Each tyrosine mutation would disrupt the interaction between the receptor and the substrate binding at this site. The juxtamembrane Tyr516 lies in the conserved NPQY motif, which is typical binding motif for PTB domain. Many adaptor proteins containing PTB domain, such as Shc and Frs2, bind to Trk receptors at this site. And we supposed that this tyrosine is also the binding site for Dok5. Another key amino residue, Lys572, is essential for tyrosine kinase activity of the receptor. The mutation of this lysine would produce kinase inactive receptor. As expected, the yeast two-hybrid results proved that NPQY motif is the binding site for Dok5, because the mutant receptor Y516F lost the interaction with Dok5. Other tyrosine mutations have no obvious influence on the interaction. And the kinase inactive TrkB (K572A) had no interaction with Dok5 either, which implied that the kinase activity is essential for the interaction (Fig. 3B). There was similar result between Dok5 and TrkC. By Co-IP experiment, we further proved that Dok5 bind to NPQY motif of TrkB depending on phosphorylation. The result showed that wild type TrkB receptors could be immunoprecipitated by Dok5 upon BDNF stimulation; while the mutant receptor, TrkBM1 (K572A) or TrkBM2 (Y516F), could not be co-immunoprecipitated by Dok5 with the stimulation. The unrelated tyrosine mutation, TrkBM6 (Y817F), did not affect the interaction. Similar results were obtained between Dok5 and TrkC with stimulation of NT-3 (Fig. 3C). Moreover, as substrates of TrkB and TrkC, Dok5 phosphorylation level was greatly enhanced upon neurotrophin stimulation (Fig. 3D). These results strongly suggested that Dok5 binds to NPQY motif of TrkB/C receptors in a kinase-activity-dependent manner.
To confirm the results of yeast two-hybrid assay, we used GST pull-down assay to examine whether there are interaction between Dok5 and Trk receptors in vitro. The result proved that Dok5 PTB domain can bind to the activated TrkB and TrkC receptors (Fig. 2A). Upon neurotrophin stimulation, Trk receptors became highly phosphorylated comparing with those without stimulation. And only activated TrkB or TrkC receptors can be pulled down by GST-Dok5-PTB fusion protein. However, GST-Dok5-PTB
3.3. Dok5 binds to NPQY motif of TrkB and TrkC in a kinaseactivity-dependent manner
3.4. Dok5 co-localizes with TrkB and TrkC receptors in differentiated PC12 cells
Fig. 5. Dok5 and N-shc compete for TrkB (A) or TrkC (B) receptors binding. HEK293 cells were transfected with Trk, Dok5 and N-shc and then treated with 100 ng/ml BDNF or NT-3 for 10 min. Lysates were immunoprecipitated with anti-Flag antibody and the precipitated proteins were detected with anti-V5 and anti-Flag antibody. Cell lysates were detected with anti-V5 and anti-Myc antibody to examine the expression of Trk and N-shc.
PC12 cell line has been widely used as a good cell model for researches on neurotrophin and Trk receptors. With induction of NGF, PC12 cells would manifest characteristics of neuronal cells. So we choose PC12 cells for Dok5 and Trk receptors localization study. We constructed EGFPN1-Dok5 and mRFPN1-TrkB/C clones and co-transfected them into PC12 cells. The cells were
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
2001
Fig. 6. Dok5 mediate the activation of MAPK pathway upon neurotrophin stimulation in HEK293 cells. HEK293 cells were transfected with the constructs as indicated. The cells were treated with BDNF (A) or NT-3 (B) and harvested at 0 min, 5 min, 15 min, 30 min and 60 min. Cell lysates were examined by Western blotting with antibody indicated.
induced with 50 ng/ml BDNF or NT-3 in low concentration serum medium (1% equine serum) for 2 days and then analyzed by confocal microscopy. Obviously, Dok5 co-localize with TrkB and TrkC receptors in differentiated PC12 cells (Fig. 4). The colocalization of Dok5 and Trk receptors provided another evidence for their interaction. 3.5. Dok5 competes with N-shc for interaction with the same site of Trk receptor As well known, Shc can bind to the NPQY motif of Trk receptors through its PTB domain. In this study, we showed that Dok5 PTB domain also bind to the same site of Trk receptor. So it is possible that Dok5 competes with Shc for Trk receptor binding. Shc adaptor protein family comprises three members; ShcA, B and C. Among them, ShcC/N-shc is exclusively expressed in brain and also interact with Trk receptors [33–35]. We subcloned N-shc into pcDNA4Myc vector and performed a competition experiment. In competition assay, pcDNA3.1V5Trk, pcDNA3.1Flag-Dok5 and pcDNA4Myc-N-shc were cotransfected into HEK293 cells. The plasmid amounts of Trk and Dok5 were kept equal, with the plasmid amount of N-shc increasing. As indicated in Fig. 5, with the increasing expresion of N-shc, the amount of Trk receptors that interacted with Dok5 reduced gradually (Fig. 5). When there was a huge amount of Nshc, almost no Trk receptors were co-immunoprecipitated by
Dok5. This result provided a good competitive model for Dok5 and N-shc interacting with Trk receptor. 3.6. Dok5 is involved in the activation of MAPK signal pathway induced by BDNF and NT-3 To make clear the function of Dok5 in neurotrophin signaling, we detected its effects on MAPK signal pathway. We cotransfected Dok5 and Trk into HEK293 cells, another group the empty vector and Trk were used as control. Twenty-four hours after transfection, cells were treated with BDNF or NT-3 and then harvest cells at indicated time and detected MAPK pathway. Obviously, the existence of Dok5 enhanced the phosphorylation level of Erk1/2 compared with the control group. Meanwhile Dok5 also prolonged the activation time of MAPK pathway (Fig. 6). Previous studies showed that the C-terminal region of Dok protein was essential for signal transduction. Dok1/2 C-terminal region have multiple tyrosine residues and PXXP motifs that can serve as docking site for SH2/SH3 containing signal proteins [13,36]. The deletion of C-terminal region would disrupt their functions. So we were motivated to construct Dok5ΔC mutant and detect MAPK pathway again. As shown in Fig. 6, Dok5ΔC could inhibit MAPK pathway by reducing the strength and shorten the activation time. Thus, the result demonstrated that Dok5 participated in MAPK pathway activation induced by BDNF and NT-3.
2002
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003
4. Discussion In this study, by using yeast two-hybrid and pull-down assay, we identified Dok5 as substrate of TrkB and TrkC receptors. There was no obvious interaction between Dok5 and TrkA. As members of same family, TrkA, B and C share high homology in structure and they all have juxtamembrane NPQY motif for PTB domain binding. By aligning the amino acid sequence around NPQY motif, we found there were several amino acid residue deletions in TrkA compared with TrkB and TrkC receptors (Fig. 2B). These amino residues at the N-terminal or C-terminal of NPQY motif might provide the recognition specificity for the substrates. Interestingly, study from Shi et al. illustrated the structural basis of recognition between Dok1 PTB domain and c-Ret, and they also showed that Dok1 PTB domain did not recognize phosphopeptide including NPQY motif from TrkA [37]. The structural information from Dok1 may provide the clue for Dok5 specific recognition among Trk receptors. Furthermore, Dok5 selectively binding to TrkB and TrkC may suggest its distinct function in BDNF and NT-3 signal pathway. Though evidences showed that there might be functional redundancy among Trk signal pathways, studies also proved TrkA/B/C had nonoverlapping functions in neural development. The significant loss of neurons in BDNF/TrkB and NT-3/TrkC knockout mouse demonstrated their indispensable functions during sensory and motor neurons development [38]. The overlapped expression patterns of Dok5 and TrkB/C receptors and their interaction imply that Dok5 may have important functions in sensory and motor neurons development and survival. Moreover, we proved that Dok5 binds to TrkB and TrkC receptors NPQY motif in a kinase-activity-dependent manner and Dok5 is phosphorylated by Trk receptors upon neurotrophin stimulation. Co-localization of Dok5 and TrkB or TrkC receptors in differentiated PC12 cells provides another evidence for their physical interaction. The identification of Dok5 as substrate of TrkB/C receptors will enrich the signal network of neurotrophin pathway. Besides Dok5, many other adaptor proteins containing PTB domain have been reported to bind to the NPQY motif of Trk receptors, for example Shc, IRS and FRS2 [4,39]. It is hard to imagine that so many kinds of adaptor proteins bind to the receptor at the same site and at the same time. The most possibility is that they compete with each other for receptor binding. And our result established a competitive model between Dok5 and N-shc for Trk binding. This might be a good explanation for several adaptor proteins binding to the receptors at the same site to mediate different downstream signal pathway. But actual situation in vivo may be much more complicated. The different expression tissues and cell types of these adaptor proteins, and their different binding affinity with the receptor and many other factors should be considered. Unlike Dok1 and Dok2 as negative regulator in MAPK pathways [22,40,41], Dok5 mediated GDNF-induced MAPK activation [16]. Here we also showed that overexpression of Dok5 could augment MAPK activation induced by BDNF or NT-3; while the deletion of Dok5 C-terminal region disrupted this effect. This indicates that Dok5 C-terminal region is res-
ponsible for cascading the signal to downstream. However, till now downstream signal molecules that associate with Dok5 Cterminal region have not been identified yet. Compared with Dok1/2, Dok5 has a much shorter C-terminal region. Many SH2/ SH3 containing proteins that associated with Dok1/2, including RasGAP and NCK, showed no interaction with Dok5 [16,27]. Crowder et al. tried to find downstream signal proteins coupled with Dok4/5/6 subgroup by bioinformatics database search, but no functional known proteins were obtained [17]. So the studies of identifying the downstream partners of Dok5 become very important for explaining its function in signal transduction. In summary, we identified Dok5 as the new substrate of TrkB and TrkC receptors and it was involved in MAPK activation induced by BDNF and NT-3. Dok5 bind to the NPQY motif of Trk receptors in a kinase-activity-dependent manner through its PTB domain. These results enrich the neurotrophin signaling network and supply important information about Dok5 function in nervous system. Our studies as well as previous researches have showed that Dok5 was involved in TrkB/C, c-Ret, and insulin signal transduction, however the detailed regulatory mechanism and comprehensive functions of Dok5 in these pathways are yet unknown, more studies need to be done. Acknowledgements This work was supported by grants from the National Program for the Key Basic Research Project (‘973’ Nos. 2001CB510206, 2004CB518604 and 2005CB522507), and the National Sciences Foundation of China (Nos. 30421003 and 30430200). References [1] E.J. Huang, L.F. Reichardt, Annu. Rev. Neurosci. 24 (2001) 677. [2] E.J. Huang, L.F. Reichardt, Annu. Rev. Biochem. Allied Res. India 72 (2003) 609. [3] A. Obermeier, R. Lammers, K.H. Wiesmuller, G. Jung, J. Schlessinger, A. Ullrich, J. Biol. Chem. 268 (1993) 22963. [4] S.O. Meakin, J.I. MacDonald, E.A. Gryz, C.J. Kubu, J.M. Verdi, J. Biol. Chem. 274 (1999) 9861. [5] J.I. MacDonald, E.A. Gryz, C.J. Kubu, J.M. Verdi, S.O. Meakin, J. Biol. Chem. 275 (2000) 18225. [6] N. Jones, S.H. Chen, C. Sturk, Z. Master, J. Tran, R.S. Kerbel, D.J. Dumont, Mol. Cell Biol. 23 (2003) 2658. [7] T.B. van Dijk, E. van Den Akker, M.P. Amelsvoort, H. Mano, B. Lowenberg, M. von Lindern, Blood 96 (2000) 3406. [8] X. Liang, D. Wisniewski, A. Strife, Shivakrupa, B. Clarkson, M.D. Resh, J. Biol. Chem. 277 (2002) 13732. [9] M.J. Wick, L.Q. Dong, D. Hu, P. Langlais, F. Liu, J. Biol. Chem. 276 (2001) 42843. [10] N. Jones, D.J. Dumont, Curr. Biol. 9 (1999) 1057. [11] N. Carpino, D. Wisniewski, A. Strife, D. Marshak, R. Kobayashi, B. Stillman, B. Clarkson, Cell 88 (1997) 197. [12] Y. Yamanashi, D. Baltimore, Cell 88 (1997) 205. [13] A. Di Cristofano, N. Carpino, N. Dunant, G. Friedland, R. Kobayashi, A. Strife, D. Wisniewski, B. Clarkson, P.P. Pandolfi, M.D. Resh, J. Biol. Chem. 273 (1998) 4827. [14] N. Jones, D.J. Dumont, Oncogene 17 (1998) 1097. [15] S. Lemay, D. Davidson, S. Latour, A. Veillette, Mol. Cell Biol. 20 (2000) 2743. [16] J. Grimm, M. Sachs, S. Britsch, S. Di Cesare, T. Schwarz-Romond, K. Alitalo, W. Birchmeier, J. Cell Biol. 154 (2001) 345. [17] R.J. Crowder, H. Enomoto, M. Yang, E.M. Johnson Jr., J. Milbrandt, J. Biol. Chem. 279 (2004) 42072.
L. Shi et al. / Cellular Signalling 18 (2006) 1995–2003 [18] R. Gugasyan, C. Quilici, S.T. I., D. Grail, A.M. Verhagen, A. Roberts, T. Kitamura, A.R. Dunn, P. Lock, J. Cell Biol. 158 (2002) 115. [19] I. Tamir, J.C. Stolpa, C.D. Helgason, K. Nakamura, P. Bruhns, M. Daeron, J.C. Cambier, Immunity 12 (2000) 347. [20] P. Van Slyke, M.L. Coll, Z. Master, H. Kim, J. Filmus, D.J. Dumont, Mol. Cell Biol. 25 (2005) 3831. [21] J.D. Robson, D. Davidson, A. Veillette, Mol. Cell Biol. 24 (2004) 2332. [22] T. Yasuda, M. Shirakata, A. Iwama, A. Ishii, Y. Ebihara, M. Osawa, K. Honda, H. Shinohara, K. Sudo, K. Tsuji, H. Nakauchi, Y. Iwakura, H. Hirai, H. Oda, T. Yamamoto, Y. Yamanashi, J. Exp. Med. 200 (2004) 1681. [23] M. Niki, A. Di Cristofano, M. Zhao, H. Honda, H. Hirai, L. Van Aelst, C. Cordon-Cardo, P.P. Pandolfi, J. Exp. Med. 200 (2004) 1689. [24] A. Di Cristofano, M. Niki, M. Zhao, F.G. Karnell, B. Clarkson, W.S. Pear, L. Van Aelst, P.P. Pandolfi, J. Exp. Med. 194 (2001) 275. [25] A. Bedirian, C. Baldwin, J. Abe, T. Takano, S. Lemay, J. Biol. Chem. 279 (2004) 19335. [26] S. Itoh, S. Lemay, M. Osawa, W. Che, Y. Duan, A. Tompkins, P.S. Brookes, S.S. Sheu, J.I. Abe, J. Biol. Chem. 280 (2005) 26383. [27] D. Cai, S. Dhe-Paganon, P.A. Melendez, J. Lee, S.E. Shoelson, J. Biol. Chem. 278 (2003) 25323. [28] H. Yamashita, S. Avraham, S. Jiang, I. Dikic, H. Avraham, J. Biol. Chem. 274 (1999) 15059. [29] R.M. Stephens, D.M. Loeb, T.D. Copeland, T. Pawson, L.A. Greene, D.R. Kaplan, Neuron. 12 (1994) 691.
2003
[30] N. Shi, W. Zhou, K. Tang, Y. Gao, J. Jin, F. Gao, X. Peng, M. Bartlam, B. Qiang, J. Yuan, Z. Rao, Acta Crystallogr., D Biol. Crystallogr. 58 (Pt 12) (2002) 2170. [31] B. Margolis, J. Lab. Clin. Med. 128 (1996) 235. [32] M.T. Uhlik, B. Temple, S. Bencharit, A.J. Kimple, D.P. Siderovski, G.L. Johnson, J. Mol. Biol. 345 (2005) 1. [33] T. Nakamura, R. Sanokawa, Y. Sasaki, D. Ayusawa, M. Oishi, N. Mori, Oncogene 13 (1996) 1111. [34] T. Nakamura, S. Muraoka, R. Sanokawa, N. Mori, J. Biol. Chem. 273 (1998) 6960. [35] H.Y. Liu, S.O. Meakin, J. Biol. Chem. 277 (2002) 26046. [36] F. Cong, B. Yuan, S.P. Goff, Mol. Cell Biol. 19 (1999) 8314. [37] N. Shi, S. Ye, M. Bartlam, M. Yang, J. Wu, Y. Liu, F. Sun, X. Han, X. Peng, B. Qiang, J. Yuan, Z. Rao, J. Biol. Chem. 279 (2004) 4962. [38] R. Klein, FASEB J. 8 (1994) 738. [39] C. Miranda, A. Greco, C. Miele, M.A. Pierotti, E. Van Obberghen, J. Cell Physiol. 186 (2001) 35. [40] H. Shinohara, T. Yasuda, Y. Yamanashi, Genes Cells 9 (2004) 601. [41] H. Shinohara, A. Inoue, N. Toyama-Sorimachi, Y. Nagai, T. Yasuda, H. Suzuki, R. Horai, Y. Iwakura, T. Yamamoto, H. Karasuyama, K. Miyake, Y. Yamanashi, J. Exp. Med. 201 (2005) 333.